Avoiding Common ICU Errors, 1st Edition (2007)

Authors:
Marcucci, Lisa; Martinez, Elizabeth A. ; Haut, Elliott R. ; Slonim,

Anthony D. ; Suarez, Jose I.
T itle: A voi di ng Common I CU Errors, 1st Edi ti on
Copyright ©2007 Lippincot t Williams & Wilkins
> Fr ont of B ook > Author s
Authors
Lisa Marcucci MD
Assist ant Prof essor of Surgery
Di vi si on of Cri ti cal Care and Trauma Surgery, Department of Surgery, Thomas
Jef f erson Uni versi ty, Phi l adel phi a, Pennsyl vani a
Elizabeth A. Martinez MD, MHS

Cardi ac Sci ences, Assi stant Prof essor, Anesthesi ol ogy/ Cri ti cal Care Medi ci ne
and Surgery, Johns Hopki ns Uni versi ty, Medi cal Di rector, Adul t Post Anesthesi a
Care Uni ts, The Johns Hopki ns School of Medi ci ne, Bal ti more, Maryl and
Elliott R. Haut MD
Surgery, Assi stant Prof essor of Surgery and Anesthesi ol ogy and Cri ti cal Care
Medi ci ne, Di vi si on of Cri ti cal Care and Trauma Surgery, The Johns Hopki ns
Uni versi ty School of Medi ci ne, The Johns Hopki ns Hospi tal , Bal ti more,
Maryl and
Anthony D. Slonim MD, DrPH

Medi ci ne, Executi ve Di rector, Center f or Cl i ni cal Ef f ecti veness, Chi l dren's
Nati onal Medi cal Center, Associ ate Prof essor and Vi ce Chai rman, Department
of Pedi atri cs, The G eorge Washi ngton Uni versi ty School of Medi ci ne,
Washi ngton, Di stri ct of Col umbi a
Jose I. Suarez MD
Neurosci ences, Associ ate Prof essor, Department of Neurol ogy/ Neurosurgery,
Case Western Reserve Uni versi ty; Di rector, Neurosci ences Cri ti cal Care,
Department of Neurol ogy/ Neurosurgery, Uni versi ty Hospi tal s of Cl evel and,
Cl evel and, O hi o
Secondary Editors
Brian Brown
Acquisit ions Edit or
Nicole T. Dernoski
Managing Edit or
Angela Panetta
Market ing Manager
Bridgett Dougherty
Product ion Edit or
Benjamin Rivera
Senior Manuf act uring Manager
Risa Clow
Design Coordinat or
TechBooks
Composi tor
R. R. Donnelley, Crawfordsville
Pri nter
Associate Editors
Eric M. Bershad MD
Fellow
Neurosci ences Cri ti cal Care, Department of Neurol ogy, Case Western Reserve
Uni versi ty, Uni versi ty Hospi tal s of Cl evel and, Cl evel and, O hi o
Eliahu S. Feen MD
Fellow
Heidi L. Frankel MD
Associat e Prof essor of Surgery
Department of Surgery, UT Southwestern, Chi ef , Surgi cal Cri ti cal Care,
Parkl and Memori al Hospi tal , Dal l as, Texas
James H. Holmes IV MD
Chi ef , Burn Servi ces, Wake Forest Uni versi ty, Wi nston-Sal em, North Carol i na
Edward T. Horn PharmD

Clinical Associat e Prof essor
Uni versi ty of Maryl and School of Pharmacy, Cl i ni cal Pharmacy Speci al i st,
Surgi cal Intensi ve Care, Department of Pharmacy Servi ces, The Johns Hopki ns
Hospi tal , Bal ti more, Maryl and
Nirav G . Shah MD
Crit ical Care Fellow
Cri ti cal Care Medi ci ne Department, Nati onal Insti tutes of Heal th, Bethesda,
Maryl and
Bradford D. Winters MD, PhD

Assist ant Prof essor
Anesthesi ol ogy and Cri ti cal Care Medi ci ne, The Johns Hopki ns Uni versi ty
School of Medi ci ne, Bal ti more, Maryl and
Contributors
Yasir M. Akmal MD
Resident
Department of Surgery, G ei si nger Medi cal Center, Danvi l l e, Pennsyl vani a
Laith R. Altaweel MD
Department of Cri ti cal Care Medi ci ne, Nati onal Insti tutes of Heal th, Bethesda,
Maryl and
Rahul G . Baijal MD
Resident
Departments of Anesthesi a and Cri ti cal Care Medi ci ne, Johns Hopki ns,
Bal ti more, Maryl and
Amisha V. Barochia MD, CCMD

Post G raduat e Fellow
Cri ti cal Care Medi ci ne, Nati onal Insti tutes of Heal th, Bethesda, Maryl and
M. Craig Barrett PharmD

Surgery Clinical Specialist
Department of Pharmacy, Carol i nas Medi cal Center, Charl otte, North Carol i na
Mazen I. Bedri MD
Plast ic Surgery Resident
Department of Surgery, The Johns Hopki ns Uni versi ty School of Medi ci ne,
Johns Hopki ns Hospi tal , Bal ti more, Maryl and
Sean M. Berenholtz MD, MHS, FCCM

Anesthesi ol ogy & Cri ti cal Care Medi ci ne, Johns Hopki ns School of Medi ci ne,
Lauren C. Berkow MD
Anesthesi a & Cri ti cal Care Medi ci ne, The Johns Hopki ns School of Medi ci ne,
Staf f Anesthesi ol ogi st, Di rector of Di f f i cul t Ai rway Educati on, Anesthesi a &
Cri ti cal Care Medi ci ne, Johns Hopki ns Medi cal Insti tuti on, Bal ti more, Maryl and
Adam R. Berliner MD
Fellow in Nephrology
Department of Medi ci ne, Johns Hopki ns Uni versi ty School of Medi ci ne,
Eric M. Bershad MD
Fellow
Rachel Bluebond-Langner MD
Resident
Pl asti c & Reconstructi ve Surgery, The Johns Hopki ns Uni versi ty School of
Medi ci ne, Johns Hopki ns Hospi tal , Bal ti more, Maryl and
Aaron Bransky MD
Department of Surgery, Uni versi ty of Texas, Southwestern, Parkl and Hospi tal ,
Dal l as, Texas
Benjamin Braslow MD
Department of Surgery, Uni versi ty of Pennsyl vani a School of Medi ci ne, Di vi si on
of Traumatol ogy/ Surgi cal Cri ti cal Care, Department of Surgery, Hospi tal of the
Uni versi ty of Pennsyl vani a, Phi l adel phi a, Pennsyl vani a
E. David Bravos MD
Resident
Department of Anesthesi ol ogy and Cri ti cal Care Medi ci ne, Johns Hopki ns
Uni versi ty, Johns Hopki ns Hospi tal , Bal ti more, Maryl and
Benjamin S. Brooke MD
Post doct oral Fellow
Department of Surgery, Johns Hopki ns Uni versi ty, Resi dent, Department of
Surgery, The Johns Hopki ns Hospi tal , Bal ti more, Maryl and
Daniel R. Brown MD, PhD, FCCM

Chai r, Di vi si on of Cri ti cal Care Medi ci ne, Department of Anesthesi ol ogy, Mayo
Cl i ni c, Rochester, Mi nnesota
Brandon R. Bruns MD
Surgery Resident
Department of G eneral Surgery, Uni versi ty of Texas Southwestern, Parkl and
Hospi tal , Dal l as, Texas
William R. Burns MD
Resident
Department of Surgery, Johns Hopki ns Uni versi ty, The Johns Hopki ns Hospi tal ,
Christina L. Cafeo RN, MSN

Nurse Manager
Department of Surgery, Johns Hopki ns Hospi tal , Cardi ac Surgi cal Intensi ve
Care Uni t, Cardi ac Progressi ve Care Uni t, Bal ti more, Maryl and
Melissa S. Camp MD
Resident
Department of Surgery, Johns Hopki ns, Bal ti more, Maryl and
Molly B. Campion MS, CCC-SLP

Speech-Language Pat hologist
Physi cal Medi ci ne & Rehabi l i tati on, Johns Hopki ns Hospi tal , Bal ti more,
Maryl and
David J. Caparrelli MD
Resident
Di vi si on of Cardi ac Surgery, The Johns Hopki ns Uni versi ty School of Medi ci ne,
The Johns Hopki ns Medi cal Insti tuti ons, Bal ti more, Maryl and
Brendan G . Carr MD, MA

Fellow
Di vi si on of Trauma and Surgi cal Cri ti cal Care, Department of Surgery,
Instructor, Department of Emergency Medi ci ne, Hospi tal of the Uni versi ty of
Pennsyl vani a, Phi l adel phi a, Pennsyl vani a
Brett M. Cascio MD
Chief Resident
O rthopedi c Surgery, Johns Hopki ns, Bal ti more, Maryl and
Alan Cheng MD
Assist ant Prof essor of Medicine
Department of Medi ci ne-Cardi ovascul ar El ectrophysi ol ogy, Johns Hopki ns
Uni versi ty School of Medi ci ne, Bal ti more, Maryl and
Sara E. Cosgrove MD, MS

Di vi si on of Inf ecti ous Di seases, Johns Hopki ns Uni versi ty School of Medi ci ne,
Di rector, Anti bi oti c Management Program, Associ ate Hospi tal Epi demi ol ogi st,
Bryan A. Cotton MD
Department of Surgery, Vanderbi l t Uni versi ty School of Medi ci ne, Trauma &
Surgi cal Cri ti cal Care Attendi ng, Department of Surgery, Vanderbi l t Uni versi ty
Medi cal Center, Nashvi l l e, Tennessee
Peter F. Cronholm MD, MSCE

Fami l y Medi ci ne and Communi ty Heal th, Uni versi ty of Pennsyl vani a, Hospi tal of
the Uni versi ty of Pennsyl vani a, Phi l adel phi a, Pennsyl vani a
G regory Dalencourt MD
Resident
Department of Surgery, G ei si nger Medi cal Center, Danvi l l e, Pennsyl vani a
Constantine A. Demetracopoulos BS
Medical St udent
The Johns Hopki ns School of Medi ci ne, Bal ti more, Maryl and
Cameron Dezfulian MD
Clinical Fellow
Maryl and, Cl i ni cal Fel l ow, Pedi atri c Anesthesi a and Cri ti cal Care, Medi ci ne
Di vi si on, Johns Hopki ns Hospi tal , Bal ti more, Maryl and
J. Christopher DiG iacomo MD

Direct or of Trauma
Department of Surgery, Jersey Ci ty Medi cal Center, Jersey Ci ty, New Jersey
Michael J. Dorsi MD
Resident
Neurosurgery, Johns Hopki ns School of Medi ci ne, Bal ti more, Maryl and
Lesia K. Dropulic MD
Department of Medi ci ne, Johns Hopki ns Uni versi ty School of Medi ci ne, Co-
Di rector, Transpl ant & Inf ecti ous Di seases, Department of Medi ci ne, Johns
Hopki ns Hospi tal , Bal ti more, Maryl and
Muhammad I. Durrani MD
Fellow
Cardi ac Anesthesi ol ogy, The Johns Hopki ns Hospi tal , Bal ti more, Maryl and
David T. Efron MD
Department of Surgery, Johns Hopki ns School of Medi ci ne and Nursi ng, Johns
Michael J. Faulkner MD
Division Chief
Cardi ac Anesthesi ol ogy, Department of Anesthesi a and Peri operati ve Medi ci ne,
Wi l l i am Beaumont Hospi tal , Royal O ak, Mi chi gan
Eliahu S. Feen MD
Fellow
Derek M. Fine MD
Department of Medi ci ne, Di vi si on of Nephrol ogy, The Johns Hopki ns, School of
Medi ci ne, The Johns Hopki ns Hospi tal , Bal ti more, Maryl and
Heidi L. Frankel MD
Department of Surgery, UT Southwestern, Chi ef , Surgi cal Cri ti cal Care,
Parkl and Memori al Hospi tal , Dal l as, Texas
Frank J. Frassica MD
Robert A. Robinson Prof essor and Chair
Department of O rthopaedi cs, Johns Hopki ns Uni versi ty, Bal ti more, Maryl and
Iosifina G iannakikou MD
Anest hesiologist
Department of Cardi ac Anesthesi a, Metropol i tan Hospi tal , Athens, G reece
B. Robert G ibson MD
Anaesthesi a–Cri ti cal Care Medi ci ne, The Johns Hopki ns Hospi tal , Bal ti more,
Maryl and
Ashita G oel MD
Resident
Anesthesi ol ogy and Cri ti cal Care Medi ci ne, The Johns Hopki ns Uni versi ty,
Sherita Hill G olden MD, MHS

Assist ant Prof essor of Medicine and Epidemiology
Department of Medi ci ne, Di vi si on of Endocri nol ogy and Metabol i sm, The Johns
Hopki ns School of Medi ci ne, Chai rperson, G l ucose Control Task Force, The
Kelly L. G rogan MD
Department of Anesthesi ol ogy and Cri ti cal Care Medi ci ne, Johns Hopki ns
Michael D. G rossman MD, FACS

Associat e Clinical Prof essor
Department of Surgery, Uni versi ty of Pennsyl vani a School of Medi ci ne, Chi ef ,
Di vi si on of Trauma/ Surgi cal , Cri ti cal Care, St. Luke's Hospi tal , Bethl ehem,
Pennsyl vani a
Rajan G upta MD
Department of Surgery, Dartmouth Medi cal School , Department of Surgery,
Dartmouth Hi tchcock Medi cal Center, Lebanon, New Hampshi re
Jacob T. G utsche MD
Fellow
Surgi cal Cri ti cal Care, Department of Anesthesi ol ogy and Cri ti cal Care,
Uni versi ty of Pennsyl vani a School of Medi ci ne, Phi l adel phi a, Pennsyl vani a
Barbara Haas MD
Resident
Department of Surgery, Uni versi ty of Toronto, Toronto, O ntari o, Canada
Ala' S. Haddadin MD
At t ending Anest hesiologist
Assi stant Prof essor, Department of Anesthesi a, Yal e Uni versi ty School of
Medi ci ne, New Haven, Connecti cut
David N. Hager MD
Fellow
Departments of Pul monary and Cri ti cal Care Medi ci ne, Johns Hopki ns Medi cal
Insti tuti ons, Bal ti more, Maryl and
Nancy Sokal Hagerman MD

Fellow
Pedi atri c Anesthesi ol ogy, Department of Anesthesi a, Ci nci nnati Chi l dren's
Hospi tal , Medi cal Center, Ci nci nnati , O hi o
Nadia N. Hansel MD, MPH

I nst ruct or of Medicine
Department of Medi ci ne, The Johns Hopki ns School of Medi ci ne, Bal ti more,
Maryl and
Elliott R. Haut MD
Assist ant Prof essor of Surgery and Anest hesiology and Crit ical Care Medicine
Di vi si on of Cri ti cal Care and Trauma Surgery, Department of Surgery, The
Johns Hopki ns School of Medi ci ne, The Johns Hopki ns Hospi tal , Bal ti more,
Maryl and
Michael Joel Haut MD

Clinical Prof essor
Department of Medi ci ne, Uni versi ty of Pennsyl vani a School of Medi ci ne,
Hematol ogi st/ O ncol ogi st, Department of Medi ci ne, Pennsyl vani a Hospi tal , Joan
Karnel l Cancer Center, Phi l adel phi a, Pennsyl vani a
Awori J. Hayanga MD
Resident
G eneral Surgery, Department of G eneral Surgery, Uni versi ty of Mi chi gan Heal th
Systems, Ann Arbor, Mi chi gan
Eugenie S. Heitmiller MD, FAAP

Associat e Prof essor
Anesthesi ol ogy and Cri ti cal Care Medi ci ne, The Johns Hopki ns, School of
Medi ci ne, Vi ce Chai rman f or Cl i ni cal Af f ai rs, Anesthesi ol ogy and Cri ti cal Care
Medi ci ne, Johns Hopki ns Uni versi ty Hospi tal , Bal ti more, Maryl and
J. G regory Hobelmann MD
Fellow
Department of Anesthesi ol ogy and Cri ti cal Care Medi ci ne, Johns Hopki ns,
Deborah B. Hobson BSN

Surgical I nt ensive Care Nurse
Coach f or Center f or Innovati on i n Q ual i ty Pati ent Care, Johns Hopki ns
William S. Hoff MD
Clinical Associat e Prof essor of Surgery
Uni versi ty of Pennsyl vani a Medi cal Center, Phi l adel phi a, Pennsyl vani a, Trauma
Program Medi cal Di rector, St. Luke's Hospi tal , Bethl ehem, Pennsyl vani a
Department of Surgery, Wake Forest Uni versi ty School of Medi ci ne, Medi cal
Di rector, Burn Center, Department of Surgery, Wake Forest Uni versi ty Bapti st
Medi cal Center, Wi nston-Sal em, North Carol i na
Edward T. Horn PharmD, BCPS

Uni versi ty of Maryl and School of Pharmacy, Cl i ni cal Pharmacy Speci al i st,
Surgi cal Intensi ve Care, Department of Pharmacy Servi ces, The Johns Hopki ns
Leo Hsiao DO
Housest aff
Department of Anesthesi ol ogy and Cri ti cal Care, Johns Hopki ns Uni versi ty, The
David G . Hunt RN, BSN

Clinical Nurse Educat or
Department of Cardi ac Surgery, The Johns Hopki ns Hospi tal , Bal ti more,
Maryl and
Elizabeth A. Hunt MD, MPH

Anesthesi ol ogy & Cri ti cal Care Medi ci ne, The Johns Hopki ns School of
Medi ci ne, Di rector, Johns Hopki ns Si mul ator Center, Bal ti more, Maryl and
Jacqueline Janka MD
Inf ecti ous Di seases and Cri ti cal Care, Nati onal Insti tutes of Heal th, Bethesda,
Maryl and
Praveen Kalra MD
Department of Anesthesi a, O kl ahoma Uni versi ty Heal th Sci ence, Center,
O kl ahoma Ci ty, O kl ahoma
Andrew J. Kerwin MD
Medi cal Di rector, Surgi cal Intensi ve Care Uni t, Department of Surgery,
Uni versi ty of Fl ori da Heal th Sci ences Center – Jacksonvi l l e, Jacksonvi l l e,
Fl ori da
Suneel Khetarpal MD
Department of Surgery, Uni versi ty of Fl ori da Heal th Sci ences, Center –
Jacksonvi l l e, Jacksonvi l l e, Fl ori da
Patrick K. Kim MD
Department of Surgery, Uni versi ty of Pennsyl vani a School of Medi ci ne,
Attendi ng Surgeon, Department of Surgery, Hospi tal of the Uni versi ty of
Benjamin Kratzert MD
Resident
Anesthesi ol ogy, Department of Anesthesi ol ogy, Uni versi ty of Cal i f orni a, San
Di ego, San Di ego, Cal i f orni a
Lee Ann R. Lau MD

Chief Resident
Department of Surgery, Uni versi ty of Texas Southwestern, Dal l as, Texas
Noah Lechtzin MD, MHS

Maryl and
John J. Lewin III PharmD, BCPS

Clinical Assist ant Prof essor
Uni versi ty of Maryl and School of Pharmacy, Cl i ni cal Speci al i st, Neurosci ences
Cri ti cal Care, Department of Pharmacy Servi ces, The Johns Hopki ns Hospi tal ,
Jayme E. Locke MD
Resident
Department of Surgery, The Johns Hopki ns Hospi tal , Bal ti more, Maryl and
Ying Wei Lum MD

Resident
Shelley Sylvester Magill MD

Department of Medi ci ne, The Johns Hopki ns Uni versi ty School of Medi ci ne,
Acti ve Staf f , Di vi si on of Inf ecti ous, Di sease, Department of Medi ci ne, The
Warren R. Maley MD
Department, of Surgery, Di vi si on of Transpl antati on, The Johns Hopki ns School
of Medi ci ne, Bal ti more, MD
D. Joshua Mancini MD
Resident
Department of G eneral Surgery, Dartmouth Hi tchcock Medi cal Center, Lebanon,
New Hampshi re
Benjamin A. Mandel MD
Resident
Department of Surgery, Uni versi ty of Wi sconsi n, Uni versi ty of Wi sconsi n
Hospi tal and Cl i ni cs, Madi son, Wi sconsi n
G ary T. Marshall MD
I nst ruct or of Surgery
Trauma and Cri ti cal Care, Vanderbi l t Uni versi ty Medi cal School ; Instructor of
Surgery, Trauma and Cri ti cal Care, Vanderbi l t Uni versi ty Medi cal Center,
Nashvi l l e, Tennessee

Anesthesi ol ogy/ Cri ti cal Care Medi ci ne and Surgery, The Johns Hopki ns School
of Medi ci ne, Medi cal Di rector, Adul t Post Anesthesi a Care Uni ts, Johns Hopki ns
Medi ci ne, Bal ti more, Maryl and
Susanna Lovell Matsen MD

Halst ed Resident
Department of Surgery, Johns Hopki ns Uni versi ty, The Johns Hopki ns Hospi tal ,
Madhavi Meka
Christian Merlo MD, MPH

I nst ruct or
Maryl and
William G . Merz PhD

Prof essor of Pat hology (Microbiology)
Department of Pathol ogy, The Johns Hopki ns School of Medi ci ne; Di rector of
the Mycol ogy Lab, Co-Di rector of The Mol ecul ar Epi demi ol ogy Lab, Department
of Pathol ogy, The Johns Hopki ns Hospi tal , Bal ti more, Maryl and
Robert K. Michaels MD, MPH

Resident
Anesthesi ol ogy and Cri ti cal Care Medi ci ne, The Johns Hopki ns Hospi tal ,
Jennifer Miles-T homas MD

Department of Urol ogy, The Johns Hopki ns Uni versi ty, Bal ti more, Maryl and
Stephen M. Milner MD, FACS

Direct or
Department of Burn Surgery, Johns Hopki ns Medi cal Insti tuti on, Bal ti more,
Maryl and
Anushirvan Minokadeh MD
Assist ant Clinical Prof essor
Department of Anesthesi ol ogy, Uni versi ty of Cal i f orni a, San Di ego Medi cal
Center, San Di ego, Cal i f orni a
Ardalan Minokadeh BS
Tul ane Uni versi ty Heal th Sci ences Center, MD/ PhD Candi date, New O rl eans,
Loui si ana
T imothy M. Moore MD, PhD

Resident
Departments of Anesthesi ol ogy and Cri ti cal Care Medi ci ne, Johns Hopki ns
Medi cal Insti tuti ons, Bal ti more, Maryl and
Michael J. Moritz MD
Chief
Secti on of Transpl antati on Servi ces, Lehi gh Val l ey Hospi tal , Al l entown,
Pennsyl vani a
Dana Y. Nakamura O T, CLT, CLMC

O ccupati onal Therapy Cl i ni cal , Speci al i st—Burns, Department of Rehabi l i tati on
Medi ci ne, Uni versi ty of Washi ngton, Burn Center, Harborvi ew Medi cal Center,
Seattl e, Washi ngton
Hari Nathan MD
Research Fellow
Department of Surgery, The Johns Hopki ns School of Medi ci ne, House Staf f ,
Beverly J. Newhouse MD
Resident
Department of Anesthesi ol ogy, UCSD Medi cal Center, San Di ego, Cal i f orni a
Shaytone Nichols MD
Kelly O rlino MD
Resident
Department of Surgery, Johns Hopki ns Hospi tal , Bal ti more, Maryl and
Lawrence O sei MD
Clinical Fellow
Cri ti cal Care Medi ci ne, Nati onal Insti tutes of Heal th, Bethesda, Maryl and
Mehmet S. O zcan MD
Department of Anesthesi ol ogy, Uni versi ty of O kl ahoma, Attendi ng
Anesthesi ol ogi st, Department of Anesthesi ol ogy, O kl ahoma Uni versi ty Heal th
Sci ences Center, O kl ahoma Ci ty, O kl ahoma
Derek F. Papp MD
Resident Physician
O rthopaedi c Surgery, The Johns Hopki ns School of Medi ci ne, The Johns
B. Lauren Paton MD
Resident
G eneral Surgery, Carol i nas Medi cal Center, Charl otte, North Carol i na
Ronald W. Pauldine MD
Anesthesi ol ogy and Cri ti cal Care Medi ci ne, Johns Hopki ns Uni versi ty, Vi ce
Chai r, Anesthesi ol ogy and Cri ti cal Care Medi ci ne, Johns Hopki ns Bayvi ew
Medi cal Center, Bal ti more, Maryl and
Travis L. Perry MD
Surgery Crit ical Care Fellow
Department of Surgery, The Uni versi ty of Texas Medi cal Branch, Burn Surgery
Fel l ow, Department of Burn Surgery, Shri ners Hospi tal f or Chi l dren, G al veston,
Texas
Julius Cuong Pham MD

Department of Anesthesi a/ Cri ti cal Care Medi ci ne, Department of Emergency
Medi ci ne, The Johns Hopki ns School of Medi ci ne, The Johns Hopki ns Hospi tal ,
Myron S. Powell MD
G eneral Surgery Resident
Department of G eneral Surgery, Wake Forest Uni versi ty, Bapti st Medi cal
Center, Wi nston-Sal em, North Carol i na
Peter J. Pronovost MD, PhD

Prof essor
Department of Anesthesi ol ogy and Cri ti cal Care Medi ci ne, The Johns Hopki ns,
Juan N. Pulido MD
Chief Resident Associat e
Resi dent, Department of Anesthesi ol ogy, Mayo Cl i ni c Col l ege of Medi ci ne,
Rochester, Mi nnesota
Jeremy W. Pyle MD
Resident
Department of Pl asti c and Reconstructi ve Surgery, Wake Forest Uni versi ty,
Bapti st Medi cal Center, Wi nston-Sal em, North Carol i na
Melvin K. Richardson MD
Resident
Department of Anesthesi a and Cri ti cal Care Medi ci ne, The Johns Hopki ns
Jose Manuel Rodriguez-Paz MD

Anesthesi ol ogy and Cri ti cal Care Medi ci ne, Johns Hopki ns Uni versi ty, Johns
Hopki ns Medi cal Insti tuti ons, Bal ti more, Maryl and
Frank Rosemeier MD
Department of Anesthesi ol ogy, Peri operati ve Medi ci ne and Pai n Management,
Uni versi ty of Mi ami Mi l l er School of Medi ci ne, Jackson Memori al Hospi tal ,
Mi ami , Fl ori da
Andrew L. Rosenberg MD
Chi ef , Di vi si on of Cri ti cal Care, Departments of Anesthesi ol ogy, Cri ti cal Care
and Internal Medi ci ne, Uni versi ty of Mi chi gan Medi cal Center, Ann Arbor,
Mi chi gan
Tuhin K. Roy MD, PhD

Assist ant Prof essor of Anest hesiology
Department of Anesthesi ol ogy, Mayo Cl i ni c Col l ege of Medi ci ne; Consul tant
Department of Anesthesi ol ogy, Mayo Cl i ni c, Rochester, Mi nnesota
Deba Sarma MD
Resident
Johns Hopki ns Uni versi ty, Bal ti more, Maryl and
Prasert Sawasdiwipachai MD
I nst ruct or
Department of Anesthesi ol ogy, Mahi dol Uni versi ty, Instructor, Department of
Anesthesi ol ogy, Si ri raj Hospi tal , Bangkok, Thai l and
Patrick Schaner MD
Jef f erson Medi cal Col l ege, G eneral Surgery Resi dent, Department of Surgery,
Thomas Jef f erson Uni versi ty Hospi tal , Phi l adel phi a, Pennsyl vani a
Dorry L. Segev MD
I nst ruct or
Surgi cal Staf f , Department of Transpl ant Surgery, The Johns Hopki ns Hospi tal ,
Ashish S. Shah MD
Di vi si on of Cardi ac Surgery, The Johns Hopki ns School of Medi ci ne, The Johns
Nirav G opal Shah MD

Maryl and
Richard Wong She MHB (Hons), MBChB, FRACS (Plastics)

Consult ant Burn and Plast ic Surgeon
Nati onal Burn Centre and Regi onal Centre f or Reconstructi ve Pl asti c
Maxi l l of aci al and Hand Surgery, Mi ddl emore Hospi tal , Auckl and, New Zeal and
Kristin Shipman MD
Assist ant I nst ruct or of Surgery
UT Southwestern Medi cal Center, Trauma/ Surgi cal Cri ti cal Care Fel l ow,
Parkl and Hospi tal , Di vi si on of Burns, Trauma, Cri ti cal Care, Dal l as, Texas
Angela D. Shoher MD
Resident
Department of Surgery, Johns Hopki ns Uni versi ty, Bal ti more, Maryl and
Carrie A. Sims MD, MS

Di vi si on of Surgery, Uni versi ty of Pennsyl vani a, Di vi si on of Trauma and
Surgi cal Cri ti cal Care, Hospi tal at the Uni versi ty of Pennsyl vani a, Phi l adel phi a,
Pennsyl vani a
Ronald F. Sing DO
Department of Surgery, Uni versi ty of North Carol i na at Chapel Hi l l , Facul ty,
Department of Surgery, Carol i nas Medi cal Center, Charl otte, North Carol i na
Harjot K. Singh MD
Clinical Fellow
Department of Inf ecti ous Di sease, The Johns Hopki ns Hospi tal , Bal ti more,
Maryl and
Vijay Anand Singh MD

Burn Fellow
Department of Burn Surgery Johns Hopki ns Medi cal Insti tute, Bal ti more,
Maryl and
Tammy Slater CRNP
Acut e Care Nurse Pract it ioner

Execut ive Direct or
Center f or Cl i ni cal Ef f ecti veness, Chi l dren's Nati onal Medi cal Center, Associ ate
Prof essor and Vi ce Chai rman, Department of Pedi atri cs, The G eorge
Washi ngton Uni versi ty School of Medi ci ne, Washi ngton, DC
Christopher J. Sonnenday MD, MHS

Clinical Lect urer
Department of Surgery, The Uni versi ty of Mi chi gan, Fel l ow i n Transpl antati on,
Department of Surgery, The Uni versi ty of Mi chi gan Hospi tal , Ann Arbor,
Mi chi gan
Konstantinos Spaniolas MD
Research Fellow
Department of Surgery, Harvard Medi cal School , Research Fel l ow, Department
of Surgery, Massachusetts G eneral Hospi tal , Boston, Massachusetts
Jason L. Sperry MD
Trauma Crit ical Care Fellow
Department of Surgery, Uni versi ty of Texas Southwestern Medi cal Center,
Dal l as, Texas
Dimitris Stefanidis MD, PhD

Fellow
G eneral Surgery, Carol i nas Medi cal Center, Charl otte, North Carol i na
Joseph B. Straton MD, MSCE

Fami l y Medi ci ne and Communi ty Heal th, Uni versi ty of Pennsyl vani a,
Phi l adel phi a, Pennsyl vani a
Michael B. Streiff MD
Departments of Medi ci ne and Hematol ogy, Johns Hopki ns Medi cal Insti tuti ons,
Assi stant Prof essor, Attendi ng Physi ci an, Department of Medi ci ne, The Johns
Jose I. Suarez MD
Department of Neurol ogy/ Neurosurgery, Case Western Reserve Uni versi ty,
Di rector, Neurosci ences Cri ti cal Care, Department of Neurol ogy/ Neurosurgery,
Uni versi ty Hospi tal s of Cl evel and, Cl evel and, O hi o
Aruna K. Subramanian MD
Di vi si on of Inf ecti ous Di seases, Johns Hopki ns Uni versi ty, Co-Di rector,
Transpl ant Inf ecti ous Di seases, Department of Medi ci ne, The Johns Hopki ns
Sandra M. Swoboda RN, MSN

Senior Research Program Coordinat or
Department of Surgery, The Johns Hopki ns School of Medi ci ne, Bal ti more,
Maryl and
Meghan C. Tadel MD
Resident Physician
Anesthesi ol ogy and Cri ti cal Care Medi ci ne, Johns Hopki ns Uni versi ty,
Peter G . T homas DO
Trauma and Surgical Crit ical Care Fellow
Department of Surgery, Uni versi ty of Pennsyl vani a, Phi l adel phi a, Pennsyl vani a
G len T inkoff MD
Department of Surgery, Thomas Jef f erson Medi cal Col l ege, Phi l adel phi a,
Pennsyl vani a, Medi cal Di rector, Trauma Program, Chri sti ana Care Heal th
System, Newark, Del aware
Meredith S. T inti MD
Fellow
Departments of Trauma and Surgi cal Cri ti cal Care, Hospi tal of the Uni versi ty of
O liver A. Varban MD
Resident
Department of Surgery, Wake Forest Uni versi ty, Wi nston-Sal em, North Carol i na
G eorge C. Velmahos MD, PhD, MSEd

Prof essor of Surgery
Harvard Medi cal School , Chi ef , Trauma, Emergency Surgery and Surgi cal
Cri ti cal Care, Massachusetts G eneral Hospi tal , Boston, Massachusetts
Patricia M. Veloso MD
I nst ruct or
Department of Anesthesi ol ogy & Cri ti cal Care Medi ci ne, The Johns Hopki ns
Tonya N. Walker MD
House St aff
Emergency Department, New York Presbyteri an Hospi tal , Cornel l and Col umbi a,
New York, New York
Eric S. Weiss MD
Resident in G eneral Surgery
Department of Surgery, The Johns Hopki ns School of Medi ci ne, The Johns
Matthew J. Weiss MD
Resident
James F. Weller MD
St aff Anest hesiologist
Department of Anesthesi ol ogy, Bethesda North Hospi tal , Ci nci nnati , O hi o
Kathleen A. Williams RN, MSN CRNP

I npat ient Diabet es Nurse Pract it ioner
Department of Medi ci ne, Di vi si on of Endocri nol ogy, The Johns Hopki ns

Anesthesi ol ogy and Cri ti cal Care Medi ci ne, The Johns Hopki ns School of
Medi ci ne, Bal ti more, Maryl and
John Zannis MD
Resident
Pl asti c & Reconstructi ve Surgery, Wake Forest Uni versi ty, Wi nston-Sal em,
North Carol i na
Faramarz Zarfeshanfard RPh

Pharmacist
Department of Pharmacy, The Johns Hopki ns Hospi tal , Bal ti more, Maryl and
Authors: Marcucci, Lisa; Martinez, Elizabeth A. ; Haut, Elliott R. ; Slonim,
Copyright ÂŠ2007 Lippincot t Williams & Wilkins
> Fr ont of B ook > D edic ation
Dedication
To Bret t , f or his support , love, and inspirat ion and Jayne and Arenal, t he best
lit t le f amily ever and Mariana and Ana Maria
> Fr ont of B ook > P r efac e
Preface
I n t he six years since t he I nst it ut e of Medicine released it s landmark report “To

Err is Human, ” progress t ow ard improving pat ient saf et y has been slow and
arduous. Clinicians and researchers st ruggle t o advance t he science of pat ient
saf et y, underst and it s epidemiology, clarif y priorit ies, implement scient if ically
sound yet f easible int ervent ions, and develop measures t o evaluat e progress. As
Robert Frost said, “… w e have miles t o go bef ore w e sleep. ”
As errors have become more visible and our pat ient s cont inue t o suff er
prevent able harm, pat ient s, regulat ors, accredit at ors, and caregivers have grow n
f rust rat ed. While t here is broad consensus t hat f ault y syst ems rat her t han f ault y
people cause most errors, healt hcare w orkers st ruggle t o f ind pract ical and
sound w ays t o address and mit igat e hazards.
I nt ensive care unit s provide lif e-sust aining t herapies, but are also a place w here
harm lurks. I CUs are highly complex environment s in w hich clinicians make t ime-
pressured decisions f or pat ient s w it h limit ed physiological reserves. This
combinat ion of condit ions makes I CU care at t imes hazardous f or vulnerable
pat ient s.
To make progress t ow ard reducing t his harm, w e need t o expose t he mist akes,
develop st rat egies t o reduce t hem, and evaluat e our progress. This book moves
us in t hat direct ion. I t includes a broad array of commonly-made I CU errors and
expert review regarding st rat egies t o mit igat e t hem. These review s are w rit t en
by pract icing clinicians w ho underst and t he science of improving pat ient saf et y.
The inf ormat ion cont ained in t his book is t imely, accurat e, and pract ical. I magine
w hat I CU care w ould look like if w e could eliminat e all t he hazards ident if ied in
t his book. As such, reading t his w ell-w rit t en book is a f irst st ep on a journey t o
improve pat ient saf et y.
Pet er Pronovost MD, PhD
Balt imore, Maryland
The edit ors w elcome comment s and suggest ions regarding t his book and request
t hat t hey be sent t o: lisa. marcucci@gmail. com
FRONT OF BOOK ↑
[+] Authors
[+] Editors
- Dedication
- Preface
TABLE OF CONTENTS ↑
I. Medications
[+] 1 - Monitor Patients who have Received Intrathecal Preservative-Free Morphine
[+] 2 - Know the Characteristics of the Narcotics You Prescribe
[+] 3 - Consider Clonidine to Combat Effects of Drug Withdrawal
[+] 4 - Strongly Consider Prophylaxis for Alcohol Withdrawal
[+] 5 - Avoid Concomitant Use of Steroids, Neuromuscular Blockade, and
Aminoglycosides to Lessen the Risk of Critical Illness Myopathy
[+] 6 - Use Prophylaxis for the Immediate Side Effects of Steroids
[+] 7 - Specifically Query for Previous Steroid Use
[+] 8 - Do Not Use Succinylcholine in Patients with Burns, Paralysis, or Other High
Potassium States
[+] 9 - Consider Using Cisatracurium for Neuromuscular Paralysis in Patients with Hepatic
and Renal Failure
[+] 10 - Remember that there are two “Neos”
[+] 11 - Treat Neuroleptic Malignant Syndrome as an Emergency and Remember its
Presentation may not be Dose Dependent
[+] 12 - Remember that Malignant Hyperthermia may not have Hyperthermia
[+] 13 - Remember That Amiodarone Causes Hypothyroidism
[+] 14 - Do not use Amiodarone to Rate Control Chronic Atrial Fibrillation
[+] 15 - Exercise Care in the use of Amiodarone and Alternative Antiarrhythmics for the
Treatment of Atrial Fibrillation
[+] 16 - Be Aware that Furosemide Contains a Sulfa Moiety
[+] 17 - Carefully Observe the Clinical Response to Intermittent Furosemide Dosing when
Deciding on Additional Doses
[+] 18 - Do not Administer Methylene Blue if there is Possible Gastrointestinal Absorption
[+] 19 - Be Alert for Metabolic Acidosis in Patients on Lorazepam Drips
[+] 20 - Be Alert for the Development of Cyanide Toxicity when Administering
Nitroprusside.
[+]
21 - Administer Beta-Blockade Initially Before Administering “Other” Antihypertensives
When Treating Aortic Dissection
[+] 22 - Do not use Vasopressin in Patients with Heart Failure or Mesenteric Ischemia
[+] 23 - Remember that Cardiac Pressors do not Work in a Low-pH Environment
[+] 24 - Consider Thrombolytics in ST-Elevation Myocardial Infarction if Percutaneous
Coronary Intervention is not Available or Delayed
[+] 25 - Consider the Use of Glycoprotein IIb/IIIa Inhibitors in Unstable Coronary
Syndromes
[+] 26 - Strongly Consider Low-Molecular-Weight Heparin in the Treatment of Unstable
Angina and Non-ST-Elevation Myocardial Infarction
[+] 27 - Strongly Consider Using Glycoprotein IIb/IIIa Inhibitors as an Added Treatment to
Stenting in Acute Myocardial Infarction
[+] 28 - Consider Nesiritide in Acutely Decompensated Heart Failure
[+] 29 - Be Aggressive in Considering Reperfusion Therapy in Acute Myocardial Infarction
[+] 30 - Strongly Consider Starting an Angiotensin-Converting Enzyme Inhibitor or
Angiotensin Receptor Blocker after Myocardial Infarction
[+] 31 - Use Prophylaxis for Erosive Gastritis in the Appropriate Patient
[+] 32 - Beware of Metabolites
[+] 33 - Do not use Erythromycin as a Prokinetic Agent in Patients on Tacrolimus (Or
Cyclosporine)
[+] 34 - Consider use of Enoxaparin over Unfractionated Heparin in Trauma Patients
[+] 35 - Maintain Tight Glucose Control in the Intensive Care Unit
[+] 36 - Do not use Subcutaneous Insulin in the Intensive Care Unit Population
[+] 37 - Do not use Insulin Glargine in the Intensive Care Unit Without Also Giving A
Shorter-Acting Insulin Form
[+] 38 - Remember that Patients with Insulin Deficiency Require Basal Insulin Even When
They Are Nil Per Os
[+] 39 - Look for Medication-Induced Causes of Hyperglycemia in Intensive Care Patients
[+] 40 - Do not use Midazolam and Lorazepam Interchangeably in the Intensive Care Unit
[+] 41 - Try to Avoid Using Benzodiazepines for Sleep in the Intensive Care Unit,
Especially in the Elderly
[+] 42 - Remember that Activated Protein C is not as Useful for Sepsis as Once Hoped
[+] 43 - Know the Alternate Routes for Administration of Cardiopulmonary Resuscitation
Medications
[+] 44 - Alkalinize the Urine in Tricyclic Antidepressant Overdose
[+] 45 - Check Triglyceride Level in Patients on Propofol Drips
[+] 46 - Be Alert for Drug-Related Pancreatitis in HIV/AIDS Patients And Consider a
Period of Proximal Bowel Rest
[+] 47 - Consider the Use of Fluconazole Prophylaxis in Intensive Care Patients with
Severe Pancreatitis, Abdominal Sepsis, or Need for Multiple Abdominal Surgeries
[+] 48 - Have a High Threshold for Administering Vitamin K Intravenously
[+] 49 - Do Not Use Benzocaine Spray: It Increases the Risk of Methemoglobinemia
[+] 50 - Know Which Weight to Use When Dosing Medications
[+] 51 - Beware of Intensive Care Unit Medications that can Increase Serum Potassium
and Cause Hyperkalemia
[+] 52 - Administer Acetazolamide (Diamox) On A One-Time Dose Schedule Only
[+] 53 - Check an Electrocardiogram for Long Qt Interval Before Giving Haloperidol
[+] 54 - Do not use Ipratropium in Meter-Dose Inhaler Form in Patients with Nut Allergies
[+] 55 - Be Cautious in Using Ketorolac in Patients with Marginal Urine Output or Renal
Function
[+] 56 - Do not Crush Sevelamer Hydrochloride (Renagel) to Place Down a Nasogastric or
Feeding Tube
[+] 57 - Remember that Many Commonly used Intensive Care Unit Drugs Should not be
Used in Porphyria
II. Devices/Lines/Tubes/Catheters/Drains/Procedures
[+] 58 - Do not Insert, Change, or Remove A Central Line With the Patient Sitting Up
[+] 59 - Avoid Placement of Central Access In The Right Internal Jugular in Cardiac
Transplant Patients if Possible
[+] 60 - Place the Tip of a Central Venous Catheter at the Junction of the Superior Vena
Cava and Atrium
[+] 61 - Do not Remove the Intravenous Catheter Used for Plasmapheresis Immediately
after the Last Treatment
[+] 62 - Never Thread a Triple-Lumen through a Cordis
[+] 63 - Be Meticulous in Technique when Inserting and Caring for Central Venous Access
Catheters in the Intensive Care Unit to Lower the Incidence of Infection
[+] 64 - In a Patient with a Previously Placed Vena Cava Filter, do not Use the J-tip on the
Guidewire When Using the Seldinger Technique to Place a Central Venous Catheter
[+] 65 - Do not “Whip the Tip” when Testing a Pulmonary Catheter before Insertion
[+] 66 - Remove Kinked or Coiled Pulmonary Artery Catheters
[+] 67 - Remove a Pulmonary Artery Catheter in Patients with New-Onset Ventricular
Arrhythmia and in most Atrial Arrhythmias
[+] 68 - Remember that in Normal Physiology the Wedge Pressure is less than the
Pulmonary Artery Diastolic Pressure
[+] 69 - Do not Use Pulmonary Artery Catheter Measurements in Tricuspid Regurgitation
[+] 70 - Always Turn on a Pacing Pulmonary Artery Catheter Prior to Floating The Device
[+] 71 - Make Sure any Cordis in Place is the Correct Size when Emergently Floating a
Transvenous Pacing Wire
[+] 72 - Try Placing a Skin Lead to Act as a Ground Wire if Pacing Wires Are not
Functioning Correctly Postoperatively
[+] 73 - Know the Backup Mechanism of Action and Requirements for Electrical
Cardioversion and Anticoagulation for Ventricular Assist Devices
[+] 74 - Place the Defibrillator in Synchronous Mode when Cardioverting
[+] 75 - Turn off the Synchronization Mode on the Defibrillator if there is no Recognized R
Wave
[+] 76 - Do not Rule Out the Presence of A Myocardial Infarction by a Normal
Electrocardiogram
[+] 77 - Do not use the 3 or 5 Lead ECG Monitor as a 12-lead Electrocardiogram
[+] 78 - Put a Board under the Patient when doing Chest Compressions
[+] 79 - Know the Pacemaker Alphabet Soup
[+] 80 - Do not use Cooling Blankets to Cool
[+] 81 - Do not Place Femoral Arterial Lines or Venous Catheters Above the Inguinal
Ligament or Below the Thigh Crease
[+] 82 - Consider Respiratory Variation on the Arterial-Line Monitor Tracing as a Sign of
Hypovolemia
[+] 83 - Do Not Use Low-Molecular-Weight Heparin in Patients with Epidurals
[+] 84 - Administer Epidural test Doses
[+] 85 - Consider an Intravenous Naloxone Drip for Treatment of Pruritus Associated with
Epidural Analgesia
[+] 86 - Be Meticulous when Dosing Bupivacaine in Patients with Both Epidural and Pleural
Catheters
[+] 87 - Do not Dismiss Rib Fractures as Trivial and Consider an Epidural Catheter for
Pain Control in Multiple Rib Fractures
[+] 88 - Use a Two-Step Technique with Radiographic Confirmation when Placing a
Feeding Tube
[+] 89 - Perform Doppler Ultrasound before Placing Sequential Compression Devices
[+] 90 - Consider Changing the Foley Catheter when Patient has a Urinary Tract Infection
[+] 91 - Do Not Flush Ureteral Stents if a Urological Consultation is Available
[+] 92 - Make Sure the Cuff (Pilot Balloon) is Completely Deflated on a Cuffed
Tracheostomy Tube Before a Passy-Muir Speaking Valve is Placed
[+] 93 - Know The Clinically Important Issues With Using an Intra-Aortic Balloon Pump
[+] 94 - Remember that Right Heart Failure is a Common And Important
Complication/Management Challenge Following Placement of a Left Ventricular Assist
Device
[+] 95 - Know the Common Problems Associated With Cardiopulmonary Support Using
Extracorporeal Membrane Oxygenation
[+] 96 - Treat “201C;wrinkling” in the Abiomed Diaphragm as a Possible Sign of
Hypovolemia
[+] 97 - Administer Antibiotics Before Placing a Chest Tube in Trauma Patients
[+] 98 - Be Suspicious of a Large Persistent Air Leak in a Chest Tube
[+] 99 - Treat Any Milky Fluid Coming From the Chest or Abdomen as Chylous Until
Proven Otherwise
[+] 100 - Know the Conditions that Cause an Inaccurate Pulse Oximetry Reading
[+] 101 - Remember that Pulse Oximetry is Inaccurate at Lower Saturation Levels
[+] 102 - Consider Placing an Intracranial Pressure Monitor in Patients with Glasgow
Coma Scale ≤8
[+] 103 - Know How to Use the Licox System to Measure Brain Tissue Oxygenation
III. Ventilators/Airway/Intubation/Extubation
[+] 104 - Preoxygenate Patients Before Intubation
[+] 105 - Use Cricoid Pressure when Performing Rapid Sequence Intubation or Bag-Mask
Ventilation
[+] 106 - Do not use The Presence of End-Tidal CO2 to Rule Out Esophageal Intubation
[+] 107 - Have a Low Threshold for Contacting the Most Experienced Available Airway
Professional in Patients with Diseases Associated with Difficult Airways
[+] 108 - Always Use a Vertical Incision when Performing a Cricothyroidotomy
[+] 109 - Use Bronchoscopic Guidance for Bedside Percutaneous Dilatational
Tracheostomy (PDT)
[+] 110 - Consider Early Tracheostomy in Select Patients
[+] 111 - Position the Tip of the Endotracheal Tube 4 Centimeters Above the Carina
[+] 112 - Remember that Even a Fully Inflated Cuff on the Endotracheal Tube is not
Adequate Protection Against Aspiration
[+] 113 - Do not Overinflate the Endotracheal Cuff
[+] 114 - Check for a Cuff Leak in Patients who Might have Tracheal EDEMA Before
Extubation
[+] 115 - Know how to Measure Plateau Pressure when Using Pressure-Regulated
Volume Control Ventilation Mode and Know what to do with the Value Once it is Obtained
[+] 116 - Use Plateau or Mean Pressure as a more Accurate Assessment of Barotrauma
than Peak Pressure
[+] 117 - Consider Airway Pressure Release Ventilation for Delivering an Open-Lung
Strategy
[+] 118 - Remember that Static Compliance of the Respiratory System is not the same
Thing as Dynamic Compliance
[+] 119 - Remember that Recruitment of Alveoli Using an Increased Level of Positive End-
Expiratory Pressure can Take 6 to 12 Hours
[+] 120 - Have a High Suspicion of Auto Positive End-Expiratory Pressure when
Attempting to Wean Patients with Chronic Obstructive Pulmonary Disease
[+] 121 - Be Cautious in Using Positive End-Expiratory Pressure After Single-Lung
Transplants
[+] 122 - Select an Initial Pressure Setting that is Just Slightly Higher than the Patient's
Peak Pressure When Attempting a Pressure Support Wean
[+] 123 - Do Not Reverse Neuromuscular Blockade Unless the Patient is Warm
[+] 124 - Use Glycopyrrolate Before Using Neostigmine when Reversing Neuromuscular
Blockade
[+] 125 - Do Not Attempt to Reverse Neuromuscular Blockade if there are no Twitches
[+] 126 - Remove Continuous Positive Airway Pressure and Bilevel Positive Airway
Pressure Masks Periodically
[+] 127 - Empirically Cover the Common Nosocomial Microbes in Ventilator-Associated
Pneumonia until the Cultures are Returned
[+] 128 - Keep the Head of Bed Elevated at Least 30 degrees for Intubated Patients if no
Contraindications Exist
[+] 129 - Treat Ventilator-Associated Pneumonia for 8 days
[+] 130 - Do not Routinely Extubate on Clinical Picture Alone
[+] 131 - Consider Using Heliox in the Mechanically Vented Asthmatic Patient
IV. Infectious Disease
[+] 132 - Consider Parvovirus B19 Infection in Patients with Anemia or Pancytopenia
[+] 133 - Consider Prophylactic Antibiotics when Leeches are Applied to Free Flaps
[+] 134 - Treat Methicillin-Resistant Staphylococcus Aureus with a Minimum of 14 days of
Antibiotics
[+] 135 - Be Alert for Thrombocytopenia and Neutropenia with Linezolid
[+] 136 - Have a High Threshold for Using Caspofungin and Voriconazole in Patients with
Liver Disease
[+] 137 - Do Not Use Caspofungin or Voriconazole to Treat Yeast in the Urine Because a
Very Small Amount of These Drugs are Excreted in the Urine
[+] 138 - Administer a Dose of Antibiotic Before the Bile System is Instrumented or
Manipulated
[+] 139 - Administer an Antibiotic Before Urinary Tract Obstruction is Relieved
[+] 140 - Remember that Enterococcus is a Rare Invasive Pulmonary Tract Infection
[+] 141 - Know How to Calculate the Clinical Pulmonary Infection Score
[+] 142 - Remember that Lack of Positive Blood Cultures does not Rule Out Bacterial
Endocarditis
[+] 143 - Treat Black Lips or a Black Spot on the Oral Mucosa as a Surgical Emergency
[+] 144 - Check for Cryptosporidium in Immunosuppressed Patients with Chronic, Severe,
or Refractory Diarrhea
[+] 145 - Pay Attention to the Morphology Reported By the Microbiology Lab for Fungal
Cultures
[+] 146 - Consider possible fungal infection in patients with hypothermia and bradycardia
[+] 147 - Give Special Consideration to the Extended-Spectrum Beta-Lactamase-
Producing Organisms Before Administering Antibiotics
[+] 148 - Have a High Threshold for Thoracentesis When Looking for a Source of Infection
[+] 149 - Aim for a Peak of Ten Times the Minimum Inhibitory Concentration (MIC) to Kill
Pseudomonas When Using an Aminoglycoside
[+] 150 - Know the Definition of a Catheter-Related Bloodstream Infection
[+] 151 - Strongly consider Stopping Prophylactic Antibiotics after 24 hours in Penetrating
Abdominal Trauma
[+] 152 - Use Clindamycin in Necrotizing Fasciitis to Cover Group a Streptococcus
[+] 153 - Be Cautious in Using Antibiotics for Uninfected Pancreatitis
[+] 154 - Wash Your Hands
[+] 155 - Consider Ventriculoperitoneal Shunt Infection in Patients with Sepsis
[+] 156 - Recognize that Vancomycin has very Poor Central Nervous System Penetration
[+] 157 - Be Alert for Seizures with Imipenem Use
[+] 158 - Remember if there is a Normal Platelet Count it Cannot be Hantavirus Pulmonary
Syndrome
[+] 159 - Consider Empiric Helicobacter Pylori Treatment when Gastric or Duodenal
Ulcers are Found
V. Shock/Fluids/Electrolytes
[+] 160 - Administer Empiric Broad-Spectrum Antibiotics when a Patient may be in Septic
Shock
[+] 161 - Do not Administer the Cosyntropin test within 24 Hours of Using Etomidate
[+] 162 - Switch from Hydrocortisone to Dexamethasone if the Cosyntropin Stimulation
Test is to be Administered
[+] 163 - Use Vasopressors Instead of Large-Volume Resuscitation in the Treatment of
Shock from Massive Pulmonary Embolism
[+] 164 - Consider the Diseases that Mimic Septic Shock in the Differential of this
Condition
[+] 165 - Be Alert for a Large Systemic Inflammatory Response after Back Surgery
[+] 166 - Be Alert for the Development of Abdominal Compartment Syndrome
[+] 167 - Treat Abdominal Pain Out of Proportion To Physical Exam as Mesenteric
Ischemia Until Proven Otherwise
[+] 168 - Do not Bolus Fluids that Contain Dextrose
[+] 169 - Remember that Diuresis May not be the Best Treatment for Hyperkalemia in the
Early Post Cardiopulmonary Bypass Period
[+] 170 - Remember that Most Patients Receive Mannitol When Going On the Pump so
Postoperative Urine Output is not a Marker for Volume Status or Perfusion After Cardiac
Surgery
[+] 171 - Do not use Urine Output as a Measure of Volume Status in Patients who are
Cold
[+] 172 - Be Careful to Not Overhydrate Postoperative Liver-Transplant Patients
[+] 173 - Remember that Uremia Alone Rarely Causes An Anion Gap to be Greater Than
[+] 174 - Do not Replete Calcium in the Setting of High Phosphorus or Phosphorus in the
Setting of High Calcium
[+] 175 - Check Postoperative and Serial Serum Levels of Phosphorus and Aggressively
Replete
[+] 176 - Consider Electrolyte Disturbances When There is a Change of Mental Status
[+] 177 - Keep the Serum Potassium at High or Normal Levels When Attempting to
Correct a Metabolic Alkalosis
[+] 178 - Do Not Replete Calcium in Rhabdomyolysis Unless a Patient is in Tetany
[+] 179 - Consider Excess Chloride as a Cause of an Unexplained Non-Anion-Gap
Metabolic Acidosis
[+] 180 - Consider Hyperchloremic Metabolic Acidosis to be a Renal Tubular Acidosis Until
Proven Otherwise if Obvious Sources of Bicarbonate Losses Like Diarrhea, Urinary
Diversions, and the Administration of Chloride are not Present
VI. Neuro
[+] 181 - Be Alert for New-Onset Cauda Equina Syndrome in Patients with Sacral or
Spinal Fractures or Surgery and Obtain Emergent Neurosurgical Consultation if Suspected
[+] 182 - Keep Patients with Dural Tears Flat for 24–48 Hours
[+] 183 - Know The Status of Cervical, Thoracic, and Lumbar Spine Stability on All
Postoperative and Trauma Patients
[+] 184 - Apply Appropriate Deep Vein Thrombosis Prophylaxis To Patients With Spinal
Cord Injury
[+] 185 - Be Alert For Autonomic Dysreflexia in Intensive Care Unit Patients With a Spinal
Cord Injury
[+] 186 - Consider the Use of Steroids for Neurological Trauma in Blunt Spinal Cord Injury
Only
[+] 187 - Start an Early Bowel Regimen in Patients After Spinal Cord Injury
[+] 188 - Consider Moderate Hypothermia After Cardiac Arrest
[+] 189 - Be Vigilant for Blunt Cerebrovascular Injury After Trauma
[+] 190 - Have an Extremely High Threshold in Giving Antihypertensives in Head Trauma
[+] 191 - Do Not Give More than 7 Days of Antiseizure Medication in Head Trauma
[+] 192 - Calculate the Glasgow Coma Scale Using the Best Motor Response
[+] 193 - Use Magnetic Resonance Imaging (Not Head Computed Tomography) as the
Gold Standard Test for Diffuse Axonal Injury
[+] 194 - Remember that there is Usually an Upward Drift in Some Intracranial-Pressure
Monitor Readings The Longer they Have Been in Place
[+] 195 - Be Aware that Increasing Positive End-Expiratory Pressure may Result in
Increasing Intracranial Pressure
[+] 196 - Obtain a Computed Tomography Scan of the Head Immediately after any
Craniotomy or Intracranial Procedure if Patient's Neurologic Examination is Different from
Preoperative Assessments
[+] 197 - Consider Last-Ditch Maneuvers to Lower Intracranial Hypertension in Impending
Herniation
[+] 198 - Remember that Patients Undergoing Barbiturate Coma must have Adequate
Electrophysiological Monitoring
[+] 199 - Be Alert for Conversion of Nonhemorrhagic (Ischemic) Stroke to Hemorrhagic
Stroke
[+] 200 - Obtain a Head Computed Tomography Scan and Lumbar Puncture in Patients
with Human Immunodeficiency Virus or Acquired Immunodeficiency Syndrome and New-
Onset Mental Status Changes
[+] 201 - Have a Low Threshold for Obtaining an Initial and Repeat Head Computed
Tomography Scan after Subarachnoid Hemorrhage
[+] 202 - Be Alert for Rebleeding in Patients with Subarachnoid Hemorrhage
[+] 203 - Remember that New Electrocardiogram Changes in a Patient with a
Subarachnoid Bleed may be a Sign of Progression of the Bleed
[+] 204 - Consider the Use Of Thrombolytic Agents for Treatment of Acute Ischemic
Stroke
[+] 205 - Consider Absence of Withdrawal to Pain at 24 hours and Absence of Eye
Reflexes at 72 hours Post–Cardiac Arrest to be Highly Correlated with Permanent Coma
[+] 206 - Remember that Failure to Recognize Pituitary Apoplexy Can Result in a
Neurologic Catastrophe
[+] 207 - Do not Perform a Lumbar Puncture on Patients with Posterior Fossa Masses
[+] 208 - Evaluate for Guillain-Barré Syndrome in Patients with Acute Paralysis or
Respiratory Failure and Areflexia
[+] 209 - Start Early Plasmapheresis or Immunoglobulins in Guillain-Barré Patients
[+] 210 - Remember that Patients with Myasthenia Gravis Exacerbation Usually Look Well
Until Just Before they Require Intubation
[+] 211 - Avoid Using Incentive Spirometry in Myasthenia Gravis Patients; use
Intrapulmonary Percussive Ventilation (IPPV) or Flutter Valve Instead
[+] 212 - Remember that not all Seizures are Convulsive and Obvious
[+] 213 - Treat Status Epilepticus as a Medical Emergency
[+] 214 - Know the Potential Adverse Effects of Valproic Acid
[+] 215 - Learn the Cranial Nerve Examination as One Can Obtain a Lot of Information
Even in Comatose and Poorly Cooperative Patients
[+] 216 - Do not Squeeze the Toenails or Administer a Painful Stimulus to the foot to Test
Response to Pain in Comatose Patients as this may Trigger a Spinal Reflex that may be
seen even in Brain-Dead Patients
VII. Laboratory
[+] 217 - Do not Ascribe an Increased Serum Lactate Level to Renal Insufficiency
[+] 218 - Do not Treat Lactic Acidosis with Bicarbonate
[+] 219 - Remember that Postoperative Hypertension can be a Result of Increased PCO2
[+] 220 - Do not use a Normal Arterial Blood Gas to rule out a Pulmonary Embolism
[+] 221 - Remember that Argatroban Increases International Normalized Ratio But does
Not Affect Coagulation in the Extrinsic System
[+] 222 - Check Serial Methemoglobin Levels in Patients on Inhaled Nitric Oxide
[+] 223 - Remember that Troponin Levels are Inaccurate as a Measure of Cardiac
Damage in Renal Insufficiency
[+] 224 - Ask the Laboratory for a Synergy Panel in Resistant Pseudomonas Infections
[+] 225 - Obtain Blood Used for Mixed Venous Oxygen Testing From the Distal Pulmonary
Artery Catheter Port
[+] 226 - Check Thyroid Function in Critically Ill Patients
[+] 227 - Use an Empty Lab Tube to Check the Viability of a Stoma
[+] 228 - Never Retest Low Serum Glucose; Treat Immediately and Then Prevent Further
Episodes
[+] 229 - D-Dimer Levels Can Rule Out But Not Rule in Pulmonary Embolism in
Postoperative Patients
[+] 230 - Remember That Diabetic Ketoacidosis Often Begins With an Anion-Gap
Metabolic Acidosis
[+] 231 - Consider Obtaining A Serum B-Type Natriuretic Peptide Level in the Acutely
Dyspneic Patient
[+] 232 - Consider Hypomagnesemia as a Cause of Refractory Hypokalemia
VIII. Nutrition
[+] 233 - Be Aware that Enteral Feeds can Lower Phenytoin Levels
[+] 234 - Consider Early Enteral Feeding
[+] 235 - Be Alert for Overfeeding
[+] 236 - Be Alert for the Development of Refeeding Syndrome
[+] 237 - Consider Using Elemental or Semielemental Feeds in Patients with Albumin Less
than 2.5 Grams Per Deciliter
[+] 238 - Consider Enteral Feedings in Pancreatitis and Enterocutaneous Fistulae
[+] 239 - Use The Metabolic Cart only when Patients are on Low Vent Settings
[+] 240 - Use a Dedicated, Upper-Body, Single-Lumen Central Venous Catheter for
Administration of Parenteral Nutrition
IX. Renal
[+] 241 - Be Alert for Hypophosphatemia in the Intensive Care Unit Patient on Dialysis
[+] 242 - Know the Drugs that Must be Redosed after Dialysis
[+] 243 - Remember that Continuous Venovenous Hemodialysis can Obscure a
Temperature Spike
[+] 244 - Clamp the Dialysate Line in Continuous Venovenous Hemodialysis Immediately if
it Becomes Pink Tinged
[+] 245 - Do not use Continuous Venovenous Hemodialysis in the Setting of Angiotensin-
Converting Enzyme 2 and Vice Versa
[+] 246 - Do Not Give Fludrocortisone to Patients on Dialysis
[+] 247 - Avoid the Subclavian Vein for Central Access of any Type in a Dialysis Patient or
Possible Dialysis Patient
[+] 248 - Use Caution when Using Milrinone in Renal Failure
[+] 249 - Decrease the Dose of Ganciclovir in Renal Insufficiency
[+] 250 - Remember that Trimethoprim-Sulfamethoxazole (Bactrim) Crystals can
Precipitate in the Kidney and Cause Renal Damage and Failure
[+] 251 - Be Aware that Lipid-Based Amphotericin Products are Associated with Less
Renal Toxicity than Regular Amphotericin but can Still Cause Renal Injury
[+] 252 - Have a High Level of Suspicion for Drug-Induced Acute Interstitial Nephritis
[+] 253 - Consider Rhabdomyolysis in the Patient who Develops Oliguric Renal Failure
after a Prolonged Surgery where Muscle Compression may have Occurred
[+] 254 - Aim for 2 Milliliters Per Kilogram Per Hour of Urine Output in Rhabdomyolysis
[+] 255 - Do not Attempt to Convert Oliguric to Nonoliguric Renal Failure with Diuretics
[+] 256 - Consider N-Acetylcysteine or Sodium Bicarbonate Prophylaxis Along with
Adequate Hydration to Combat Contrast-Induced Nephropathy
X. Blood
[+] 257 - Remember that Transfusion-Related Acute Lung Injury is Not Dose Dependent
[+] 258 - Know the Signs of a Transfusion Reaction
[+] 259 - Have a High Threshold in Transfusing Platelets, Especially in Nonbleeding
Patients who are not Preoperative
[+] 260 - Do not Administer Platelets in Immune Thrombocytopenic Purpura
[+] 261 - Do Not Administer Platelets in Type 2 Heparin-Induced Thrombocytopenia
[+] 262 - Remember that Reticulocyte Count is not Accurate After Blood Transfusion
[+] 263 - Consider Leukocyte-Depleted Blood in Patients who are Immunosuppressed
[+] 264 - Administer Octreotide in Variceal Bleeding While Waiting for Endoscopy
[+] 265 - Consider Bleeding Around a Chest Tube to be a Sign of Bleeding in the Chest
Cavity Until Proven Otherwise
[+] 266 - Start a Proton Pump Inhibitor Infusion for Gastric and Duodenal Bleeding
[+] 267 - Remember that Bleeding Associated with Direct Thrombin Inhibitors is not
Correctable with Protamine, Fresh-Frozen Plasma, or Platelets
[+] 268 - Attempt to Decrease Phlebotomy
[+] 269 - Consider the Use of Factor viia to Treat Medical Bleeding in a Surgical or
Trauma Patient
[+] 270 - Consider Angiography as an Adjunct in Controlling Solid-Organ Bleeding After
Damage Control Surgery
[+] 271 - Consult Surgery Emergently if a Patient With a Bleeding Peptic Ulcer Rebleeds
after Endoscopic Control
[+] 272 - Call for a Sengstaken-Blakemore or Minnesota Tube When a Cirrhotic Patient
has an Upper Gastrointestinal Bleed
XI. Imaging and Tests
[+] 273 - Treat Loss of Doppler Signals in a Free Flap as a Surgical Emergency
[+] 274 - Remember When Reviewing Doppler Ultrasound Results that the Superficial
Femoral Vein is a Component of the Deep Venous System
[+] 275 - Post a Sign Outside of the Door of Patients Who Have Received a Tagged
White Blood Cell Scan Warning of the Pregnancy Risk
[+] 276 - Obtain an Echocardiogram to Rule Out Bacterial Endocarditis in Gram-Positive
Bacteremia
[+] 277 - Diagnose Tamponade Based on Clinical Findings and not Echocardiogram
[+] 278 - Consider Pulsus Alternans as a Sign of Impending Tamponade
[+] 279 - Do Not Use a Negative Focused Assessment With Sonography for Trauma
(Fast) Exam to Rule Out Bowel Injury or Injury to the Retroperitoneum or as the Only Test
in Penetrating Trauma
[+] 280 - Consider Using Computed Tomography Reconstructions of the
Chest/Abdomen/Pelvis to Rule Out Thoracic and Lumbar Fractures and Dislocations
[+] 281 - Do not Rule Out Cervical Spine or Spinal Cord Injury on Bony Films or Computed
Tomography Alone
[+] 282 - Use the Position of the Mediastinum as a CLue to the Diagnosis of A White-out
on Chest Radiograph
[+] 283 - Consider Angiography for Blunt Solid-Organ Injury
[+] 284 - Be Alert for Compartment Syndromes
[+] 285 - Be Alert for Acalculous Cholecystitis
[+] 286 - Avoid Giving Intravenous Contrast Dye more Often Than Every 48 Hours if
Possible
XII. Pregnancy
[+] 287 - Place Pregnant Patients with Right Side Elevated 15 Degrees
[+] 288 - Avoid the Use of Drugs Harmful to the Fetus if at all Possible
[+] 289 - Strongly Consider the Use of an Electronic Fetal Monitor in Caring for a Patient
at Greater than 24 Weeks' Gestation in the Intensive Care Unit
[+] 290 - Know the Normal Physiologic Changes and Associated Laboratory Values that
Occur in Pregnancy
XIII. Burns
[+] 291 - Do not Administer Prophylactic Antimicrobials to Burn Patients
[+] 292 - Remember that not All Fever in the Burn Patient is Due to Infection
[+] 293 - Never Underestimate the Severity of an Electrical Burn
[+] 294 - Do not Use Hyperbaric Oxygen Therapy in Burns
[+] 295 - Do not Use Parenteral Nutrition, if at all Possible, in Burn Patients
[+] 296 - Strongly Consider Escharotomy in Circumferential Burns
[+] 297 - Be Aware of the Pitfalls in the Management of Stevens-Johnson Syndrome and
Toxic Epidermal Necrolysis
[+] 298 - Adopt a Philosophy of Early Excision and Grafting of Burn Wounds
[+] 299 - Be Aggressive in Mobilizing Burn Wounds
[+] 300 - Have a Low Threshold for Intubating a Patient with an Inhalation Burn Injury
[+] 301 - Do not Under-Resuscitate a Burn Patient
[+] 302 - Avoid the Pitfalls of Vascular Access in Burn Patients
[+] 303 - Know How to Estimate Burn Size and Depth
XIV. Miscellanceous
[+] 304 - Do not Talk to Families about Organ Donation
[+] 305 - Alert the Transplant Team Emergently if There is an Acute Decrease in Urine
Output after a Kidney Transplant
[+] 306 - Remember that Cardiac Output is not the Same Thing as Ejection Fraction
[+] 307 - Do Not “Rock the Pelvis” in a Fracture
[+] 308 - Consider Treatment for Heterotopic Ossification After Trauma
[+] 309 - Look For Missed Extremity Fractures in Patients with a Diagnosed Extremity
Fracture
[+] 310 - Have a Working Knowledge of Intensive Care Unit Scoring Systems
[+] 311 - Have a Working Knowledge of the Emergency Medical Treatment and Active
Labor Act as it Applies to the Intensive Care Unit
[+] 312 - Know What the Basic Statistical Terms Mean
[+] 313 - Consider Emboli When There is a Change in Mental Status After an Invasive
Procedure
[+] 314 - Know the Noninfectious Causes of Fever in the Intensive Care Unit
[+] 315 - Cardiovert Unstable Tachycardias (Both Narrow and Wide Complex)
[+] 316 - Be Concerned About Chest Pain Even if it is Found to be Noncardiac in Nature
[+] 317 - Know the Difference Between mmHg and cm H2O
[+] 318 - Be Alert for Lupus Flares
BACK OF BOOK ↑
[+] Index
> Table of C ontents > Medic ations > 1 - Monitor P atients w ho have R ec eived Intr athec al P r es er vative-
Fr ee Mor phine
1
Monitor Patients who have Received Intrathecal
Preservative-Free Morphine
The int rat hecal administ rat ion of opioids has emerged as a popular and eff ect ive
f orm of post operat ive pain cont rol. I nt rat hecal opioids are able t o provide long-
last ing analgesia af t er a single inject ion. They w ork by binding t o t he µ opioid
recept ors, w hich are locat ed in t he subst ant ia gelat inosa of t he dorsal horn of
t he spinal cord. These recept ors are concent rat ion dependent and are t ypically
not act ivat ed by syst emic doses of opioids. Unlike int rat hecal local anest het ics,
int rat hecal opioids provide analgesia w it hout disrupt ing sensory, mot or, or
sympat het ic f unct ions.
Because of it s hydrophilic propert ies and pot ent recept or aff init y, preservat ive-
f ree morphine (i. e. , Duramorph or Ast ramorph) is t he ideal opioid f or int rat hecal
use. The onset of analgesic eff ect s is direct ly proport ional t o t he lipid solubilit y
of t he opioid. Preservat ive-f ree morphine (along w it h hydromorphone and
meperidine) has a relat ively low lipid solubilit y and it s onset of act ion is delayed
f or t ypically 20 t o 40 minut es af t er administ rat ion. The hydrophilic nat ure of t he
opioid also det ermines it s durat ion of act ion. Preservat ive-f ree morphine is very
hydrophilic and poorly lipid soluble, w hich ext ends it s durat ion of analgesic eff ect
up t o 12 t o 24 hours. Because of it s poor lipid solubilit y, int rat hecal morphine
remains in t he cerebrospinal f luid (CSF) f or a prolonged period of t ime. I t is
circulat ed t hrough cerebral spinal bulk f low and event ually rises rost rally t o
supraspinal levels. I nt rat hecal morphine, t heref ore, has bimodal analgesic
eff ect s. The f irst peak is soon af t er administ rat ion and is due t o spinal opiat e
recept or binding. The second peak occurs 12 t o 24 hours lat er and is due t o
supraspinal binding as t he drug is circulat ed.
Compared w it h syst emic dosing of morphine, int rat hecal administ rat ion is
eff ect ive in providing analgesia at a f ract ion of t he syst emic dose (0. 25–0. 5 mg)
and t hus has a much low er side-eff ect prof ile. The side eff ect s, how ever, are
import ant t o recognize and t reat . Respirat ory depression can be delayed up t o
24 hours af t er administ rat ion and is due t o t he cephalad spread of int rat hecal
morphine t o t he opioid recept ors in t he medullary cent ers of t he brain st em.
Thus, pat ient s receiving int rat hecal morphine must be closely monit ored f or up t o
24 hours af t erw ard f or signs of respirat ory depression.
What Not to Do
Pat ient s w it h post operat ive pain despit e having received int rat hecal morphine
present a management dilemma. G iving t he pat ient addit ional syst emic opioids
must be done caut iously, as it may increase and pot ent iat e t he risk of
respirat ory depression. G enerally, pat ient s w ho have received int rat hecal
morphine should not be placed on a pat ient -cont rolled analgesia machine and
should be given only int ermit t ent doses of short -act ing narcot ics unt il t he
int rat hecal morphine analgesic eff ect occurs. Nonopioid analgesics can also be
considered if not cont raindicat ed af t er surgery.
O t her pot ent ial side eff ect s of int rat hecal morphine are similar t o side eff ect s of
syst emic morphine and include prurit us, nausea, vomit ing, and urine ret ent ion.
These eff ect s are dose relat ed and may be reversed w it h naloxone.
Suggested Readings
Raj PP, ed. Pract ical Management of Pain. 3rd Ed. St Louis: Mosby;
2000: 180.
Rat hmell JP, Lair TR, Nauman B. The role of int rat hecal drugs in t he
t reat ment of acut e pain. Anest h Analg 2005; 101: S30–S43.
Waldman SD, ed. Interventi onal Pai n Management. 2nd Ed. Philadelphia: WB
Saunders; 2001: 621–622.
> Table of C ontents > Medic ations > 2 - K now the C har ac ter is tic s of the Nar c otic s You P r es c r ibe
2
Know the Characteristics of the Narcotics You
Prescribe
Nirav G . Shah MD
Pain cont rol is an import ant aspect of crit ical care medicine and t he use of
opiat e medicat ions has t ypically been an import ant met hod of achieving cont rol in
t he int ensive care unit (I CU) set t ing. The brain has f our opiat e recept ors t hat
include t he mu, kappa, delt a, and sigma recept ors. Current ly used opioid
analgesics bind t o t he mu recept or and init iat e t he pharmacological eff ect s of
analgesia, miosis, respirat ory depression, euphoria, and physical dependence.
The rout e of opiat e administ rat ion is of t en int ravenous, w it h bolus administ rat ion
f or mild t o moderat e pain, cont inuous inf usion f or moderat e t o severe pain, and
pat ient -cont rolled analgesia f or t he post -surgical pat ient w ho can part icipat e in
his or her care. Like ot her I CU int ervent ions w here benef it s must be balanced
against risks, adequat e analgesia must caref ully be balanced w it h t he side
eff ect s of opiat e t herapy, part icularly in t he crit ically ill pat ient . These include
respirat ory depression, hypot ension, emesis, f lushing, bronchospasm, and
const ipat ion.
What to Do
The t hree most commonly prescribed analgesics in t he I CU are morphine,
hydromorphone, and f ent anyl. While morphine is w idely used in t he I CU set t ing,
f ent anyl has addit ional benef it s f or t he crit ically ill pat ient because of it s
increased pot ency, lipid solubilit y, and hemodynamic st abilit y. Morphine dosing
usually begins at 2 mg and is t hen t it rat ed up by 1 mg t o 2 mg every f ew hours if
given in int ermit t ent bolus f orm and 1 mg/ hour if given cont inuously. Morphine
met abolism occurs in t he liver w it h excret ion occurring in t he kidney. Theref ore,
t he dose should be reduced if t he pat ient has a glomerular f ilt rat ion rat e less
t han 30 mL/ min in order t o prevent accumulat ion of it s act ive met abolit e.
Hydromorphone, a semisynt het ic opiat e agonist , is markedly more pot ent and
quicker in onset t han morphine. Dosing begins at 0. 2 t o 0. 6 mg w it h repeat ed
doses every 2 t o 3 hours. This dose may need t o be increased in pat ient s w ho
have had marked exposure t o opiat es in t he past . I n addit ion, if given as an
int ravenous cont inuous inf usion, t he dose should be 0. 5 t o 1 mg per hour af t er
t he bolus dose. Like morphine, hydromorphone is met abolized by t he liver. I n
cont rast t o morphine, how ever, t he met abolit es are all inact ive. Theref ore, dose
adjust ment should be considered f or hepat ic f ailure.
Fent anyl is a synt het ic opiat e agonist and is 100 t imes more pot ent t han
morphine. Many I CUs have increasingly used f ent anyl as t he analgesic of choice
because of it s rapid onset and pot ency. When administ ered in int ermit t ent
boluses, t he dose is usually 25 t o 75 mcg every hour. How ever, it is more
eff ect ively used t o prevent pain if given cont inuously at a rat e of 25 t o 50
mcg/ hour af t er t he bolus dose. Fent anyl accumulat es in adipose t issue and if
administ ered f or longer t han 5 days may have prolonged sedat ion eff ect s upon
it s discont inuat ion. Fent anyl is met abolized t o inact ive compounds by t he liver
and excret ed by t he kidneys; t hus t he dose may need t o be adjust ed f or liver
f ailure, but not renal f ailure.
Physical dependence can occur w it h any of t hese medicat ions and w ill result in
w it hdraw al if st opped abrupt ly. I n addit ion, many pat ient s develop t olerance t o
t he opiat es and require increasing dosages t o achieve t he same level of pain
cont rol. O ne met hod of overcoming t olerance is t o use adjuvant t herapies t hat
enhance t he eff ect of opiat es. For example, t he concomit ant use of
benzodiazepines w ill assist w it h anxiolysis and improve t he response t o
analgesia. How ever, it w ill also cont ribut e t o t he side eff ect of respirat ory
depression already seen w it h opiat e medicat ions.
Suggested Readings
Hamill-Rut h RJ. Evaluat ion of pain in t he int ensive care unit . Crit Care Clin
1999; 15: 35â 5 4.
Joranson DE, Ryan KM, G ilson AM, et al. Trends in medical use and abuse of
opioid analgesics. JAMA 2000; 283: 1710.
> Table of C ontents > Medic ations > 3 - C ons ider C lonidine to C om bat E ffec ts of D r ug W ithdr aw al
3
Consider Clonidine to Combat Effects of Drug
Withdrawal
Clonidine is an alpha-2 recept or agonist t hat dow n regulat es t he sympat het ic
nervous syst em. By st imulat ing alpha-2 adrenergic recept ors in t he brain st em,
clonidine act ivat es inhibit ory pat hw ays in t he cent ral nervous syst em (CNS),
w hich result s in reduced cat echolamine release and reduced sympat het ic out f low
f rom t he CNS. This eff ect ively causes a decrease in blood pressure, heart rat e,
peripheral resist ance, and renal vascular resist ance.
Clonidine's mechanism of act ion makes it a usef ul ant ihypert ensive agent but
also enables it t o combat eff ect s of drug w it hdraw al, especially nicot ine and
opioids (including heroin and met hadone). Bot h nicot ine and opioid w it hdraw al
t ypically involve cat echolamine release w it h such varied sympt oms as pupillary
dilat at ion, lacrimat ion, rhinorrhea, piloerect ion, yaw ning, sneezing, anorexia,
nausea, vomit ing, and diarrhea. Clonidine has been f ound t o be eff ect ive in
count eract ing t hese sympat het ic-mediat ed sympt oms t hrough it s CNS inhibit ory
mechanisms.
What to Do
For w it hdraw al, clonidine should be st art ed via t he oral rout e. Clonidine is
t ypically given at 0. 1 mg by mout h (PO ) 2 t o 4 t imes per day. An alt ernat e
approach f or rapid det oxif icat ion, using clonidine in conjunct ion w it h nalt rexone,
is clonidine 6 mcg/ kg/ day PO divided int o 3 doses t he f irst day, increased t o 11
mcg/ kg/ day PO divided int o 3 doses on day t w o, t hen t apered t o 0. 6 mcg/ kg/ day
PO divided int o 3 doses on t he t hird day. I ncreased side eff ect s of clonidine use
are likely w it h higher doses, and include ort host at ic hypot ension, sedat ion, dry
mout h, and const ipat ion.
Clonidine can be cont inued PO and gradually t apered over a 10-day period or
convert ed t o a 7-day t ransdermal pat ch, using t he f ollow ing conversion prot ocol:
Day 1: Place t ransdermal clonidine. Addit ionally, administ er 100% of oral

dose.
Day 2: Pat ch remains; administ er 50% of oral dose.
Day 3: Pat ch remains; administ er 25% of oral dose.
Day 4: Pat ch remains; no f urt her oral clonidine necessary.
Abrupt discont inuat ion of clonidine may cause a hypert ensive rebound and
sympt oms of sympat het ic overact ivit y. Severe hypert ension can be seen 12 t o 36
hours af t er t he last dose, especially in pat ient s receiving higher doses.
Clonidine, t heref ore, should be t apered gradually over several days w it h close
monit oring f or t he signs and sympt oms of clonidine w it hdraw al. I t should be
remembered t hat during clonidine w it hdraw al, concomit ant bet a blockade may
w orsen rebound hypert ension.
Suggested Readings
Davison R, Kaplan K, Fint el D, et al: The eff ect of clonidine on t he cessat ion
of cigaret t e smoking. Clin Pharmacol Ther 1988; 44: 265.
Franz DN, Hare BD, McCloskey KL: Spinal sympat het ic neurons: Possible
sit es of opiat e-w it hdraw al suppression by clonidine. Science 1982; 215: 1643.
G old MS, Pot t ash AC, Sw eeney DR, Klever HD: O piat e w it hdraw l using
clonidine. A saf e, eff ect ive, and rapid nonopiat e t reat ment . JAMA
1980; 243: 343.
Hughes JR: Clonidine, depression, and smoking cessat ion. JAMA

1988; 259: 2901.
> Table of C ontents > Medic ations > 4 - S tr ongly C ons ider P r ophylaxis for Alc ohol W ithdr aw al
4
Strongly Consider Prophylaxis for Alcohol
Withdrawal
Wit hdraw al f rom alcohol is a serious and common complicat ion associat ed w it h
hospit alizat ion t hat can result in t he condit ion know n as delirium t remens. Many
pat ient s w ho consume alcohol cease drinking on or prior t o admission f or bot h
volunt ary and involunt ary reasons. A pat ient does not need t o be an alcoholic by
medical def init ion t o be at risk f or t his alcohol w it hdraw al syndrome. Pat ient s
w ho drink as lit t le as t w o drinks per day have been know n t o develop delirium
t remens. Addit ionally, chronic benzodiazepine users are also at risk f or t he same
syndrome, since t his class of drugs aff ect s t he same recept ors in t he brain.
Signs And Symptoms

Delirium t remens is a const ellat ion of signs and sympt oms t hat include conf usion,
agit at ion, delirium, combat iveness, hallucinat ions (commonly visual changes
involving bright light s and color), and pot ent ial seizure act ivit y. These responses
put pat ient s at risk f or self -injury as w ell as injury t o st aff and ot hers.
Addit ionally, t hese pat ient s exhibit surges of sympat het ic out put result ing in
t achycardia, hypert ension, prof use sw eat ing, and mydriasis. Pat ient s w it h
concomit ant cardiovascular disease are at risk f or myocardial inf arct ion,
int racerebral hemorrhage, and st roke. Even w it h t reat ment , delirium t remens
carries a mort alit y risk of approximat ely 10%.
What to Do
O ne w ay t o prevent delirium t remens is t o have t he pat ient cont inue alcohol
consumpt ion. While it is desirable t o t reat someone f or alcoholism, abrupt
cessat ion is not t he answ er in t he acut e hospit alizat ion period w hen ot her
medical or surgical concerns are paramount . I f t he pat ient is able t o drink liquids
or has an ent eral f eeding t ube (e. g. , nasogast ric t ube, orogast ric t ube,
percut aneous endoscopic gast ronomy t ube) and t here are no ot her
cont raindicat ions t o ent eral f eeding, t he easiest w ay t o prevent delirium t remens
is simply t o give t he pat ient alcohol. Many hospit als st ock beer or spirit s eit her
in t heir pharmacy or as part of t heir f ood service syst em and t hese may be
prescribed f or t he pat ient deemed at risk of developing delirium t remens. Limit ed
amount s w ill
suff ice, generally one t o t w o beers or t he equivalent in spirit s or w ine (one 12-
ounce beer is equal t o 5. 5 ounces of w ine or 1. 5 ounces of spirit s such as vodka
or w hiskey assuming a proof of 80) w it h meals. This w ill not lead t o overt
int oxicat ion, part icularly in someone w ho normally consumes much larger
amount s. Blood alcohol levels can be measured, how ever, if t here is such a
concern. I f ent eral f eeding is not an opt ion, et hanol can be inf used int ravenously.
I t is usually ordered as 10% ETO H t o be run at a rat e of 20 t o 40 cc/ hr. Like
ent eral alcohol, t his in most cases w ill prevent w it hdraw al and delirium t remens
w it hout leading t o overt int oxicat ion.
Alt ernat ively, if t he pharmacy cannot or w ill not make an et hanol inf usion and
ent eral et hanol is not an opt ion, benzodiazepines are t he pref erred alt ernat ive
met hod of prevent ing et hanol w it hdraw al. These may be given ent erally or
int ravenously. Lorazepam is commonly employed because of it s lack of act ive
met abolit es and an int ermediat e durat ion of act ion, but most drugs in t his class
w ill w ork eff ect ively. Benzodiazepines may be given on an around-t he-clock
(ATC) or as needed (prn) basis. Doses commonly used are 2 t o 4 mg every 4 t o
8 hours or prn. Recent lit erat ure suggest s t hat w hile t he ATC met hod is most
commonly employed, t reat ing t he emergence of sympt oms prn w it h
benzodiazepines is just as eff ect ive at prevent ing and t reat ing delirium t remens
and w ill minimize complicat ions w hile at t he same t ime reducing lengt h of st ay.
Despit e prophylaxis, a cert ain percent age of pat ient s may st ill progress t o
w it hdraw al and delirium t remens. O nce t his has occurred, any et hanol being
given ent erally or by int ravenous inf usion should be abandoned (et hanol is not
eff ect ive in t reat ing t he sympt oms of delirium t remens) and benzodiazepines
subst it ut ed. The int ravenous f orm is pref erred t o ensure administ rat ion since
w it hdraw al may include sympt oms of nausea and vomit ing, impeding ent eral
absorpt ion. O nce t he delirium t remens has st art ed, doses of benzodiazepines
equal t o or higher t han prophylaxis are used. Somet imes, very large doses or
cont inuous inf usions are necessary. O nce t he w it hdraw al syndrome is under
cont rol, t hese should be w eaned slow ly so as t o avoid reemergence of
w it hdraw al.
Addit ional agent s t o consider are clonidine and haloperidol. Clonidine 0. 1 mg
orally every 8 hours is very eff ect ive in blunt ing t he sympat het ic surge
associat ed w it h et hanol w it hdraw al and as such may help prevent complicat ions
such as myocardial inf arct ion in t hose at risk. Bet a-1 specif ic blockers may also
be used t o block end-organ responses if t achycardia is a t hreat but t his may
have lit t le eff ect on blood pressure secondary t o alpha-1 agonism, w hich may
need t o be t reat ed
w it h separat e agent s. Haloperidol is eff ect ive in providing sedat ion and
ameliorat ing some of t he hallucinat ions t hat occur. An elect rocardiogram should
be checked each day t o w at ch f or long Q T syndrome w hen t he pat ient is get t ing
haloperidol. Treat ment f or t his is cessat ion of haloperidol and magnesium
supplement at ion.
O ne f inal not e is t hat many pat ient s hide or do not admit t o t heir alcohol use and
t heir f amilies may not be aw are of t he ext ent of t heir use. Delirium t remens
should be in t he diff erent ial f or any pat ient experiencing ment al st at us change,
especially w hen coupled w it h signs of sympat het ic act ivit y.
Suggested Readings
Ebell MH. Benzodiazepines f or alcohol w it hdraw al. Am Fam Physician
2006; 73: 1191.
McI nt yre J, Hill KR, Woodside J Jr. Alcohol w it hdraw al syndrome. Am Fam
Physician 2004; 69: 1443.
Sanouri I , Dikin M, Soubani AO . Crit ical care aspect s of alcohol abuse. Sout h
Med J 2005; 98: 372â 3 81.
> Table of C ontents > Medic ations > 5 - Avoid C onc om itant Us e of S ter oids , Neur om us c ular B loc kade,
and Am inoglyc os ides to Les s en the R is k of C r itic al Illnes s Myopathy
5
Avoid Concomitant Use of Steroids,
Neuromuscular Blockade, and Aminoglycosides
to Lessen the Risk of Critical Illness Myopathy
T imothy M. Moore MD, PhD
Acut e myopat hy has been increasingly recognized as a signif icant complicat ion
of pat ient s cared f or in t he int ensive care unit (I CU). The t erm cri ti cal i l l ness
myopathy (CI M) is now used t o describe a general syndrome of muscle
dysf unct ion occurring in t he crit ically ill pat ient , w it h subt ypes of CI M also being
def ined. The major f eat ure of CI M is diff use, f laccid w eakness of limb, neck, and
f acial muscles, as w ell as t he diaphragm. O pht halmoplegia may be present and
t endon ref lexes are of t en depressed. The t iming of onset of CI M is diff icult t o
det ermine, but by def init ion w eakness must present af t er t he onset of crit ical
illness. Current ly, t he overall occurrence of CI M is unknow n because of
nonunif ormit y in st udies’ pat ient case mixes, diagnost ic crit eria used, and t iming
of evaluat ion f or CI M.
Signs and Symptoms

I t is only during a syst emat ic w orkup f or generalized w eakness and vent ilat ory
f ailure in t he crit ically ill pat ient t hat a diagnosis of CI M can be reached. A
combinat ion of elect rophysiologic st udies and hist opat hologic f indings is
required. Crit eria f or t he diagnosis of CI M include sensory nerve act ion pot ent ial
amplit udes >80% of t he low er limit of normal; needle elect romyogram (EMG )
w it h short -durat ion, low -amplit ude mot or unit pot ent ials w it h early or normal f ull
recruit ment (w it h or w it hout f ibrillat ion pot ent ials); absence of a decrement al
response on repet it ive nerve st imulat ion; muscle biopsy f indings of myopat hy
w it h myosin loss; and compound muscle act ion pot ent ial amplit udes <80% of t he
low er limit of normal in t w o or more nerves w it hout conduct ion block. Elevat ed
serum creat inine kinase and demonst rat ion of muscle inexcit abilit y are also
diagnost ic f eat ures w hen t aken t oget her w it h t he ot her f indings. Furt her
ident if icat ion of t he subt ype of CI M (t hick f ilament myosin loss,
rhabdomyolysis, necrot izing myopat hy of int ensive care, or cachect ic myopat hy)
may aid in prognost icat ion, since only necrot izing myopat hy of int ensive care is
associat ed w it h a poor prognosis f or ret urn of muscle st rengt h.
The def init ive pat hophysiology of CI M is st ill elusive, but most clinicians and
scient ist s t o dat e agree t hat t he syst emic inf lammat ory response syndrome
(SI RS) account s f or t he muscle organ f ailure, w it h CI M similar t o ot her organ
f ailures seen in pat ient s w it h SI RS. This is due t o t he increased cyt okine burden
having a st rong pot ent ial f or mediat ing sepsis-induced prot eolysis of myof ibrillar
prot eins in muscle. I n addit ion t o cyt okine mediat ion, ext ernal f act ors can t rigger
t he onset of CI M, w it h t he most import ant being high-dose glucocort icoids.
Addit ionally, t he prot ein degradat ion pot ency of st eroids is increased by
impaired neuromuscular t ransmission, w hich can be induced by surgical st ress,
drugs, or an underlying crit ical illness neuropat hy. Taken t oget her, sepsis,
st eroids, and impaired neuromuscular t ransmission act synergist ically t o
st imulat e muscle prot eolysis charact erist ic of CI M. The def ined subt ypes of CI M
may t heref ore ref lect t he balance of prot eolyt ic input s in a dose-dependent
manner, exposure-dependent manner, or bot h, and t he severit y of CI M may be
cont rolled t hrough limit ing t hese input s.
I n t erms of alt ering neuromuscular t ransmission, more t han 50 drugs, not ably
neuromuscular blocking agent s (i. e. , pancuronium and vecuronium),
aminoglycosides, clindamycin, and colist in cause pharmacologic muscle
denervat ion. I n addit ion, denervat ion also causes a rise in glucocort icoid
recept ors in t he cyt osol of skelet al muscle result ing in increased sensit ivit y of
muscle t o st eroids. Thus, t he pot ent ial exist s f or a vicious cycle t o develop w hen
neuromuscular blockade, aminoglycosides, and st eroids are administ ered
simult aneously in t he crit ically ill I CU pat ient .
There is no current t reat ment f or CI M, alt hough int ensive physiot herapy seems
promising. The current mainst ay of CI M management is direct ed at prevent ion
t hrough judicious use of drugs associat ed w it h t he development of t he myopat hy.
I t should be not ed, how ever, t hat ot her myt ot oxic f act ors responsible f or
producing myopat hies in t he crit ically ill have been ident if ied, independent of
sepsis or glucocort icoids or neuromuscular blocker administ rat ion. Most
subt ypes of CI M st ill carry a good prognosis f or recovery of muscle st rengt h
over t ime, w it h t he except ion of necrot izing myopat hy.
Suggested Readings
Bolt on CF. Neuromuscular manif est at ions of crit ical illness. Muscle Nerve
2005; 32: 140—163.
Friedrich O , Fink RHA, Hund E. Underst anding crit ical illness myopat hy:
approaching t he pat homechanism. J Nut r 2005; 135: 1813S—1817S.
Lat ronico N, Peli E, Bot t eri M. Crit ical illness myopat hy and neuropat hy. Curr
O pin Crit Care 2005; 11: 126—132.
Lat ronico N, Shehu I , Seghelini E. Neuromuscular sequelae of crit ical illness.

Curr O pin Crit Care 2005; 11: 381—390.
> Table of C ontents > Medic ations > 6 - Us e P r ophylaxis for the Im m ediate S ide E ffec ts of S ter oids
6
Use Prophylaxis for the Immediate Side Effects
of Steroids
Rajan G upta MD
G lucocort icoids are most commonly used t o mit igat e an inf lammat ory response.
They are used in a w ide variet y of pat ient populat ions seen in t he int ensive care
unit (I CU). Unf ort unat ely, t hese agent s have many side eff ect s, w hich of t en limit
t heir use. I n t he I CU set t ing, t he f unct ions most commonly aff ect ed are
endocrine regulat ion, immune response, and gast roint est inal (G I ) int egrit y, as
w ell as skin and w ound healing.
Watch Out For

Derangement in glucose met abolism can be caused by glucocort icoids t hrough
decreased insulin product ion, increased insulin resist ance, and alt ered glucose
synt hesis. St eroid-induced diabet es should be managed similarly t o glucose
cont rol used f or ot her pat ient s in t he I CU. I nsulin inf usion t herapy t o cont rol
hyperglycemia is of t en necessary. Hyperglycemia secondary t o glucocort icoid
t reat ment t ypically recedes w it hin 48 hours of discont inuat ion of t he
glucocort icoids.
The immune-modulat ing eff ect s of glucocort icoids are cent ral t o t heir t herapeut ic
eff ect s as w ell as one of t heir great est limit at ions. They can cause a
leukocyt osis in t he absence of syst emic inf ect ion. I mmune suppression occurs
f rom eff ect s on inhibit ion of inf lammat ory cells as w ell as inhibit ion of t he release
of cyt okines and proinf lammat ory modulat ors. Some have suggest ed t hat
syst emic f ungal inf ect ion is a cont raindicat ion t o t he inst it ut ion of st eroid
t herapy.
St eroids impair w ound healing by inhibit ing t he early inf lammat ory phase and t he
at t ract ion of cells cent ral t o t he process of w ound healing. They also inhibit new
prot ein synt hesis, leading t o decreased collagen deposit ion by f ibroblast s.
Wounds event ually heal w it h t he same t ensile st rengt h but t ake longer t o do so.
Limit ed case report s suggest t hat vit amin A administ rat ion can help ameliorat e
some of t hese delet erious eff ect s on w ound healing, and supplement at ion may
be considered in t hese pat ient s.
Wit hin t he gast roint est inal syst em, glucocort icoid t herapy can result in increased
acid secret ion, decreased gast ric mucous product ion, and pariet al cell
hyperplasia. This leads t o an increased incidence of pept ic ulcer disease and
upper G I bleeding in pat ient s receiving t his
t herapy. I CU pat ient s of t en carry ot her risk f act ors f or gast roint est inal st ress
ulcerat ion as w ell. Pat ient s receiving st eroids should have st ress ulcer
prophylaxis in t he f orm of H2-blockers or prot on pump inhibit ors (PPI s). St udies
have not show n a clear benef it of PPI s over H2-blockers. Early ent eral f eeding
may also help reduce t he incidence of st ress ulcers in t he I CU pat ient
populat ion.
Long-t erm glucocort icoid t herapy may lead t o suppression of t he hypot halamus-
pit uit ary-adrenal axis. Sudden w it hdraw al of st eroid t herapy can cause severe
adrenal insuff iciency and hemodynamic inst abilit y. Adrenal insuff iciency can also
result af t er a pat ient experiences increased st ress f rom inf lammat ion, surgery,
or t rauma w it hout an increase in t he exogenous dose of st eroids. Thus, st ress-
dose t herapy should be considered f or pat ient s on previous st eroid t herapy
undergoing surgery. Similarly, a pat ient suff ering f rom t he st ress of surgery,
t rauma, or inf lammat ion w ho is not receiving exogenous st eroid t herapy is also
at risk t o develop relat ive adrenal insuff iciency. This is usually due t o an inabilit y
of t he hypot halamus-pit uit ary-adrenal axis t o appropriat ely respond t o t he
increased need f or circulat ing cort icoids, result ing in a physiologic def iciency.
Thus, t he dosing f or replacement t herapy in such pat ient s is signif icant ly low er
t han st ress or t herapeut ic dosing, and it remains unclear if t he side eff ect s of
t his low er dosing are as pronounced. Adrenal insuff iciency should also be a part
of t he diff erent ial diagnosis in I CU pat ient s w it h hypot ension and hemodynamic
inst abilit y ref ract ory t o f luid and vasopressor management .
Suggested Readings
Brit t RC, Devine A, Sw allen KC, et al. Cort icost eroid use in t he int ensive care
unit : at w hat cost ? Arch Surg 2006; 141: 145—149.
Daley RJ, Rebuck JA, Welage LS, et al. Prevent ion of st ress ulcerat ion:
current t rends in crit ical care. Crit Care Med 2004; 32: 2008—2013.
Schacke H, Docke WD, Asadullah K. Mechanisms involved in t he side eff ect s

of glucocort icoids. Pharmacol Ther 2002; 96: 23—43.
> Table of C ontents > Medic ations > 7 - S pec ific ally Q uer y for P r evious S ter oid Us e
7
Specifically Query for Previous Steroid Use
Lisa Marcucci MD
Human st eroids are produced by t he adrenal gland and are under t he direct or
indirect cont rol of t he hypot halamus, pit uit ary, and adrenal glands. The t w o major
classes of st eroids t hat have signif icant clinical met abolic eff ect s are t he
glucocort icoids (mainly cort isol), w hich regulat e glucose and ot her anabolic
cascades, and t he mineralocort icoids (mainly aldost erone), w hich handle Na–K
equilibrium.
Pat ient s can become def icient in st eroid product ion t hrough primary Addison's
disease (e. g. , adrenal cort ex dest ruct ion, hemorrhage) or secondary Addison's
disease (e. g. , t hrough a def iciency of cort icot ropin or adrenocort icot ropic
hormone [ ACTH] or t hrough exogenous st eroid administ rat ion). Primary Addison's
disease is rare, but secondary Addison's disease is not uncommon. Exogenous
glucocort icoids are used in a variet y of diseases including organ and bone
t ransplant s, rheumat oid art hrit is, syst emic lupus eryt hemat osus and ot her
collagen vascular disorders, psoriasis, chronic obst ruct ive pulmonary disease,
inf lammat ory bow el disease, and many hemat ological diseases such as
idiopat hic t hrombocyt openic purpura (I TP).
What to Do
Alt hough t he indicat ions and dosages of st ress-dose st eroids f or t hese pat ient s
undergoing surgery is an area of act ive discussion, it is imperat ive t hat t he
ast ut e clinician know s st eroid hist ory so t hat an inf ormed decision can be made.
Acut e adrenal insuff iciency is a morbid and somet imes f at al condit ion t hat can be
manif est ed by circulat ory collapse, f ever, hypoglycemia, and depressed ment al
st at us. Because of t he seriousness of t his condit ion and t he relat ive ease of
adequat ely t reat ing it prophylact ically, all pat ient s should be specif ically queried
as t o w het her t hey have ever been on st eroids and if so, how much and w hen.
Some experienced clinicians f eel t he level of adrenal f unct ioning remains
decreased if an equivalent of 5 mg of prednisone w as administ ered f or at least 2
w eeks in t he previous year.
O ne f inal not e is t hat 1 mg of dexamet hasone equals 5 mg of prednisone, w hich
equals 25 mg of hydrocort isone (w hich is similar t o nat ural cort isol).
Suggested Readings
Annane D. Eff ect of t reat ment w it h low doses of hydrocort isone and
f ludrocort isone on mort alit y in pat ient s w it h sept ic shock. JAMA
2002; 288: 862–871.
Cooper MS. Cort icost eroid insuff iciency in acut ely ill pat ient s: current
concept s. N Engl J Med 2003; 348: 727–734.
Salem M. Perioperat ive glucocort icoid coverage: a reassessment 42 years

af t er emergence of a problem. Ann Surg 1994; 219: 416–425.
> Table of C ontents > Medic ations > 8 - D o Not Us e S uc c inylc holine in P atients w ith B ur ns , P ar alys is ,
or O ther High P otas s ium S tates
8
Do Not Use Succinylcholine in Patients with
Burns, Paralysis, or Other High Potassium
States
Eugenie S. Heitmiller MD
Succinylcholine is a depolarizing muscle relaxant used f or urgent endot racheal
int ubat ion in t he operat ing room, int ensive care unit (I CU), and emergency
depart ment . The int ravenous (I V) dose is 0. 5 t o 1. 5 mg/ kg. Time t o eff ect is 1
minut e and durat ion of act ion in pat ient s w it h normal pseudocholinest erase
act ivit y is approximat ely 2 minut es w it h complet e recovery in 5 minut es. I f a
w orking I V cat het er is not available, succinylcholine may be given 3 t o 4 mg/ kg
int ramuscularly (I M) (maximum dose 150 mg) w it h an onset of 2 t o 3 minut es.
Muscle f asciculat ions w ill of t en be seen prior t o complet e muscle relaxat ion.
Know n side eff ect s of succinylcholine are bradycardia and muscle pain.
Succinylcholine is indicat ed w hen rapid sequence induct ion is needed t o quickly
secure t he airw ay w it h an endot racheal t ube. Alt hough succinylcholine may
increase int racranial pressure, it can be used in pat ient s w it h acut e head t rauma
because t he rapid onset , superior int ubat ing condit ions, and reversibilit y
out w eigh t he risk associat ed w it h it s use. Succinylcholine in low er doses is also
used t o t reat laryngospasm.
What Not to Do
Succinylcholine is cont raindicat ed in pat ient s af t er t he acut e phases of major
t rauma or burns, ext ensive denervat ion of skelet al muscle, upper or low er mot or
neuron injury, and severe inf ect ions, part icularly clost ridia, bot ulism, and t et anus,
because succinylcholine in t hese pat ient s may result in severe hyperkalemia and
cardiac arrest . The risk of hyperkalemia in t hese pat ient s increases over t ime
and usually peaks at 7 t o 10 days af t er t he injury, alt hough t he precise t ime of
onset and t he durat ion of t he risk period are unknow n. I t is also cont raindicat ed
in pat ient s w it h a personal or f amily hist ory of malignant hypert hermia or skelet al
muscle myopat hies and in pat ient s w it h disuse at rophy because acut e severe
rhabdomyolysis may occur w it h subsequent hyperkalemia, vent ricular
arrhyt hmias, and cardiac arrest result ing.
The mechanism by w hich succinylcholine use result s in hyperkalemia is relat ed t o
it s eff ect on muscle nicot inic acet ylcholine recept ors. I n t he condit ions at risk f or
hyperkalemia list ed previously,
t here is an increase of muscle nicot inic acet ylcholine recept ors, w hich w hen
depolarized by succinylcholine leads t o eff lux of int racellular pot assium int o t he
plasma, leading t o acut e hyperkalemia.
Alt ernat ive agent s are nondepolarizing agent s such as vecuronium and
rocuronium, but t hese do not result in as rapid a relaxat ion.
Suggested Readings
Koenig KL. Rapid-sequence int ubat ion of head t rauma pat ient s: prevent ion of
f asciculat ion w it h pancuronium versus minidose succinylcholine. Ann Emerg
Med 1992; 21: 929–932.
Mart yn JAJ, Richt sf eld M. Succinylcholine-induced hyperkalemia in acquired

pat hologic st at es. Anest hesiology 2006; 104: 158–169.
Naguib M, Samarkandi AH, El-Din ME, et al. The dose of succinylcholine

required f or excellent endot racheal int ubat ing condit ions. Anest h Analg
2006; 102: 151–155.
> Table of C ontents > Medic ations > 9 - C ons ider Us ing C is atr ac ur ium for Neur om us c ular P ar alys is in
P atients w ith Hepatic and R enal Failur e
9
Consider Using Cisatracurium for
Neuromuscular Paralysis in Patients with
Hepatic and Renal Failure
Muhammad I. Durani MD
The ideal neuromuscular blocking (NMB) drug w ould be rapid in onset , have a
predict able off set , be nont oxic, lack delet erious cardiovascular or aut onomic
eff ect s, undergo a def ined means of met abolism and excret ion pref erably
independent of end-organ f unct ion, and be inexpensive. Many of t hese
charact erist ics are f ound in clinically available drugs like cisat racurium.
To brief ly review, neuromuscular blockade occurs via one of t w o diff erent
pharmacologic modes. Drugs t hat act as a prolonged agonist at t he nicot inic
acet ylcholine (nACh) recept or are called depolarizing agent s (e. g. ,
succinylcholine). Succinylcholine at t aches t o each of t he alpha subunit s of t he
nACh recept or and mimics t he act ion of acet ylcholine, t hus depolarizing t he
post junct ional membrane. Neuromuscular blockade develops because a
depolarized post junct ional membrane cannot respond t o subsequent release of
acet ylcholine.
A second group of WMB agent s t hat bind noncovalent ly and compet it ively t o
nACh recept ors and inhibit neuromuscular t ransmission are called
nondepolarizing NMB drugs. I n high doses, t hese drugs may act by blocking t he
ion recept or channels. O ccupat ion of as many as 70% of t he nACh recept ors
does not produce evidence of neuromuscular blockade. Neuromuscular
t ransmission, how ever, f ails w hen 80% t o 90% of t he recept ors are blocked.
Nondepolarizing NMB drugs may be classif ied on t he basis of t heir st ruct ure and
durat ion of act ion. Clinically available nondepolarizing drugs can be grouped int o
t w o basic st ruct ural cat egories. The benzylisoquinolinium drugs (at racurium,
cisat racurium, mivacurium, t ubocurarine) t end t o be pot ent (and t heref ore slow er
in onset ) NMB drugs t hat are eliminat ed by t he kidneys or by Hof mann
eliminat ion and may t rigger hist amine release. Conversely, t he aminost eroid
compounds (pancuronium, vecuronium, rocuronium) are less pot ent , have a
f ast er onset of act ion, are eliminat ed by t he liver w it h act ive met abolit es, and
lack signif icant hist amine release or aut onomic int eract ions.
NMB drugs also may be classif ied as short (succhinylcholine, mivacurium),
int ermediat e (at racurium, cisat racurium, vecuronium,
rocuronium), or long-act ing (t ubocurarine, pancuronium) on t he basis of t heir

durat ion of act ion.
Cisatracurium
This is t he purif ied f orm of one of t he t en st ereoisomers of at racurium.
Cisat racurium has an ED95 of 50 µg/ kg and has an onset of act ion of 3 t o 5
minut es and durat ion of neuromuscular blockade last ing 20 t o 35 minut es.
Neuromuscular blockade is easily maint ained at a st able level by inf usion at a
const ant rat e and does not diminish over t ime. I n cont rast t o vecuronium, t he
rat e of spont aneous recovery f rom cisat racurium-induced neuromuscular
blockade is not inf luenced by lengt h of inf usion in pat ient s requiring mechanical
vent ilat ion.
Cisat racurium undergoes spont aneous nonenzymat ic degradat ion at normal body
t emperat ure and pH by a base-cat alyzed react ion t ermed Hof mann el i mi nati on,
t o f orm laudanosine and monoquat ernary acrylat e. Hof mann eliminat ion
represent s a chemical mechanism of eliminat ion account ing f or 77% of t he
clearance of cisat racurium, w hereas renal clearance is responsible f or anot her
16%. The organ-independent clearance of cisat racurium means t hat t his
nondepolarizing NMB drug can be administ ered t o pat ient s w it h hepat ic or renal
dysf unct ion w it hout a change in it s neuromuscular blocking prof ile. The
pharmacokinet ics of cisat racurium is only marginally inf luenced by advanced age.
The met abolit es of cisat racurium by Hoff man eliminat ion are inact ive at t he
neuromuscular junct ion (NMJ). I n cont rast t o at racurium, plasma concent rat ions
of laudanosine af t er administ rat ion of 2 × ED95 dose of cisat racurium are
f ivef old less t han t hat present af t er a 1. 5 × ED95 dose of at racurium (Lien et al. ,
1996). Cisat racurium, in cont rast t o at racurium, is devoid of hist amine-releasing
eff ect s such t hat cardiovascular changes do not accompany t he rapid
int ravenous (I V) administ rat ion of even large doses (8 × ED95) of cisat racurium.
Cisat racurium administ ered t o adult neurosurgical pat ient s produces less
cerebral hemodynamic changes compared w it h equipot ent doses of at racurium.
Pancuronium
This is a long-act ing aminost eroid nondepolarizing NMB drug w it h an ED95 of 70
µg/ kg t hat has an onset of act ion in 3 t o 5 minut es and a durat ion of
neuromuscular blockade last ing 60 t o 90 minut es. An est imat ed 80% of a single
dose of pancuronium is eliminat ed unchanged in t he urine. I n renal f ailure, t he
plasma clearance
is decreased 33% t o 50%. An est imat ed 10% t o 40% undergoes hepat ic
deacet ylat ion. The 3-desacet ylpancuronium met abolit e is approximat ely 50% as
pot ent as pancuronium at t he NMJ. Pat ient s w it h t ot al biliary obst ruct ion and
hepat ic cirrhosis have an increased volume of dist ribut ion, decreased plasma
clearance, and prolonged eliminat ion half -t ime of pancuronium.
Vecuronium
This is an int ermediat e-act ing aminost eroid nondepolarizing NMB drug w it h an
ED95 of 50 µg/ kg t hat produces an onset of act ion in 3 t o 5 minut es and a
durat ion of neuromuscular blockade last ing 20 t o 35 minut es. Vecuronium
undergoes bot h hepat ic met abolism and renal excret ion. The 3-
desacet ylvecuronium met abolit e is approximat ely one half as pot ent as t he
parent compound. The eliminat ion half -t ime is prolonged in pat ient s w it h renal
f ailure. I n pat ient s w it h cholest asis, hepat ic cirrhosis, or alcoholic liver disease,
t he administ rat ion of vecuronium, 0. 2 mg/ kg, result s in a prolonged eliminat ion
half -t ime and increased durat ion of act ion.
Rocuroni um is an int ermediat e-act ing aminost eroid w it h an ED95 of 0. 3 mg/ kg
t hat has an onset of act ion in 1 t o 2 minut es and a durat ion of neuromuscular
blockade last ing 20 t o 35 minut es. Rocuronium is largely excret ed unchanged (up
t o 50% in 2 hours) in t he bile. Liver disease increases t he volume of dist ribut ion
of rocuronium and could result in a longer durat ion of act ion, especially w it h
repeat ed doses or prolonged I V administ rat ion. Renal excret ion of rocuronium
may be >30% in 24 hours, and administ rat ion of t his drug t o pat ient s in renal
f ailure may produce a modest ly prolonged durat ion of act ion.
Suggested Readings
Lien CA, Schmit h VD, Belmont MR, Abalos A, Kisor DF, and Savarese JJ.
Pharmacokinet ics of cisat racurium in pat ient s receiving nit rous
oxide/ opioid/ barbit urat e anest hesia. Anesthesi ol ogy 84: 300–308, 1996.
Murray MJ, Cow en J, DeBlock H, Erst ad B, G ray AW, Jr. , Tescher AN, McG ee
WT, Prielipp RC, Susla G , Jacobi J, Nasraw ay SA, Jr. , and Lumb PD. Clinical
pract ice guidelines f or sust ained neuromuscular blockade in t he adult crit ically
ill pat ient . Crit Care Med 30: 142–156, 2002.
St oelt ing RK, ed. Pharmacology and Physiology in Anest het ic Pract ice. 3rd
Ed. Philadelphia: Lippincot t Williams & Wilkins; 1999: 182–223.
> Table of C ontents > Medic ations > 10 - R em em ber that ther e ar e tw o “ Neos ”
10
Remember that there are two “Neos”
I n our current cult ure of pat ient saf et y, it is import ant t o avoid commonly used
medical abbreviat ions so t hat medical errors can be prevent ed. For example, t he
t erm neo is of t en used in t he operat ing room and int ensive care unit (I CU) t o
ref er t o t he drug Neo-Synephrine; how ever, it could also reasonably be
int erpret ed as neost igmine, anot her commonly used drug in t he operat ing room
and I CU. To avoid t his conf usion and t o decrease t he risk of a medicat ion error,
t he use of t he brand name Neo-Synephrine should be abandoned.
Neo-Synephrine
Neo-Synephrine is t he t rade name f or phenylephrine. I t is a pure alpha recept or
agonist and has bot h venous and art erial const rict ive eff ect s. BecauseΑ1
recept ors have been discovered in t he myocardium, it is also possible t hat it has
posit ive inot ropic eff ect s. Acut ely, phenylephrine causes an increase in venous
ret urn (preload) due t o it s venous const rict ive eff ect s; it increases af t er load as
w ell. I n normal individuals, it does not aff ect cardiac out put . How ever, in pat ient s
w it h ischemic heart disease, it can decrease cardiac out put . O t her uses f or
phenylephrine include reversing right -t o-lef t shunt f low in pat ient s w it h t et ralogy
of Fallot and t erminat ing supravent ricular t achycardias (SVTs) as it can cause
ref lex vagal st imulat ion in response t o elevat ed blood pressure. I n t his last
circumst ance, phenylephrine is part icularly usef ul as it t reat s bot h t he arrhyt hmia
and t he hypot ension.
Phenylephrine is given as eit her an int ravenous (I V) bolus or inf usion; bolus
doses are 1 t o 10 mcg/ kg, or in 50- t o 100-mcg boluses in adult s. As an I V
inf usion, it is usually mixed as 10 t o 15 mg in 250 mL and dosed as 0. 15 t o 0. 75
mcg/ kg/ min. For use in pediat ric pat ient s w it h t et ralogy of Fallot , I V bolus doses
are 5 t o 50 mcg/ kg.
Neostigmine
Neost igmine (market ed under t he t rade name Prost igmin) is a reversible
acet ylcholinest erase inhibit or. I t is used clinically in t he t reat ment of myast henia
gravis, glaucoma, and at ony of t he gast roint est inal and urinary t ract s. I n t he
operat ing room and I CU set t ings, it is commonly used t o reverse nondepolarizing
neuromuscular blockade.
Alt hough it ant agonizes neuromuscular blocking agent s at nicot inic recept ors, it
has muscarinic eff ect s as w ell. I t s adverse eff ect s result f rom it s act ion at t hese
recept ors: bradycardia and bradyarrhyt hmias; increased salivat ion; increased
bow el mot ilit y; and possibly bronchospasm. To count eract t hese adverse eff ect s,
neost igmine is t ypically given in conjunct ion w it h an ant icholinergic agent –eit her
at ropine or glycopyrrolat e.
The dose of int ravenous neost igmine required t o reverse neuromuscular blockade
is great er in long-act ing neuromuscular blockers t han short -act ing blockers. For
example, 40 t o 50 mcg/ kg of neost igmine is required t o reverse a 90% block
produced by pancuronium or d-t ubocurarine, w hereas 20 t o 30 mcg/ kg is needed
f or at racurium, vecuronium, and rocuronium, and 5 mcg/ kg is needed f or
mivacurium. Because neost igmine dosage has a ceiling eff ect , administ ering
great er t han 0. 07 mg/ kg has lit t le benef it .
Suggested Readings
Barash PG , Cullen BF, St oelt ing RK, ed. Clinical Anest hesia. 4t h Ed.
Philadelphia: Lippincot t Williams & Wilkins; 2001: 296–297, 439–441.
Hensley FA Jr. , Mart in DE, G ravlee G P, ed. A Practi cal Approach to Cardi ac
Anesthesi a. 3rd Ed. Philadelphia: Lippincot t Williams & Wilkins; 2003: 43.
> Table of C ontents > Medic ations > 11 - Tr eat Neur oleptic Malignant S yndr om e as an E m er genc y and
R em em ber its P r es entation m ay not be D os e D ependent
11
Treat Neuroleptic Malignant Syndrome as an
Emergency and Remember its Presentation may
not be Dose Dependent
Eliahu S. Feen MD
Jose I. Suarez MD
Neurolept ic malignant syndrome (NMS) is a clinical syndrome consist ing of f our
primary f eat ures: rigidit y, alt ered ment al st at us, hypert hermia, and aut onomic
inst abilit y. I t occurs in t he set t ing of t he use of dopamine-blocking agent s or t he
w it hdraw al of dopamine-enhancing medicat ions.
Epidemiology
I ncidence of NMS is est imat ed at 0. 1% t o 2%. There is a preponderance of
young men w it h NMS, but t his may be because of t he increased f requency of
schizophrenia and aff ect ive disorders in t his group and t he subsequent increased
use of neurolept ics. O t her risk f act ors may include neurolept ic-induced
cat at onia, dehydrat ion or malnut rit ion as a precipit at ing cause, and a hist ory of
elevat ed serum creat inine kinase (CK) during psychot ic episodes not in
associat ion w it h NMS.
Clinical Presentation
The mot or sympt oms consist most commonly of parkinsonian-t ype sympt oms
such as “lead-pipe” rigidit y, but ot her sympt oms include a t remor superimposed
on t he rigidit y (“cogw heel rigidit y”), akinesia, bradykinesia, and dyst onia (e. g. ,
blepharospasm, opist hot onus, oculogyric crises, t rismus, and orobuccal
dyskinesia). The alt ered ment al st at us ranges f rom delirium t o st upor or even
coma. A f ever is seen in almost all cases (unlike malignant hypert hermia) and is
usually great er t han 38°C and of t en great er t han 41°C. Arrhyt hmias, blood
pressure f luct uat ions, and respirat ory abnormalit ies const it ut e t he main,
pot ent ially lif e-t hreat ening aut onomic f eat ures of NMS. Rare clinical
manif est at ions include seizures, at axia, and nyst agmus. NMS usually evolves
over 24 t o 72 hours. O ccasionally t here can be a slow er progression over days,
but in t he case of depot neurolept ics (e. g. , int ramuscular f orm of f luphenazine),
progression over a couple of w eeks has been report ed. The clinical course last s
7 t o 10 days, w it h a longer t ime if t he incit ing agent w as depot neurolept ics,
because of slow er
clearance. Laborat ory abnormalit ies f ound in associat ion w it h NMS are
leukocyt osis in t he range of 10 t o 40, 000 cells/ α L and elevat ed serum creat inine
kinase (CK) in t he range of 200 t o several t housand I U/ L.
Cases of NMS w it h prominent medical complicat ions and f at alit ies are w idely
described. Some of t he causes f or t he morbidit y and mort alit y of NMS relat e t o
it s secondary complicat ions. Because of t he associat ed rhabdomyolysis, severe
dehydrat ion and prerenal acut e renal f ailure can occur. This is predict ive of
mort alit y, w it h aggressive hydrat ion absolut ely necessary (urine alkalinizat ion is
cont roversial). Because of rigidit y and consequent immobilit y, as w ell as
act ivat ion of t he coagulat ion cascade t o met abolic condit ions creat ed by t he
rhabdomyolysis, venous t hromboembolism has been report ed. Pulmonary
embolism has been report ed t o cause almost one quart er of f at alit ies. I n severe
cases, aspirat ion pneumonia, respirat ory f ailure, cardiac arrhyt hmias, and
dysaut onomia have been report ed.
Some import ant clinical syndromes w it h a similar clinical pict ure f orm a
diff erent ial f or NMS. Malignant hypert hermia appears in t he set t ing of exposure
t o cert ain anest het ic agent s such as halot hane, isof lurane, sevof lurane, and
desf lurane, and depolarizing muscle relaxant s like succinylcholine. Acut e let hal
cat at onia involves hypert hermia, akinesia, and rigidit y, but t his is usually
preceded in t he previous couple of w eeks by behavioral changes. The serot onin
syndrome occurs in t he set t ing of an overexposure t o select ive serot onin
reupt ake inhibit ors (SSRI s) or t he combinat ion use of SSRI s t oget her w it h
monoamine oxidase inhibit ors, t ricyclic ant idepressant s, or meperidine.
Precipitating Factors
NMS is clearly medicat ion relat ed (Table 11. 1). For t he neurolept ic medicat ion-
relat ed cases, t he onset is not dose relat ed and can occur many mont hs af t er
init iat ion of t herapy. For pat ient s w ho have been on dopamine agonist s (most
t ypically Parkinson pat ient s), NMS t ypically occurs in t he set t ing of a sudden
w it hdraw al of t he agonist agent or t he change of dosage or change t o a diff erent
agonist alt oget her. Most of t he Parkinson pat ient s have had sympt oms of t heir
disease f or more t han 8 years. The perioperat ive period is a classic scenario f or
onset of NMS because of t he alt erat ions in serum levels of t he agonist relat ed t o
changes w hen pat ient s t ake t heir medicat ions or t o met abolic alt erat ions. For
pat ient s w ho have been on dopamine-blocking agent s, such as haloperidol, NMS
can occur even af t er init ial exposure. I ncreased dosages of ,
changes in t he part icular agent , or parent eral administ rat ion of t he neurolept ic
are risk f act ors f or t he development of NMS in t his group of pat ient s. The
et iology of NMS is consequent ly t hought t o be relat ed t o t he acut e blockade of
t he nigrost riat al and hypot halamic dopamine pat hw ays in t he brain.
TABLE 11-1 M EDICATIONS ASSOCIATED WITH THE

DEVELOPM ENT OF NM S
DOPAMINE
DOPAMINERGIC/DOPAMINE
ANTAGONIST /NEUROLEPT I
AGONIST S WIT HDRAWAL
ADMINIST RAT ION
Neuroleptics: phenothiazines
butyrophenones, thiothixanes
Atypical antipsychotics:
clozapine, olanzapine,
risperidone, quetiapine
Levodopa Antiemetics: metoclopramide
COMT inhibitors: tolcapone, droperidol, prochlorperazine,
entacapone promethazine
Dopamine agonists: Others: reserpine,
bromocriptine, pergolide, carbamazepine
ropinirole, pramipexole, Rare reported cases in:
cabergoline, apomorphine overdose of some tricyclic
Amantadine and SSRI antidepressants,
overdose of citalopram,
loxapine, diatrizoate, lithium
Rare reported cases in drug
abuse: cocaine,
amphetamines
M anagement
Treat ment involves f irst and f oremost reversing t he incit ing cause—eit her
discont inuing t he neurolept ic/ dopamine-mediat ing agent or reinst it ut ing t he
dopaminergic t herapy t hat may have been st opped. Support ive care is essent ial
and possibly lif esaving, because of t he secondary complicat ions. Aggressive
cooling, caref ul monit oring of cardiovascular f unct ions, and high-volume
int ravenous f luid t herapy necessit at e int ensive care hospit alizat ion f or all but t he
mild cases. Pharmacologic t herapy includes bromocript ine (a dopamine agonist )
and/ or dant rolene (a skelet al muscle relaxant ) in addit ion t o support ive care.
These t w o medicat ions are used independent ly or t oget her. Pharmacologic
t herapy improves t he prognosis of NMS, decreasing t ime of resolut ion of
sympt oms f rom 16 days w it h support ive t herapy alone t o about 9 days.
Medicat ion administ rat ion is st art ed immediat ely and cont inued f or 10 days af t er
t he resolut ion of sympt oms. How ever, residual parkinsonian or cat at onic
sympt oms can persist f or w eeks af t er t he acut e episode resolves.
Elect roconvulsive t herapy (ECT) has been report ed t o be benef icial in cases of
NMS ref ract ory
t o ot her t herapies. For such cases ECT has been report ed t o resolve sympt oms
af t er t hree t o f our t reat ment s and is generally used f or 6 days t ot al.
Recurrence of NMS has been report ed. Some of t he risk f act ors f or recurrence
w hen neurolept ics are rest art ed include higher pot ency of neurolept ics,
rest art ing w it hin 2w eeks of an episode of NMS, higher init ial dose, and use of
concomit ant lit hium.
Suggested Readings
Bhanushali MJ, Tuit e PJ. The evaluat ion and management of pat ient s w it h
neurolept ic malignant syndrome. Neurol Clin 2004; 22: 389–411.
Chan TC, Evans SD, Clark RF. Drug-induced hypert hermia. Crit Care Clin
1997; 13: 785–808.
Sachdev PS. Neurolept ic-induced movement disorders: an overview. Psychiat r

Clin Nort h Am 2005; 28: 255–274.
> Table of C ontents > Medic ations > 12 - R em em ber that Malignant Hyper ther m ia m ay not have
Hyper ther m ia
12
Remember that Malignant Hyperthermia may not
have Hyperthermia
Malignant hypert hermia (MH) is a hypermet abolic disorder of skelet al muscle
t hat is t riggered in suscept ible individuals by several inhalat ion anest het ic agent s
(sevof lurane, desf lurane, isof lurane, halot hane, enf lurane, and met hoxyf lurane)
and succinylcholine. These anest het ic t riggers cause int racellular hypercalcemia
in skelet al muscle by decreasing t he upt ake of calcium by t he sarcoplasmic
ret iculum. The int racellular hypercalcemia act ivat es met abolic pat hw ays t hat
result in adenosine t riphosphat e (ATP) deplet ion, acidosis, membrane
dest ruct ion, and ult imat ely cell deat h. Suscept ibilit y t o MH is inherit ed as an
aut osomal dominant disorder and most suscept ible individuals are complet ely
asympt omat ic unt il exposed t o t riggering agent s.
Episodes of MH occur most of t en, but not exclusively, in children. The mean age
f or MH is 15 years, alt hough cases have been report ed in inf ant s as w ell as t he
elderly. The incidence of MH ranges f rom approximat ely 1 in 10, 000 t o 1 in
50, 000 individuals w ho are exposed t o t he t riggering agent s.
Signs and Symptoms

I n most cases, t he f irst signs and sympt oms of t he disorder are evident in t he
operat ing room. How ever, MH can also occur in t he recovery room or even af t er
t he pat ient has been t ransf erred t o t he pat ient f loor. I t is import ant t o remember
t hat none of t he signs and sympt oms occur in all cases. I nit ial signs usually
include t achycardia (90% of cases) and t achypnea (80% of cases) due t o t he
sympat het ic nervous syst em response t o t he underlying hypermet abolism and
hypercarbia (80% of cases). I n t he paralyzed pat ient under general anest hesia,
t he f irst sign of MH can be hypercarbia t hat is resist ant t o adjust ment s in t he
vent ilat or set t ing. Next , an increase in blood pressure, cardiac dysrhyt hmias, and
muscle rigidit y (80% of cases) are of t en seen. Pat ient s may t hen become
hypert hermic (70% of cases), w it h a rise of 1 t o 2 degrees Celsius every 5
minut es. An art erial blood gas w ill usually show respirat ory and met abolic
acidosis. O t her lab abnormalit ies include hyperkalemia, hypercalcemia, lact ic
acidemia and myoglobinuria. Creat inine kinase levels rise t o 20, 000 or more
w it hin 12 t o
24 hours, put t ing t he pat ient at risk f or myoglobinuric renal f ailure. I solat ed
myoglobinuria in t he early post operat ive period should make t he anest hesiologist
and surgical t eam suspicious f or MH. Masset er muscle rigidit y short ly af t er t he
administ rat ion of succinylcholine has also been associat ed w it h MH as w ell.
MH is a t reat able disorder. Wit h early diagnosis and appropriat e t reat ment , t he
mort alit y rat e approaches zero. I f MH is suspect ed, all t riggering agent s should
be discont inued immediat ely. I n t he operat ing room, t he pat ient should be
vent ilat ed w it h 100% oxygen at a f low of 10 L/ minut e. I f a general anest het ic
must be cont inued, it is saf e t o administ er barbit urat es, benzodiazepines,
opioids, and propof ol. The pat ient should be int ubat ed as soon as possible.
Cardiac dysrhyt hmias, hyperkalemia, acidosis, and ot her medical problems
should be managed appropriat ely. The pat ient should be given dant rolene as
soon as possible. Dant rolene act s by inhibit ing t he release of calcium f rom t he
sarcoplasmic ret iculum. The init ial dose is a 2. 5mg/ kg int ravenous (I V) bolus
f ollow ed by 1 mg/ kg I V every 6 hours f or at least 24 hours. Dant rolene's side
eff ect s include nausea, phlebit is, and w eakness f or approximat ely 24 hours af t er
t he drug is discont inued. Each vial of dant rolene cont ains 20 mg of dant rolene
and 3 mg of mannit ol and must be mixed w it h 60 mL of st erile w at er. The init ial
dose f or a 70-kg adult is 175 mg or 9 vials. Assist ance should be sought in
mixing t he dant rolene as it is poorly soluble. Dant rolene 2. 5 mg/ kg should be
given every 5 t o 10 minut es unt il t here is a f all in heart rat e, normal cardiac
rhyt hm, a reduct ion in muscle t one, and a decline in body t emperat ure.
Maint enance of normal volume st at us is essent ial in t he set t ing of
rhabdomyolysis. Hyperkalemia may require t reat ment w it h insulin and glucose.
Caut ion should be excercised w it h pot assium-losing diuret ics as hypovolemia
should be avoided. Furt hermore, administ rat ion of calcium should be avoided
unless t here is sympt omat ic hypocalcemia because of t he expect ed
hypercalcemia during t he recovery phases of rhabdomyolysis. Dysrhyt hmias can
be t reat ed w it h bet a-blockers and lidocaine. Calcium channel blockers should be
avoided. The pat ient 's t emperat ure should be managed w it h ice packs, cold I V
f luids, and cold w at er dow n t he nasogast ric t ube. Pat ient s should recover in t he
int ensive care unit (I CU) t o receive monit oring of art erial blood gas, elect rolyt es,
creat inine kinase, myoglobin, and lact at e levels. Pat ient s should also be
monit ored f or a recrudescence of MH t hat can occur w it hin hours of t he init ial
episode, as w ell as disseminat ed int ravascular coagulat ion (DI C) and
myoglobinuric renal f ailure, f or w hich t he pat ient s should receive ample volume
replacement .
The Malignant Hypert hermia Associat ion of t he Unit ed St at es (MHAUS) is a

nonprof it organizat ion t hat w as creat ed t o educat e t he public, t o counsel MH
pat ient s and t heir relat ives about t heir condit ion, and t o educat e anest hesia and
operat ing room personnel about t he management of MH. Ahot line is available t o
advise clinicians in t he acut e management of MH. Addit ional educat ional
mat erials can be obt ained by calling t he hot line number, 1-800-MHHY PER (1-
800-644-9737).
Suggested Readings
Barash, Paul G . , Cullen Bruce F. , St oelt ing, Robert K. ed. Cl i ni cal Anesthesi a
4t h ed. Lippincot t t Williams and Wilkins, Philadelphia, 2001 pp. 521–530.
Collins, Chad P. , Beirne, O . Ross. Concept s in t he Prevent ion and

Management of Malignant Hypert hermia. Journal of O ral and Maxi l l of aci al
Surgery 61: 1340–1345; 2003.
McCart hy, E. Jane. Malignant Hypert hermia: Pat hophysiology, Clinical

Present at ion, and Treat ment . Ameri can Associ ati on of Cri ti cal -Care Nurses
15: 231–237; 2004.
> Table of C ontents > Medic ations > 13 - R em em ber That Am iodar one C aus es Hypothyr oidis m
13
Remember That Amiodarone Causes
Hypothyroidism
Amiodarone has become a popular drug in t he int ensive care unit (I CU) because
of it s ut ilit y in t he management of a broad spect rum of vent ricular and at rial
arrhyt hmias. As a class I I I ant iarrhyt hmic agent , it s mechanism of act ion is t o
delay repolarizat ion and increase t he durat ion of t he act ion pot ent ial t hrough
inhibit ion of myocardial pot assium ion channels. While very eff ect ive as an
ant iarrhyt hmic, amiodarone t herapy is also associat ed w it h a w ide array of
adverse side eff ect s involving t he cornea, lungs, liver, skin, and t hyroid. Alt hough
t he majorit y of amiodarone's adverse eff ect s on t hese organs result f rom
deposit ion of t he drug in t he parenchyma, it s eff ect on t he t hyroid gland is quit e
unique.
Watch Out For

Amiodarone bears a st ruct ural resemblance t o t he t hyroid hormones t hyroxine
(T 4 ) and t riiodot hyronine (T3 ), w hile cont aining approximat ely 37% iodine by
w eight . Met abolism of amiodarone causes deiodinat ion, w hich result s in t he
release of up t o 50- t o 100-f old excess iodine compared w it h normal daily
int ake. Amiodarone-induced t hyroid dysf unct ion result s f rom t his massive
increase in t ot al iodine, as w ell as alt erat ions in hormone met abolism caused by
t he drug. While amiodarone can cause eit her hyper- or hypo-t hyroidism, it
t ypically induces hypot hyroidism in iodine-suff icient areas such as t he Unit ed
St at es w here t he incidence is est imat ed t o range up t o 30%. I n part icular,
pat ient s w it h preexist ing t hyroid aut oimmunit y are at increased risk f or t he
development of hypot hyroidism w hile receiving t he drug. The acut e eff ect s
observed af t er init iat ing t herapy is an increase in t hyroid-st imulat ing hormone
(TSH) levels by 20% t o 50%, w it h a decrease in serum T3 levels by 15% t o 20%
w it hin t he f irst 2 w eeks of t herapy. Amiodarone blocks t he peripheral conversion
of T4 t o T3 and inhibit s t he ent ry of t hese hormones int o peripheral t issue. The
result ing sympt oms of hypot hyroidism include cold int olerance, dry skin, w eight
gain, and f at igue. I n t he I CU, signs of hypot hyroidism may also manif est w hen a
pat ient is having diff icult y w eaning f rom mechanical vent ilat ion. O nce pat ient s
are diagnosed, t reat ment w it h
levot hyroxine should be init iat ed and amiodarone may be cont inued if a suit able
replacement ant iarrhyt hmic is not available.
Bef ore st art ing pat ient s in t he I CU on amiodarone t herapy, a caref ul examinat ion
of t he t hyroid should be perf ormed along w it h baseline measurement s of serum
TSH, f ree T4 , T3 , and t hyroid peroxidase and t hyroglobulin ant ibodies. This w ill
help det ect underlying t hyroid dysf unct ion and ident if y pat ient s w ho may be more
predisposed t o develop t hyroid dysf unct ion w hile on amiodarone. Thyroid
f unct ion should t hereaf t er be checked every 3 t o 6 mont hs w hile pat ient s are
cont inued on t herapy.
Suggested Readings
Basaria S, Cooper DS. Amiodarone and t he t hyroid. Am J Med 2005; 118: 706–
714.
Mart ino E, Bart alena L, Bogazzi F, et al. The eff ect s of amiodarone on t he
t hyroid. Endocr Rev 2001; 22: 240–254.
> Table of C ontents > Medic ations > 14 - D o not us e Am iodar one to R ate C ontr ol C hr onic Atr ial
Fibr illation
14
Do not use Amiodarone to Rate Control Chronic
Atrial Fibrillation
Chronic at rial f ibrillat ion (A f ib) has an incidence of 0. 4% in t he general
populat ion (approximat ely 10% of t hose more t han 60 years of age) and many of
t hese pat ient s are admit t ed t o int ensive or int ermediat e care unit s f or reasons
unrelat ed t o t his diagnosis. Chronic A f ib needs t o be clearly diff erent iat ed f rom
new -onset A f ib since t he t reat ment and management st rat egies w ill diff er.
Since chronic A f ib or paroxysmal A f ib has a very low likelihood of convert ing t o
sinus rhyt hm and remaining in sinus, t he primary st rat egy is rat e cont rol w it h
agent s t hat slow conduct ion t hrough t he at riovent ricular (AV) node so t hat t he
pat ient does not experience a rapid vent ricular response t hat may severely
compromise cardiac out put . Pat ient s w it h A f ib already have lost approximat ely
20% of t heir diast olic f illing secondary t o t he loss of coordinat ed “at rial kick” at
end diast ole. A rapid heart rat e t hat compromises diast olic f illing t ime f urt her can
t hus compromise cardiac out put t o t he point of shock. This may especially be
t rue w it h pat ient s w ho have diast olic dysf unct ion secondary t o aort ic st enosis
and chronic hypert ension. Commonly used agent s include bet a-blockers or
calcium channel blockers. Digoxin w as a mainst ay of t reat ment in t he past but is
much less commonly used current ly.
Watch Out For

Pharmacological rhyt hm cont rol, t hat is, convert ing A f ib t o sinus rhyt hm, is not
desirable in chronic A f ib. Pat ient s w it h chronic A f ib may develop clot in t he
f ibrillat ing at ria, part icularly in t he at rial appendage. Sudden conversion t o sinus
may lead t o an embolus t hat can result in a st roke, myocardial inf arct ion, gut
ischemia, or limb inf arct ion. For t his reason, pat ient s w it h chronic A f ib,
especially if it is paroxysmal in nat ure, are usually on ant icoagulat ion t herapy t o
reduce t he risk of developing at rial clot . Agent s t hat promot e conversion t o sinus
rhyt hm such as amiodarone and procainamide are not used t o acut ely t reat
chronic A f ib.
Treat ment and management of new -onset A f ib f alls int o t w o primary approaches
and pract it ioners vary as t o w hich t hey subscribe t o. The f irst , like chronic A f ib,
is rat e cont rol and t he second is rhyt hm
cont rol. So long as t he pat ient has not been in A f ib f or an ext ended period of
t ime (generally more t han 48 hours), t he risk of having developed clot is judged
t o be low and conversion may be desirable. This may be achieved elect rically
w it h synchronized cardioversion or pharmacologically w it h amiodarone or ot her
agent s. How ever, t he nat ural hist ory of new -onset A f ib is such t hat most
pat ient s (80%) w ill spont aneously convert on t heir ow n, usually w it hin 2 w eeks.
Several f act ors cont ribut e t o t his, most not ably t he reasons most pat ient s
develop new -onset A f ib. Post cardiac surgical pat ient s have some of t he highest
rat es of new -onset A f ib (report edly as high as 60%), presumably secondary t o
t issue damage associat ed w it h cannulat ion of t he at rium during surgery. G eneral
surgical pat ient s also have a high incidence of t his condit ion (as high as 40% in
noncardiac t horacic pat ient s) secondary t o a variet y of f act ors including at rial
st ret ch induced by hypervolemia and circulat ing cat echolamines. O nce t hese
pert urbat ions have resolved, so most ly does t he A f ib. Because of t his, many
advocat e t hat rat e cont rol is perf ect ly adequat e in new -onset A f ib since it is
usually a self -limit ing phenomenon. Amiodarone, bet ablockers, and calcium
channel blockers are good choices. Rat e cont rol may be t he goal (amiodarone
w ill slow t he rat e) but conversion, w het her spont aneous or pharmacological, is
likely t o be sust ained at w hich t ime t he drug may be w eaned off . For t he unusual
pat ient w ho remains in A f ib, one must t hen det ermine w het her t o sw it ch t o
chronic rat e cont rol w it h long-t erm ant icoagulat ion or consider
elect rophysiological st udies and possibly ablat ion t herapy.
Suggested Readings
De Denus S, Sanoski CA, Carlsson J, et al. Rat e vs rhyt hm cont rol in pat ient s
w it h at rial f ibrillat ion: a met a-analysis. Arch I nt ern Med 2005; 165: 258–262.
Nat t el S, O pie LH. Cont roversies in at rial f ibrillat ion. Lancet 2006; 367: 262–
272.
> Table of C ontents > Medic ations > 15 - E xer c is e C ar e in the us e of Am iodar one and Alter native
Antiar r hythm ic s for the Tr eatm ent of Atr ial Fibr illation
15
Exercise Care in the use of Amiodarone and
Alternative Antiarrhythmics for the Treatment of
Atrial Fibrillation
Alan Cheng MD
At rial f ibrillat ion is a common arrhyt hmia in t he int ensive care unit (I CU) set t ing.
The mainst ay of t herapy is rat e cont rol, w it h bet a-blockers being t he f irst -line
agent . How ever, f requent ly t he decision is made t o use rhyt hm-convert ing
agent s, such as amiodarone. Amiodarone is a complex ant iarrhyt hmic agent
(predominant ly class I I I ) t hat shares at least some of t he propert ies of each of
t he ot her t hree Vaughn-Williams classes of ant iarrhyt hmics. Amiodarone is
commonly used f or t he t reat ment and prevent ion of persist ent at rial and
vent ricular t achyarrhyt hmias, alt hough it is only FDA approved f or management
of vent ricular arrhyt hmias. I t is one of t he f ew agent s t hat can be used saf ely in
individuals w it h congest ive heart f ailure. Cont raindicat ions t o amiodarone include
severe sinus node dysf unct ion w it h marked sinus bradycardia or syncope,
second- or t hird-degree heart block, know n hypersensit ivit y t o it s cont ent s,
cardiogenic shock, possibly severe chronic lung disease.
Amiodarone is highly lipid soluble, ext ensively dist ribut ed in t he body, and highly
concent rat ed in many t issues, especially in t he liver and lungs. Af t er variable
(30% t o 50%) and slow gast roint est inal absorpt ion, amiodarone is very slow ly
eliminat ed, w it h a half -lif e of about 25 t o 110 days. The onset of act ion af t er oral
administ rat ion is delayed and a st eady-st at e drug eff ect may not be est ablished
f or several mont hs unless large loading doses are used. I t undergoes ext ensive
hepat ic met abolism t o t he pharmacologically act ive met abolit e
deset hylamiodarone (DEA). Amiodarone is not excret ed by t he kidneys but rat her
by t he lacrimal glands, t he skin, and t he biliary t ract . Neit her amiodarone nor
DEA is dialyzable.
Amiodarone is bot h an ant iarrhyt hmic and a pot ent vasodilat or. Amiodarone
lengt hens t he eff ect ive ref ract ory period by prolonging t he act ion pot ent ial
durat ion in all cardiac muscles, including bypass t ract s (class I I I act ivit y). I t also
has a pow erf ul class I ant iarrhyt hmic eff ect t hat w orks by inhibit ing inact ivat ed
sodium channels at high st imulat ion f requencies. Amiodarone slow s phase 4
depolarizat ion of t he sinus node as w ell as conduct ion t hrough t he
at riovent ricular (AV) node. I t also decreases Ca2+ current (class I V eff ect ) and
t ransient
out w ard delayed rect if ier and inw ard rect if ier K+ current s. Amiodarone
noncompet it ively blocks α - and β -adrenergic recept ors (class I I eff ect ); t his
eff ect is addit ive t o compet it ive recept or inhibit ion by bet a-blockers.
Watch Out For

There is a risk of hypot ension w it h int ravenous amiodarone administ rat ion, w hich
is more common w it h rapid administ rat ion. O t her not able acut e eff ect s include
bradycardia, hypokalemia, int eract ions w it h medicat ions such as w arf arin
(Coumadin) and digoxin, and rarely torsades de poi ntes. There is a risk of
pulmonary t oxicit y w it h high doses st art ing w it h pneumonit is and leading t o
pulmonary f ibrosis. O t her organ syst ems aff ect ed by amiodarone t herapy include
t hyroid (hypot hyroidism or hypert hyroidism), cent ral nervous syst em (proximal
muscle w eakness, peripheral neuropat hy, and neural sympt oms), gast roint est inal
(nausea 25%, elevat ed liver f unct ions), t est icular dysf unct ion, corneal
microdeposit ion, and phot osensit ive slat e-gray or bluish skin discolorat ion.
Agent s t hat can be considered as alt ernat ives t o amiodarone f or t he t reat ment of
at rial f ibrillat ion depend upon t he pat ient 's cardiac hist ory, as individuals w it h
reduced lef t vent ricular syst olic f unct ion are especially prone t o t he
proarrhyt hmic eff ect s of cert ain ant iarrhyt hmic agent s. Some commonly used
alt ernat ives include ibut ilide, dof et ilide, and sot alol. St rong considerat ion should
be made f or obt aining consult at ion w it h a cardiac elect rophysiologist prior t o
init iat ing t hese agent s.
I but ilide prolongs repolarizat ion by inhibit ion of t he delayed rect if ier pot assium
current (Ikr) and by select ive enhancement of t he slow inw ard sodium current .
This drug is eff icacious in t he t erminat ion of at rial f ibrillat ion (AF) and f lut t er w it h
bot h single and repeat ed int ravenous inf usions. I t is as eff ect ive as amiodarone
in cardioversion of at rial f ibrillat ion. The proarrhyt hmic eff ect result ing in
torsades de poi ntes is higher in individuals w it h heart f ailure, t hose w it h
bradycardia, nonw hit e subject s, w omen, and t hose given t he drug f or at rial
f lut t er rat her t han at rial f ibrillat ion. The risk of t his is great est during or short ly
af t er t he inf usion of t he drug (w it hin 1 hour) and rapidly w anes af t er
administ rat ion since t he half -lif e (2 t o 12 hours) of t his agent is short . The
pat ient should be monit ored f or at least 4 hours af t er t he st art of t he inf usion.
Dof et ilide prolongs t he act ion pot ent ial and Q TC in a concent rat ion-relat ed
manner. Dof et ilide exert s it s eff ect s solely by inhibit ion of t he rapid component
of t he delayed rect if ier pot assium
current Ikr. Dof et ilide has st ronger evidence in it s f avor f or acut e cardioversion
of at rial f ibrillat ion t han f or maint enance t hereaf t er, according t o a met a-
analysis. I t can be given t o pat ient s w it h depressed f unct ion but needs t o be
init iat ed w hile being cont inuously monit ored on t elemet ry f or t he f irst 3 days of
t herapy since t his t oo has a risk of proarrhyt hmia-like ibut ilide. The risk of
torsades de poi ntes can be reduced by normal serum pot assium and magnesium
levels, predose adjust ment of renal f unct ion, and post dose reduct ion based on
Q T C (ideally baseline Q TC below 429 milliseconds). Administ rat ion of dof et ilide
requires t hat t he hospit al and t he prescriber be t rained as conf irmed
administ rat ors.
Sot alol has combined class I I and class I I I propert ies and is act ive against a
variet y of arrhyt hmias and has t he abilit y t o produce prof ound bradycardia or
prolongat ion of t he Q T int erval. O f t he many indicat ions, sot alol is most
commonly used f or maint enance of sinus rhyt hm af t er cardioversion f or at rial
f ibrillat ion and reducing vent ricular t achyarrhyt hmias. Despit e it s abilit y t o
prevent t achyarrhyt hmias, sot alol can also be proarrhyt hmic (as w it h any ot her
ant iarrhyt hmic) t hrough it s abilit y t o prof oundly prolong t he Q T int erval. As a
result , init iat ion of t his drug should occur w hile t he pat ient is closely monit ored.
Sot alol is cont raindicat ed in pat ient s w it h reduced creat inine clearance (below
40 mL/ minut e) and ast hma. I t should be avoided in pat ient s w it h serious
conduct ion def ect s, in bronchospast ic disease, and w hen t here are evident risks
of proarrhyt hmia.
Suggested Readings
Miller MR, McNamara RL, Segal JB, et al. Eff icacy of agent s f or
pharmacologic conversion of at rial f ibrillat ion and subsequent maint enance of
sinus rhyt hm: a met a-analysis of clinical t rials. J Fam Pract 2000; 49: 1033–
1046.
O pie LH, G ersh BJ, eds. Drugs f or t he Heart . 6t h Ed. Philadelphia: Elsevier;
2005: 218–274.
Tikosyn. Pf izer. ht t p: / / w w w. t ikosyn. com

> Table of C ontents > Medic ations > 16 - B e Aw ar e that Fur os em ide C ontains a S ulfa Moiety
16
Be Aware that Furosemide Contains a Sulfa
Moiety
Ardalan Minokadeh BS
Furosemide (Lasix) is a member of t he general class of l oop di ureti cs t hat inhibit
t he Na+ / K + / 2Cl - cot ransport carrier w it hin t he ascending limb of t he loop of Henle.
The decrease in sodium and chloride reabsorpt ion seems t o come f rom t he f act
t hat f urosemide compet es f or t he chloride posit ion on t he carrier. Alt hough all
loop diuret ics have t he same primary mechanism of act ion, f urosemide has been
show n t o addit ionally inhibit sodium and chloride reabsorpt ion in t he dist al renal
t ubule. I t is import ant t o not e t he addit ional decrease in calcium reabsorpt ion
because calcium t ransport in t he loop of Henle is direct ly relat ed t o t he gradient
est ablished by t he movement of sodium and chloride. I t has a w ide range of
clinical uses including t reat ment of hypert ension, congest ive heart f ailure, and
hyperkalemia.
Signs and Symptoms

Clinically, a cont roversial issue in t he administ rat ion of f urosemide is t he
incidence of cross-react ivit y w it h sulf onamide-cont aining ant ibiot ics because of
t he sulf onamide moiet y t hat f urosemide cont ains (Fig. 16. 1). The clinical issue
cent ers on t he administ rat ion of f urosemide t o a pat ient w it h document ed sulf a
allergy. O pinions diff er on t he believed hypersensit ivit y cross-react ion t hat may
manif est it self as anaphylaxis, dermat it is, urt icaria, eczema, St evens-Johnson
syndrome, rash, or f ever w it hin 30 days, but current lit erat ure st at es t hat t he
lack of published clinical evidence f or t he allergic react ions is not ew ort hy. The
aut hors of a recent ly published review of t he lit erat ure on t his t opic st at e t hat
much of t he evidence used t o provide support f or t he t heory is in f act not
conclusive and does not implicat e appropriat e causat ion. Det ract ors of t he belief
of a legit imat e cross-react ivit y also cit e t hat sulf onamide nonant ibiot ics lack t he
aromat ic amine group at t he N4 posit ion of sulf onamide-cont aining ant ibiot ics,
believed t o be t he cause of sulf onamide ant ibiot ic hypersensit ivit y.
A ret rospect ive st udy w as perf ormed t o det ermine if f urosemide produced
allergic cross-react ions in pat ient s w ho report ed sulf a allergies. The aut hors did
not f ind suff icient evidence t o implicat e a report ed sulf a allergy in producing
severe react ions. I t is also t he opinion
of some aut hors t hat t he risk of react ions is so low t hat if alt ernat ive nonsulf a
t reat ment s are unavailable, administ rat ion of a drug such as f urosemide w it h
monit oring is w arrant ed. Physicians w ho w ish t o provide an alt ernat e drug t o
pat ient s about w hom t hey may be concerned could consider et hacrynic acid, a
loop diuret ic t hat does not cont ain a sulf a moiet y.
FIG URE 16. 1. Chemical St ruct ure
Suggested Readings
Lee A, Anderson R, Kardon R, et al. Presumed “sulf a allergy” in pat ient s w it h
int racranial hypert ension t reat ed w it h acet azolamide or f urosemide: cross
react ivit y, myt h or realit y? Am J O pht hal 2004; 138: 114–118.
Johnson K, G reen D, Rif e J, et al. Sulf onamide cross-react ivit y: f act or

f ict ion? Ann Pharmacot her 2005; 39: 290–301.
Sullivan TJ. Cross-react ions among f urosemide, hydrochlorot hiazide, and

sulf onamides [ Let t er] . JAMA 1991; 265: 120–121.
> Table of C ontents > Medic ations > 17 - C ar efully O bs er ve the C linic al R es pons e to Inter m ittent
Fur os em ide D os ing w hen D ec iding on Additional D os es
17
Carefully Observe the Clinical Response to
Intermittent Furosemide Dosing when Deciding
on Additional Doses
Melissa S. Camp MD
Furosemide is a loop diuret ic t hat act s by inhibit ing t he reabsorpt ion of sodium
and chloride (via t he Na/ K/ 2Cl cot ransport er) in t he t hick ascending limb of t he
loop of Henle. I t enhances t he excret ion of sodium, pot assium, calcium, chloride,
and w at er. Furosemide is used f requent ly in t he int ensive care unit (I CU) set t ing
f or diuresis.
I nt ravenous f urosemide begins t o w ork in 5 minut es, peaks at 30 minut es, and
last s f or about 2 hours. The eliminat ion half -lif e is approximat ely 30 t o 120
minut es. A dosing int erval of 6 hours allow s f or f our t o f ive half -lives of
eliminat ion. Typically, w it h t he appropriat e dose of I V f urosemide, a maximal
response w ill be seen w it hin t he f irst hour and t he increased urine out put w ill
cont inue in a t apering f ashion f or up t o 6 hours. A large init ial response f ollow ed
by a rapid t apering of urine out put may be an indicat ion t hat t he pat ient is not
quit e ready f or diuresis.
The dose of f urosemide is t it rat ed t o eff ect , but t he dose depends on renal
f unct ion. Higher doses are needed w it h w orsening renal f unct ion. Furosemide
act s in t he loop of Henle, so t he eff ect seen af t er a dose of f urosemide depends
on t he rat e at w hich blood is f ilt ered t hrough t he glomerulus. As creat inine
clearance decreases (implying low er glomerular f ilt rat ion rat e [ G FR] and w orse
renal f unct ion), a higher dose of f urosemide is needed t o achieve a t imely
response. I f t he dose of f urosemide administ ered is t oo low f or a given
creat inine clearance, t hen a delayed diuret ic response can be seen.
Watch Out For

The main t oxicit ies of f urosemide include ot ot oxicit y, elect rolyt e abnormalit ies,
and allergic react ions. O t ot oxicit y is more common w it h very high doses of
f urosemide and t he mechanism is t hought t o be due t o t he presence of a similar
Na/ K/ 2Cl cot ransport er in t he inner ear. Elect rolyt e abnormalit ies f rom t he
diuresis induced by f urosemide include hypokalemia, hyponat remia, and
met abolic alkalosis. O verly aggressive diuresis can lead t o elevat ed blood urea
nit rogen (BUN) and creat inine as w ell as hypot ension. The presence of allergic
react ions t o
f urosemide do occur; how ever, many clinicians doubt t he exist ence of a t rue
sulf a allergic react ion. Purport ed allergic react ions can be as mild as a rash or
as serious as anaphylaxis or St evens-Johnson syndrome.
O ne f inal not e is t hat w hen t hinking about t he clinical eff ect s of f urosemide, t he
ast ut e clinician w ill remember t hat t he brand name w as chosen in part based on
t he durat ion of t he medicat ion last ing six hours; hence, t he name La-si x
Suggested Readings
Micromedex. Mart indaleâ T he Complet e Drug Ref erence: Furosemide. USP DI
Drug I nf ormat ion f or t he Healt h Care Prof essional: Diuret ics, Loop
ht t p: / / w w w. t homsonhc. com/ hcs/ librarian.
Rose B. O pt imal dosage and side eff ect s of loop diuret ics. UpToDat e. 2005.
ht t p: / / w w w. upt odat e. com/ physicians/ pulmonology t oclist . asp
> Table of C ontents > Medic ations > 18 - D o not Adm inis ter Methylene B lue if ther e is P os s ible
Gas tr ointes tinal Abs or ption
18
Do not Administer Methylene Blue if there is
Possible Gastrointestinal Absorption
Beverly J. Newhouse MD
Met hylene blue is a w at er-soluble blue t hiazine dye used most commonly as a
t reat ment f or met hemoglobinemia or as an indicat or dye. I t s ut ilit y as an
indicat or dye has been applied t o several clinical sit uat ions including
ident if icat ion of aspirat ion or placement of nasogast ric t ubes in crit ically ill
pat ient s; localizat ion of parat hyroid adenomas; t est ing of t he int egrit y of t he
biliary syst em during hepat ic surgery; and t est ing of t he int egrit y of uret eral or
bladder anast omoses. Addit ional applicat ions of met hylene blue include it s use
as a urinary ant isept ic, reversal f or chemot herapy-induced encephalopat hy, a
t opical agent t o phot oinact ivat e viruses, an experiment al cyt ot oxic agent f or
t umor cells, and a vasopressor in pat ient s w it h sept ic or anaphylact ic shock or
f or pat ient s undergoing t he reperf usion phase of liver t ransplant at ion.
Treat ment of met hemoglobinemia is t he most common clinical use of met hylene
blue. Met hemoglobin is t he oxidized f orm of hemoglobin, creat ed w hen t he iron
moiet y changes f rom Fe2+ t o Fe3+ . This f orm of hemoglobin cannot bind O2 or
CO 2 and t heref ore loses it s oxygen-carrying capacit y and t ransport f unct ion,
predisposing t he pat ient t o hypoxemia. Drug-induced causes of
met hemoglobinemia include local anest het ics (e. g. , prilocaine and benzocaine)
and nit rat es (e. g. , nit roglycerin and nit roprusside). Met hylene blue, administ ered
slow ly at a dose of 1 t o 2 mg/ kg int ravenously (I V) w it h a maximum dose of 7
mg/ kg, is convert ed in vivo t o leukomet hylene blue, w hich reduces
met hemoglobin back t o hemoglobin. Nicot inamide adenine dinucleot ide phosphat e
(NADPH) is essent ial in t he conversion t o leukomet hylene blue and t hus, I V
met hylene blue administ rat ion is cont raindicat ed in t hose w it h low endogenous
NADPH (i. e. , glucose-6-phosphat e dehydrogenase [ G 6PD] def iciency). Wit hout
t his NADPH-dependent conversion, t he use of met hylene blue can lead t o
hemolyt ic anemia and exacerbat ion of met hemoglobinemia. Similarly, if t oo high a
dose of met hylene blue is given or it is given t oo quickly, high concent rat ions can
accumulat e and sat urat e t he reducing pat hw ay such t hat met hylene blue w ill act
as an oxidizing agent and paradoxically creat e more met hemoglobin.
Signs and Symptoms

Alt hough t he risks of I V administ rat ion of met hylene blue w ere w ell know n, t he
use of t his subst ance ent erally w as t hought t o be relat ively benign. How ever, t he
U. S. Food and Drug Administ rat ion (FDA) has recent ly become concerned about
ent eral exposure t o met hylene blue and has issued a w arning against using it in
t his sit uat ion. This is based, in part , on t he report ed cases of t oxicit y (including
deat h) associat ed w it h ent eral exposure of a very similar dye, FD&C Blue No. 1
(Blue 1). I n most of t hese cases, t he pat ient s had a hist ory of sepsis. The FDA
concluded t hat pat ient s at risk of increased int est inal permeabilit y (e. g. , pat ient s
w it h cancer, sepsis, burns, renal f ailure, celiac sprue, or inf lammat ory bow el
disease) w ere at risk f or mit ochondrial t oxicit y. Signs and sympt oms of
mit ochondrial t oxicit y include nausea, vomit ing, abdominal pain, f ever, sudden
w eight loss, dizziness, headache, precordial pain, t achycardia, hypert ension,
dyspnea, prof use sw eat ing, rest lessness, t remors, and conf usion. Unt reat ed,
mit ochondrial t oxicit y can result in myopat hy, neuropat hy, lipoat rophy, hemolyt ic
anemia, hypoxemia, arrhyt hmias, hepat ic st eat osis, liver f ailure, lact ic acidosis,
and ref ract ory hypot ension.
Suggested Readings
Clif t on J, Leikin JB. Met hylene blue. Amer J Ther 2003; 10: 289–291.
FDA Public Healt h Advisory: Report s of blue discolorat ion and deat h in
pat ient s receiving ent eral f eedings t int ed w it h t he dye, FD&C Blue No. 1. U. S.
Food and Drug Administ rat ion; 2003. ht t p: / / w w w. cf san. f da. gov/ ~dms/ col-
lt r2. ht ml
Nguyen ST, Cabrales RE, Bashour CA, et al. Benzocaine-induced

met hemoglobinemia. Anest h Analg 2000; 90(2): 369–371.
Reed DN, Vit ale G C, Wright son WR, et al. Decreasing mort alit y of bile leaks
af t er elect ive hepat ic surgery. Am J Surg 2003; 185(4): 316–318.
> Table of C ontents > Medic ations > 19 - B e Aler t for Metabolic Ac idos is in P atients on Lor az epam
D r ips
19
Be Alert for Metabolic Acidosis in Patients on
Lorazepam Drips
Yasir M. Akmal MD
Lorazepam (At ivan) is a f requent ly used sedat ive in t he int ensive care unit (I CU).
I t is a benzodiazepine t hat can be administ ered ent erally, as int ermit t ent
parent eral doses, or as a cont inuous inf usion. The benef icial eff ect s include
ant erograde amnesia and an opioid-sparing eff ect via a moderat ion of t he
ant icipat ory pain response. How ever, it is not a benign drug and side eff ect s
include respirat ory depression, t olerance, w it hdraw al, paradoxical agit at ion, and
prolonged eliminat ion.
Signs and Symptoms

O ne of t he carrier molecules of lorazepam is propylene glycol. This has been
implicat ed as t he cause of hyperosmolar met abolic acidosis in I CU pat ient s on
cont inuous inf usions of lorazepam. About 12% t o 45% of propylene glycol is
excret ed unchanged in t he urine, w it h t he remainder met abolized by t he liver.
Theref ore, caut ion should be exert ed w hen administ ering lorazepam in pat ient s
w it h renal or hepat ic dysf unct ion. Toxicit y f rom propylene glycol has also been
seen w it h administ rat ion of ot her drugs t hat use it as a carrier or solvent such as
phenyt oin, et omidat e, nit roglycerin, and diazepam (Table 19. 1). O t her eff ect s of
propylene glycol include renal dysf unct ion, int ravascular hemolysis, cardiac
arrhyt hmias, seizures, and cent ral nervous syst em (CNS) depression.
Fort unat ely, t hese met abolic derangement s appear t o correct af t er discont inuing
t he inf usion of lorazepam.
I t is import ant t o not e t hat t he upper limit of lorazepam dosing is 0. 1 mg/ kg/ hr,
w hich is about 7 t o 10 mg/ hr as an inf usion f or a 70- t o 100-kg pat ient . Each
millilit er of lorazepam cont ains about 0. 8 mL (830 mg) of propylene glycol, w hich
has a recommended daily allow ance of 25 mg/ kg/ day. The dose of propylene
glycol required f or t oxicit y has not been w ell est ablished and accumulat ion has
been described over a w ide range of lorazepam doses. Also, t oxicit y has been
seen w it h serum propylene glycol concent rat ions of 12 t o 130 mg/ dL and inf usion
periods ranging f rom 2 t o 24 days. The main f act ors t hat appear t o be relat ed t o
t oxicit y are rat e of inf usion, lengt h of inf usion, and degree of osmolar gap (Table
19. 2).
TABLE 19-1 PROPYLENE GLYCOL CONTENT IN

SELECTED DRUGS
AMOUNT OF PROPYLENE
DRUG
GLYCOL (% V/V)
Lorazepam, 2 mg/mL 80
Phenobarbital, 30-130
67.8–75
mg/mL
Diazepam, 5 mg/mL 40
Phenytoin, 50 mg/mL 40
Trimethoprim-
sulfamethoxazole, 16:80 40
mg/mL
Etomidate, 2 mg/mL 35
Nitroglycerin, 5 mg/mL 30
Modified from Arroliga AC, Shehab N, McCarthy K, et

al. Relationship of continuous infusion lorazepam to
serum propylene glycol concentration in critically ill
adults. Crit Care Med. 2004; 32(8):1709–1714. Table 1.
TABLE 19-2 CALCULATION OF OSM OL GAP
Osmol gap = (measured osmolality) - (calculated

osmolality)
Calculated osmolality = (12 × sodium) + (glucose/18) +

(blood urea nitrogen/2.8)
TABLE 19-3 DOSING AND HALF-LIFE OF ACTION O

SELECTED SEDATIVES
ONSET
HALF-LIFE INFU
AFT ER INT ERMIT T ENT
DRUG OF PARENT DOS
IV IV DOSE
COMPOUND RAN
DOSE
5–20 0.02–0.06 0.01

Lorazepam 8–15 hr
min mg/kg q 2-6 hr mg/
2–5 0.03–0.1 mg/kg

Diazepam 20–120 hr —
min q 0.5–6 hr
0.02–0.08
2–5 0.04
Midazolam 3–11 hr mg/kg q 0.5–2
min g/kg
hr
1–2 5–8
Propofol 26–32 hr —
min µg/k
3–20 0.03–0.15 0.04
Haloperidol 18–54 hr mg/kg q 0.5–6 0.15
min
hr mg/
Modified from Jacobi J, Fraser GL, Coursin DB, et al. Clinic

practice guidelines for the sustained use of sedatives and
analgesics in the critically ill adult. Crit Care Med.
2002;30(1):119–141. Table 3.
Alt ernat ive sedat ive drugs t o lorazepam include midazolam, propof ol, and
haloperidol. The half -lif e and dosing regimens of some of t hese drugs are show n
in Table 19. 3. Because of t he side eff ect s and of t en unpredict able w ake-up t ime
associat ed w it h benzodiazepines and ot her sedat ives, t he agent used should be
t ailored t o each individual pat ient .
Suggested Readings
Arroliga AC, Shehab N, McCart hy K, et al. Relat ionship of cont inuous inf usion
lorazepam t o serum propylene glycol concent rat ion in crit ically ill adult s. Crit
Care Med 2004; 32(8): 1709–1714.
Jacobi J, Fraser G L, Coursin DB, et al. Clinical pract ice guidelines f or t he

sust ained use of sedat ives and analgesics in t he crit ically ill adult . Crit Care
Med 2002; 30: 119–141.
Wilson KC, Reardon C, Theodore AC, et al. Propylene glycol t oxicit y: a

severe iat rogenic illness in I CU pat ient s receiving I V benzodiazepines: a case
series and prospect ive, observat ional pilot st udy. Chest 2005; 128: 1674–1681.
> Table of C ontents > Medic ations > 20 - B e Aler t for the D evelopm ent of C yanide Toxic ity w hen
Adm inis ter ing Nitr opr us s ide.
20
Be Alert for the Development of Cyanide Toxicity
when Administering Nitroprusside.
E. Dave Bravos MD
Hypert ension is commonly seen in t he perioperat ive set t ing as w ell as t he
int ensive care unit (I CU). There are many agent s t o low er blood pressure w it h
diff erent mechanisms of act ion. O ne of t he most pot ent vasodilat ors is sodium
nit roprusside, w hich causes art eriolar and venous smoot h muscle relaxat ion via
nit ric oxide-mediat ed mechanisms. I t has a rapid onset and relat ively short
durat ion of act ion, making it a readily t it rat able agent . How ever, it s dose must
be limit ed because of it s pot ent ial t oxic side eff ect s.
Toxicit y f rom sodium nit roprusside is due t o t he cyanide groups released f rom
met abolism of t he nit roprusside molecule. Af t er gaining an elect ron f rom t he iron
moiet y of hemoglobin, t he sodium nit roprusside produces an unst able radical and
met hemoglobin. The unst able nit roprusside radical produces f ive cyanide ions,
w hich can have one of t hree f at es. They can int eract w it h met hemoglobin t o
produce cyanomet hemoglobin. They can produce t hiosulf at e and it s end product
t hiocyanat e. Addit ionally, cyanide ions can bind cyt ochrome oxidase and
ult imat ely inhibit oxidat ive met abolism, leading t o cyanide t oxicit y. O rgans most
suscept ible t o t he eff ect s of loss of oxidat ive met abolism are t he heart and
brain.
Clinically, pat ient s w it h cyanide t oxicit y may exhibit alt ered ment al st at us,
cardiovascular inst abilit y, and an anion gap met abolic acidosis. I nit ially, pat ient s
may present w it h sinus t achycardia t hat may progress t o sinus bradycardia or
vent ricular dysrhyt hmias and even asyst ole. Pat ient s begin t o have rest lessness
and agit at ion w hen t he cent ral nervous syst em is aff ect ed. Wit h w orsening
t oxicit y, convulsions may occur and can ult imat ely lead t o encephalopat hy and
coma. Wit h loss of aerobic met abolism t here is an increase in lact at e, leading t o
an anion gap met abolic acidosis.
What to Do
Treat ment of cyanide t oxicit y begins w it h discont inuat ion of nit roprusside
administ rat ion. Pat ient s should concomit ant ly be placed on supplement al oxygen
and may even require mechanical vent ilat ion. Addit ionally, t hiosulf at e and sodium
nit rat e (t o convert hemoglobin t o met hemoglobin) can be given
t o increase t he kinet ics of t he ot her t w o met abolic pat hw ays, result ing in less
cyanide t o bind t o cyt ochrome oxidase. Prevent ion of cyanide t oxicit y can be
achieved by limit ing t he dose of nit roprusside t o less t han 8. 0 micrograms/ kg/ min
over 1 t o 3 hours. Addit ionally, t hiosulf at e can be given concomit ant ly w it h t he
nit roprusside.
I t is import ant t o not e t hat nit roprusside has mult iple eff ect s on organ syst ems.
G iven t hat nit roprusside relaxes art eriolar venous smoot h muscle, t here is a
decrease in af t erload as w ell as preload. This can ref lexively cause an increase
in heart rat e and cont ract ilit y. How ever, cardiac out put is usually not aff ect ed
given t he decrease in preload as w ell. An int racoronary st eal syndrome may
occur in areas w here t here is coronary art ery disease and inabilit y t o f urt her
dilat e t hese diseased vessels compared w it h nondiseased vessels. Eff ect s on
t he cerebral vasculat ure can cause an increase in cerebral blood f low and
ult imat ely in int racranial pressure. Dilat at ion of t he pulmonary vasculat ure also
occurs. This can increase shunt f ract ion by prevent ing hypoxic pulmonary
vasoconst rict ion. Renal act ivat ion of t he reninangiot ensin syst em and
cat echolamine release can occur w it h a decrease in art erial blood pressure,
w hich can cause rebound hypert ension w it h discont inuat ion of nit roprusside.
Finally, alt ernat e agent s t o nit roprusside include ot her vasodilat ors such as
nit roglycerin and hydralazine. Addit ionally, ot her classes of agent such as ACE
inhibit ors, alpha-blockers, or calcium channel blockers may be used t o cont rol
hypert ension.
Suggested Readings
Morgan G E. Clinical Anest hesiology. 3rd Ed. New York: Lange/ McG raw -Hill;
2002: 225–227.
St oelt ing RK, ed. Pharmacology and Physiology in Anest het ic Pract ice. 4t h
Ed. Philadelphia: Lippincot t -Raven; 2006: 321–337.
> Table of C ontents > Medic ations > 21 - Adm inis ter B eta- B loc kade Initially B efor e Adm inis ter ing
“ O ther ” Antihyper tens ives W hen Tr eating Aor tic D is s ec tion
21
Administer Beta-Blockade Initially Before
Administering “Other” Antihypertensives When
Treating Aortic Dissection
Eric S. Weiss MD
Signs and Symptoms

Case
A 44-year-old man w hose only past medical hist ory consist s of hypert ension and
unilat eral cat aract s present s t o his local emergency depart ment w it h t he acut e
onset of severe “t earing” chest pain radiat ing t o his back. O n exam t he physician
not es t hat t he pat ient is 6 f eet 8 inches t all w it h long, lanky f ingers. The pat ient
st at es t hat everyone in his f amily has been t all and lanky like him. His init ial
pulse rat e is 105 and his blood pressure reads 185/ 95. His elect rocardiogram
show s no ST-segment changes and init ial laborat ory t est s show no abnormalit ies
w it h a blood urea nit rogen level of 11 mg/ dL and creat inine of 0. 8 mg/ dL. A
comput ed t omography (CT) w it h int ravenous cont rast show s a 6-cm dilat ed
descending t horacic aort ic aneurysm w it h a dissect ion f lap present . He is
immediat ely t ransport ed t o t he surgical int ensive care unit (SI CU) f or
management . I n t he SI CU, t he on-call resident orders hydralazine f or blood
pressure cont rol. His heart rat e climbs t o 126 and his syst olic blood pressure
drops t o 145 mm Hg. O ver t he next 6 hours, despit e t hese maneuvers, his pain
does not improve and he becomes progressively t achycardic and hypot ensive,
init ially responding t o f luid but t hen requiring vasopressor support . Repeat
laborat ory analysis show s a serum lact at e of 5. 0 and a rising serum creat ine t o
1. 8 mg/ dL. During t his t ime he st ops making urine and his abdomen becomes
t ender and progressively dist ended. He also develops a met abolic acidosis w it h
a pH of 7. 20, requiring int ubat ion and mechanical vent ilat ion f or his increased
w ork of breat hing. I mmediat e abdominal CT show s propagat ion of t he dissect ion
t o involve t he abdominal aort a w it h visceral vessel involvement and pneumat osis
of t he small int est ine. He is immediat ely t aken t o t he operat ing room but at t his
point requires massive vasopressor support and is prof oundly acidot ic. Direct
visualizat ion inside t he abdomen show s diff usely necrot ic bow el. Approximat ely
45 minut es int o t he operat ion
t he pat ient suff ers a cardiac arrest and despit e cardiopulmonary resuscit at ion
and administ rat ion of code drugs, t he pat ient expires on t he operat ing room
t able.
Discussion
The preceding case illust rat es t he necessit y of t reat ing acut e aort ic dissect ions
w it h bet a-blockade concomit ant ly w it h vasodilat ory ant ihypert ensives. The
pat ient described had t he inherit ed connect ive t issue disorder Marf an syndrome
w it h t he classic present at ion of an acut e aort ic dissect ion heralded by t he acut e
onset of chest or back pain. The pat ient received adequat e ant ihypert ensive
t herapy but cont inued t o det eriorat e because of t he f ailure t o init iat e bet a-
blockade.
To brief ly review, aort ic dissect ions can be classif ied according t o t he St anf ord
classif icat ion as t ype A (t hose involving t he ascending aort a) and t ype B (t hose
involving only t he descending aort a).
Elect ive operat ive repair is pref erable as mort alit y rat es f rom urgent operat ive
procedures f or aort ic dissect ion range bet w een 2% and 21%. I ndicat ions f or
operat ive repair include t ype A dissect ions w it h or w it hout ulcers and t ype B
dissect ions w it h visceral involvement or large, open f ree-f low ing f alse lumens.
When urgent surgery is not indicat ed, management consist s of pain cont rol and
blood pressure low ering t o a t arget of 110 mm Hg syst olic. This is usually
accomplished w it h sodium nit roprusside, but bet a-blockade is i mperati ve
because t he pulsat ile f orce must also be minimized w it h appropriat e bet a-
blockade so as t o not propagat e t he dissect ion, and bet a-blockers decrease w all
st ress. Vasodilat ors by t hemselves can lead t o ref lex t achycardia, w hich can
increase shear st ress and propagat e t he int imal t ear.
Suggested Readings
Mukherjee D, Eagle KA. Aort ic dissect ion: an updat e. Curr Probl Cardiol
2005; 30: 287–325.
> Table of C ontents > Medic ations > 22 - D o not us e Vas opr es s in in P atients w ith Hear t Failur e or
Mes enter ic Is c hem ia
22
Do not use Vasopressin in Patients with Heart
Failure or Mesenteric Ischemia
Vasopressin is t he exogenous, parent eral f orm of ant idiuret ic hormone (ADH)
t hat may be used f or several condit ions in t he int ensive care unit (I CU), including
t he t reat ment of cent ral diabet es insipidus and as a vasopressor agent in shock.
Endogenous vasopressin is an import ant st ress hormone and plays a crit ical role
in maint aining volume hemost asis. How ever, it s benef icial physiological eff ect s
may be off set w hen t here is excess hormone in t he circulat ion. I t is import ant t o
underst and vasopressin's mechanism of act ion and t o be able t o recognize w hich
I CU pat ient s may suff er adverse eff ect s f rom t he use of t his drug.
Endogenous vasopressin is synt hesized in neurosecret ory cells of t he
hypot halamus and excret ed by t he post erior pit uit ary gland in response t o
decreases in blood volume and increases in serum osmolalit y. Recept ors f or
vasopressin are primarily locat ed in vascular smoot h muscle (V1 ) and t he renal
collect ing duct (V2 ) and act t o st imulat e vasoconst rict ion and increase f ree w at er
reabsorpt ion, respect ively. These are normal compensat ory mechanisms
designed t o maint ain vascular t one and circulat ory homeost asis during periods of
low cardiac out put .
Watch Out For

I n pat ient s w it h congest ive heart f ailure, how ever, chronic act ivat ion of t his
neurohormonal axis cont ribut es t o t he progression of disease t hrough an
increase in af t erload and w at er ret ent ion. Moreover, giving vasopressin inf usions
t o pat ient s w it h exist ing heart f ailure w ill cause an acut e increase in syst emic
vascular resist ance and pulmonary capillary w edge pressure, possibly result ing
in a f urt her decrease in cardiac out put . Theref ore, caut ion should be used w hen
administ ering exogenous vasopressin in pat ient s w it h know n heart f ailure in t he
I CU.
Anot her adverse eff ect of excessive vasopressin in t he circulat ion is t he pot ent ial
f or mesent eric hypoperf usion. When vasopressin st imulat es t he cont ract ion of
vascular smoot h muscle, t his eff ect is of t en most prominent in t he capillaries and
small art erioles or venules of t he splanchnic circulat ion. This vasoconst rict ion
can cause a major redist ribut ion of blood f low aw ay f rom t he bow el mucosa t hat
can result in
ischemia and produce it s associat ed signs, including hyperlact at emia. Along w it h

t he concurrent decrease in cardiac out put and global oxygen delivery, high doses
of exogenous vasopressin may produce or exacerbat e mesent eric ischemia and
has been show n t o increase gut pCO 2 in t onomet ry st udies.
I n order t o avoid det riment al reduct ions in mesent eric blood f low associat ed w it h
vasopressin, ot her vasopressors should be considered f irst -line agent s in shock.
This includes norepinephrine, epinephrine, and dopamine, w hich do not
compromise regional blood f low or t issue oxygenat ion at normal levels.
Vasopressin should be used only as a supplement ary vasopressor t o t hese
agent s at a cont inuous low -dose inf usion (≤0. 04 U/ min) and should not be
t it rat ed as a single agent .
O ne f inal not e is t hat vasopressin analogue desmopressin (DDAVP) is t he
t reat ment of choice f or cent ral diabet es insipidus, given t hat it has a longer
durat ion of act ion and causes less st imulat ion of smoot h muscle t han
vasopressin.
Suggested Readings
Chat t erjee K. Neurohormonal act ivat ion in congest ive heart f ailure and t he
role of vasopressin. Am J Cardiol 2005; 95: 8B-13B.
Holmes CL, Walley KR. Vasopressin in t he I CU. Curr O pin Crit Care
2004; 10: 442–448.
> Table of C ontents > Medic ations > 23 - R em em ber that C ar diac P r es s or s do not W or k in a Low - pH
E nvir onm ent
23
Remember that Cardiac Pressors do not Work in
a Low-pH Environment
Crit ically ill pat ient s of t en require inot ropic and/ or pressor support t o maint ain
adequat e cardiac out put and adequat e blood pressure t o sust ain end-organ
perf usion. Because end-organ perf usion has already likely been compromised
and may cont inue t o be problemat ic despit e use of t hese agent s, anaerobic
met abolism rat her t han aerobic met abolism is likely t o be generat ing a limit ed
amount adenosine t riphosphat e (ATP) in t he hypoperf used t issues. The
consequence is lact ic acid product ion and acidosis. Addit ionally, crit ically ill
pat ient s may have ot her causes of acidosis cont ribut ing t o t he overall acidot ic
st at e including renal f ailure, hyperchloremia, or ket oacidosis. The acidosis may
be severe w it h pH values w ell below 7. 0.
Binding of t he inot ropic or pressor agent s t o t heir recept ors is inf luenced by pH,
along w it h ot her f act ors such as t emperat ure and concent rat ion. Presumably, t he
great er t he deviat ion in eit her direct ion f rom t he opt imal pH f or t he drug-ligand
int eract ion, t he less binding t hat w ill occur and hence, t he less t he eff ect of t he
drug. This has led t o t he w idely held opinion t hat inot ropes and vasopressors
don't w ork at t he acidic pH values of t en encount ered in crit ically ill pat ient s. The
act ual relat ionship is much more complex since t he t arget of t he inot ropes and
vasopressors, t he alpha and bet a adrenergic recept ors, includes several
subt ypes w hose individual responsiveness t o t hese agent s is quit e variable under
acidic condit ions.
Watch Out For

The variabilit y in responsiveness st ems f rom not only changes in aff init y f or
binding t o t he recept ors but also because acidic condit ions have been show n t o
change recept or numbers on cell surf aces as w ell as alt er t he dow nst ream
regulat ion mediat ed by G -coupled prot eins. Some recept ors are upregulat ed
w hile t heir binding aff init y may drop. O t hers exhibit no change in t heir aff init y or
overall responsiveness t o inot ropes and vasopressors. St ill ot hers increase t heir
responsiveness. Diff erent blood vessels in diff erent t issue beds also are highly
variable in t heir responsiveness since t he predominant populat ion of t he recept or
subt ype present varies w it h t he sit e and even caliber of t he vessel. The degree
of pH change inf luences t hese result s dramat ically. Mild
acidemia act ually st imulat es t he sympat het ic nervous syst em out put , increasing
vent ricular f unct ion and vasomot or t one. As t he acidemia becomes more severe,
changes in ligand binding and pharmacological eff ect of circulat ing
cat echolamines may become more prominent . The f inal clinical eff ect w ill be a
balance bet w een t he t w o compet ing phenomena. O verall, pH values as low as
7. 15 do not have an appreciable clinical eff ect on t he act ivit y of t hese drugs.
Below t his value, how ever, reduct ions in overall eff ect iveness may become
clinically apparent .
I t is of t en not ed t hat w hen bicarbonat e is given, blood pressure improves and
vasopressors can be t it rat ed. O ne explanat ion f or t his benef it is t hat an ampule
of sodium bicarbonat e act s as a hypert onic f luid bolus expanding plasma volume.
Addit ionally, t he change in pH alone can have eff ect s on vasomot or t one
separat e f rom any eff ect s on drug binding and t his may be prominent w hen a
cert ain pH t hreshold is crossed. Treat ment of an acidot ic pH w it h bicarbonat e,
w it h t he goal of improving inot rope and vasopressor eff ect iveness, should be
reserved f or w hen t he pH is below 7. 1. O t her reasons f or w ant ing t o correct t he
acidosis w hen t he pH is above t his value may exist (such as f acilit at ing w eaning
f rom a mechanical vent ilat or) but signif icant improvement in t he eff ect iveness of
inot ropes and vasopressors w it h bicarbonat e administ rat ion should not be
expect ed at less acidot ic pH values. As alw ays, simply relying on bicarbonat e
inf usion t o correct pH should not alt er an aggressive search f or and t reat ment of
t he underlying cause of t he acidosis.
Suggested Readings
Cooper DJ, Walley KR, Wiggs BR, et al. Bicarbonat e does not improve
hemodynamics in crit ically ill pat ient s w ho have lact ic acidosis: a prospect ive
cont rolled clinical st udy. Ann I nt ern Med 1990; 112: 492.
Hall JB, Schmidt G A, Wood LD. Principles of Crit ical Care Medicine. 3rd Ed.
New York: McG raw -Hill; 2005: 1954â 1 965.
> Table of C ontents > Medic ations > 24 - C ons ider Thr om bolytic s in S T- E levation Myoc ar dial Infar c tion
if P er c utaneous C or onar y Inter vention is not Available or D elayed
24
Consider Thrombolytics in ST-Elevation
Myocardial Infarction if Percutaneous Coronary
Intervention is not Available or Delayed
Case
Mr. AMI is a 66-year-old male w ho is post operat ive day number one af t er hip
replacement surgery and complains of acut e onset of chest pain and short ness
of breat h. An elect rocardiogram (ECG ) is perf ormed t hat show s new ST-segment
elevat ion in leads I I , I I I , and aVF. You are alone in Communit y G eneral Hospit al,
w hich does not have a percut aneous coronary int ervent ion (PCI ) t eam. As you
begin t reat ment , you recall t hat G ood Heart Universit y Hospit al has a PCI t eam
and t hat door-t o-cat het erizat ion-t able t ime f rom your hospit al t o t heirs is 90
minut es. You administ er oxygen, nit roglycerin, morphine, aspirin, met oprolol, and
heparin and cont emplat e your next st ep.
Discussion
ST-elevat ion myocardial inf arct ion (STEMI ) is a myocardial inf arct ion usually
associat ed w it h acut e plaque rupt ure and occlusion of a coronary art ery. I n
addit ion t o t he t herapies described earlier, pat ient s w it h STEMI require
immediat e reperf usion t herapy. This can be achieved w it h eit her t hrombolyt ic
t herapy or percut aneous coronary int ervent ion. Upon ident if icat ion of STEMI , t he
pract it ioner should consider t he f ollow ing.
General Guidelines
1. G enerally, t hrombolyt ics are more eff ect ive if given earlier. I f present at ion is
w it hin 3 hours f rom sympt om onset , t here is no pref erence bet w een
t hrombolyt ic or percut aneous coronary int ervent ion t herapy. Af t er 3 hours
PCI (w hen available) is pref erred over t hrombolyt ics.
2. G enerally, w hen t he risk of mort alit y is high, PCI is pref erred over
t hrombolyt ics.
3. The higher t he pat ient 's risk of bleeding w it h t hrombolyt ic t herapy, t he more
st rongly t he decision should f avor PCI .
4. When t he diff erences in t ime f or door t o balloon and door t o needle is less
t han 1 hour, PCI is pref erred.
Indications for Thrombolytic Therapy (All Required)

1. Sympt oms of myocardial ischemia
2. ST elevat ion great er t han 0. 1 mV in at least t w o cont iguous leads or new lef t
bundle branch block on present ing ECG
3. O nset of sympt oms w it hin 12 hours
4. Present at ion in a f acilit y w it hout capabilit y of percut aneous coronary
int ervent ion w it hin 90 minut es or prolonged t ransport (>1 hour) t o such a
f acilit y
Contraindications to Thrombolytics
1. Any prior int racranial hemorrhage
2. Know n cerebral vascular lesion
3. Know n malignant int racranial neoplasm
4. I schemic st roke w it hin 3 mont hs
5. Suspect ed aort ic dissect ion
6. Act ive bleeding or bleeding diat hesis (except menses)
7. Signif icant closed head or f acial injury w it hin 3 mont hs
Relative Contraindications to Thrombolytics

1. Hist ory of chronic, severe, or poorly cont rolled hypert ension
2. Severe uncont rolled hypert ension on present at ion (syst olic blood pressure
>180 mm Hg or diast olic blood pressure >110 mm Hg)
3. Hist ory of prior ischemic st roke great er t han 3 mont hs, dement ia, or know n
int racranial pat hology not covered in cont raindicat ions
4. Recent t raumat ic or prolonged cardiopulmonary resuscit at ion or major
surgery (<3 w eeks)
5. Recent (2 t o 4 w eeks) int ernal bleeding
6. Noncompressible vascular punct ures
7. For st rept okinase or anist replase: prior exposure (>5 days ago) or prior
allergic react ion
8. Pregnancy
9. Act ive pept ic ulcer
10. Current use of ant icoagulant s: t he higher t he int ernat ional normalized rat io
(I NR), t he higher t he risk of bleeding
Choice of Agents
Thrombolyt ic agent s are all plasminogen act ivat ors. They f unct ion enzymat ically
t o produce t he act ive compound plasmin f rom plasminogen. Plasmin f unct ions in
t he breakdow n of t hrombus. Three agent s are approved f or t reat ment of STEMI .
Bot h ret eplase and t enect eplase have t he advant age of being bolus
administ ered.
1. Alt epase: 100 mg over 90 minut es
2. Ret eplase: 10 unit s × 2, each over 2 minut es

3. Tenect eplase: 30 t o 50 mg (depending on w eight ) bolus
Af t er administ rat ion of t hrombolyt ic t herapy, t he pat t ern of ST elevat ion should
be monit ored over t he next 60 t o 90 minut es. I f signs and sympt oms of
reperf usion do not occur (relief of sympt oms, maint enance or rest orat ion of
hemodynamic st abilit y, elect rical st abilit y, reduct ion in init ial ST-segment
elevat ion), rescue PCI should be considered.
Bef ore hospit al discharge, pat ient s w ho have had a STEMI and t hrombolyt ic
t herapy require f urt her risk st rat if icat ion. This includes st ruct ural evaluat ion
echocardiography (ECG ) and f unct ional evaluat ion (st ress t est ing) t o det ermine
t he need f or cat het erizat ion and revascularizat ion.
Suggested Readings
Ant man EM, Anbe DT, Armst rong PW, et al. ACC/ AHA guidelines f or t he
management of pat ient s w it h ST-elevat ion myocardial inf arct ion: execut ive
summary. A report of t he American College of Cardiology/ American Heart
Associat ion Task Force on Pract ice G uidelines (Writ ing Commit t ee t o revise
t he 1999 guidelines f or t he management of pat ient s w it h acut e myocardial
inf arct ion). J Am Coll Cardiol 2004; 44: 671–719.
Tint inalli J, Kelen G , St apczynski S, eds. Emergency Medicine: A

Comprehensive St udy G uide. 5t h Ed. New York: McG raw -Hill; 2000: 366–374.
> Table of C ontents > Medic ations > 25 - C ons ider the Us e of Glyc opr otein IIb/IIIa Inhibitor s in Uns table
C or onar y S yndr om es
25
Consider the Use of Glycoprotein IIb/IIIa
Inhibitors in Unstable Coronary Syndromes
Current t reat ment s f or acut e coronary syndromes including unst able angina and
non-ST-elevat ion myocardial inf arct ion now include agent s aimed at inhibit ing
plat elet act ivat ion, adhesion, and aggregat ion w it h f ibrinogen cross-linking. Tw o
classes of drugs exist t hat inhibit t hese plat elet eff ect s at t he level of t he
glycoprot ein (G P) I I b/ I I I a recept or on t he plat elet surf ace membrane. The f irst
group are t he t hienopyridines (clopidogrel and t iclopidine). These drugs block G P
I I b/ I I I a complex act ivat ion by inhibit ing adenosine diphosphat e (ADP) binding t o
t he plat elet recept or. These oral agent s have a slow er onset and longer-t erm
t reat ment durat ion. The CURE, TARG ET, CREDO , and STAI G t rials have
demonst rat ed a 21% t o 46% improvement in primary cardiac end point s
at t ribut ed t o t he addit ion of a t hienopyridine. Signif icant benef it s w ere
demonst rat ed w it h t he addit ion of clopidogrel t o aspirin among pat ient s
undergoing percut aneous coronary int ervent ions (PCI ) (4. 5 % versus 6. 4%
cardiovascular deat h, myocardial inf arct ion, or urgent revascularizat ion).
The second group of drugs are t he direct G P I I b/ I I I a ant ibodies or recept or
ant agonist s (abciximab, t irof iban, and ept if ibat ide). These drugs inhibit t he f inal
pat hw ay of plat elet aggregat ion via f ibrinogen cross-linking. Abciximab is a
monoclonal ant ibody w it h a high aff init y f or bot h act ivat ed and rest ing plat elet s,
w hereas ept if ibat ide and t irof iban are nonant ibody recept or inhibit ors w it h less
aff init y f or rest ing plat elet s and much f ast er dissociat ion rat es (short er half -
lives) t han abciximab. Met a-analysis of several large clinical t rials (PRI SM,
PURSUI T, PARAG O N, CAPTURE, and G USTO -I V) has suggest ed only a modest
12% overall survival benef it at t ribut able t o t he use of direct G P I I b/ I I I a inhibit ors
w hen added t o ot her t herapeut ics in acut e coronary syndromes. Several
subpopulat ions of high-risk pat ient s, how ever, may have signif icant ly improved
out come due t o t hese agent s. These include pat ient s receiving emergent
percut aneous int ervent ions (34% reduct ion in primary end point s versus 7%
w it hout G P I I b/ I I I a), pat ient s w it h raised t roponin levels (58% reduced mort alit y
versus 5% increase), and diabet ic pat ient s, in w hom t he great est increase in
survival w as demonst rat ed. O f int erest t o physicians is
t hat w omen w ho receive G P I I b/ I I I a inhibit ors appear t o have a 14% increase in

combined end point s as compared w it h men.
TABLE 25-1 DRUG DOSING
PHARMACOKINET ICS SIDE

DRUG DOSE
AND DYNAMICS EFFEC
T HIENOPYRIDINES
9% vs.
major a
Max antiplt effect minor
after 600 mg load 300 mg to bleedin
Clopidogrel within 3 hours. 600 mg po sig ↑
(Plavix) Extensive hepatic load, 75 bleedin
metabolism, renal mg/day. after C
excretion, 8 hr t1/2 Stop 5
prior to
CABG.
2%
24- to 48-hour onset, neutrop
Ticlopidine 250
24 hr t1/2, extensive in first
(Ticlid) mg/BID po.
hepatic metabolism. months
rx.
GP IIB/IIIA INHIBITORS
0.15 to 0.3
Twofold
30 min t1/2, 10-day mg/kg IV increas
platelet bound state, bolus, then bleedin
Abciximab 80% plts inhibited in 10 ↓Plt .14
(ReoPro) 2 hrs. Sustained mcg/min to inciden
inhibition after 12 to maintain ↓BP wi
24 hr of infusion. ACT 300- infusion
500s.
180 5% to 1
mcg/kg IV major
15 min t1/2, 75%
Eptifibatide bolus, then bleedin
renal clearance, 25%
(Integrilin) 1 (same
hepatic clearance.
mcg/kg/min control
infusion. 3% ↓Pl
2% ma
0.4
bleedin
mcg/kg/min
2 hour t1/2, 65% possibl
IV bolus,
Tirofiban renal clearance, 25% reversi
then 0.1
(Aggrastat) reduced clearance in ↓Plt. R
mcg/kg/min
elderly. insuf
infusion for
prolong
24 hr.
effect.
Concom
150 to 600
heparin
85 min t1/2, 50% mcg IV
infusion
Lamifiban return of plt function bolus, 1 to
signific
(phase III) 6 hr after terminating 5 mcg/min
↑s blee
infusion. infusion for
from 0.
120 hr.
to 3%.
What to Do
The American College of Cardiology and American Heart Associat ion guidelines
and t he Sevent h American College of Chest Physicians (ACCP) Consensus
Conf erence on ant it hrombot ic t herapies suggest t hat t he t hienopyridines are as
eff ect ive as aspirin alone and recommend clopidogrel as an alt ernat ive in
pat ient s w it h aspirin insensit ivit y because of it s low er side eff ect s. I t is
pref erable t o delay giving clopidogrel t o pat ient s w ho may undergo coronary
art ery bypass graf t surgery (CABG ) because of excessive post operat ive
bleeding. G P I I b/ I I I a inhibit ors are of most benef it in pat ient s undergoing or
likely t o undergo PCI and may be of lit t le benef it in pat ient s w ho do not receive
PCI . Ept if ibat ide is used more of t en because it cost s less and t here is more
experience w it h it s dosing (Table 25. 1).
Suggested Readings
Braunw ald E, Ant man EM, Beasley JW, et al. ACC/ AHA guidelines f or t he
management of pat ient s w it h unst able angina and non-ST-segment elevat ion
myocardial inf arct ion. A report of t he American College of
Cardiology/ American Heart Associat ion Task Force on Pract ice G uidelines
(Commit t ee on t he Management of Pat ient s w it h Unst able Angina). J Am Coll
Cardiol 2000; 36: 970.
Moust apha A, Anderson V. Cont emporary view of t he acut e coronary

syndromes. J I nvasive Cardiol 2003; 15: 71–79.
Simons M. Ant iplat elet agent s in unst able angina and acut e non-ST elevat ion
(non-Q w ave) MI . I n: Rose BD, ed. UpToDat e. Walt ham, MA: UpToDat e; 2006.
> Table of C ontents > Medic ations > 26 - S tr ongly C ons ider Low - Molec ular - W eight Hepar in in the
Tr eatm ent of Uns table Angina and Non- S T- E levation Myoc ar dial Infar c tion
26
Strongly Consider Low-Molecular-Weight
Heparin in the Treatment of Unstable Angina and
Non-ST-Elevation Myocardial Infarction
Case
Mrs. ACS is a 74-year-old f emale w ho is admit t ed t o t he int ensive care unit
(I CU) f or an acut e coronary syndrome. She present s t o t he emergency
depart ment w it h subst ernal chest pain w it h radiat ion t o t he neck, diaphoresis,
and dizziness. Her elect rocardiogram (ECG ) show s T-w ave f lat t ening in t he
inf erior leads (I I , I I I , and aVF). No ST-segment elevat ions are not ed. The pat ient
is st art ed on aspirin, nit roglycerin, and met oprolol. As you w rit e f or a heparin
drip, t he nurse asks, “Why don't w e do Lovenox [ enoxaparin] , t hat w ay I don't
have t o st art a drip. ”
Discussion
Acute coronary syndrome is a relat ively new classif icat ion t erm f or a set of
acut e coronary diseases. Causes of acut e coronary syndrome include ST-
elevat ion myocardial inf arct ion (STEMI ), non-ST-elevat ion myocardial inf arct ion
(NSTEMI ), and unst able angina. STEMI is def ined as myocardial inf arct ion
involving ST-segment elevat ion of 0. 1 mV in t w o or more consecut ive leads or a
new lef t bundle branch block. NSTEMI ref ers t o acut e myocardial damage f rom
at herosclerot ic disease t hat result s in release of cardiac biomarkers w it hout
causing ST-segment elevat ion. Unstabl e angi na ref ers t o myocardial damage
t hat does not result in release of cardiac biomarkers or ST-segment elevat ion.
These def init ions are import ant because t hey relat e t o t he pat hogenesis,
severit y of disease, and management st rat egy of t he aff ect ed pat ient s.
I n t he t reat ment of acut e coronary syndrome, ant icoagulant t herapy is indicat ed
t o minimize f urt her t hrombus f ormat ion. Heparin is t he recommended parent eral
ant icoagulant in t reat ment of acut e coronary syndrome. Heparin accelerat es t he
act ion of ant it hrombin. This leads t o inact ivat ion of f act ors I I a (t hrombin), I Xa,
and Xa. This leads t o decrease t hrombus f ormat ion. Unf ract ionat ed heparin is a
mixt ure
of diff erent w eight s of heparin (5 t o 30k) t hat have varying eff ect s on f act ors I I a,
I Xa, and Xa. Low -molecular-w eight heparin (LMWH) is a subset of heparin
(w eight ~5k) t hat has act ivit y on f act or Xa, but not on I I a. LMWH has t he
advant age of less binding t o prot ein and dose-independent clearance. This leads
t o more predict able dosing and a longer half -lif e (3 t o 6 hours). Because of t he
longer half -lif e, LMWH can be administ ered subcut aneously t w ice a day as
opposed t o a cont inuous inf usion f or unf ract ionat ed heparin (half -lif e 1 t o 2
hours).
I n pat ient s w it h unst able angina or NSTEMI , enoxaparin (Lovenox) is pref erable
t o unf ract ionat ed heparin in pat ient s not scheduled f or a coronary art ery bypass
graf t w it hin 24 hours. Four t rials have evaluat ed t he benef it s of low -molecular-
w eight heparin over unf ract ionat ed heparin in t he t reat ment of unst able angina or
NSTEMI . The t w o t rials t hat used enoxaparin each show ed a reduct ion in t he
composit e end point of deat h, myocardial inf arct ion, and recurrent ischemia
w hen compared w it h unf ract ionat ed heparin. The TI MI 11B t rial show ed a 12%
relat ive risk reduct ion (19. 6% unf ract ionat ed heparin vs. 17. 3% LMWH). The
ESSENCE t rial show ed a 19% relat ive risk reduct ion (23. 3% unf ract ionat ed
heparin vs. 19. 8% LMWH). The t w o t rials t hat used anot her f orm of LMWH
(dalt eparin and nadroparin) demonst rat ed a t rend t ow ard increased risk of t he
composit e end point , alt hough t his w as not st at ist ically signif icant .
The advant ages of LMWH are t he ease of administ rat ion, absence of need f or
monit oring, and low er associat ion w it h heparin-induced t hrombocyt openia. I n all
t rials, t here w as an increased risk of minor bleeding, but no diff erence in risk of
major bleeding. Unf ract ionat ed heparin is pref erred in pat ient s w ho have a
planned percut aneous coronary int ervent ion and coronary art ery bypass graf t
because LMWH is less eff ect ively reversed w it h prot amine and does not allow
monit oring t o adjust t he level of ant icoagulat ion.
Suggested Readings
Ant man EM, McCabe CH, G urf inkle EP, et al. Enoxaparin prevent s deat h and
cardiac ischemic event s in unst able angina/ non-Q -w ave myocardial inf arct ion;
result s of t he t hrombolysis inmyocardial inf arct ion (TI MI ) 11B t rial. Circulat ion
1999; 100: 1593–1601.
Braunw ald E, Ant man EM, Beasley JW, et al. ACC/ AHA guideline updat e f or
t he management of pat ient s w it h unst able angina and non-ST-segment
elevat ion myocardial inf arct ion: 2002. Summary art icle f rom t he Commit t ee on
t he Management of Pat ient s w it h Unst able Angina. Circulat ion
2002; 106: 1893–1900.
Cohen M, Demers C, G urf inkle EP, et al. A comparison of low -molecular-
w eight heparin w it h unf ract ionat ed heparin f or unst able coronary art ery
disease. Eff icacy and Saf et y of Subcut aneous Enoxaparin in Non-Q -Wave
Coronary Event s St udy G roup. N Engl J Med 1997; 337: 447–452.
Comparison of t w o t reat ment durat ions (6 days and 14 days) of a low

molecular w eight heparin w it h a 6-day t reat ment of unf ract ionat ed heparin in
t he init ial management of unst able angina or non-Q w ave myocardial
inf arct ion: FRAX. I . S. Eur Heart J 1999; 20: 1553–1562.
Klein W, Buchw ald A, Hillis SE, et al. Comparison of low -molecular-w eight
heparin w it h unf ract ionat ed heparin acut ely and w it h placebo f or 6w eeks in
t he management of unst able coronary art ery disease. Fragmin in Unst able
Coronary Art ery Disease St udy (FRI C). Circulat ion 1997; 96: 61–68.
> Table of C ontents > Medic ations > 27 - S tr ongly C ons ider Us ing Glyc opr otein IIb/IIIa Inhibitor s as an
Added Tr eatm ent to S tenting in Ac ute Myoc ar dial Infar c tion
27
Strongly Consider Using Glycoprotein IIb/IIIa
Inhibitors as an Added Treatment to Stenting in
Acute Myocardial Infarction
Nirav G . Shah MD
Percut aneous coronary int ervent ion (PCI ) w it h st ent ing has revolut ionized t he
care f or pat ient s w it h acut e myocardial inf arct ions. I nt racoronary st ent
placement reduces t he long-t ermrecurrence of st enosis. I n addit ion, t he
combinat ion of st ent ing w it h administ rat ion of glycoprot ein I I b/ I I I a (G P I I b/ I I I a)
inhibit or agent s, w hich prohibit t he binding of f ibrinogen and prot eins t o plat elet s,
has reduced t he f requency of ischemic complicat ions f ollow ing percut aneous
coronary int ervent ions.
Watch Out For

G P I I b/ I I I a t herapy is indicat ed in pat ient s undergoing coronary angioplast y,
inpart icular t hose pat ient s w it h unst able angina or w it h ot her clinical
charact erist ics of high risk. Cont raindicat ions t o G P I I b/ I I I a inhibit ors include
act ive bleeding, a bleeding diat hesis in t he past 30 days, int racranial t umors,
int racranial hemorrhage, art eriovenous malf ormat ion, recent st roke, major
surgery or t raumain t he preceding mont h, t hrombocyt openia, severe
hypert ension, and aort ic dissect ion. The agent s in t his class include ept if ibat ide,
t irof iban, and abciximab. Ept if ibat ide and t irof iban are pref erred in pat ient s w it h
unst able angina and non-ST-elevat ion myocardial inf arct ion (NSTEMI ) managed
medically, w hile abciximab is pref erred in t hose undergoing PCI . The dose of
abciximab as an adjunct in PCI is 0. 125 mcg/ kg/ minut e and w hen used f or
unst able angina t he dose is 10 mcg/ minut e f or 18 t o 24 hours. Tirof iban is dosed
at 0. 1 mcg/ kg/ minut e and should be cont inued f or 12 t o 24 hours af t er PCI .
Ept if ibat ide is dosed at 2 mcg/ kg/ minut e af t er a 180-mcg/ kg bolus f or acut e
coronary syndromes. The side eff ect s of t hese medicat ions include int racranial
hemorrhage, gast roint est inal bleeding, t hrombocyt openia, hemat uria,
ret roperit oneal bleeding, and ot her severe bleeding diat heses.
I t is import ant t o not e t hat st ent t echnology has also improved over t he past f ew
years. The init ial t echnology of st ent s has now been replaced by st ent s w it h
bet t er clinical out comes. The various t ypes of st ent s include bare met al, self -
expandable, balloon expandable,
covered, small vessel, and drug elut ing. Current pract ice uses most ly drug-
elut ing st ent s, w hich are coat ed w it h an ant ist enot ic drug t hat is released over
14 t o 30 days af t er implant at ion. The t w o t hat are approved by t he Food and
Drug Administ rat ion are sirolimus- and paclit axel-coat ed st ent s. Drug-elut ing
st ent s should be used pref erent ially in lesions involving t he lef t -ant erior
descending art ery and f or lef t main disease.
Suggested Readings
Fischman DL, Leon MB, Baim DS, et al. A randomized comparison of
coronary-st ent placement and balloon angioplast y in t he t reat ment of coronary
art ery disease. N Engl J Med 1994; 331: 496–501.
Schomig A, Neumann FJ, Kast rat i A, et al. A randomized comparison of

ant iplat elet and ant icoagulant t herapy af t er t he placement of coronary-art ery
st ent s. N Engl J Med 1996; 334: 1084–1089.
> Table of C ontents > Medic ations > 28 - C ons ider Nes ir itide in Ac utely D ec om pens ated Hear t Failur e
28
Consider Nesiritide in Acutely Decompensated
Heart Failure
Jacqueline Janka MD
I n pat ient s w it h cardiogenic pulmonary edema due t o acut e decompensat ed heart
f ailure, plasma brain nat riuret ic pept ide (BNP) measurement s can guide t he
diagnosis, dist inguishing t his et iology of dyspnea f rom pulmonary causes.
Plasma BNP concent rat ion is elevat ed in bot h asympt omat ic and sympt omat ic
heart f ailure (lef t vent ricular dysf unct ion) and is usef ul f or bot h diagnosis and
prognosis.
BNP is a hormone abundant ly f ound in t he heart (especially t he vent ricles) but
init ially ident if ied in t he brain. I t is released f rom myocardial cells in response t o
w all st ress, volume expansion, and high f illing pressures. Physiologically,
endogenously produced BNP t arget s t he heart , blood vessels, and kidneys t o
reduce preload and af t erload t hrough vasodilat at ion, diuresis, and nat riuresis
(sodium excret ion). BNP also count ers neurohormones such as endot helin,
aldost erone, and angiot ensin I I . Murine st udies f urt her suggest t hat BNP may
prot ect against t he progressive cardiac f ibrosis associat ed w it h chronic heart
f ailure. I n w hole, BNP reduces t he w orkload on t he heart t o improve cardiac
perf ormance.
A rapid BNP f luorescence immunoassay w as approved in 2000 and can be
rapidly perf ormed in 10 t o 15 minut es at t he bedside. Pat ient s w it h heart f ailure
have signif icant ly higher levels of plasma BNP t han t hose w it h dyspnea due t o
ot her causes. Remarkably, plasma BNP is more accurat e f or predict ing heart
f ailure t han classic paramet ers such as rales, cardiomegaly, or t he Nat ional
Healt h and Nut rit ion Examinat ion Survey (NHANES) or Framingham crit eria. BNP
also correlat es w it h t he New York Heart Associat ion f unct ional class. A value
>100 pg/ mL makes t he diagnosis of heart f ailure w it h a sensit ivit y of 90%,
specif icit y of 76%, and a predict ive value of 83%. Conversely, low BNP has a
st rong negat ive predict ive value t o rule out heart f ailure. The rapid assay cost s
approximat ely $20 per t est .
Despit e having elevat ed levels of cardioprot ect ive BNP, pat ient s w it h heart
f ailure are generally paradoxically vasoconst rict ed and sodium avid. O f int erest
t o t he clinicians is t hat , w hile appearing minimally responsive t o endogenous
BNP, heart f ailure pat ient s w ill eff ect ively respond t o exogenously administ ered
BNP (nesirit ide).
What to Do
Nesirit ide is a recombinant f orm of human B-t ype nat riuret ic pept ide (hBNP),
approved in 2001 f or t he int ravenous (I V) t reat ment of pat ient s w it h acut ely
decompensat ed congest ive heart f ailure. Nesirit ide is given as an init ial I V bolus
of 2 mcg/ kg, f ollow ed by a cont inuous inf usion at 0. 01 mcg/ kg/ min. The dose
may be increased t o a maximum rat e of 0. 03 mcg/ kg/ min. There is no dose
adjust ment f or renal insuff iciency. Nesirit ide promot es sodium excret ion, reduced
cardiac f illing pressures (reduced pulmonary capillary w edge pressures), and
improved cardiac indices and clinical st at us (e. g. , dyspnea, f at igue) in pat ient s
w it h heart f ailure. These posit ive hemodynamic eff ect s are seen w it hin t he f irst
hour of administ rat ion and are maint ained t hroughout t he inf usion period and up
t o 6 t o 24 hours af t er discont inuat ion.
Watch Out For

The most common side eff ect of nesirit ide is dose-relat ed hypot ension, w hich
may be enhanced w hen given w it h ot her vasodilat ors, such as angiot ensin-
convert ing enzyme (ACE) inhibit ors. While nesirit ide appears f airly saf e and is
w ell t olerat ed, t here are dat a t hat have prompt ed quest ions about it s eff ect s on
renal f unct ion and survival. For example, one analysis suggest ed t hat pat ient s
t reat ed w it h nesirit ide may have been more likely t o develop progressive renal
insuff iciency, alt hough t his may be due t o hypoperf usion rat her t han direct renal
t oxicit y. Anot her ret rospect ive analysis suggest ed a t rend (nonsignif icant ) t ow ard
a higher 30-day mort alit y in pat ient s t reat ed w it h nesirit ide.
I n summary, nesirit ide is an eff ect ive t herapy f or pat ient s hospit alized w it h acut e
decompensat ed heart f ailure, especially t hose requiring inot ropic support . I t may
be part icularly usef ul in heart f ailure pat ient s w it h t achycardia or arrhyt hmias,
w hich w ould ot herw ise limit t he use of t radit ional agent s such as dopamine or
dobut amine. Nesirit ide may be used in combinat ion w it h ot her heart f ailure
medicat ions (e. g. , diuret ics, ACE inhibit ors, dopamine, dobut amine, or milrinone)
w hen t olerat ed. Close monit oring of hemodynamics, urine out put , and renal
f unct ion is essent ial.
Suggested Readings
Colucci WS, Elkayam U, Hort on DP, et al. I nt ravenous nesirit ide, a nat riuret ic
pept ide, in t he t reat ment of decompensat ed congest ive heart f ailure.
Nesirit ide St udy G roup. N Engl J Med 2000; 343: 246.
Mills RM, Hobbs RE. How t o use nesirit ide in t reat ing decompensat ed heart
f ailure. Clev Clin J Med 2002; 69: 252.
Nesirit ide f or decompensat ed congest ive heart f ailure. Med Let t Drugs Ther
2001; 43: 100.
> Table of C ontents > Medic ations > 29 - B e Aggr es s ive in C ons ider ing R eper fus ion Ther apy in Ac ute
Myoc ar dial Infar c tion
29
Be Aggressive in Considering Reperfusion
Therapy in Acute Myocardial Infarction
Cardiovascular disease is a dominant cause of morbidit y and mort alit y in t he
Unit ed St at es and acut e myocardial inf arct ion (MI ) is one of t he major underlying
et iologies. Acut e MI is a medical emergency t hat is precipit at ed w hen coronary
occlusion leads t o ischemia and t hen necrosis of cardiac myocyt es. The necrosis
w ill of t en precipit at e a cardiac arrhyt hmia (e. g. , vent ricular f ibrillat ion), w hich is
a major cause of deat h f or pat ient s present ing in t he acut e phase. The diagnosis
of an acut e MI is based on t he pat ient 's hist ory and diagnost ic t est ing, including
an elect rocardiogram (ECG ) and serum enzymes. The f ast er t he diagnosis can
be made and t he occlusion reversed, t he more likely t he pat ient is t o salvage
myocardial t issue and benef it f rom improved out comes.
Watch Out For

I n nondiabet ic pat ient s, t he major sympt om of acut e MI is pain. This pain is
usually described as severe. I t is locat ed in a ret rost ernal area and may radiat e
t o eit her t he arms or neck. The charact erist ic qualit y of t he pain is described as
pressurelike or bandlike; how ever, ot her descript ions including burning, aching,
and crushing have also been used. The pain usually last s beyond 20 t o 30
minut es and does not dissipat e. Associat ed sympt oms include nausea, vomit ing,
short ness of breat h, dizziness, and diaphoresis.
I n t he set t ing of a possible acut e MI , t he ECG is a readily available, noninvasive,
easy-t o-obt ain diagnost ic t est t hat has excellent sensit ivit y and specif icit y. The
ECG provides inf ormat ion on t he dist ribut ion of changes and t he impact on
cardiac rhyt hm. The pat t ern of ST-segment elevat ion represent ing a “current of
injury” is usually associat ed w it h acut e MI and implies a coronary occlusion. This
is helpf ul f or t riaging f or st rat egies of reperf usion.
I n addit ion t o t he ECG , cardiac enzymes are part icularly sensit ive f or det ect ing
myocardial damage. These prot eins are released f rom t he myocardial cells t hat
have sust ained damage. Troponins are part icularly import ant f or t he diagnosis of
acut e MI since under normal condit ions, t hey are not f ound in t he serum. These
enzymes f irst
become det ect able f rom 2 t o 4 hours af t er injury and persist f or days. Creat ine
kinase (CK) and t he MB subt ype have f or many years dominat ed t he clinical
arena as t he major cardiac marker in t he set t ing of acut e MI now being
supplant ed by t roponins. These t est s are import ant early markers of myocardial
necrosis but t ake longer t o be det ect ed and disappear f rom t he serum more
rapidly t han t roponins. I t must be not ed t hat in post operat ive pat ient s, creat ine
phosphokinase (CPK) and t roponin have not been show n t o be as sensit ive
markers f or cardiac damage as t hey are in nonsurgical pat ient s.
What to Do
The management of pat ient s present ing w it h chest pain suggest ive of acut e MI
should t rigger a number of immediat e event s. The longest delay in t herapy is
usually relat ed t o t he pat ient 's delay in seeking care. Denial of t he pain and t he
hope t hat it w ill go aw ay are import ant underlying reasons f or t his phenomenon.
When t he pat ient present s t o t he emergency depart ment w it h chest pain, a
f ocused hist ory, physical examinat ion, and elect rocardiogram should be
obt ained. I f t he appropriat e clinical sympt oms and ECG f indings suggest an
acut e MI , aggressive management should include int ravenous access, oxygen,
sublingual nit roglycerin, morphine, and bet a-blocker t herapy. I f t he pat ient
present s w it hin 12 hours of sympt oms, aggressive met hods t o achieve
reperf usion should be considered.
The recommended st rat egy f or achieving reperf usion is cardiac cat het erizat ion
w it h angioplast y if t he inst it ut ion is capable of achieving a door-t o-balloon t ime
of <90 minut es. This st rat egy begins w it h t he administ rat ion of glycoprot ein
I I b/ I I I a inhibit ors in addit ion t o t he more generic t herapies described previously.
This st rat egy has t he advant age of achieving higher recanalizat ion rat es,
reducing residual st enosis, and improving t he out comes in t he set t ing of
cardiogenic shock. When an inst it ut ion is unable t o provide primary angioplast y
st rat egies, t hrombolyt ic t herapy can be at t empt ed. A number of cont raindicat ions
t o t hrombolyt ic t herapy need t o be considered bef ore providing t his int ervent ion
and include bleeding, a hist ory of st roke, severe hypert ension, recent t rauma or
surgery, pregnancy, aort ic dissect ion, and int racranial or spinal cord neoplasms.
Suggested Readings
G oldman L, Ausiello D, eds. Cecil Textbook of Internal Medi ci ne. 22d ed.
Philadelphia: WB Saunders; 2004: 410–424.
Keely EC, Boura JA, G rines CL. Primary angioplast y versus int ravenous
t hrombolyt ic t herapy f or acut e myocardial inf arct ion. A primary review of 23
randomized t rials. Lancet . 2003; 3612: 13–20.
> Table of C ontents > Medic ations > 30 - S tr ongly C ons ider S tar ting an Angiotens in- C onver ting
E nz ym e Inhibitor or Angiotens in R ec eptor B loc ker after Myoc ar dial Infar c tion
30
Strongly Consider Starting an Angiotensin-
Converting Enzyme Inhibitor or Angiotensin
Receptor Blocker after Myocardial Infarction
Ying Wei Lum MD
The renin-angiot ensin-aldost erone syst em (RAAS) regulat es aldost erone release
in t he body via t he act ion of angiot ensin I I on angiot ensin I recept ors.
Angiot ensin I I by it self is also a pot ent vasoconst rict or (even more t han
norepinephrine), leading t o hypert ension (Fig. 30. 1). I t s ot her harmf ul eff ect s
include t he promot ion of at herosclerosis and hypert rophic changes in t he blood
vessels and lef t vent ricle. Angiot ensin-convert ing enzyme (ACE) inhibit ors, by
blocking t he conversion of angiot ensin I t o angiot ensin I I , are believed t o
at t enuat e vent ricular dilat at ion and remodeling. Furt hermore, ACE inhibit ors
block t he degradat ion of bradykinin, w hich is t hought t o have ant ihypert ensive
eff ect s. The accumulat ion of bradykinin is believed t o be t he cause of t he chronic
cough associat ed w it h use of ACE inhibit ors. Angiot ensin I I recept or blockers
(ARBs), on t he ot her hand, w ork f urt her dow nst ream in t he RAAS by blocking
angiot ensin I I at t he recept or sit e.
I n general, pat ient s init iat ed on ACE inhibit ors during myocardial inf arct ion (MI )
have demonst rat ed a reduct ion in overall mort alit y by about 20% t o 30%. This
benef it is most prof ound in pat ient s w it h sympt omat ic congest ive heart f ailure
and t heir t herapy should be cont inued indef init ely. How ever, t herapy in pat ient s
w it hout evidence of lef t vent ricular dysf unct ion may be ceased af t er 4 t o 6
w eeks. These pat ient s should t hen be re-evaluat ed in 4 t o 6 mont hs t o reassess
vent ricular f unct ion. Pat ient s w ho are int olerant t o ACE inhibit ors (because of
severe cough, et c. ) may t ake an ARB as a suit able alt ernat ive. Valsart an w as
recent ly compared w it h capt opril and f ound t o be as eff ect ive at reducing
mort alit y, heart f ailure hospit alizat ions, and reinf arct ion.
O f int erest t o clinicians is t hat pat ient s t aking chronic ACE inhibit or t herapy have
demonst rat ed circulat ing angiot ensin I I levels similar t o t hat prior t o init iat ing
t herapy. This “ACE escape” syndrome is incomplet ely underst ood but t hought t o
be associat ed w it h ot her synt het ic pat hw ays f or angiot ensin I I . Hence, maximal
blockade of t he RAAS w ould encompass blocking bot h t he f ormat ion of
angiot ensin I I and act ion of angiot ensin I I on t he angiot ensin I recept or.
I n congest ive heart f ailure pat ient s, t his combinat ion t herapy has been show n t o
decrease t he incidence of hospit alizat ions f or heart f ailure. How ever,
combinat ion t herapy has yet t o show any survival benef it in pat ient s during
myocardial inf arct ion, w it h or w it hout heart f ailure.
FIG URE 30. 1. Mechanism of Act ion
What to Do
Barring hypot ension (syst olic blood pressure <100 mm Hg), t reat ment w it h an
ACE or ARB should be commenced on t he f irst post -MI day (as early as t he f irst
12 hours). Therapy should be init iat ed only af t er t he rout ine recommended
administ rat ion of t hrombolyt ic t herapy, aspirin, and bet a-blockers. Capt opril
off ers t he advant age of being t he drug w hose use w it hin t he f irst 24 hours has
been best st udied (SAVE and I SI S-4 t rial). How ever, various ACE inhibit ors
including lisinopril (G I SSI -3 t rial), ramipril (AI RE t rial), and t randolapril (TRACE
t rial) have also received FDA approval f or t reat ment of heart f ailure af t er
myocardial inf arct ion. Enalapril (CO NSENSUS-I I t rial) and zof enopril (SMI LE
t rial), t hough not current ly FDA approved f or t reat ment during myocardial
inf arct ion, are also f elt t o be equivalent in t heir benef icial eff ect on mort alit y.
Current ly, valsart an (VALI ANT t rial) is t he only ARB t hat has been FDA approved
f or use during myocardial inf arct ion.
Watch Out For

The most common cause f or discont inuing t herapy in t he acut e peri-MI period is
hypot ension. This is unusual, but it can occur in up t o 2% of heart f ailure pat ient s
w ho have undergone vigorous int ravascular deplet ion f rom diuret ic use.
I nt ravascular volume deplet ion can in t urn cause decreased renal perf usion and
azot emia. I n general, renal insuff iciency Cr >2. 5 mg/ dL is a cont raindicat ion in
init iat ing or cont inuing t herapy, as is renal art ery st enosis. Hyperkalemia (K >5. 5
mEq/ mL) can occur as a result of decreased aldost erone levels. O t her rare
adverse react ions include dry cough, rash, and angioedema.
Suggested Readings
Pf eff er MA. ACE inhibit ion in acut e myocardial inf arct ion. N Engl J Med
1995; 332: 118–120.
Voors AA, van Veldhuisen DJ. Role of angiot ensin recept or blockers in
pat ient s w it h lef t vent ricular dysf unct ion: lessons f rom CHARM and VALI ANT.
I nt J Cardiol 2004; 97; 3: 345–348.
> Table of C ontents > Medic ations > 31 - Us e P r ophylaxis for E r os ive Gas tr itis in the Appr opr iate
P atient
31
Use Prophylaxis for Erosive Gastritis in the
Appropriate Patient
G ast roint est inal (G I ) prophylaxis in t he int ensive care unit (I CU) is import ant in
t he prevent ion of st ress gast rit is. The incidence of clinically signif icant bleeding
in crit ically ill pat ient s is 2% t o 15%. How ever, it is import ant t o underst and t he
act ual indicat ions f or prophylaxis. Most experienced clinicians f eel it should be
limit ed t o pat ient s at high risk f or st ress-relat ed mucosal diseases, w hich include
t he f ollow ing condit ions: mechanical vent ilat ion longer t han 48 hours;
coagulopat hy; endoscopic or radiographic diagnosis of pept ic ulcer or gast rit is;
hist ory of an upper G I bleed less t han 6 w eeks prior t o admission; signif icant
burns (great er t han 15% of t ot al body surf ace area); t raumat ic brain injury; and
large doses of glucocort icoids (e. g. , >50mg hydrocort isone/ day).
Prophylact ic medicat ions t o prevent st ress gast rit is include ant acids, sucralf at e,
hist amine-2 recept or ant agonist s (H2-blockers), and prot on pump inhibit ors
(PPI s). Cook et al. , in a mult icent er randomized double-blind cont rolled t rial,
demonst rat ed t hat H2-blockers (ranit idine) compared w it h sucralf at e decreased
clinically signif icant bleeding w it h no diff erence in vent ilat or-associat ed
pneumonia. To dat e t here are no st udies t hat have prospect ively evaluat ed t he
abilit y of int ravenous PPI s t o reduce clinically signif icant bleeding in high-risk
crit ically ill pat ient s; how ever, PPI s have been show n t o raise and maint ain an
elevat ed gast ric pH. When deciding bet w een H2-blockers and PPI s, side eff ect s
and cost should be considered. I f t he pat ient requires int ravenous medicat ions,
H2-blockers are t he most cost -eff ect ive. How ever, if t he pat ient is able t o
t olerat e oral medicat ions, PPI s are a good choice given t heir abilit y t o maint ain
gast ric pH f or a sust ained period of t ime. The overall need f or prophylaxis should
be assessed w hen pat ient s are able t o meet t heir nut rit ional requirement s by
mout h.
Watch Out For

Side eff ect s of bot h drugs should be kept in mind. H2-blockers have been
associat ed w it h t achyphylaxis (i. e. , t olerance and t hus decrease in gast ric pH),
int erst it ial nephrit is, conf usion, inhibit ion of t he cyt ochrome oxidase enzyme
syst em (cimet idine only), and dose adjust ment in pat ient s
w it h creat inine clearance of less t han 50 mL/ min. PPI s are generally saf e but
rare side eff ect s include diarrhea, nausea, and prurit us.
Suggested Readings
Cook D. A comparison of sucralf at e and ranit idine f or t he prevent ion of upper
gast roint est inal bleeding in pat ient s requiring mechanical vent ilat ion. N Engl J
Med 1998; 338: 791–817.
Cook DJ, Fuller HD, G uyat t G H, et al. Risk f act ors f or gast roint est inal
bleeding in crit ically ill pat ient s. N Engl J Med 1994; 330: 377–381.
St einberg K. St ress relat ed mucosal disease in t he crit ically ill pat ient : Risk
f act ors and st rat egies t o prevent st ress-relat ed bleeding in t he int ensive care
unit . Crit Care Med 2002; 30: S362–S364.
> Table of C ontents > Medic ations > 32 - B ew ar e of Metabolites
32
Beware of Metabolites
Bradford D. Winters MD, PHD
Many drugs are met abolized t o act ive met abolit es. This is usually carried out by
t he liver in t he f orm of reduct ions, oxidat ions, and t he addit ion of met hyl, acet yl,
and ot her groups. While met abolit es are usually a f ract ion as pot ent as t he
parent compound, in some sit uat ions t hey may be more pot ent and may
accumulat e t o signif icant concent rat ions under t he right condit ions. O f t en t he
met abolit es of lipophilic drugs are made more w at er soluble t o f acilit at e renal
excret ion. How ever, int ensive care unit (I CU) pat ient s of t en have renal and/ or
hepat ic impairment alt ering t he clearance of t he drugs and t heir met abolit es.
Thus, w hile a part icular drug may be hepat ically met abolized, renal insuff iciency
may impair it s excret ion and lead t o buildup of an act ive met abolit e. O t her
met abolit es may not be act ive in t erms of t he original eff ect of t he drug but are
act ually more t oxic t han t he parent compound. Bot h of t hese sit uat ions may have
dire consequences.
Meperidine is a prime example of t his problem. This is a commonly used narcot ic
in anest hesiology and ot her set t ings including t he I CU and it s half -lif e is
approximat ely 3 hours. I t is hepat ically met abolized t o normeperidine (half -lif e of
15 t o 20 hours), w hich can build up in t he f ace of renal impairment and promot e
seizure act ivit y w hen it reaches high enough concent rat ions. These seizures are
not reversed by naloxone, w hich is usually used t o reverse t he delet erious
eff ect s of most narcot ics.
Morphine is anot her very commonly used narcot ic in t he I CU t hat is met abolized
t o t w o act ive met abolit es, 3- and 6-glucuronides. The half -lif e of morphine is
usually 2 t o 3 hours. The 3-glucuronide is not act ive as an analgesic but does
cause sedat ion and can precipit at e seizures. The 6-glucuronide is act ive as an
analgesic, in f act , more so t han morphine it self . Bot h of t hese met abolit es are
dependent on renal excret ion and may build up in crit ically ill pat ient s w it h renal
insuff iciency, leading t o prolonged sedat ion and/ or respirat ory depression or
possibly seizures like normeperidine.
Benzodiazepines are very commonly used in t he I CU f or sedat ion, t o assist w it h
mechanical vent ilat ion, and t o provide anxiolysis during painf ul procedures and
crit ical illness. Many of t he drugs in t his class are hydroxylat ed t o act ive
met abolit es, including midazolam
and diazepam. Midazolam is normally a short -act ing drug (durat ion of act ion is
approximat ely 1 hour), but t his is dependent on normal excret ion of t he
hydroxylat ed met abolit e. While t his met abolit e has only about 10% of t he act ivit y
of t he parent drug, renal impairment may lengt hen t he eff ect of t he drug. Thus a
short -act ing drug may become a longer-act ing drug especially w hen used in large
doses. A long-act ing benzodiazepine such as diazepam (durat ion of act ion is 3 t o
5 hours) can become very long act ing w hen it s act ive met abolit e is not cleared.
Diazepam has several act ive met abolit es, some of w hich are prepared and
market ed as separat e drugs, including t emazepam and oxazepam.
I t should be not ed t hat t he accumulat ion of act ive met abolit es secondary t o
hepat ic and/ or renal insuff iciency is of t en compounded by t he act of repet it ive
dosing and cont inuous inf usions of t he parent drugs. As a result a pat ient may
not aw aken f rom sedat ion f or prolonged periods of t ime, leading t o more t ime
spent on a vent ilat or and an unnecessary bat t ery of t est s being perf ormed t o
assess t he pat ient 's decreased ment al st at us. These problems are addit ionally
compounded by t he f act t hat many of t hese drugs are not used in isolat ion. Many
pat ient s are on several drugs at t he same t ime, especially t he benzodiazepines
and narcot ics. I f caref ul at t ent ion t o dosage is maint ained in t he f ace of
pot ent ially impaired clearance, t hese problems may be minimized. Bet t er yet ,
drugs t hat have no act ive met abolit es may be t he pref erred choice assuming all
ot her charact erist ics (side eff ect s, et c. ) are equal. I nst ead of using diazepam or
midazolam, one may choose lorazepam, w hich has no act ive or t oxic met abolit es.
While t his drug has a long durat ion of act ion (6 hours), paradoxically it s eff ect
may be short er in I CU pat ient s compared w it h t he ones w it h act ive met abolit es.
Similarly, Fent anyl may be used as t he pref erred narcot ic since it s met abolit es
are not t oxic or act ive.
Suggested Readings
Horn E, Nesbit SA. Pharmacology and pharmacokinet ics of sedat ives and
analgesics. G ast roint est Endosc Clin N Am 2004; 14: 247–268.
Shaf er A. Complicat ions of sedat ion w it h midazolam in t he int ensive care unit
and a comparison w it h ot her sedat ive regimens. Crit Care Med 1998; 26: 947–
956.
> Table of C ontents > Medic ations > 33 - D o not us e E r ythr om yc in as a P r okinetic Agent in P atients on
Tac r olim us ( O r C yc los por ine)
33
Do not use Erythromycin as a Prokinetic Agent
in Patients on Tacrolimus (Or Cyclosporine)
Matthew J. Weiss MD
Eryt hromycin is a macrolide ant ibiot ic t hat inhibit s bact erial prot ein synt hesis and
is eff ect ive against t he same organisms as penicillin G . The drug is commonly
used t o t reat communit y-acquired pulmonary inf ect ions in penicillin-allergic
pat ient s. Eryt hromycin also has prokinet ic eff ect s on t he gast roint est inal syst em.
At low doses, eryt hromycin induces act ivit y in t he gast ric ant rum, w hich
propagat es t o t he small int est ines. At higher doses, eryt hromycin induces a
prolonged period of ant ral act ivit y by st imulat ing mot ilin recept ors t o accelerat e
gast ric empt ying. Alt hough t ransplant recipient s f requent ly suff er f rom
opport unist ic bact erial inf ect ions and/ or gast roint est inal dysmot ilit y, eryt hromycin
should not be given t o recipient s t aking FK506 (t acrolimus, Prograf ) or
cyclosporine.
FK506 is st ruct urally similar t o eryt hromycin. I t is a macrolide t hat exhibit s
st rong immunosuppressive propert ies by inhibit ing t he t ranscript ion of int erleukin-
2 during T-cell act ivat ion. Like FK506, cyclosporine is a calcineurin inhibit or and
bot h drugs are rout inely given f or prophylaxis of organ reject ion. Bot h FK506 and
cyclosporine have narrow t herapeut ic w indow s and bot h require daily lab draw s
in t he perioperat ive period t o ensure eff icacious drug levels. Bot h drugs are
met abolized by t he liver's P450 CY P 3A4 enzyme pat hw ay (t he most common
pat hw ay f or drug met abolism). Eryt hromycin is a st rong inhibit or of t his pat hw ay
and concurrent administ rat ion of eryt hromycin and FK506 (in part icular) or
cyclosporin can result in elevat ed immunosuppression drug levels and severe
t oxicit y even af t er one or t w o doses of eryt hromycin. This t oxicit y can manif est
as severe renal dysf unct ion, t remors, seizures, brit t le diabet es, or corneal
ulcerat ions.
O t her common int ensive care unit medicat ions can also induce t he P450 CY P
3A4 syst em and t hus decrease serum levels of FK506 and cyclosporine. These
include phenobarbit al, carbamazepine, phenyt oin, and rif ampin. Drugs
met abolized t hrough t his pat hw ay should alw ays be double-checked f or drug
int eract ions.
Suggested Readings
Padhi D, Long P, Basha M, et al. I nt eract ion bet w een t acrolimus and
eryt hromycin. Ther Drug Monit 1997; 19: 120–122.
Shaeff er MS, Collier D, Sorrell MF. I nt eract ion bet w een FK506 and
eryt hromycin. Ann Pharmacot her 1994; 28: 280–281.
> Table of C ontents > Medic ations > 34 - C ons ider us e of E noxapar in over Unfr ac tionated Hepar in in
Tr aum a P atients
34
Consider use of Enoxaparin over Unfractionated
Heparin in Trauma Patients
Suneel Khetarpal MD
Barbara Haas MD
Trauma pat ient s are at high risk of deep venous t hromboembolism, w it h t he
majorit y of pat ient s exhibit ing no sympt oms. Alt hough met a-analyses report t he
overall incidence of deep vein t hrombosis (DVT) in t rauma pat ient s t o be
approximat ely 12%, t he incidence of DVT in pat ient s receiving no prophylaxis has
been report ed t o be as high as 58%. Alt hough a number of f act ors have been
purport ed t o increase t he incidence of DVT, including long bone f ract ures, pelvic
f ract ures, and head injuries, met a-analysis suggest s t hat only spinal f ract ures
and spinal cord injuries independent ly increase t he incidence of DVT among
t rauma pat ient s. Specif ically, spinal cord injury has been show n t o increase t he
risk of DVT t hreef old.
Various modalit ies have been used in an at t empt t o decrease t he incidence of
DVT and associat ed pulmonary embolism (PE) in t rauma pat ient s, including bot h
low -dose unf ract ionat ed heparin and low -molecular-w eight heparin (LMWH).
Unf ract ionat ed heparin act s by binding t o ant it hrombin I I I , w hich accelerat es it s
abilit y t o inact ivat e several molecules in t he coagulat ion cascade, including
f act or Xa and t hrombin. LMWH act s primarily by inhibit ing f act or Xa and it s
act ivit y is correlat ed t o f act or Xa levels. LMWH has improved bioavailabilit y and
decreased incidence of bleeding as compared w it h unf ract ionat ed heparin.
Alt hough low -dose unf ract ionat ed heparin has been show n t o be very eff ect ive in
reducing t he incidence of DVT among pat ient s undergoing elect ive surgery, a
number of st udies suggest t hat it s eff ect in t rauma pat ient s is limit ed and t hat
t his int ervent ion may not decrease t he incidence of DVT in t his populat ion. I n
cont rast , having originally been show n t o decrease t he incidence of DVT in
pat ient s undergoing ort hopedic procedures, LMWH has been show n t o be saf e
f ollow ing t rauma, and several report s suggest it is eff ect ive in reducing t he
occurrence of DVT and associat ed complicat ions in t he t rauma populat ion. O ne
landmark st udy demonst rat ed a decrease of 30% in t he incidence of DVT and a
decrease of 58% in t he incidence of proximal vein t hrombosis w it h t he use of
LMWH. Suggest ed dosage f or enoxaparin is 30 mg inject ed subcut aneously t w ice
a day. I t is import ant t o not e,
how ever, t hat a recent review of published st udies examining LMWH suggest s
t hat even t his t herapy may not off er signif icant benef it , and bet t er qualit y st udies
are required t o set t le t his debat e.
What not to Do
Cert ain subgroups of t rauma pat ient s cannot receive eit her heparin or LMWH
prophylaxis because of ongoing risk of hemorrhage. G enerally, most clinicians
avoid t he use of heparin in t he presence of brain and spinal cord injuries, solid
organ injuries t hat are being managed nonoperat ively, and ret roperit oneal
bleeding requiring t ransf usion. Nevert heless, pat ient s w it h t hese
cont raindicat ions t o heparin t herapy can st ill f orm t hrombi in t he ext remit ies and
require an alt ernat ive mode of PE prevent ion. I n t hese pat ient s, inf erior vena
caval (I VC) f ilt ers are f requent ly considered an accept able alt ernat ive. While
t here is convincing evidence f or t he use of I VC f ilt ers in pat ient s w ho have know n
DVTs and eit her cannot receive heparin or have DVTs or PE despit e f ull
ant icoagulat ion, t he t iming and t arget populat ion of f ilt ers insert ed f or
prophylaxis remain debat able. I VC f ilt ers can cause signif icant complicat ions,
including caval penet rat ion, and t he evidence f or t he use of t hese f ilt ers as
prophylaxis in t he general t rauma populat ion is equivocal. Their use is support ed
in specif ic pat t erns of injury, including spinal cord injury, pelvic f ract ure combined
w it h long bone f ract ure, severe head injury combined w it h long bone f ract ure,
and complex long bone injuries. More recent ly, t emporary f ilt ers have been
st udied in t he t rauma populat ion and t he lit erat ure suggest s t hey are a saf e
opt ion in t hese pat ient s. How ever, report ed ret rieval rat es are as low as 35%,
and long-t erm f ollow -up is required t o assess t he proport ion of t hese devices
t hat is ult imat ely ret rieved.
O ne f inal not e is t hat t he use of enoxaparin in t he set t ing of renal f ailure is
problemat ic. The manuf act urer's package insert recommends t he dose of
enoxaparin t o be 30 mg sq qday if t he creat inine clearance is <30. How ever,
many experienced I CU pharmacist s and clinicians recommend a dose of
unf ract ionat ed heparin q8 hours f or pat ient s receiving hemodialysis or cont inuous
venovenous hemodialysis.
Suggested Readings
G eert s WH, Jay RM, Code KI , et al. A comparison of low -dose heparin w it h
low -molecular-w eight heparin as prophylaxis against venous t hromboembolism
af t er major t rauma. N Engl J Med 1996; 335: 701–707.
Knudson MM, I kossi DG , Khaw L, et al. Thromboembolism af t er t rauma: an
analysis of 1602 episodes f rom t he American College of Surgeons Nat ional
Trauma Dat a Bank. Ann Surg 2004; 240: 490–496; discussion 496–498.
> Table of C ontents > Medic ations > 35 - Maintain Tight Gluc os e C ontr ol in the Intens ive C ar e Unit
35
Maintain Tight Glucose Control in the Intensive
Care Unit
Hyperglycemia and insulin resist ance are common in crit ically ill pat ient s even if
t hey did not have diabet es bef ore t heir illnesses. Hyperglycemia in t he int ensive
care unit (I CU) set t ing has been show n t o be associat ed w it h an increased
mort alit y, and rest orat ion of nor-moglycemia using int ensive insulin t herapy has
decreased mort alit y f or t he f ollow ing populat ions: myocardial inf arct ion, burn, and
st roke. Specif ically, in t he surgical I CU set t ing, t he use of int ensive insulin
t herapy t o achieve nor-moglycemia has result ed in a low er incidence of
nosocomial inf ect ions.
The most compelling st udy by Van den Berghe et al. show ed t hat , f or
hyperglycemic pat ient s in t he surgical I CU set t ing on t he vent ilat or, int ensive
insulin t herapy t o achieve nor-moglycemia (80 t o 110 mg/ dL = 4. 4 t o 6. 1 mmol/ L)
result ed in a 43% reduct ion in mort alit y (8% compared w it h 4. 6%) w hen
compared w it h pat ient s t reat ed w it h sliding scale insulin w it h t he aim of
achieving a glucose level bet w een 180 and 200 mg/ dL (10. 0 t o 11. 1 mmol/ L).
The low ered mort alit y w as largely account ed f or in t he group w it h an I CU st ay of
more t han 5 days w hose mort alit y f rom mult iple organ f ailure w as low ered.
Addit ional out comes associat ed w it h nor-moglycemia included f ew er
bact eremias, reduced requirement f or hemodialysis, f ew er t ransf usions, and
short er durat ion on t he vent ilat or and in t he I CU. Alt hough t heir I CU populat ion
w as skew ed t ow ard post operat ive cardiac surgery pat ient s, t he-I CU-st ay
populat ion in w hom t he mort alit y reduct ion w as more st riking, and t heref ore a
bet t er balanced group.
Accordingly, t he current ly available evidence f avors aiming f or nor-moglycemia
(blood glucose level of less t han 110 mg/ dL = 6. 1 mmol/ L) t hrough int ravenous
insulin inf usion in adult surgical I CU pat ient s. I t has not been proven t hat t he
same improved out comes w ill be f ound in nonadult or nonsurgical I CU pat ient s.
The use of int ensive insulin inf usion can result in hypoglycemia eit her f rom
excess insulin or f rom improvement in t he pat ient s’ condit ions (w it h a f all in
insulin resist ance). Many I CUs count eract t his risk of hypoglycemia by also
placing t he pat ient on an inf usion of dext rose during insulin administ rat ion.
Frequent blood glucose monit oring (on t he
order of hourly) is an import ant part of t he t reat ment prot ocol in insulin inf usion
t o avoid hypoglycemia.
Suggested Readings
Coursin DB, Connery LE, Ket zler JT. Perioperat ive diabet ic and hyperglycemic
management issues. Crit Care Med 2004; 32: S116–S125.
Van den Berghe G , Wout ers P, Weekers F, et al. I nt ensive insulin t herapy in
t he surgical int ensive care unit . N Engl J Med 2001; 345: 1359–1367.
> Table of C ontents > Medic ations > 36 - D o not us e S ubc utaneous Ins ulin in the Intens ive C ar e Unit
P opulation
36
Do not use Subcutaneous Insulin in the
Intensive Care Unit Population
Kristin Shipman MD
Heidi L. Frankel MD
Regular insulin should be administ ered int ravenously and not subcut aneously in
t he int ensive care unit (I CU). This is because t he rat e of subcut aneous insulin
absorpt ion is highly variable. I nsulin absorpt ion is slow er w it h high insulin
concent rat ions and in t he t ypically high dose volumes t hat are of t en required t o
cont rol blood glucose in t he crit ically ill. I n addit ion, decreased capillary surf ace
area decreases insulin absorpt ion; because of poor perf usion, crit ically ill
pat ient s of t en experience t his condit ion secondary t o cardiogenic shock,
hypot ension, or vasopressor t herapy. Last ly, poor perf usion of t he subcut aneous
t issues may lead t o slow or errat ic absorpt ion of subcut aneous insulin w it h
result ing poor glucose cont rol.
I n cont rast t o subcut aneous administ rat ion, int ravenous insulin administ rat ion
leads t o reliably eff ect ive absorpt ion and prompt t reat ment of elevat ed blood
glucose levels. I nt ravenous insulin has an onset of act ion in 10 t o 30 minut es
w it h a durat ion of act ion of 1 hour; subcut aneous insulin has an onset of act ion
of 30 t o 60 minut es and peaks at 2 t o 4 hours and last s up t o 8 hours.
I t cannot be overst at ed t hat st rict glucose cont rol reduces mort alit y. I nt ensive
insulin t herapy is recommended f or crit ical care pat ient s t o maint ain blood
glucose bet w een 80 and 110 mg/ dL. The risk reduct ion f or mort alit y has been
show n t o be as high as 45%. I nt ensive insulin t herapy reduces overall hospit al
mort alit y by 34%, bloodst ream inf ect ions by 46%, acut e renal f ailure requiring
dialysis or hemof ilt rat ion by 41%, median number of red blood cell t ransf usions
by 50%, and crit ical illness polyneuropat hy by 44%. This is best achieved w it h
insulin inf usion and t it rat ed f or blood glucose levels bet w een 80 and 110 mg/ dL.
Suggested Readings
Mizock BA. Alt erat ions in carbohydrat e met abolism during st ress: a review of
t he lit erat ure. Am J Med 1995; 98: 75–84.
Van den Berghe G , Wout ers P, Weekers F, et al. I nt ensive insulin t herapy in
crit ically ill pat ient s. N Engl J Med 2001; 345: 1359–1365.
Van den Berghe G , Wilmer A, Hermans G , et al. I nt ensive insulin t herapy in

t he medical I CU. N Engl J Med 2006; 354: 516–518.
> Table of C ontents > Medic ations > 37 - D o not us e Ins ulin Glar gine in the Intens ive C ar e Unit W ithout
Als o Giving A S hor ter - Ac ting Ins ulin For m
37
Do not use Insulin Glargine in the Intensive Care
Unit Without Also Giving A Shorter-Acting
Insulin Form
Kristin Shipman MD
Heidi L. Frankel MD
Lant us is a long-act ing insulin analogue also know n as insulin glargine. Similar t o
human insulin, it is produced using recombinant DNA t echniques. I nsulin glargine
subst it ut es glycine f or asparagine at posit ion A21 of t he insulin molecule and
adds t w o arginine molecules at posit ion B30. The isoelect ric point of insulin
glargine is close t o a pH of 7. 0. Alt ering t hese amino acids makes insulin
glargine precipit at e in subcut aneous t issue. Zinc is added t o st abilize
int rahexamer cont act t o prolong act ivit y.
Lant us is given subcut aneously as a depot inject ion. The onset of act ion is 2 t o 4
hours. There is no peak act ion. The durat ion of act ion is 20 t o 24 hours. Wit h a
long durat ion of act ion, Lant us lends it self t o once-a-day dosing in t he morning—
an at t ract ive f ormulat ion f or t hose not crit ically ill.
Lant us should be used only as a basal insulin replacement . Prandial insulin
requirement s should be covered w it h short -act ing insulin and t he pat ient should
be f ollow ed closely as insulin needs of t en decrease as pat ient s become less ill.
Lant us should not be used in t hose in w hom blood glucose levels may vary
great ly (e. g. , crit ically ill pat ient s w hose levels may f luct uat e as pressors are
t it rat ed and sepsis occurs) or in w hom int errupt ion of parent eral nut rit ion and
t ube f eeds may occur (as it does in virt ually all I CU pat ient s f or many reasons).
I t should be not ed t hat in crit ically ill pat ient s, t he long durat ion of Lant us can
cause prof ound hypoglycemia w it hout concomit ant administ rat ion of glucose
calories.
Suggested Readings
Hirsch I . I nsulin analogues. N Engl J Med 2005; 352: 174–183.
Y ki-Jarvinen H, Ziemen M, Dressler A. Less noct urnal hypoglycemia and
bet t er post dinner glucose cont rol w it h bedt ime insulin glargine compared w it h
bedt ime NPH insulin during insulin combinat ion t herapy in t ype 2 diabet es.
Diabet es Care 2000; 23: 1130–1136.
> Table of C ontents > Medic ations > 38 - R em em ber that P atients w ith Ins ulin D efic ienc y R equir e
B as al Ins ulin E ven W hen They Ar e Nil P er O s
38
Remember that Patients with Insulin Deficiency
Require Basal Insulin Even When They Are Nil
Per Os
Kathleen A. Williams RN, MSN CRNP
Maint aining euglycemia (def ined as glucose <110 mg/ dL in int ensive care unit
set t ings and <180 mg/ dL in non–int ensive care unit set t ings by American College
of Endocrinology Consensus G uidelines) has been show n t o great ly reduce
morbidit y and mort alit y. This has creat ed a paradigm shif t in t he inpat ient set t ing
w here t he goal is now t o prevent hyperglycemia rat her t han t reat hyperglycemia
af t er it develops. I nsulin, in bot h int ravenous and subcut aneous regimens, is
being used in t he inpat ient set t ing t o achieve opt imum glucose cont rol. The
clinical challenge w it h t he use of insulin includes recognizing w hen t o hold or
decrease insulin and in w hich t ype of diabet ic pat ient insulin should be adjust ed
in t he nil per os (NPO ) set t ing.
Considerable conf usion has been caused by t he previous diabet es classif icat ions
of insulin-dependent diabet es mellit us (I DDM) and non–insulin-dependent
diabet es mellit us (NI DDM). Because of t he obesit y epidemic, NI DDM pat ient s
of t en progress t o requiring insulin, causing healt h care providers t o t hen ref er t o
t hese pat ient s as having I DDM. Yet , t he NI DDM pat ient w ho progressed t o
requiring insulin is physiologically very diff erent f rom t he pat ient w ho w as
diagnosed init ially as an I DDM pat ient . While bot h pat ient s require insulin, t he
physiological cause of t heir insulin requirement and t he response t o exogenous
insulin is signif icant ly diff erent .
I n response t o t his issue, diabet es is now classif ied based on t he et iology of t he
respect ive disease process. These classif icat ions are as f ollow s:
a. Type I Diabet es: Charact erized by absolut e insulin def iciency. The pat ient is
physiologically unable t o make endogenous insulin. This inabilit y is due t o
eit her an immune-mediat ed process or severe injury or resect ion of t he
pancreas. I n rare cases, t here is an idiopat hic process leading t o bet a cell
dest ruct ion.
b. Type I I Diabet es: Varying degree of insulin resist ance and relat ive insulin
def iciency. This can occur in bot h children and adult s and is most of t en
associat ed w it h obesit y. I t is of t en t reat ed init ially
w it h oral agent s but w ill usually progress t o requiring insulin f or adequat e

glucose cont rol.
TABLE 38-1 FEATURES OF INSULIN DEFICIENCY
Features of insulin deficiency include the following:
• Type I diabetic patients
• Type II diabetic patients who have required insulin

therapy for >5 years
• Type II diabetic patients or who have been diagnosed

as a type II diabetic for >10 years
• Patients who have undergone a total pancreatectomy
I t is recommended t hat pat ient s w ho are insulin def icient (Table 38. 1) must be
given some f orm of basal insulin eit her as an insulin inf usion or as a
subcut aneous int ermediat e (NPH) or long-act ing (glargine) insulin t o prevent
diabet ic ket oacidosis, even in t he pat ient w ho is NPO .
I t is of t en diff icult t o ascert ain w hich pat ient s are insulin def icient . I f it is unclear
w hich t ype of diabet es a pat ient may have, t he f ollow ing physiological indict ors
of insulin def iciency should be considered:
Posit ive serum or urine ket ones

Wide f luct uat ions in glucose values
Hist ory of requiring insulin t herapy t o cont rol blood sugar since init ial
diagnosis
Previous admissions f or diabet ic ket oacidosis (DKA) or hyperosmolar
hyperglycemia nonket ot ic syndrome (HHNK)
Suggested Readings
Clement S, Brait hw ait e SS, Magee MF, et al. Management of diabet es and
hyperglycemia in hospit als. Diabet es Care 2004; 27: 553–591.
Coursin DB. Perioperat ive diabet ic and hyperglycemia management issues.

Crit Care Med 2004; 32: S116–S125.
Dagogo-Jack S. Management of diabet es in surgical pat ient s. Diabet es Spect

2002; 15: 44–48.
> Table of C ontents > Medic ations > 39 - Look for Medic ation- Induc ed C aus es of Hyper glyc em ia in
Intens ive C ar e P atients
39
Look for Medication-Induced Causes of
Hyperglycemia in Intensive Care Patients
Hyperglycemia, def ined as blood glucose levels >110 mg/ dL, has been show n in
mult iple st udies t o adversely impact out comes in crit ically ill pat ient s. Pat ient s
w ho are diabet ic are at an increased risk f or developing hyperglycemia
post operat ively. Wit h t he harm of hyperglycemia now know n, it is import ant t o
search out pat ient f act ors t hat can cont ribut e t o t his. Pat ient s need t o have t heir
int ravenous (I V) f luids assessed. For example, inf using f luids t hat cont ain
dext rose 5% in w at er (D5W) at 125 cc/ hr can deliver as much as 150 grams of
dext rose, or more t han 500 kcal of carbohydrat e f uel. The I V f luids t hat are used
t o administ er medicat ions also need t o be examined. For example,
sulf amet hoxazole-t rimet hoprim (Bact rim) requires a large amount of D5W t o be
inf used (at least 250 mL per dose). Hyperglycemia can also be seen w it h
perit oneal dialysat e solut ions t hat cont ain high (>2. 5%) concent rat ions of
dext rose. I nt olerance t o parent eral nut rit ion can manif est as hyperglycemia.
I n addit ion t o int ravenous f luids, many medicat ions t hat are used in t he crit ically
ill can exacerbat e hyperglycemia. Medicat ions can int erf ere w it h glucose
met abolism t hrough mult iple mechanisms, including increased insulin resist ance,
decreased insulin secret ion, and increased glucagon product ion. The major
organs t hat are inf luenced by medicat ions are t he pancreas, liver, and skelet al
muscle. Medicat ions can inhibit insulin secret ion f rom t he bet a cells of t he
pancreas (e. g. , gat if loxacin), as w ell increase glycogen breakdow n in t he liver
(e. g. , diazoxide), and cause peripheral insulin resist ance in t he skelet al muscle
(pent amidine). O ct reot ide inhibit s bot h glucagon and insulin secret ion, result ing
in eit her hypo- or hyperglycemia, but t he lat t er is more common clinically. O t her
medicat ions (e. g. , prot ease inhibit ors, at ypical ant ipsychot ics) can cause w eight
gain and f at redist ribut ion, w hich can lead t o t ype I I diabet es mellit us. Table
39. 1 cont ains examples of medicat ions t hat can cause glycemic dysregulat ion.
Management st rat egies include avoiding (if possible) t hese agent s in pat ient s at
risk f or t he development of hyperglycemia (diabet ics, obese pat ient s) and
monit oring serum glucoses closely. I nsulin should be init iat ed f or agent s t hat
aff ect insulin secret ion and cause an increase
in glycogen breakdow n. I nsulin-sensit izing agent s (met f ormin, t hiazolidinediones)

can be st art ed in pat ient s w ho w ill require t herapy f or hyperglycemia as
out pat ient s, alt hough t hese agent s carry t heir ow n risks (lact ic acidosis,
exacerbat ion of peripheral edema) and should not be rout inely st art ed in t he
crit ically ill. Met f ormin should especially be avoided if t he pat ient is in renal
f ailure or is going t o have procedures done t hat require I V dye because of t he
risk of lif e-t hreat ening lact ic acidosis. I nsulin t herapy, w it h regular insulin, may
be pref erred w hile pat ient s are in t he int ensive care unit .
TABLE 39-1 M EDICATION EFFECTS ON GLUCOSE

HOM EOSTASIS
SIT E OF MEDICAT ION

MECHANISM
DYSREGULAT ION IMPLICAT ED
Thiazide
diuretics
Beta-agonists
Diazoxide
Inhibition of Glucocorticoids
Pancreas
insulin secretion Cyclosporine
Tacrolimus
Pentamidine
Epinephrine
Gatifloxacin
Thiazide
diuretics
Increased Beta-agonists
Liver
glycogenolysis Diazoxide
Glucocorticoids
Ethanol
Thiazide
diuretics
Decreased Beta-agonists
Peripheral tissues insulin Diazoxide
sensitivity Glucocorticoids
Cyclosporine
Pentamidine
Suggested Readings
Luna B, Feinglos MN. Drug-induced hyperglycemia. JAMA 2001; 286: 1945–
1948.
Pandit MK. Drug induced disorders of glucose t olerance. Ann I nt ern Med
1993; 118: 529–539.
Van Den Berghe G , et al. I nt ensive insulin t herapy in t he crit ically ill pat ient s.
N Engl J Med 2001; 345: 1359–1367.
> Table of C ontents > Medic ations > 40 - D o not us e Midaz olam and Lor az epam Inter c hangeably in the
Intens ive C ar e Unit
40
Do not use Midazolam and Lorazepam
Interchangeably in the Intensive Care Unit
Ronald F. Sing DO
A common misconcept ion regarding benzodiazepines used in t he int ensive care
unit (I CU) is t hat t hey are int erchangeable agent s producing t he same eff ect s.
This approach does not acknow ledge t he pharmacologic diff erences bet w een
midazolam and lorazepam, w hich is primarily in t heir lipophilic propert ies.
Lipophilicit y direct ly aff ect s t he onset and durat ion of act ion of a drug. I n
addit ion, it is import ant t o not e t hat t he pharmacokinet ics of each drug changes
considerably w hen t hey are given as a single bolus versus an inf usion.
As a single bolus, midazolam has a short er onset and durat ion of act ion t han
lorazepam, w hich makes it opt imal f or acut e agit at ion or sedat ion prior t o a
procedure. The reason is t hat midazolam undergoes a t ransf ormat ion t o a highly
lipid st at e af t er administ rat ion. The highly lipophilic benzodiazepines, such as
midazolam and diazepam, are able t o cross t he blood–brain barrier rapidly,
causing a quick onset of act ion; midazolam has an onset of act ion of 0. 5 t o 5
minut es. I n cont rast , lorazepam is less lipophilic, w hich delays t he crossover int o
t he blood– brain barrier af t er a single bolus; lorazepam has an onset of act ion of
15 t o 20 minut es. Being highly lipophilic also result s in short durat ion of act ion
af t er a single bolus. Agent s like midazolam are able t o readily leave t he cent ral
nervous syst em (CNS) and move back int o t he plasma compart ment (midazolam
durat ion is 2 hours), w hereas t he more polar lorazepam is t rapped in t he CNS,
result ing in a longer durat ion of act ion (6 t o 10 hours).
Watch Out For

How ever, t his t ime course of act ion is reversed in t he set t ing of cont inuous
inf usions. I t is import ant t o not e t hat a cont inuous inf usion of midazolam result s
in drug accumulat ion in peripheral t issues. The general rule is t hat t he longer t he
midazolam inf usion and t he higher t he dose, t he longer it w ill t ake f or t he pat ient
t o aw aken once t he inf usion has been discont inued. I n essence, t he prolonged
and/ or high-dose inf usions creat e a midazolam depot in t he peripheral
compart ment . O nce t he inf usion has been discont inued, peripheral t issues begin
t o release midazolam int o plasma, w here it reaches t he CNS and crosses
t he blood–brain barrier. The clinical eff ect t hat occurs is a delayed and
unpredict able w ake t ime. As discussed earlier, lorazepam is less lipophilic,
result ing in minimal peripheral accumulat ion w it h const ant inf usion. The clinical
eff ect is a more predict able, alt hough st ill prolonged, w ake t ime, w hich is
proport ional t o t he dose administ ered.
The clinical relevance of t his inf ormat ion is t hat midazolam should be t he
pref erred benzodiazepine f or pat ient s w ho need t o be sedat ed f or less t han 48
t o 72 hours w hile lorazepam should be reserved f or t hose pat ient s requiring
sedat ion longer t han 72 hours. O ne st rat egy is t o init iat e sedat ion w it h
midazolam f or t he f irst 48 t o 72 hours and t o sw it ch t o lorazepam if t herapy is
prolonged. There is no consensus on equivalent doses f or lorazepam and
midazolam, but an easy-t o-remember rule of t humb is a 2: 1 dose of midazolam
t o lorazepam. Convert ing back t o midazolam in preparat ion f or ext ubat ion should
be a considerat ion.
Anot her common error in t he I CU is t he misconcept ion t hat pat ient s receiving
lorazepam inf usions should receive lorazepam boluses f or acut e agit at ion.
Midazolam should be t he agent of choice f or acut e agit at ion f or pat ient s
receiving eit her midazolam or lorazepam inf usions because of midazolam's more
rapid onset of act ion. I n addit ion, bolus doses need t o be given w it h each
inf usion rat e increase. The bolus is int ended t o alleviat e acut e agit at ion unt il t he
inf usion reaches it s new st eady st at e (f ive drug half -lives).
O ne f inal not e is t hat prolonged high-dose inf usions of lorazepam have been
report ed t o cause reversible acut e t ubular necrosis, lact ic acidosis, and
hyperosmolar st at es. The lorazepam solvent s polyet hylene glycol (PEG ) and
propylene glycol (PG ) have been suggest ed as t he cause of t hese eff ect s and
t hese high doses should be avoided if possible.
Suggested Reading
Jacobi J, Fraser G L, Coursin DB, et al. Societ y of Crit ical Care Medicine,
American Societ y of Healt h-Syst em Pharmacist s, American College of Crit ical
Care Medicine. Clinical pract ice guidelines f or t he sust ained use of sedat ives
and analgesics in t he crit ically ill adult . Crit Care Med 2002; 30: 119–141.
> Table of C ontents > Medic ations > 41 - Tr y to Avoid Us ing B enz odiaz epines for S leep in the Intens ive
C ar e Unit, E s pec ially in the E lder ly
41
Try to Avoid Using Benzodiazepines for Sleep in
the Intensive Care Unit, Especially in the Elderly
Bryan A. Cotton MD
During hospit alizat ion, repeat ed arousals disrupt sleep cont inuit y. These
aw akenings are secondary t o laborat ory draw s, daily chest radiographs, vit al
sign recording, and numerous aw akenings by nursing st aff , ancillary personnel,
medical st udent s, and varying levels of physicians in t raining. Several st udies
have not ed t he f requency of sleep int errupt ions t o be as of t en as every 20
minut es. I n addit ion, t he absence of diurnal light cycles is a major source of
int ensive care unit (I CU) sleep disrupt ion and f requent ly result s in cognit ive
dist urbances. Almost 50% of sleep experienced by pat ient s in t he I CU occurs in
t he dayt ime. This f orm of sleep, how ever, lacks delt a w ave (deep) and rapid eye
movement (REM) sleep, w hich are t he “rest f ul” and physiologically st able f orms
of sleep. I n addit ion, sleep disrupt ion is associat ed w it h many f requent ly used
I CU medicat ions such as bet a-blockers, diuret ics, benzodiazepines, and opiat es.
Watch Out For

Sleep disrupt ion result s in numerous delet erious eff ect s on an already f ragile
and dist urbed physiology. Pulmonary consequences include decreased f unct ional
vit al capacit y, blunt ed hypercapnic and hypoxic responses, and decreased
respirat ory muscle endurance. As a result of t he aut onomic imbalances t hat
f ollow sleep deprivat ion, increased hypert ensive episodes, more f requent
arrhyt hmias, and increased risk of acut e myocardial ischemia have been
described. I n addit ion, sleep deprivat ion promot es a negat ive nit rogen balance,
increased rest ingenergy expendit ure, and immunological depression via
suppression of ant ibody and cell-mediat ed responses. Most obvious t o t he
physician, how ever, are t he neurocognit ive sequelae. Delirium, anxiet y,
hallucinat ions, and mood disorders have been associat ed w it h sleep deprivat ion
in t he I CU. I n f act , healt hy volunt eers subject ed t o I CU-like sleep dist urbances
develop irrit abilit y, disorient at ion, and slurred speech.
Nonpharmacologic opt ions t o improve sleep hygiene include placing pat ient s in
privat e or single rooms w it h several w indow s, minimizing night t ime conversat ions
by hospit al personnel, placing alarms
out side of pat ient rooms, and scheduling bat hs, linen changes, rout ine laborat ory
draw s, and radiographs during t he dayt ime. St udies t hat have evaluat ed t he
impact of noise reduct ion in t he I CU have not ed marked improvement in sleep
qualit y. I mproved REM sleep has been demonst rat ed w it h t he implement at ion of
earplugs in I CU pat ient s. Addit ionally, light s should be t urned off or dimmed at
night t o maint ain circadian light cycles. Conversely, keeping t he light s on during
t he dayt ime has been show n t o reset circadian cycles and improves sleep at
night , specif ically in t he elderly. I n t he high-risk populat ion of elderly I CU
pat ient s, physical t herapy and increased act ivit y have been show n t o promot e
sleep w it h short er lat ency and deeper levels achieved.
Pharmacological agent s are of t en used t o improve sleep in t he I CU. The ideal
hypnot ic w ould have a short half -lif e, have f ew drug-drug int eract ions, have f ew
cardiac or pulmonary eff ect s, and maint ain sleep archit ect ure. Alt hough t his drug
has yet t o be creat ed, several agent s current ly available come close and should
be considered f or t he crit ically ill pat ient w it h sleep dist urbances. O ne of t he
oldest and least expensive medicat ions available f or sleep in t his populat ion is
chloral hydrat e. I t is f requent ly used t o init iat e sleep f or polysomnography
st udies, sedat e children f or procedures and diagnost ic evaluat ions, as w ell as
induce deeper levels of sleep in t he pediat ric I CU set t ing. Chloral hydrat e
induces sleep w it hout changes in respirat ory rat e, PaCO 2, PaO 2, or t idal
volumes. This is part icularly saf e and usef ul in t he elderly in t he induct ion and
maint enance of sleep.
O t her drugs somet imes used are zolpidem, t razodone, and mirt azapine. Alt hough
zolpidem is chemically unrelat ed t o t he benzodiazepines, it binds t o omega
recept ors in t he brain and has been show n t o preserve t he deep st ages of sleep,
as w ell as REM. This agent (or similar drugs in it s grow ing class) may prove t o
be t he ideal hypnot ic, but it has yet t o be st udied in t he I CU. Tw o ot her agent s
w ort hy of considerat ion are t he ant idepressant s t razodone and mirt azapine. Wit h
t he increasing recognit ion of mood disorders associat ed w it h prolonged I CU
st ays, t hese agent s have seen increased ut ilizat ion in t hose crit ically ill pat ient s
w it h sleep dist urbances and post t raumat ic st ress disorder (PTSD) or sit uat ional-
appropriat e mood disorders. Trazodone has been ident if ied as a very eff ect ive
agent f or sleep in depressed pat ient s and t hose w it h PTSD. I n addit ion,
t razodone is w ell t olerat ed and associat ed w it h decreased aw akenings and
deeper levels of sleep achieved. Mirt azapine has a superior eff ect on t he
sympt oms of insomnia, poor appet it e, and anxiet y. I t has been show n t o improve
sleep lat ency, increase deeper st ages of sleep, and increase t ot al t ime of sleep.
What not to Do
Several drugs should be avoided in t he I CU pat ient , especially w hen used as
hypnot ics. Benzodiazepines have been previously prescribed f or sleep promot ion
and are f requent ly t he only opt ion know n t o many house off icers, as w ell as
at t endings. How ever, t his class of drugs may have delet erious eff ect s on t he
respirat ory syst em (increased hypercapnia, upper airw ay hypot onia) and on
sleep archit ect ure. Benzodiazepines are associat ed w it h a signif icant hangover
eff ect and cognit ive dist urbances, especially in t he elderly. Benzodiazepine use
in t he elderly has been associat ed w it h increased incidence of delirium, likely as
a result of increased suscept ibilit y t o cognit ive decline, as w ell as less
predict able plasma drug levels. Moreover, t he apolipoprot ein E (APO E) 4 allele,
w hich is associat ed w it h increased risk of development of Alzheimer disease,
has been ident if ied as a pot ent ial explanat ion as t o w hy many elderly pat ient s
respond t o benzodiazepines w it h a paradoxical agit at ion and/ or conf usion
f ollow ing exposure. How ever, diphenhydramine, an ant ihist amine and sedat ing
drug w it h st rong ant icholinergic propert ies, is t he agent most likely t o result in
adverse event s f ollow ing it s use as a sleep agent . Paradoxical excit ement , visual
dist urbances, and delirium are a result of it s ant icholinergic propert ies. The
f requency of t hese event s, especially in t he elderly, should limit t he use of
diphenhydramine in t he I CU t o anaphylact ic react ions.
O ne f inal not e is t hat alt hough pot ent ially eff ect ive in t he out pat ient set t ing,
melat onin has f ailed t o show improvement in sleep in t he I CU pat ient .
Suggested Readings
G abor JY, Cooper AB, Hanly PJ. Sleep disrupt ion in t he int ensive care unit .
Curr O pin Crit Care 2001; 7: 21–27.
Part hasarat hy S, Tobin MJ. Sleep in t he int ensive care unit . I nt ens Care Med
2004; 30: 197–206.
Peruzzi WT. Sleep in t he int ensive care unit . Pharmacot herapy 2005; 25: 34S–
39S.
> Table of C ontents > Medic ations > 42 - R em em ber that Ac tivated P r otein C is not as Us eful for
S eps is as O nc e Hoped
42
Remember that Activated Protein C is not as
Useful for Sepsis as Once Hoped
Act ivat ed prot ein C (APC) exert s a number of biological eff ect s including ant i-
inf lammat ory, ant it hrombot ic, and prof ibrinolyt ic propert ies. Prof ibrinolyt ic
eff ect s include inact ivat ion of plasminogen act ivat or inhibit or. Ant it hrombot ic
eff ect s include inact ivat ion of f act or Va and VI I I a. These eff ect s are t hought t o
aid in rest orat ion and maint enance of microcirculat ory blood f low, re-
est ablishment of coagulat ion homeost asis, and preservat ion of t he
microcirculat ion. Tissue injury and organ dysf unct ion is t hought t o be at t enuat ed
by ant i-inf lammat ory propert ies including decreasing t hrombin-relat ed
inf lammat ion, inhibit ing t umor necrosis f act or (TNF) and macrophage migrat ion
inhibit ory f act or product ion, blocking adhesion of leukocyt es t o select ins, and
decreasing nuclear f act or–Kβ act ivat ion.
The init ial opt imism associat ed w it h recombinant human act ivat ed prot ein C
(rhAPC), (drot recogin alf a [ act ivat ed] ) has been t empered by subsequent clinical
t rials. Dat a f rom t he Recombinant Human Act ivat ed Prot ein C Worldw ide
Evaluat ion in Severe Sepsis (PRO WESSS) t rial demonst rat ed an absolut e
reduct ion of deat h f rom all causes at 28 days of 6. 1%. The t rial observed an
increase in serious bleeding f rom 2. 0% in t he placebo group t o 3. 5% in t he APC
group. How ever, t his eff ect w as not st at ist ically signif icant (it must be not ed t hat
t his st udy w as not pow ered t o det ect signif icance). Subgroup analysis revealed
t hat t he great est risk reduct ion occurred in pat ient s w it h t he great est severit y of
disease as assessed by APACHE I I scores. Subsequent ly t he drug w as approved
by t he Food and Drug Administ rat ion (FDA) f or use in pat ient s w it h APACHE I I
scores >25. The recommendat ion t o limit t herapy t o t he sicker group of pat ient s
w as conf irmed in a t rial of APC in pat ient s w it h APACHE I I scores < 25 or w it h
single organ f ailure. This t rial, t he Administ rat ion of Drot recogin Alf a (Act ivat ed)
in Early St age Severe Sepsis (ADDRESS), w as discont inued af t er enrolling
2, 640 pat ient s because of an inabilit y t o demonst rat e reduced mort alit y in t his
less severely ill group w it hAPC. Adding f urt her t o t he argument against using
APC is t hat t he risk of serious bleeding w it h APC may be great er t han originally
report ed in PRO WESS. An uncont rolled open label t rial (ENHANCE) report ed
incidence of serious bleeding at 6. 5%. O t her open-label t rials have also

demonst rat ed lack of benef it in pediat ric pat ient s.
Watch Out For

Cont raindicat ions t o t reat ment w it h RhAPC include condit ions associat ed w it h
risk of bleeding such as act ive bleeding, recent hemorrhagic st roke (w it hin 3
mont hs), recent int racranial or spinal surgery or recent severe head t rauma
(w it hin 2 mont hs), t rauma w it h increased risk of lif e-t hreat ening bleeding,
presence of an epidural cat het er, or int racranial neoplasm, mass lesion, or
cerebral herniat ion. An FDA w arning w as added early in 2005 not ing higher all-
cause mort alit y in pat ient s w it h recent surgery (w it hin 30 days) and single organ
dysf unct ion based on dat a f rom pat ient s enrolled in t he PRO WESS t rial. The
ADDRESS t rial also f ound higher all-cause mort alit y in pat ient s w it h recent
surgery and APACHE I I score <25 w it h single organ dysf unct ion. Based on t his
dat a, many expert s recommend considerat ion of rhAPC f or surgical pat ient s only
if t w o or more organ f ailures are present .
O ne f inal not e is t hat a 4-day course of rhAPC cost s approximat ely $6, 800.
Using dat a f rom PRO WESS, t he cost per qualit y-adjust ed lif e year f or pat ient s
w it h APACHE I I scores >25 is $27, 400. rhAPC is administ ered via a cont inuous
inf usion at 24 mcg/ kg/ hr f or 96 hours based on act ual body w eight .
Suggested Readings
Abraham E, Lat erre PF, G arg, R, et al. Drot recogin alf a (act ivat ed) f or adult s
w it h severe sepsis and a low risk of deat h. N Engl J Med 2005; 353: 1332–
1341.
Bernard G R, Vincent JL, Lat erre PF, et al. Eff icacy and saf et y of recombinant
human act ivat ed prot ein C f or severe sepsis. N Engl J Med 2001; 344: 699–
709.
Fourrier F. Recombinant human act ivat ed prot ein C in t he t reat ment of severe
sepsis: an evidence-based review. Crit Care Med 2004; 32: S534–S541.
Rice TW, Bernard G R. Drot recogin alf a (act ivat ed) f or t he t reat ment of
severe sepsis and sept ic shock. Am J Med Sci 2004; 328: 205–214.
Vincent JL, Bernard G R, Beale R, et al. Drot recogin alf a (act ivat ed) t reat ment
in severe sepsis f rom t he global open-label t rial ENHANCE; f urt her evidence
f or survival and saf et y and implicat ions f or early t reat ment . Crit Care Med
2005; 33: 2266–2277.
> Table of C ontents > Medic ations > 43 - K now the Alter nate R outes for Adm inis tr ation of
C ar diopulm onar y R es us c itation Medic ations
43
Know the Alternate Routes for Administration of
Cardiopulmonary Resuscitation Medications
Eric M. Bershad MD
Jose I. Suarez MD
I n a cardiac arrest , pharmacologic support may be needed t o help rest ore
spont aneous circulat ion. These medicat ions may include at ropine, epinephrine,
lidocaine, vasopressin, dopamine, naloxone, and ot hers. I n some pat ient s, a
cent ral line or peripheral I V are not readily available, t hus ot her met hods of drug
delivery must be used. The alt ernat ive met hods include int raosseous (I O ) and
endot racheal (ET) administ rat ion.
What to Do
The American Heart Associat ion recommends t he use of int raosseous (I O )
cannulat ion bef ore resort ing t o endot racheal (ET) administ rat ion of
cardiopulmonary resuscit at ion (CPR) medicat ions. This is based on mult iple
st udies in children and adult s document ing saf e and eff ect ive f luid resuscit at ion,
drug delivery, and blood sampling w it h t he I O approach. Furt hermore, I O
cannulat ion enables drug delivery comparable w it h t hat achieved by using a
cent ral line. This may be in part relat ed t o t he noncollapsible venous plexus
accessed in t he I O approach. Any drug t hat can be given f or resuscit at ion
int ravenously can also be given I O . I n adult s t he st ernum or proximal t ibia has
been successf ully used f or I O drug delivery. There are commercially available
kit s t hat f acilit at e I O access in adult s.
I f I V or I O access is not available, t he ET rout e should be used. The medicat ions
t hat can be given ET include t he NAVEL drugs: naloxone; at ropine; vasopressin;
epinephrine; and l idocaine (Table 43. 1). Addit ionally, vasopressin may be
administ ered via t he ET rout e. Blood levels of drugs given t he ET rout e are low er
t han comparable blood levels w hen drugs are given int ravenously. The opt imal
dose of ET-administ ered medicat ions is unknow n, but generally should be given
at least 2 t o 2. 5 t imes great er t han t he I V dose, w it h t w o st udies suggest ing t he
required dose of epinephrine given ET should be 3 t o 10 t imes higher t han t he
equipot ent I V dose. The administ ered endot racheal medicat ions should be given
in 5 t o 10 mL of w at er or normal saline and t hen f lushed w it h several brisk
vent ilat ions w it h a bag-mask valve. Some published st udies examining ET-
administ ered
epinephrine and lidocaine show ed t hat dilut ion w it h w at er, rat her t han normal
saline, may achieve bet t er absorpt ion.
TABLE 43-1 STANDARD DOSES OF CPR DRUGS
ET (EXACT
DRUG IV OR IO DOSING
UNKNOWN)
Multiply
1 to 1.5 mg/kg, can repeat
IV/IO dose
0.5 to 0.75 mg/kg q 5 to 10
Lidocaine by 2.5 to
minutes to a maximum
get ET
dose of 3 mg/kg
dose.
1 mg q 3 to 5 minutes up
Atropine to maximum dose of 0.04 Same
mg/kg
0.4 to 2.0 mg initially, can

Naloxone repeat q 2 to 3 minutes up Same
to 10 mg
2 to 2.5 mg
Epinephrine 1 mg q 3 to 5 minutes q 3 to 5
minutes
Vasopressin 40 units × 1 dose

Suggested Readings
American Heart Associat ion. 2005 American Heart Associat ion guidelines f or
CPR and ECC. Circulat ion 2005; 112: I V-1–I V-203.
I serson K. I nt raosseous inf usions in adult s. J Emerg Med 1989; 7: 587–591.
Macnab A, Christ enson J, Findlay J, et al. A new syst em f or st ernal

int raosseous inf usion in adult s. Prehosp Emerg Care 2000; 4: 173–177.
> Table of C ontents > Medic ations > 44 - Alkaliniz e the Ur ine in Tr ic yc lic Antidepr es s ant O ver dos e
44
Alkalinize the Urine in Tricyclic Antidepressant
Overdose
Eliahu S. Feen MD
Jose I. Suarez MD
Tricyclic ant idepressant s (TCAs) are t hree-ringed organic compounds t hat have
been used f or decades in t he t reat ment of depression and cert ain ot her
psychiat ric condit ions. O t her common condit ions f or w hich TCAs are used include
chronic pain syndromes and migraine prophylaxis. Current ly, TCAs are less
commonly used f or t he t reat ment of depression and ot her mood disorders and
psychiat ric condit ions because of t he rise of new er t herapies, especially t he
int roduct ion of select ive serot onin reupt ake inhibit ors (SSRI s).
M echanism of Action
The mechanism of act ion of TCAs relat es t o t he inhibit ion of t he reupt ake of bot h
serot onin and norepinephrine f rom t he synapt ic clef t of neurons in t he cent ral
nervous syst em. I nhibit ion of reupt ake has an acut e onset ; how ever, t he desired
clinical eff ect s can t ake w eeks. Thus, TCAs likely induce post synapt ic neuron
regulat ory changes t hat ult imat ely provide t heir clinical benef it s. I n addit ion,
TCAs have ant icholinergic eff ect s and inhibit ory eff ect s at hist amine and α1 -
adrenergic recept ors, w hich explains t he TCA side-eff ect prof ile.
Clinical M anifestation of TCA Overdose

Compared w it h ot her ant idepressant s (SSRI s or monoamine oxidase inhibit ors
[ MAO I s] ), TCAs carry a great er risk of deat h due t o overdose. TCAs have more
t han t w ice t he risk of deat h associat ed w it h overdoses of MAO I s and more t han
f ive t imes t he risk associat ed w it h SSRI s and at ypical ant idepressant
medicat ions. O f TCA overdoses, 2% t o 3% result in deat h, and most of t hese are
due t o cardiac complicat ions. The clinical manif est at ions of TCA overdose
include psychomot or depression, seizures, t achycardia, and cardiac conduct ion
def ect s (specif ically, prolonged PR, Q RS, and Q T int ervals). I n addit ion, t he
ant icholinergic propert ies of TCAs w ill induce dry skin and mucous membranes,
blurred vision, f lushing, urinary ret ent ion, const ipat ion, and pot ent ially aut onomic
inst abilit y—most not ably hypot ension.
M anagement of TCA Overdose

Treat ment of TCA overdose, like any poisoning, necessit at es immediat e at t ent ion
t o t he pat ient 's ABCs (airw ay, breat hing, and circulat ion) and prompt cont act
w it h a local poison cont rol cent er, if available. G ast ric lavage using a
nasogast ric t ube may be helpf ul if t he pat ient present s w it hin 12 hours of
ingest ion. Act ivat ed charcoal administ rat ion is highly recommended. How ever,
syrup of ipecac is not recommended because of aspirat ion risk, since pat ient s
are usually let hargic. Hemodialysis and similar met hods (charcoal
hemoperf usion, perit oneal dialysis, and exchange t ransf usion) are not know n t o
be signif icant ly eff ect ive in removing TCA f rom t he serumand t issues.
A cent ral t enet of t reat ment of TCA overdose is t hat all pat ient s should be
alkalinized. I t must be not ed t hat w hile alkalinizing t he urine in order t o prevent
renal reabsorpt ion of cert ain ionic subst ances is helpf ul in some poisonings, it
does not specif ically assist w it h inhibit ion of renal reabsorpt ion in TCA. However,
t he reason f or urine alkalinizat ion in TCA overdose is because of t he
cardiovascular t oxicit y of TCAs. Administ rat ion of int ravenous (I V) sodium
bicarbonat e and/ or hypervent ilat ion (in int ubat ed pat ient s) can reverse t oxic
eff ect s on t he myocardium due t o TCAs. Alkalinizat ion has been show n t o
decrease Q RS prolongat ion and repress arrhyt hmias. Typically 1 t o 2 mEq per
kg of sodium bicarbonat e in normal saline (or D5W, if avoiding a saline load is
desired) is given as a bolus, f ollow ed by 150 mEq in a lit er of normal saline (or
D5W) as a cont inuous inf usion. The goal is t o achieve an art erial pH of bet w een
7. 45 and 7. 55. Pat ient s w it h persist ent Q RS prolongat ion may be bolused w it h
sodium bicarbonat e t o achieve a Q RS durat ion of less t han 0. 16 seconds.
Suggested Readings
Henry JA, Alexander CA, Sender EK. Relat ive mort alit y f rom overdose of
ant idepressant s. BMJ 1995; 310: 221–224.
Piment el L, Trommer L. Cyclic ant idepressant overdoses. Emerg Med Clin

Nort h Am 1994; 12: 533–547.
Sarko J. Ant idepressant s, old and new : a review of t heir adverse eff ect s and
t oxicit y in overdose. Emerg Med Clin Nort h Am 2000; 18: 637–654.
> Table of C ontents > Medic ations > 45 - C hec k Tr iglyc er ide Level in P atients on P r opofol D r ips
45
Check Triglyceride Level in Patients on Propofol
Drips
Amisha Barochia MD
Propof ol is a lipid-soluble alkylphenol agent t hat is f requent ly used t o sedat e
mechanically vent ilat ed pat ient s in t he int ensive care unit (I CU). I t has many
desirable propert ies of an ideal sedat ive. I t crosses t he blood-brain barrier
rapidly and has a rapid onset of act ion. I t is met abolized quickly t o inact ive
met abolit es, giving it a short durat ion of act ion. I t is easy t o t it rat e and no dose
adjust ment is necessary f or hepat ic or renal insuff iciency. I t is not associat ed
w it h t olerance or w it hdraw al sympt oms. Propof ol decreases cerebral met abolism
and cerebral blood f low, w hich result s in a decrease in int racranial pressure. I t
has been used w it h benef it in t rauma pat ient s w it h brain injury. I t has also been
used f or seizure cont rol in st at us epilept icus w hen ot her t herapeut ic agent s have
f ailed. I n addit ion, propof ol can be used as a general anest het ic at higher doses
and can provide some ant egrade amnesia. How ever, it does not have analgesic
propert ies and it s amnest ic propert ies are not reliable, so it is of t en used in
conjunct ion w it h benzodiazepines or opiat es.
I n comparison w it h midazolam, propof ol seems t o be a bet t er sedat ive in many
I CU pat ient s. I t is more easily t it rat able, it can be reversed more quickly w it hout
t he development of w it hdraw al or t olerance, and it s use may decrease t he need
f or muscle relaxant s w hen propof ol is used as a f irst -line agent f or sedat ion. O f
not e, a reduced t ime t o w eaning f rom t he vent ilat or has been show n w it h
propof ol use as compared w it h midazolam. Propof ol and midazolam have also
been used t oget her (w it h or w it hout narcot ics) t o minimize t he short comings of
each agent used alone.
Watch Out For

As w it h most ot her sedat ives, t here are some draw backs t o using propof ol.
Hypot ension is a f requent side eff ect , especially w it h bolus administ rat ion of t he
drug. A 25% t o 40% decrease in blood pressure may be seen. This is a dose-
dependent phenomenon and is usually t ransient . I t can of t en be avoided by
caref ul pat ient select ion and adequat e volume replet ion.
O f more concern clinically t han t his easily managed hypot ension is t he propof ol
inf usion syndrome, w hich w as f irst described in 1992
in children and has since been report ed in adult s. The syndrome is usually
associat ed w it h high-dose inf usions of propof ol, and it has been hypot hesized
t hat concomit ant st eroid or cat echolamine use, sepsis, syst emic inf lammat ory
response syndrome (SI RS), and brain injury may cont ribut e t o it s occurrence.
Abnormalit ies in t he mit ochondrial and f at t y acid met abolism are t hought t o be
part of t he pat hogenesis of t his syndrome. The clinical aspect s of t he syndrome
include met abolic acidosis, hypert riglyceridemia, rhabdomyolysis, hypot ension,
bradycardia, and event ually asyst ole and deat h. Alt hough a causal relat ionship
bet w een t his syndrome and propof ol inf usion has not been f irmly est ablished,
vigilance f or t his spect rum of clinical f indings is advised w hen using propof ol,
since cardiac impairment and deat h of t en occur w it hout w arning and t he mort alit y
rat e is high. I f suspect ed, propof ol should be immediat ely discont inued and
support ive measures inst it ut ed. Therapies such as cont inuous hemof ilt rat ion or
hemodialysis might improve out comes.
I n addit ion, it is import ant t o remember t hat propof ol is emulsif ied in a soybean-
based lipid f ormulat ion, w hich provides about 1. 1 kCal/ mL as f at calories. When
inf used at high doses it can be a signif icant source of calories f or t he pat ient ,
w hich should be t aken int o account w hen calculat ing t he pat ient 's nut rit ional
requirement s. As a signif icant calorie source, it may increase CO2 product ion
and necessit at e a great er minut e vent ilat ion t o maint ain acid-base balance.
Anot her concern w it h propof ol inf usions is t he associat ion w it h increased levels
of t riglycerides of up t o 500 t o 600 mg/ dL in some st udies, w hich may
predispose pat ient s t o acut e pancreat it is. I f propof ol is used at high dosages or
f or prolonged periods, t riglyceride levels should be monit ored. The lipid carrier
may also predispose t o t hrombosis by int erf ering w it h prot hrombin t imes, but
inf ormat ion about t his complicat ion is limit ed.
There have been report ed cases of blood-st ream inf ect ions secondary t o t he
inf usion of cont aminat ed propof ol. Propof ol preparat ions are now available w it h
a choice of preservat ives like et hylenediaminet et raacet ic acid (EDTA) and
met abisulf it e. Current recommendat ions st at e t hat propof ol should not be used
f or more t han 6 hours af t er it has been removed f rom it s packaging and t hat
inf usion bot t les and t ubing should be changed f requent ly t o decrease t he risk of
inf ect ion.
Finally, propof ol inf usions can irrit at e peripheral veins and can be very painf ul
during administ rat ion. Lidocaine inject ion prior t o
or w it h t he propof ol may be helpf ul but is not alw ays eff ect ive in ameliorat ing t he
pain.
Suggested Readings
Angelini G , Ket zler JT, Coursin DB. Use of propof ol and ot her
nonbenzodiazepine sedat ives in t he int ensive care unit . Crit Care Clin
2001; 17(4): 863–880.
Riker R, Fraser G L. Adverse eff ect s associat ed w it h sedat ives, analgesics,

and ot her drugs t hat provide pat ient comf ort in t he int ensive care unit .
Pharmacot herapy 2005; 25: 8S–18S.
> Table of C ontents > Medic ations > 46 - B e Aler t for D r ug- R elated P anc r eatitis in HIV/AID S P atients
And C ons ider a P er iod of P r oxim al B ow el R es t
46
Be Alert for Drug-Related Pancreatitis in
HIV/AIDS Patients And Consider a Period of
Proximal Bowel Rest
Lawerence O sei MD
Acut e pancreat it is is an inf lammat ory condit ion of t he pancreas, w hich is
manif est ed clinically as abdominal pain w it h elevat ed pancreat ic enzymes—
amylase and lipase. I t is pot ent ially lif e t hreat ening. Pancreat it is is caused by a
variet y of insult s t o t he pancreas. I t is est imat ed t hat in t he Unit ed St at es,
gallst ones and chronic alcoholism account f or approximat ely 75% of acut e
pancreat it is cases. The incidence of pancreat it is in non-human immunodef iciency
virus (HI V) pat ient s is relat ively low and ranges f rom 17 t o 30 cases per 100, 000
populat ion. How ever, it is considerably higher in t he U. S. HI V populat ion. O ne
st udy of prevalence in t he HI V populat ion f ound incidence rat es as high as 14
cases per 100 pat ient s over a 1-year period. This is t hought t o be due t o
comorbid condit ions like et hanol use, medicat ions f requent ly used in HI V pat ient s
(e. g. , didanosine, st avudine, cort icost eroids, sulf onamides, isoniazid,
ket oconazole, met ronidazole), and opport unist ic inf ect ions like crypt osporidiosis,
mycobact eria, and cyt omegalovirus (CMV) disease.
HI V-posit ive pat ient s are f requent ly admit t ed t o t he int ensive care unit (I CU). I n
addit ion t o pancreat it is, some of t he leading indicat ions f or I CU admissions are
pneumococcal pneumonia or meningit is, crypt ococcal meningit is, t oxoplasmosis,
liver f ailure f rom co-inf ect ion w it h hepat it is B or C, or shock f rom eit her
inf ect ions or drugs (abacavir hypersensit ivit y). The crit ically ill pat ient w it h
pancreat it is on ant iret roviral drugs is a challenge and I CU physicians t ypically
base t reat ment decisions on physician experience rat her t han dat a f rom
cont rolled st udies (t hat largely do not exist ). Clearly, if t he present ing condit ion
is pancreat it is relat ed t o t he HI V t herapy, t he drugs should be st opped and a
regimen of proximal bow el rest should be undert aken. How ever, it is of t en
diff icult t o subst it ut e w it h alt ernat ive regimes because of overlapping t oxicit ies,
previous drug resist ance, or diff icult ies in administ ering t he drug. I n addit ion, in
pat ient s w ho are nil per os (NPO ), abrupt ly st opping ant iret roviral t herapy may
cause a sudden f all in CD4 (+) cell count and a rise in viral load. Some pat ient s
may even develop an acut e seroconversion illness.
Complicat ingmat t ers f or t he HI V pat ient s w it h pancreat it is is t hat most
ant iret roviral drugs can only be administ ered t o pat ient s w it h a normally
f unct ioning gast roint est inal t ract (G I ) w ho can be f ed orally or t hrough a
nasogast ric t ube. There are f ew dat a on t he absorpt ion and pharmacokinet ics of
any of t he ant iret rovirals w hen given t hrough a gast ric or a post pancreat ic jejunal
f eeding t ube. Errat ic absorpt ion f rom a nonf unct ioning G I t ract means t he
possibilit y of pat ient s being on a drug at subt herapeut ic levels, w hich can
promot e resist ance t o highly act ive ant iret roviral t herapy (HAART). How ever,
w here it w ill adversely impact care, it might be prudent t o discont inue
ant iret roviral t herapy unt il t he crit ical illness is over.
When rest art ing HAART af t er a period of I CU management f or pancreat it is, care
must be t aken t o review possible drug-drug int eract ions w it h t he ant iret roviral
regimen, especially w hen enzyme inducers, such as phenyt oin and rif ampicin,
have been st art ed. There are st ill relat ively f ew dat a on t he int eract ions bet w een
many of t he drugs used in t he I CU and t hose used in HI V t herapy. The prot ease
inhibit ors and t he nonnucleosides can act as cyt ochrome P450 3A4 inhibit ors and
aff ect t he met abolism of many of t he drugs used in I CU, such as midazolam or
opiat es. List ed in Tabl e 46. 1 are some drug int eract ions t o check f or.
TABLE 46-1 DRUG-DRUG INTERACTIONS

ASSOCIATED WITH ANTIRETROVIRALS/PROTEASE
INHIBITORS
DRUG PRECAUT IONS
Amiodarone Do not use with ritonavir.
Statins Increases AUC for statins.
Simvastatin,
Do not use.
lovastatin
Atorvastatin Increase 5.8-fold.

Pravastatin AUC increased 30%.
Decreases AUC of lopinavir and

Rifampin
ritonavir 75%.
Lopinavir and ritonavir raises rifabutin

Rifabutin
level 3 times.
Ergot alkaloids Unpredictable ergot levels.
Voriconazole Bidirectional.
Decreases AUC for lopinavir and

Methadone
amprenavir (not ritonavir).
Oral
Decrease ethinyl estradiol 42%.
contraceptives
Sildenafil Ritonavir increases AUC 11 times.
Desipramine Increase AUC.
Propylene Oral fosamprenavir has propylene

glycol glycol.
AUC, area under the curve.
Suggested Readings
Dut t a SK, Ting CD, Lai LL. St udy of prevalence, severit y, and et iological
f act ors associat ed w it h acut e pancreat it is in pat ient s inf ect ed w it h human
immunodef iciency virus. Am J G ast roent erol 1997; 92: 2044–2048.
Soni N, Pozniak A. Cont inuing HI V t herapy in t he I CU. Crit Care 2001; 5: 247–
248.
> Table of C ontents > Medic ations > 47 - C ons ider the Us e of Fluc onaz ole P r ophylaxis in Intens ive
C ar e P atients w ith S ever e P anc r eatitis , Abdom inal S eps is , or Need for Multiple Abdom inal S ur ger ies
47
Consider the Use of Fluconazole Prophylaxis in
Intensive Care Patients with Severe Pancreatitis,
Abdominal Sepsis, or Need for Multiple
Abdominal Surgeries
Lisa Marcucci MD
There is a grow ing aw areness of t he increasing role of f ungal inf ect ions in
morbidit y and mort alit y of int ensive care unit (I CU) pat ient s. The use of t he
azoles in prophylaxis f or f ungal inf ect ion inimmunocompromised pat ient s is w ell
described and has proven eff icacy. I n addit ion, t here is an increasing body of
w ork t hat azole prophylaxis may be of benef it in cert ain pat ient populat ions as
w ell. I n bot h pat ient populat ions, inf ect ions caused by Candi da species are
t hought t o develop f rom endogenous colonizat ion, yet t he value of f ungal
surveillance cult ures in crit ically ill pat ient s is uncert ain and lacks a high posit ive
predict ive value.
Fungal inf ect ions may develop in up t o 30% t o 35% of pat ient s w it h necrot izing
pancreat it is, w it h Candi da al bi cans being t he most f requent ly isolat ed f ungal
species by f ar. Tw o recent st udies show ed a signif icant decrease in f ungal
inf ect ions in a f luconazole prophylaxis group compared w it h a cont rol group. I n
pat ient s w it h sept ic shock f rom abdominal nonpancreat ic sources, use of empiric
f luconazole show ed a decreased incidence of candidemia and f ungal-relat ed
deat hs in t hree recent st udies, alt hough reduct ion in overall mort alit y w as less
cert ain. Some caut ion has been voiced concerning t he prophylact ic use of
f luconazole in “moderat ely ill” immunocompet ent pat ient s because of t he risk of
developing drug-resist ant (resist ant t o azole drugs) f ungal st rains, especially
Candi da gl abrata, alt hough Sw oboda et al. have report ed no shif t t o nonalbicans
pat hogens w it h a decreased risk of mort alit y using prophylaxis.
O f recent int erest t o crit ical care physicians is t he emerging lit erat ure on t he
ant i-inf lammat ory propert ies of f luconazole. I t appears t hat , separat e f rom it s
ant if ungal propert ies, use of f luconazole may improve out comes in sept ic
pat ient s because of blunt ing of t he syst emic inf lammat ory response t hat t ypically
occurs. I n light of t his eff ect and t he high mort alit y of candidemia, it seems
reasonable t o inst it ut e a short (10- t o 14-day) course of prophylact ic oral azole
t herapy (init ial bolus of 800 mg and t hen a minimum of 200 and pref erably
400 mg/ day of f luconazole) in pat ient s w it h severe pancreat it is, abdominal
perf orat ion, or anast omot ic breakdow n, and ot her abdominal cat ast rophes.
Suggested Readings
De Waele JJ, Vogelaers D, Blot S, et al. Fungal inf ect ions in pat ient s w it h
severe acut e pancreat it is and t he use of prophylact ic t herapy. J I nf ect Dis
2003; 37: 208–213.
He Y M, Lu XS, Ai ZL, et al. Prevent ion and t herapy of f ungal inf ect ion in
severe acut e pancreat it is: a prospect ive clinical st udy. World J G ast roent erol
2003; 9: 2619–2621.
Sw oboda SN, Merz WG , Lipset t PA. Candidemia: t he impact of ant if ungal

prophylaxis in a surgical int ensive care unit . Surg I nf ect 2003; 4: 345–354.
Sypula WT, Kale-Pradhan PB. Therapeut ic dilemma of f luconazole prophylaxis

in int ensive care. Ann Pharmacot her 2002; 36: 155–159.
> Table of C ontents > Medic ations > 48 - Have a High Thr es hold for Adm inis ter ing Vitam in K
Intr avenous ly
48
Have a High Threshold for Administering Vitamin
K Intravenously
Anaphylact oid react ions in pat ient s receiving int ravenously (I V) administ ered
vit amin K have been w idely report ed. A recent review of t he lit erat ure along w it h
t he US Food and Drug Administ rat ion (FDA) adverse drug react ion dat abase
uncovered a t ot al of 155 cases, 27 of w hich w ere f at al and w it h t he t rue number
no doubt being much higher. The manuf act urer has suff icient concern over t he
saf et y of I V administ rat ion of vit amin K t hat it w as volunt arily removed f rom t he
Canadian market . Anaphylact ic react ions and f at alit ies have occurred even w hen
I V vit amin K w as given at low doses and by slow dilut e inf usion. O f t he report ed
155 cases, 21 cases w it h f our f at alit ies occurred in pat ient s w ho received doses
smaller t han 5 mg of I V vit amin K. React ions in pat ient s receiving vit amin K by a
non-I V rout e do occur but are much less common.
Watch Out For

The pat hogenesis of t his react ion is unknow n and may be due t o vasodilat at ion
relat ed t o t he solubilizing vehicle or immune-mediat ed (i. e. , allergic) processes.
The solubilizing agent is polyet hoxylat ed cast or oil (Cremophor EL). Despit e
rumors t o t he cont rary, t here has been no change in t he f ormulat ion of t he
solubilizing agent in t he last several years t hat decreases t he risk of I V use.
O t her drugs using t his agent include paclit axel, cyclosporine, and t eniposide, and
all t hree drugs, w hen administ ered by I V, have been associat ed w it h react ions
including anaphylaxis. The incidence of anaphylaxis af t er I V vit amin K appears t o
be similar t o t hat of ot her drugs know n t o cause anaphylaxis, such as penicillin
or iron dext ran.
Vit amin K is commonly used t o t reat overant icoagulat ion f rom w arf arin, w hich
can be lif e t hreat ening. The American College of Chest Physicians (ACCP)
guidelines provide recommendat ions f or vit amin K use in overant icoagulat ed
pat ient s but only f or I V (despit e t he previously described 155 cases) and oral
administ rat ion. I n addit ion t o I V and oral administ rat ion, vit amin K can be given
subcut aneously. How ever, subcut aneous vit amin K may have less reliable kinet ics
t han t he ot her rout es. I n addit ion, several report s have f ound t hat oral
vit amin K has a more rapid eff ect on low ering t he int ernat ional normalized rat io
(I NR) t han t he subcut aneous rout e.
To re-emphasize, t he use of I V vit amin K should be avoided in almost all pat ient s
w it h overant icoagulat ion and should be reserved f or t hose w it h serious
hemorrhage, inabilit y t o t ake oral vit amin K, and inabilit y t o administ er f resh
f rozen plasma. I f I V vit amin K must be given, it should be administ ered mixed
int o a minibag of 100 mL of D5W or saline and given over a 30-minut e period.
I rrespect ive of t he risk of an anaphylact oid react ion, t here is no recommendat ion
t o pret reat pat ient s w it h st eroids or ant ihist amines bef ore administ ering I V
vit amin K.
Suggested Readings
Crow t her MA, Douket is JD, Schnurr T, et al. O ral vit amin K low ers t he
int ernat ional normalized rat io more rapidly t han subcut aneous vit amin K in t he
t reat ment of w arf arin-associat ed coagulopat hy. A randomized, cont rolled t rial.
Ann I nt ern Med 2002; 137: 251–254.
Fiore LD, Scola MA, Cant illon CE, et al. Anaphylact oid react ions t o vit amin K.
J Thromb Thrombolysis 2001; 11: 175–183.
Riegert -Johnson DL, Volcheck G W. The incidence of anaphylaxis f ollow ing

int ravenous phyt onadione (vit amin K1): a 5-year ret rospect ive review. Ann
Allergy Ast hma I mmunol 2002; 89: 400–406.
Wjasow C, McNamara R. Anaphylaxis af t er low dose int ravenous vit amin K. J

Emerg Med 2003; 24: 169–172.
> Table of C ontents > Medic ations > 49 - D o Not Us e B enz oc aine S pr ay: It Inc r eas es the R is k of
Methem oglobinem ia
49
Do Not Use Benzocaine Spray: It Increases the
Risk of Methemoglobinemia
Kelly G rogan MD
Topical anest het ics are commonly used f or procedures including endot racheal
int ubat ion, endoscopy, bronchoscopy, laryngoscopy, orogast ric t ube placement ,
dent al ext ract ions, and off ice gynecological procedures. Topical anest het ics are
also w idely available in several over-t he-count er medicat ions such as t eet hing
gels and t hroat lozenges. How ever, one t opical anest het ic w hose use should be
avoided is benzocaine spray (aka Hurricaine spray) due t o t he risk of
met hemoglobinemia. Benzocaine-induced met hemoglobinemia is an uncommon
occurrence in clinical pract ice; how ever, know ledge of t his pot ent ially lif e-
t hreat ening condit ion is necessary t o t hose pract it ioners w ho perf orm
procedures requiring local, t opical anest hesia.
Physiology and Pathophysiology

Met hemoglobinemia occurs w hen an imbalance due t o eit her increased
met hemoglobin product ion or decreased met hemoglobin reduct ion is present .
Normal hemoglobin cont ains an iron molecule t hat exist s in t he divalent f errous
st at e (Fe2+ ). Met hemoglobin result s f rom t he conversion of t he iron f errous ion
(Fe 2+ ) t o t he oxidized f erric (Fe3+ ) st at e. The f erric hemes of met hemoglobin are
unable t o bind oxygen. I n addit ion, t he oxygen aff init y of any accompanying
f errous hemes in t he hemoglobin t et ramer is increased. This result s in a “lef t
shif t ” in t he oxygen dissociat ion curve and oxygen delivery t o t he t issues is
impaired (Fig. 49. 1). Theref ore, t he pat ient w it h increased concent rat ions of
met hemoglobin has a f unct ional anemia as t he circulat ing met hemoglobin-
cont aining molecules are unable t o carry oxygen and deliver it t o t he t issues.
Aut o-oxidat ion of hemoglobin t o met hemoglobin occurs spont aneously at a slow
rat e in normal individuals, convert ing 0. 5% t o 3% of t he available hemoglobin t o
met hemoglobin per day. The only physiologically import ant pat hw ay f or reducing
met hemoglobin back t o hemoglobin is t he nicot inamide adenine dinucleot ide
(NADH) reduct ase-dependent react ion cat alyzed by cyt ochrome b5 reduct ase. An
alt ernat ive pat hw ay is an enzyme ut ilizing nicot ine adenine dinucleot ide
phosphat e (NADPH) generat ed by glucose-6-phosphat e
dehydrogenase (G 6PD). There is normally no elect ron carrier present in red

blood cells t o int eract w it h NADPH met hemoglobin reduct ase, making t his
react ion a minor player in t he conversion of met hemoglobin t o hemoglobin.
How ever, it s act ivit y is markedly enhanced by elect ron accept ors or redox dyes,
such as met hylene blue.
FIG URE 49. 1. O xygen Dissociat ion Curve
Most cases of met hemoglobinemia are acquired, result ing f rom increased
met hemoglobin f ormat ion by various exogenous agent s. The mechanism involved
in t he f ormat ion of met hemoglobin is not clear but appears t o be t he direct or
indirect oxidat ion of hemoglobin t o a degree t hat overw helms t he capacit y of t he
reduct ive pat hw ay. Local anest het ics have been implicat ed as a cause of
met hemoglobinemia f or more t han 50 years. Alt hough almost all t opical
anest het ics have been associat ed, benzocaine is t he most commonly implicat ed
agent . O f t he cases report ed, more t han half involved inf ant s and t he elderly.
Cases have been report ed af t er applicat ion t o t he pharyngeal mucosa, rect al
mucosa, vaginal mucosa, and skin. The majorit y of cases of heredit ary
met hemoglobinemia are due t o a def iciency of NADPH met hemoglobin reduct ase.
This aut osomal recessive disease is most common in t he I nuit populat ion and in
Alaskan Nat ive Americans. Hemoglobin M is anot her f orm of congenit al
met hemoglobinemia charact erized by an abnormal hemoglobin molecule.
Met hemoglobinemia should be immediat ely suspect ed in any pat ient w ho has
cent ral cyanosis and a decrease in oxygen sat urat ion t hat develops af t er t he
administ rat ion of benzocaine or ot her t opical anest het ic. Clinical signs and
sympt oms depend on t he level of met hemoglobinemia. Levels great er t han 15%
are associat ed w it h cyanosis w hile headache, let hargy, t achycardia, w eakness,
and dizziness generally present at levels of 20% t o 45%. Dyspnea, cyanosis,
cardiac dysrhyt hmias, heart f ailure, seizures, and coma may occur at levels
exceeding 45%. Met hemoglobin levels above 70% are associat ed w it h a high
mort alit y rat e. The clinical eff ect s may appear earlier and be more severe in
pat ient s w it h underlying anemia or cardiopulmonary disorders.
Diagnosis
Met hemoglobinemia may be clinically suspect ed by t he presence of clinical
“cyanosis” in t he f ace of a normal art erial PO 2 (PaO 2) as obt ained by art erial
blood gases. The blood in met hemoglobinemia has been variously described as
dark red, chocolat e, or brow nish t o blue in color, and, unlike deoxyhemoglobin,
t he color does not change w it h t he addit ion of oxygen. Pulse oximet ry is
inaccurat e in monit oring oxygen sat urat ion in t he presence of
met hemoglobinemia, and it cannot be used t o make t he diagnosis of t he
disorder. Pulse oximet ry is based on diff erent ial light absorpt ion of
oxyhemoglobin and reduced hemoglobin at t w o w avelengt hs of light (t ypically 660
and 940 nm). The rat io of absorbance is used t o provide an est imat e of SpO 2,
calibrat ed so t hat a 1: 1 rat io corresponds t o a sat urat ion of 85%.
Met hemoglobin has an absorbance pat t ern similar t o reduced hemoglobin and
exceeding t hat of oxyhemoglobin at 660 nm, and great er absorbance t han eit her
reduced hemoglobin or oxyhemoglobin at 940 nm. Theref ore, in t he presence of
met hemoglobin, light absorbance is great ly increased at bot h w avelengt hs and
t ends t o drive t he absorbance rat io t ow ard unit y. Based on t hese charact erist ics,
SpO 2 t ends t o be about 85% over a w ide range of pat hologic met hemoglobin
levels. Thus, oxygen sat urat ion measurement by pulse oximet ry can over- or
underest imat e t he degree of oxygenat ion, depending on it s severit y. How ever,
t he presence of met hemoglobin can be suspect ed w hen t he oxygen sat urat ion as
measured by t he pulse oximet ry is signif icant ly less t han t he oxygen sat urat ion
f rom art erial blood gas analysis (sat urat ion gap).
The laborat ory diagnosis of met hemoglobinemia is based upon analysis of it s
absorpt ion spect ra, w hich has peak absorbance at
631 nm. A f resh specimen should alw ays be obt ained as met hemoglobin w ill
increase w it h st orage. The st andard met hod of assaying met hemoglobin ut ilizes
a microprocessor-cont rolled, f ixed w avelengt h co-oximet er. This inst rument
int erpret s all readings in t he 630-nm range as met hemoglobin; t hus f alse
posit ives may occur in t he presence of ot her pigment s including sulf hemoglobin
and met hylene blue.
Treatment
I n acquired met hemoglobinemias, t he off ending agent s should be discont inued. I n
one st udy of 138 cases of acquired met hemoglobinemia, use of dapsone
account ed f or 42% of all aff ect ed pat ient s, w it h a mean level of met hemoglobin
of 7. 6%. The most severe cases w ere seen af t er t he use of 20% benzocaine
spray f or t opical anest hesia (mean peak met hemoglobin level 44%, range 16% t o
60%).
I n lesser degrees of met hemoglobinemia, no t herapy ot her t han discont inuat ion
of t he off ending agent s may be required. I n t he absence of serious underlying
illnesses, met hemoglobin levels less t han 30% usually resolve spont aneously
over 15 t o 20 hours. How ever, t his condit ion can be lif e t hreat ening w hen
met hemoglobin const it ut es more t han 50% of t ot al hemoglobin. Blood
t ransf usions or exchange t ransf usions may be helpf ul in pat ient s w ho are in
shock.
Met hylene blue (met hylt hionine chloride), given int ravenously in a dose of 1 t o 2
mg/ kg over f ive minut es, provides an art if icial elect ron accept or f or t he reduct ion
of met hemoglobin via t he NADPH-dependent pat hw ay, great ly increasing t he
enzymat ic reduct ion of met hemoglobin. Met hylene blue is convert ed t o
leukomet hylene blue by accept ing an elect ron f rom NADPH in t he presence of
NADPH-met hemoglobin reduct ase. Leukomet hylene blue t hen donat es t his
elect ron t o met hemoglobin, result ing in it s conversion t o hemoglobin. Response
is usually rapid and w hile met hylene blue may be redosed in one hour, t his is
f requent ly not necessary. I t is import ant t o not e t hat large doses of met hylene
blue (>7 mg/ kg) can cause dyspnea, chest pain, and hemolysis. Doses exceeding
15 mg/ kg may act ually cause met hemoglobinemia by direct oxidat ion of
hemoglobin. Since co-oximet ry det ect s met hylene blue as met hemoglobin, t his is
not a usef ul met hod t o det ermine t he response of met hemoglobin levels t o
t reat ment w it h met hylene blue.
Finally, pat ient s w it h G 6PD def iciency w ho have decreased product ion of NADPH
w ill not respond t o met hylene blue.
Suggested Readings
Anderson S, Hajduczek J, Barker S. Benzocaine-induced met hemoglobinemia
in an adult : accuracy of pulse oximet ry w it h met hemoglobinemia. Anest h Analg
1988; 67: 1099–1101.
Ash-Bernal R, Wise R, Wright S. Acquired met hemoglobinemia: a

ret rospect ive series of 138 cases at 2 t eaching hospit als. Medicine
2004; 83: 265.
Rodriguez L, Smolik L, Zbehlik A. Benzocaine-induced met hemoglobinemia:

report of a severe react ion and review of t he lit erat ure. Ann Pharmacot her
1994; 28: 643–649.
> Table of C ontents > Medic ations > 50 - K now W hic h W eight to Us e W hen D os ing Medic ations
50
Know Which Weight to Use When Dosing
Medications
Volume of dist ribut ion (Vd) is a nebulous pharmacokinet ic principle t hat
describes t he hypot het ical volume of plasma int o w hich a drug dist ribut es. While
t his is not an act ual physiologic paramet er like cardiac out put or clearance, it is
an ext remely usef ul t ool t hat describes how w idely a drug dist ribut es t hroughout
body f luids and t issues.
Vd is of t en report ed in eit her lit ers or lit ers per kg t o help normalize a part icular
drug's Vd t o a pat ient 's w eight . As t he Vd increases, so does t he drug's abilit y
t o dist ribut e int o various body t issues. The body t issues t hat t he drug dist ribut es
int o w ill be based on drug charact erist ics in addit ion t o t he Vd. This is import ant
t o underst and w hen det ermining t he dose of a medicat ion t o administ er t o a
pat ient . Dose is relat ed t o Vd by t he f ollow ing equat ion:
This show s t hat as Vd increases, t he dose must increase t o achieve t he same

concent rat ion; t he converse is also t rue. As a pat ient 's w eight changes t hrough
his or her st ay in an int ensive care unit (I CU) (early on, w eight gain due t o f luid
resuscit at ion; lat er, w eight loss due t o diuresis), Vd w ill change as w ell and may
aff ect how doses result in t he t herapeut ic levels t hat are achieved. Vd can be
aff ect ed by various physiologic aberrat ions t hat occur in t he I CU pat ient because
of deranged organ f unct ion. These include decreased plasma prot eins (and
plasma prot ein binding), pH diff erences, and f luid balance abnormalit ies.
These issues w it h Vd can be relat ed t o t he w eight changes and t he various
w eight def init ions t hat are used in crit ically ill pat ient s. I n drugs w it h a larger Vd
t hat w ill dist ribut e w idely w it hin t he body, doses based on act ual w eight are
used. To brief ly review, t he various w eight def init ions are
Act ual Body Weight : The act ual w eight of t he pat ient at any given t ime;
includes f luids f or resuscit at ion and f at .
I deal Body Weight (I BW): Weight calculat ed based on height and gender.
I BW (male) = (height in inches >60 × 2. 3) + 50

I BW (f emale) = (height in inches >60 × 2. 3) + 45. 5
TABLE 50-1 GUIDELINE FOR DOSING SOM E COM M ON

DRUGS
WEIGHT TO USE WHEN CALCULAT ING DOSE
ACT UAL
ACT UAL BW
BW NORMAL NO
DRUG/CLASS >150% OF
≤150% PEAK TR
IBW
OF IBW
Adjusted
weight =
Actual 8–10 <2
Aminoglycosides IBW + 0.4
weight mcg/mL mc
(Act BW -
IBW )
Adjusted
weight =
Actual
Heparin IBW + 0.4 n/a n/a
weight
(Act BW -
IBW )
Actual
weight Actual
(use weight (use
Beta-lactams higher higher end of Varies Va
end of normal
normal doses)
doses)
Actual
weight Actual
(use weight (use
Ciprofloxacin higher higher end of
end of normal
normal doses)
doses)
Actual Actual 15
Vancomycin n/a
weight weight mc
Adjusted
weight, but
Actual may need to
Amiodarone n/a n/a
weight give
supplemental
doses
Actual Actual
Drotrecogin Alfa n/a n/a
weight a weight a
a Models what was done in large randomized controlled trials
Adjust ed Body Weight : Weight calculat ed f rom t he diff erence bet w een ideal and
act ual w eight .
Adj BW = I BW + (0. 25 t o 0. 40 [ Act ual BW - I BW] )
As ment ioned earlier, drug dosing is also impact ed by t he physiochemical
propert ies of t he drug—specif ically t he lipophilicit y or hydrophilicit y of t he agent
in quest ion. For example, aminoglycosides are very hydrophilic agent s. I n
morbidly obese pat ient s, dosing should be based on an adjust ed body w eight .
How ever, an act ual w eight should be used t o est imat e doses in a post operat ive
I CU pat ient w ho has large amount s of ext racellular f luid on board. I t should be
not ed t hat medicat ions used as inf usions should be t it rat ed t o eff ect , regardless
of w eight used.
Table 50. 1 is a guide t o help make dosing decisions f or common I CU drugs
based on w eight . As a general rule, in obese pat ient s, t he act ual w eight should
be used f or drugs t hat dist ribut e w ell int o t issues (e. g. , quinolones, vancomycin).
An adjust ed w eight should be used f or medicat ions t hat do not dist ribut e w ell
int o body t issues (e. g. , heparin, aminoglycosides). For medicat ions t hat
dist ribut e int o ext ravascular f luid w ell, like aminoglycosides, any increases in
w eight due t o f luid administ rat ion must be t aken int o considerat ion. As a general
recommendat ion, you can add t hat diff erence in w eight t o t he act ual or adjust ed
w eight t hat w ill be used. For example, imagine t he pat ient w eight s are as
f ollow s:
Act ual: 130 kg

Usual: 120 kg
I deal: 70 kg
The w eight t hat w ould be used f or aminoglycoside dosing f or t his pat ient w ould
be an adjust ed w eight (Adj w eight =70+0. 4 [ 120-70] = 90 kg) w it h t he addit ion of
t he ext ra 10 kg of f luid on board, or 100 kg t ot al. I f you w ant ed t o add
vancomycin t o t his regimen, t hen you w ould just use t he act ual w eight , since
vancomycin has a larger Vd and w ill dist ribut e int o t he f at as w ell as t he f luid on
board.
Suggested Readings
But on ME, Shaw LM, Schent ag JJ, eds. Applied Pharmacokinet ics and
Pharmcodynamics: Principles of Therapeut ic Drug Monit oring. 4t h Ed.
Philadelphia: Lippincot t Williams & Wilkins; 2006: 8–29.
Erst ad BL. Dosing of medicat ions in t he morbidly obese in t he int ensive care
unit set t ing. I nt ensive Care Med 2004; 30: 18–32.
> Table of C ontents > Medic ations > 51 - B ew ar e of Intens ive C ar e Unit Medic ations that c an Inc r eas e
S er um P otas s ium and C aus e Hyper kalem ia
51
Beware of Intensive Care Unit Medications that
can Increase Serum Potassium and Cause
Hyperkalemia
Adam R. Berliner MD
Derek M. Fine MD
Many common medicat ions can raise serum pot assium (K+ ) concent rat ion.
Alt hough t hese eff ect s are usually insignif icant in healt hy pat ient s w it h normal
renal f unct ion, t he synergy bet w een polypharmacy and t he high prevalence of
coexist ing renal dysf unct ion in t he int ensive care unit (I CU) can make t hese
eff ect s very dangerous in crit ically ill pat ient s. Table 51. 1 list s some f requent
off enders in t he I CU set t ing.
Diagnosis of Hyperkalemia
Elect rocardiographic (ECG ) changes in hyperkalemia are st riking if present and
include peaked T-w aves, varying degrees of A-V block, Q RS-int erval w idening,
or diminut ive or lost P-w aves. How ever, t he severely hyperkalemic pat ient (a)
does not progress t hrough t hese ECG changes in a predict able sequence and (b)
may not have any of t hese changes. A benign ECG can move t o vent ricular
f ibrillat ion w it h no int ervening changes. ECG changes alone are not sensit ive
enough t o use t heir absence t o dismiss t he danger of hyperkalemia. Regardless
of ECG changes, a serum K+ of ≥6. 0 meq/ L merit s prompt t herapy t o rapidly
st abilize t he myocardium t o prot ect against arrhyt hmias and t o rapidly low er
ext racellular K+ via shif t ing t ow ard int racellular st ores, w it h an eye t ow ard more
def init ive t herapy t hereaf t er. Values in t he range of 5. 0 t o 5. 9 meq/ L are also
cause f or concern if a t rend of cont inuing increase is expect ed. O t herw ise,
elevat ions t o t his degree (5. 0 t o 5. 9 meq/ L) may be t reat ed w it h def init ive
t herapies alone or simply minimizat ion of off ensive medicat ions or K+ int ake.
Treatment Options
Immedi ate Therapy (f or Cardi ac Myocyte Membrane Stabi l i zati on to Prevent
Arrhythmi a). G enerally reserved f or K+ levels ≥6. 0 meq/ L, int ravenous calcium
salt s (most commonly 10 cc [ one ampule] of 10% calcium gluconat e solut ion) are
usually given concurrent ly w it h t he short -act ing t herapies discussed next . The
eff ect is rapid (onset 1 t o 3 min) but brief (30 t o 60 min).
TABLE 51-1 COM M ON M EDICATIONS THAT CAN

CAUSE HYPERKALEM IA
MECHANISM OF
DRUG
HYPERKALEMIA
Decreased renin release (and

hence diminished aldosterone
Beta-2 blockers action) and impaired cellular
uptake of K+ by sodium-
potassium pump.
Aldosterone receptor inhibitor,

Spironolactone,
down regulates distal nephron
eplerenone
K+ excretion.
Impaired aldosterone
Heparin (IV or SQ),
metabolism leading to
ketoconazole
diminished aldosterone effect.
Angiotensin-
Suppression of aldosterone
converting enzyme
release; decreased glomerular
inhibitors/angiotensin
filtration rate (GFR).
II receptor blockers
Muscle cell depolarization-

Succinylcholine induced K+ “leak” from intra- to
extracellular space.
Nonsteroidal anti-
Decreased renin release;
inflammatory drugs
decreased GFR.
(NSAIDs)
Inhibition of sodium-potassium
Digoxin
pump.
Tacrolimus,
Decreased renin release.
cyclosporine
Blockade of collecting duct

Trimethoprim,
sodium channels, thereby
amiloride,
reducing sodium-for-potassium
triamterene,
exchange and hence K+
pentamidine
excretion.
Short-acti ng Therapy (to “Shi f t” Potassi um f rom Extracel l ul ar to Intracel l ul ar

Si tes). The f oremost “shif t ing” t herapy is int ravenous insulin (usually 10 unit s
coadminist ered w it h 1 ampule I V dext rose unless t he pat ient is already
hyperglycemic), w hich act s via st imulat ion of t he cellular sodium-pot assium
pump. The eff ect begins in 10 t o 20 minut es and last s 4 t o 6 hours. I n
persist ent ly hyperglycemic pat ient s an insulin drip may be required, since
hyperglycemia it self can promot e hyperkalemia. Nebulized bet a-2 agonist s,
usually 10 t o 20 mg albut erol (about f our t o eight t imes t he dose of a common
2. 5-mg nebulizer t reat ment ), can be given as w ell. Last ly, t he role of int ravenous
bicarbonat e in t his set t ing is debat ed. I f severe acidemia is present ,
normalizat ion of blood pH can help reduce hyperkalemia but ot herw ise
bicarbonat e by it self is t hought t o be of lit t le int rinsic benef it in shif t ing K+
int racellularly. I ndeed, t hrough t he low ering of ionized calcium by increasing
serum pH, t he use of bicarbonat e may act ually be det riment al in t he f ace of a
signif icant hyperkalemia. I n chronic dialysis pat ient s in part icular, one st udy
show ed bicarbonat e t o have very lit t le pot assium-low ering eff ect .
Def i ni ti ve Therapy (to Reduce Total Body Potassi um Stores). Sodium

polyst yrene sulf at e (Kayexalat e) is t he most commonly used anion exchange
resin used t o t reat hyperkalemia. G iven orally or as a ret ent ion enema, it
absorbs K+ and ot her luminal cat ions, w hich are t hen lost in st ool. I t carries a
f init e incidence of colonic necrosis, f requent ly in t he cont ext of an elderly pat ient
w it h w idespread at herosclerot ic disease w ho may have mesent eric
hypoperf usion due t o hypot ension or dehydrat ion and impaired gast roint est inal
mot ilit y (usually post operat ive). I t should never be given t o renal t ransplant
pat ient s.
Intravenous l oop di ureti cs at doses suff icient t o induce high urinary f low rat es
w ill clear K+ via t he urine. I n t he presence of severe underlying renal disease,
how ever, t his approach may be unsuccessf ul.
Di al ysi s is t he most rapid met hod of low ering serum K+ , t hough since t he
overw helming majorit y of bodily K+ st ores are int racellular, rebound hyperkalemia
post dialysis can be problemat ic. Rebound hyperkalemia more of t en occurs if
t here is an ongoing K+ source (e. g. , ischemic t issue) or if dialysis is preceded by
aggressive pharmacological shif t ing of K+ t o int racellular st ores.
Other Potassium Fun Facts

Af t er spironolact one w as show n t o reduce morbidit y and mort alit y in severe
heart f ailure, a Canadian st udy show ed a f our- t o f ivef old increase in
spironolact one prescript ions, a f our- t o f ive-f old increase in hospit alizat ion
f or hyperkalemia, and a sixf old increase in hyperkalemia-relat ed mort alit y.
Ext ra caut ion is required w hen considering t he init iat ion of spironolact one in
hospit alized pat ient s.
O t her causes of hyperkalemia in t he I CU include severe acidemia,
rhabdomyolysis, ischemic or gangrenous t issue, and hemolysis.
Think about pseudohyperkalemia—caused by K+ release int o t he already
collect ed blood sample—in condit ions w hen eryt hrocyt e, leukocyt e, or
plat elet numbers are pat hologically high (e. g. , leukemia, polycyt hemia). For
an accurat e assay t he K+ should be rechecked in a serum separat or t ube (or
in a heparinized t ube in t he case of severe t hrombocyt osis).
K + can hide in medicat ion component s (e. g. , penicillin G pot assium has 1. 7
mEq pot assium/ million unit s) or diet ary int ake. For approximat ion w hen
reading labeled f ood ingredient s, 39 mg pot assium is equal t o 1 mEq (in
vit ro).
Suggested Readings
Blumberg A. Eff ect of various t herapeut ic approaches on plasma pot assium
and major regulat ing f act ors in t erminal renal f ailure. Am J Med 1988; 85: 507–
512.
G reenberg A. Hyperkalemia: t reat ment opt ions. Semin Nephrol 1998; 18: 46–
57.
Palmer B. Current concept s: managing hyperkalemia caused by inhibit ors of

t he renninangiot ensin-aldost erone syst em. N Engl J Med 2004; 351: 585–592.
Pit t B, Zannad F, Remme WJ, et al. The eff ect of spironolact one on morbidit y
and mort alit y in pat ient s w it h severe heart f ailure. Randomized Aldact one
Evaluat ion St udy I nvest igat ors. N Engl J Med 1999; 342: 709–717.
> Table of C ontents > Medic ations > 52 - Adm inis ter Ac etaz olam ide ( D iam ox) O n A O ne- Tim e D os e
S c hedule O nly
52
Administer Acetazolamide (Diamox) On A One-
Time Dose Schedule Only
Ying Wei Lum MD
Met abolic alkalosis is usually cat egorized as eit her chloride responsive or
nonchloride responsive. Common causes f or chloride-responsive met abolic
alkalosis in t he int ensive care unit (I CU) set t ing include diuret ic t herapy, vomit ing
(or f rom excessive nasogast ric suct ioning), and diarrhea. Nonchloride causes
include alkali administ rat ion (w it h sodium bicarbonat e or rapid inf usion of more
t han eight unit s of blood t hat cont ain large quant it ies of cit rat e), mineralo-
cort icost eroid excess (cort icost eroid administ rat ion), chronic respirat ory
acidosis (secondary t o permissive hypercapnea f rom mechanical vent ilat ion), and
hypokalemia.
Therapy f or met abolic alkalosis in most pat ient s involves discont inuing t he
causat ive f act or f or alkalosis and undert aking volume and elect rolyt e replet ion.
Pat ient s in an edemat ous st at e, how ever, f ollow a diff erent principle since it may
be desirable t o avoid volume replet ion. Treat ment w it h acet azolamide (Diamox)
may be indicat ed in t his set t ing. Acet azolamide is a carbonic anhydrase inhibit or
t hat act s by promot ing t he renal excret ion of bicarbonat e anions and hence
reversing t he met abolic alkalosis.
Acet azolamide is commonly administ ered bet w een 250 and 500 mg every 24
hours f or 3 t o 4 doses; cont inuous dosing can cause a met abolic acidosis. A
w ide body of evidence t o support a once-daily dosing st rat egy is lacking.
How ever, t here have been report s t hat t he eff ect of acet azolamide last s 24 t o 72
hours af t er single dose administ rat ion. Marik et al. st udied t he eff ect in a
populat ion of 30 mechanically vent ilat ed (most ly surgical) pat ient s and not ed t hat
af t er t he administ rat ion of a single dose of 500 mg of int ravenous
acet azolamide, t here w as a mean reduct ion of 6. 4 mmol/ L of bicarbonat e at 24
hours. The onset of act ion w as w it hin 2 hours and peaked at a mean of 15. 5
hours, alt hough t here w as w ide variat ion.
What not to Do
Healt h care providers w ho administ er acet azolamide should be caut ious about
overcorrect ion of met abolic alkalosis, w hich can in t urn cause a hyperchloremic
acidosis. Anot her common precaut ionary measure is t he monit oring of serum
pot assium since hypokalemia can develop
as a result of increased sodium bicarbonat e delivery t o t he collect ing t ubules,

w hich in t urn can enhance sodium and/ or pot assium channels t o encourage
pot assium excret ion. Most experienced clinicians w ait f or pot assium t o be in t he
high normal range bef ore giving acet azolamide.
Suggested Readings
Marik PE, Kussman BD, Lipman J, et al. Acet azolamide in t he t reat ment of
met abolic alkalosis in crit ically ill pat ient s. Heart Lung 1991; 20: 455–459.
Mazur JE, Devlin JW, Pet ers MJ, et al. Single versus mult iple doses of
acet azolamide f or met abolic alkalosis in crit ically ill medical pat ient s: a
randomized, double-blind t rial. Crit Care Med 1999; 27: 1257–1261.
> Table of C ontents > Medic ations > 53 - C hec k an E lec tr oc ar diogr am for Long Q t Inter val B efor e
Giving Haloper idol
53
Check an Electrocardiogram for Long Qt Interval
Before Giving Haloperidol
Shaytone Nichols MD
Several drugs used in t he int ensive care unit (I CU) have t he pot ent ial t o prolong
t he Q T int erval. I f not correct ed, t his can evolve int o t he malignant arrhyt hmia of
torsades de poi ntes. The Q T int erval is t he t ot al durat ion of depolarizat ion and
repolarizat ion of t he vent ricles; prolonged Q T is due t o lengt hening of t he
repolarizat ion phase and usually due t o changes in pot assium handling by sodium
or pot assium channels. Along Q T can be inherit ed or acquired; t he acquired f orm
is usually induced by drugs or elect rolyt e abnormalit ies like hypokalemia and
hypomagnesaemia.
Drugs commonly used in t he I CU t hat pot ent iat e t his eff ect are haldol,
amiodarone, met oclopramide, ibut ilide, procainamide, azit hromycin,
clarit hromycin, cisapride, eryt hromycin, met hadone, and pent amidine. When
using t hese drugs, an elect rocardiogram (ECG ) should be checked t o ensure a
normal Q T.
What to Do
I f t he pat ient does develop a long Q T int erval, st rong considerat ion should be
given t o st opping t he medicine and correct ing any elect rolyt e abnormalit ies.
Experienced pract it ioners w ill give empiric magnesium in t he set t ing of normal
renal f unct ion. I f t he pat ient develops torsades de poi ntes, overdrive pacing,
isoprot erenol, and magnesium are t he st andard of care. The goal is t o increase
heart rat e t o short en vent ricular repolarizat ion. O verdrive pacings is t o a rat e of
100 t o 120 beat s per minut e eit her t ransvenously or ext ernally. I soprot erenol
may also be eff ect ive t o t his end.
A recent analysis of 250 people w ho had drug-induced long Q T show ed t hat in
t he majorit y of cases, predisposing f act ors can be ident if ied. These f act ors
included inhibit ion of cyt ochrome P450 by ot her concurrent medicines t he pat ient
w as t aking (47%); elect rolyt e derangement s (32%); and concomit ant
administ rat ion of more t han one Q T-prolonging medicine (23%).
O ne import ant drug class t hat f alls int o t he cat egory of pot ent iat ing t he long Q T
int erval via cyt ochrome P450 inhibit ion is t he azole class used t o t reat f ungal
inf ect ions. Prophylaxis f or f ungal inf ect ions is now st andard in many I CU set t ings
and it is import ant t o remember
t hat w hile t his class has not been def init ely linked t o t orsades, it can pot ent iat e
t his eff ect .
Suggested Reading
Roden DM. Drug-induced prolongat ion of t he Q T int erval. N Engl J Med
2004; 350: 1013â 1 022.
> Table of C ontents > Medic ations > 54 - D o not us e Ipr atr opium in Meter - D os e Inhaler For m in
P atients w ith Nut Aller gies
54
Do not use Ipratropium in Meter-Dose Inhaler
Form in Patients with Nut Allergies
I prat ropium bromide is an ant icholinergic agent t hat is commonly used in t he
int ensive care unit (I CU) and in t he t reat ment of ast hma and chronic obst ruct ive
airw ays disease (CO PD). I prat ropium is a compet it ive muscarinic acet ylcholine
recept or ant agonist t hat w hen given int ravenously is most pot ent at t he inhibit ion
of bronchial recept ors. When given by t he inhaled rout e, even in high dosage, t he
syst emic eff ect s, such as salivary, cardiac, ocular, and urinary eff ect s, are
negligible. I t produces a dose-relat ed inhibit ion of bot h subst ance-induced and
exercised-induced bronchoconst rict ion. I t is a pot ent bronchodilat or in pat ient s
w it h CO PD but is less pot ent in pat ient s w it h ast hma t hat is responsive t o bet a-2
agonist s. I n pat ient s w it h ast hma alone, t he combinat ion of iprat ropium and a
bet a-2 agonist may be synergist ic, but of t en bet a-2 agonism may be a good
solo-agent t reat ment . O t her t reat ment s f or CO PD in addit ion t o ant icholinergics
and bet a-2 agonist s are t heophylline and cort icost eroids.
The onset of act ion of iprat ropium w hen given by t he inhaled rout e is slow er t han
t hat of t he bet a-2 agonist s, (around 30 t o 60 minut es) and it s eff ect s last up t o 4
t o 6 hours w it h a maximal eff ect at 1 t o 2 hours. Very lit t le of t he drug is
absorbed f ollow ing inhalat ion or ingest ion so hepat ic or renal impairment has
lit t le eff ect on it s t herapeut ic use and t here is no evidence of t olerance w it h
prolonged use.
There are several concerns t hat must be considered w hen st art ing a pat ient on
iprat ropium. I t should be used w it h caut ion in pat ient s w ho have glaucoma,
diff icult y w it h urinat ion, an allergy t o soy or peanut s, or recreat ional use of bet el
nut s. Pat ient s w it h recreat ional use of bet el nut s are know n t o be more prone t o
allergic react ions. Pat ient s w it h allergies t o nut s or soy may have an allergic
react ion t o t he met ered-dose inhaler (MDI ) preparat ion but not t he nebulized
f orm since t he allergy is t o a compound in t he preparat ion of t he MDI . I f an
allergy is present , pat ient s may experience a const ellat ion of sympt oms
consist ing of chest pain; t achycardia; t rouble breat hing or w heezing; sw elling of
t he f ace, lips, or t ongue; and/ or a rash or hives. The f irst st ep in t herapy is t o
eliminat e t he inst igat ing f act or, namely t he MDI , and t hen use support ive
measures as f or any allergic react ion, depending on severit y.
Suggested Readings
Bet el Nut . Wikipedia. ht t p: / / en. w ikipedia. org/ w iki/ Bet el nut
Facchini G , Ant onicelli L, Cint i B, et al. Paradoxical bronchospasm and

cut aneous rash af t er met ered-dose inhaled bronchodilat ors. Monaldi Arch
Chest Dis 1996; 51: 201â 203.
Micromedex Healt hcare Series on I prat ropium. Thomson Healt hcare.

ht t p: / / w w w. t homsonhc. com/ hcs/ librarian
> Table of C ontents > Medic ations > 55 - B e C autious in Us ing K etor olac in P atients w ith Mar ginal
Ur ine O utput or R enal Func tion
55
Be Cautious in Using Ketorolac in Patients with
Marginal Urine Output or Renal Function
Angela D. Shoher MD
Ket orolac is a nonst eroidal inf lammat ory agent w it h great er syst emic analgesic
propert ies t han ant i-inf lammat ory act ivit y. The onset and eff icacy of it s analgesic
propert ies are claimed t o be comparable w it h t hose of morphine. Ket orolac
w orks by compet it ively inhibit ing t he cyclooxygenase isoenzymes (CO X-1 and
CO X-2). These enzymes cat alyze t he conversion of arachidonic acid t o
prost aglandins. Prost aglandins E and F appear t o be responsible f or sensit izing
pain recept ors, explaining t he analgesic propert ies of t he drug. Bioavailabilit y of
t he drug is 100% af t er oral, int ravenous, or int ramuscular administ rat ion. Food
only decreases t he rat e but not t he ext ent of absorpt ion. Ket orolac is also 99%
bound t o albumin. I t is met abolized by t he liver t o a f orm t hat is less t han 1% as
pot ent as t he original f orm.
What not to Do
The dosing of ket orolac should not exceed a period of 3 days. O ral use is
indicat ed only in combinat ion t herapy w it h int ravenous (I V) or int ramuscular (I M)
f orms and is rarely used in t he int ensive care unit (I CU). I n pat ient s w ho are less
t han 65 years of age, 30 mg can be given I V every 6 hours. Special
considerat ion of ket orolac includes t he renal eff ect s of t he drug. Prost aglandins
in t he kidney mediat e sodium and w at er reabsorpt ion. I n t he set t ing of
decreased volume, prost aglandins help maint ain renal blood f low. Ket orolac
administ rat ion in a pat ient w it h t enuous renal f unct ion t heref ore may lead t o
decreased renal perf usion and acut e renal f ailure. I n pat ient s w it h renal
impairment (serum creat inine >1. 9 mg/ dL), t he dose may need t o be adjust ed.
These pat ient s and t hose w ho are older t han 65 years or w eigh less t han 50 kg
should receive no more t han 15 mg every 6 hours. Ket orolac can also inhibit
plat elet aggregat ion by decreasing levels of plat elet t hromboxane A2, w hich
leads t o an increase in bleeding t ime. The plat elet inhibit ion is reversible w it hin
24 t o 48 hours. O f not e, ket orolac can also cause gast roint est inal side eff ect s
including ulcerat ion relat ed t o t he CO X-1 inhibit ion.
Suggested Readings
Kat zung BG , ed. Basic and Clinical Pharmacology. 9t h Ed. New York:
McG raw -Hill; 2004: 596â 6 24.
Wickersham RM, G remillion S, eds. Drug Fact s and Comparisons Pocket

Version. 10t h Ed. St . Louis: Wolt ers Kluw er Healt h; 2005: 473.
> Table of C ontents > Medic ations > 56 - D o not C r us h S evelam er Hydr oc hlor ide ( R enagel) to P lac e
D ow n a Nas ogas tr ic or Feeding Tube
56
Do not Crush Sevelamer Hydrochloride
(Renagel) to Place Down a Nasogastric or
Feeding Tube
Ashita G oel MD
Hyperphosphat emia is a common met abolic derangement seen in end-st age renal
disease secondary t o t he inabilit y of t he kidney t o excret e serum phosphat e.
High levels of phosphat es can result in secondary hyperparat hyroidism and
release of calcium f rom bone. This loss of calcium f rom bone can result in
ost eodyst rophy and ect opic calcium deposit ion int o sof t t issues including
deposit s in coronary or cerebral art eries.
What not to Do
Sevelamer (Renagel) is an aluminum- and calcium-f ree phosphat e binder used in
t he t reat ment of hyperphosphat emia in end-st age renal disease. Sevelamer is a
cat ionic polymer (poly[ allylamine hydrochloride] ) cross-linked w it h
epichlorohydrin, w hich binds phosphat e anions by ion exchange and hydrogen
binding. Alt hough sevelamer is hydrophilic, it is not absorbed by t he
gast roint est inal t ract . When t he polymer comes in cont act w it h gast ric or
int est inal f luid, it expands t hrough hydrat ion; it can expand up t o eight t imes it s
w eight . This gelat inous f orm binds phosphat e in t he gast roint est inal t ract and
t hen is excret ed in f eces. This polymer cannot be crushed or broken prior t o
ingest ion because an int act gelat inous mat rix is needed f or eff icient phosphat e
binding.
O f int erest t o clinicians is t hat unlike ot her phosphat e binders, sevelamer has t he
unique propert y t o sequest er bile acids in addit ion t o phosphat es. Use of t his
drug has been associat ed w it h low ered low -densit y lipoprot ein (LDL) and t ot al
serum cholest erol levels. Possibly t hrough an improved lipid prof ile and
reduct ions in hypercalcemia, sevelamer has been show n t o st abilize or ret ard
coronary and aort ic calcif icat ion. How ever, clinical correlat ions are unclear. Side
eff ect s of sevelamer include nausea, dyspepsia, const ipat ion, and w orsening
met abolic acidosis. The met abolic acidosis is more pronounced in pat ient s
receiving an acet at e-based dialysat e over bicarbonat e.
Current ly, t he most w idespread met hod of phosphorous reduct ion relies on ot her
phosphat e binders. Calcium salt s such as calcium acet at e and calcium carbonat e
are current ly t he most broadly used
agent s. When compared w it h sevelamer, calcium salt s have similar eff icacy
prof iles in reduct ion of phosphat e levels at signif icant ly low er cost s. How ever,
calcium salt s can result in hypercalcemia (w it h calcium carbonat e being t he most
pronounced off ender), adynamic bone disease, vascular calcif icat ions, and lead
t oxicit ies. O t her met allic salt s can also be used in phosphat e binding. Early on,
aluminum salt s w ere used as phosphat e binders, but t hey w ere f ound t o result in
deposit ion of aluminum in sof t t issue including t he brain. Recent ly, salt s of rare
eart h met als are under invest igat ion f or t heir abilit y t o bind phosphat e.
Lant hanum carbonat e appears t o be t he next promising aluminum and calcium-
f ree phosphat e binder. Clinical t rials have show n minimal gast roint est inal
absorpt ion of t he met al w it h phosphat e-binding eff icacy similar t o t he calcium
salt s. These new binders do not possess any cholest erol-low ering propert ies.
Suggested Reading
Loghman-Adham M. Saf et y of new phosphat e binders f or chronic renal f ailure.
Drug Saf 2003; 26: 1093â 1110.
> Table of C ontents > Medic ations > 57 - R em em ber that Many C om m only us ed Intens ive C ar e Unit
D r ugs S hould not be Us ed in P or phyr ia
57
Remember that Many Commonly used Intensive
Care Unit Drugs Should not be Used in
Porphyria
The porphyrias are a const ellat ion of congenit al and acquired def iciencies in t he
act ivit y of enzymes involved in t he biosynt hesis of heme. Abnormal amount s of
t he heme int ermediat e compounds are f ormed, result ing in t issue accumulat ion
and excessive excret ion. All of t hese int ermediat es in t he heme biosynt het ic
pat hw ay can be t oxic and can result in neurovisceral sympt oms or
phot ocut aneous sympt oms. Some porphyrias cause lif e-t hreat ening neurovisceral
abnormalit ies and are charact erized by t he presence of aminolevulinic acid (ALA)
or porphobilinogen in t he urine. O t hers cause phot osensit ivit y and skin damage
w it h exposure t o ult raviolet light w it h subsequent product ion of t issue-damaging
f ree radicals.
Watch Out For

The diagnosis of each t ype of porphyria may be made by ident if ying t he
met abolit e of excess in t he urine, f eces, or blood. The t w o most common t ypes
are porphyria cut anea t arda and acut e int ermit t ent porphyria. The diagnosis of
porphyria cut anea t arda is st rongly suggest ed by t he presence of blist ering skin
lesions on areas exposed t o ult raviolet light . This condit ion is conf irmed by a
high plasma porphyrin level and uroporphyrin in t he urine. Liver enzymes may be
elevat ed, but liver f unct ion is rarely aff ect ed. Risk f act ors f or developing
porphyria cut anea t arda include excessive alcohol use, hepat it is C, human
immunodef iciency virus (HI V), and iron overload. Treat ment consist s of
phlebot omy and low -dose chloroquine, w hich are almost alw ays eff ect ive. O t her
t han pot ent ial cut aneous scars, t here are no long-t erm sequelae of t he disease.
Care must be t aken t o pad t he skin during hospit al care t o avoid f urt her damage.
Acut e int ermit t ent porphyria is charact erized by neurologic dysf unct ion aff ect ing
t he peripheral, aut onomic, or cent ral nervous syst em. There are many sympt oms
of an acut e exacerbat ion including abdominal pain, nausea, vomit ing,
const ipat ion, diarrhea, ileus, urinary ret ent ion, incont inence, dysuria,
t achycardia, hypert ension, sw eat ing, t remor, and psychosis. Mult iple f act ors may
precipit at e an at t ack including reduced caloric int ake, st ress (including surgery
and illness),
smoking, hepat ocellular carcinoma, and many drugs and chemicals (Table 57. 1).
I nit ial t reat ment consist s of support ive care, w it hdraw al of any pot ent ially
off ending agent s, and appropriat e caloric int ake. I f t his is not eff ect ive, t he
administ rat ion of int ravenous (I V) heme preparat ions such as hemat in or heme
arginat e t o suppress non-specif ic delt a-amino-levulinat e synt hase (ALAS-N)
act ivit y or administ rat ion of cimet idine may be indicat ed. Usually at t acks last
f rom a f ew w eeks t o several mont hs, but low er mot or lesions may be permanent ,
leading t o bulbar paralysis.
TABLE 57-1 PARTIAL LIST OF DRUGS CONSIDERED

UNSAFE IN PORPHYRIA
Amitriptyline Etomidate Oxcarbazepine
Barbiturates Furosemide Phenytoin
Carbamazepine Halothane Pyrazinamide
Chloramphenicol Hydralazine Rifampin
Chlordiazepoxide Hydrochlorothiazide Spironolactone
Co-trimoxazole Hydroxyzine Sulfamethoxazole
Dapsone Ketoconazole Sulfonylureas
Dihydralazine Lidocaine Theophylline
Enalapril Lisinopril Valproic acid

Erythromycin Metoclopramide Verapamil
Ethanol Nifedipine
Suggested Readings
American Porphyria Foundat ion.
ht t p: / / w w w. porphyriaf oundat ion. com/ about _por/ drugs/ drugs02. ht ml.
Braunw ald E, Fauci AS, Kasper DL, et al. , eds. Harrison's Principles of
I nt ernal Medicine. 15t h Ed. New York: McG raw -Hill; 2001: 2264.
Miller RD. Miller's Anest hesia. 6t h Rd. New York: Elsevier; 2005: 1096â 1 097.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 58 - D o not Ins er t, C hange,
or R em ove A C entr al Line W ith the P atient S itting Up
58
Do not Insert, Change, or Remove A Central Line
With the Patient Sitting Up
Aaron Bransky MD
Heidi Frankel MD
The insert ion of cent ral venous cat het ers is an ext remely common procedure in
t he Unit ed St at es, account ing f or more t han 5 million procedures annually. There
are many diff erent complicat ions t hat have been described. O ne of t he more
underrecognized and f eared complicat ions is a venous air embolism. Clinically,
air embolism may present w it h acut e chest pain, hypoxia, dyspnea, hypot ension,
visual changes, or convulsions.
The t w o most common t imes f or air embolism t o develop are upon placement or
removal of a cent ral venous cat het er, including inst ances of int ent ional or
inadvert ent removal by t he pat ient . Upon placement of a cent ral venous cat het er,
proper care should be t aken t o prevent each lumen of t he cat het er f rom
connect ing w it h at mospheric air w it hout a column of f luid w it hin it . Furt hermore,
t o increase cent ral venous pressure, t he cat het er should be placed w it h t he
pat ient in a Trendelenburg posit ion. Removal of a cat het er should also be done
w it h caut ion, part icularly in a cachect ic pat ient or if t he cat het er has been in
place f or a sust ained t ime and a f ibrous sheat h has developed. The pat ient
should be placed in a Trendelenburg posit ion w it h his or her breat h held. Upon
removal of t he cat het er, an occlusive dressing should be placed immediat ely. The
risk of deat h is relat ed t o bot h t he volume of t he air embolus and t he rat e at
w hich it ent ers. I n humans, t he minimum inject ion rat e and volume are 100
mL/ sec and 300 t o 500 mL, respect ively.
The act ual physiological mechanism t hat causes t he hemodynamic inst abilit y f rom
an air embolus is a right vent ricular out let obst ruct ion caused by t he slurry of
churned air embolus w it h blood. Because gases rise, t he pat ient should be
placed in a posit ion w here t he gas w ill exit t he vent ricle or prevent it f rom
ent ering t he vent ricle. This is best accomplished by t he reverse Trendelenburg
posit ion w it h t he lef t side dow n. This places t he right at rium in t he least
dependent posit ion, causing air t o ascend t o t his posit ion. O nce t he pat ient is in
t his posit ion, if t he cent ral venous cat het er remains in place, an at t empt can be
made t o aspirat e t he air, alt hough t his is rarely successf ul. O t her proposed
t herapies include hyperbaric oxygen t hat addresses
t he neurological dist urbances t hat occur as a result of a pat ent f oramen ovale
(present in up t o 30% of t he populat ion) or a shunt . Hyperbaric oxygen is
successf ul if ut ilized early af t er recognit ion of neurological sympt oms.
Suggested Readings
O rebaugh SL. Venous air embolism: clinical and experiment al considerat ions.
Crit Care Med 1992; 20: 1169–1177.
Pronovost PJ, Wu AW, Sext on JB. Acut e decompensat ion af t er removing a

cent ral line: pract ical approaches t o increasing saf et y in t he int ensive care
unit . Ann I nt ern Med 2004; 140: 1025–1033.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 59 - Avoid P lac em ent of
C entr al Ac c es s In The R ight Inter nal Jugular in C ar diac Tr ans plant P atients if P os s ible
59
Avoid Placement of Central Access In The Right
Internal Jugular in Cardiac Transplant Patients if
Possible
The out come of cardiac t ransplant at ion has improved dramat ically over t he past
several decades, in large part because of improvement s in immunosuppression
and post operat ive management . During t he f irst 2 t o 3 w eeks af t er
t ransplant at ion, pat ient s are closely monit ored f or evidence of reject ion. This
includes a percut aneous t ransvenous endomyocardial biopsy t o assess f or
hist ological signs of organ reject ion and t o adjust immunosuppressive
medicat ions if necessary. The f irst cardiac biopsy is usually perf ormed 2 w eeks
post t ransplant , unless earlier signs of reject ion occur. The pref erred sit e f or
vascular access t o perf orm t his procedure is t hrough t he right int ernal jugular
vein, as t he t ransvenous biot ome used t o t ake t he biopsies does not need a
guiding cat het er or sheat h w hen insert ed t hrough t his approach. Moreover, right -
sided cardiac biopsies are associat ed w it h f ew er complicat ions.
I n order t o preserve t he right int ernal jugular sit e f or cardiac biopsy in heart
t ransplant pat ient s, caregivers should give considerat ion t o placing cent ral
venous cat het ers elsew here in cardiac t ransplant pat ient s. Cat het er-relat ed
venous t hrombosis may occur in up t o 22% of cat het erized pat ient s and t he risk
of clot f ormat ion is higher w it h int ernal jugular insert ion as compared w it h
subclavian vein insert ion. The pref erred sit e f or cent ral venous access in t hese
pat ient s is t he subclavian vein or t he lef t int ernal jugular vein. The f emoral vein
may also be used f or vascular access if necessary, but it is associat ed w it h t he
highest risk of overall inf ect ious complicat ions.
Suggested Readings
Cot t s WG , Johnson MR. The challenge of reject ion and cardiac allograf t
vasculopat hy. Heart Fail Rev 2001; 6: 227–240.
McG ee DC, G ould MK. Prevent ing complicat ions of cent ral venous
cat het erizat ion. N Engl J Med 2003; 348: 1123–1133.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 60 - P lac e the Tip of a
C entr al Venous C atheter at the Junc tion of the S uper ior Vena C ava and Atr ium
60
Place the Tip of a Central Venous Catheter at the
Junction of the Superior Vena Cava and Atrium
Jayme E. Locke MD
Cent ral venous access is of t en necessary in order t o adequat ely resuscit at e a
pat ient w it h large volumes of f luid or blood product s, monit or int ravascular
volume, perf orm hemodialysis, administ er parent eral nut rit ion, and administ er
ant ibiot ics. I nt ernal jugular and subclavian veins are most commonly used. When
access is obt ained via t he int ernal jugular (I J) vein, placement of t he cat het er on
t he pat ient 's right side is best (except in cardiac t ransplant pat ient s; see
Chapt er 59) secondary t o t he vein's superf icial posit ion and larger size in t hat
locat ion. O verall, t he I J vein is t he pref erred sit e f or dialysis access, allow s f or
easier cont rol of art erial punct ure, and is less likely t o cause cent ral st enosis.
The subclavian vein is associat ed w it h t he low est risk f or inf ect ion, but it does
come w it h a higher risk f or pneumot horax and cent ral st enosis. Femoral venous
access should be avoided unless no ot her access is available or it is an
emergency sit uat ion. Placement of cent ral access in t he groin has t he highest
risk of inf ect ion and deep venous t hrombosis.
Cent ral venous cat het ers can be cat egorized as eit her nont unneled or t unneled.
Nont unneled cat het ers can be placed and removed at t he pat ient 's bedside, can
be easily exchanged over a guidew ire, and are inexpensive. Examples of
nont unneled cat het ers include single-, double-, and t riple-lumen cat het ers, and
Cordis, Hohn, and Shiley cat het ers. Some nont unneled cat het ers such as t he
Hohn can be used f or up t o 6 mont hs. Hohn cent ral lines have a silver-
impregnat ed gelat in cuff t hat lies just beneat h t he skin surf ace at t he punct ure
sit e, w hich serves as a barrier t o inf ect ion.
I n cont rast , t unneled cat het ers t end t o be f or long-t erm use. These cat het ers are
generally t unneled f rom t he chest w all t o t he subclavian or int ernal jugular vein.
They are double cuff ed and designed t o induce scar f ormat ion, allow ing t hem t o
remain in place indef init ely. Examples include Hickman, Broviac (a small cat het er
used in children), and G roshong cat het ers (does not require heparin f lushes;
ideal f or pat ient s w it h allergies t o heparin). Specialized t unneled cat het ers,
called implant ed venous port s, are best used w hen chronic t herapy is necessary
and only int ermit t ent access is needed. The port s have
reservoirs t hat are locat ed inf raclavicular in a subcut aneous pocket . Accessing
port s requires t he special noncoring Huber needle. Mediport , I nf usaport , and
Port -a-Cat h are examples of specif ic implant ed venous port s.
What to Do
Regardless of t he t ype of cent ral venous access, proper posit ioning is
imperat ive. Cent ral cat het ers placed in eit her t he subclavian or int ernal jugular
veins should be posit ioned w it h t he t ip at t he cavoat rial junct ion. The cavoat rial
junct ion is locat ed approximat ely 5 cm below t he t racheobronchial angle, a
reliable f luoroscopic landmark. This dist ance is reproducible in all pat ient s
independent of gender and body habit us. Most cent ral venous cat het ers are
manuf act ured w it h marks denot ing t he lengt h of t he cat het er. I n general t o reach
t he cavoat rial junct ion, cat het ers should be placed at t he f ollow ing lengt hs: RI J ~
12 cm, LI J ~ 15 cm, RSC ~ 15 cm, LSC ~ 18 cm. Proper placement of cat het ers
in bot h t he int ernal jugular and subclavian veins can be conf irmed using a chest
x-ray or w it h f luoroscopy. Cent ral cat het ers placed in t he f emoral vein should
have t he t ip locat ed at t he conf luence of t he lef t and right iliac veins. The t hird
lumbar vert ebral body approximat es t his locat ion, w hich can be conf irmed on
pelvic x-ray. I mproperly placed t ips of cat het ers can creat e signif icant morbidit y
and in some cases mort alit y. Specif ically, cat het er t ips locat ed in t he right at rium
or right vent ricle can cause arrhyt hmias or perf orat e t he heart , leading t o
t amponade and deat h. Cat het er t ips placed t oo proximally, in t he subclavian vein
or brachiocephalic vein, are associat ed w it h higher rat es of t hrombus f ormat ion
and cent ral st enosis.
Suggested Reading
Klingensmit h ME, ed. The Washingt on Manual of Surgery. 4t h ed.
Philadelphia: Lippincot t Williams & Wilkins; 2005: 133â 1 39.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 61 - D o not R em ove the
Intr avenous C atheter Us ed for P las m apher es is Im m ediately after the Las t Tr eatm ent
61
Do not Remove the Intravenous Catheter Used
for Plasmapheresis Immediately after the Last
Treatment
Kelly O lino MD
Plasmapheresis is perf ormed t hrough eit her cent rif ugat ion or membrane plasma
separat ion. Cent rif ugat ion t echnique is t he same t echnology used by blood banks
t o process blood and can be done w it h eit her int ermit t ent or cont inuous f low.
Blood component s are separat ed out according t o densit y, and plasma or cells
can be removed. I nt ermit t ent f low requires only one venipunct ure sit e but can be
associat ed w it h more hypot ension since larger volume shif t s are needed.
Cont inuous f low requires t w o venous access sit es or a double-lumen cat het er
capable of handling high f low st at es. Membrane f ilt rat ion t echnique can be used
w it h hemodialysis equipment and uses a special f ilt er t hat allow s only plasma t o
pass t hrough pores. Diff erent f ilt ers or hemadsorpt ion columns can be used
depending on t he indicat ion.
A plasmapheresis t reat ment session usually removes 1 t o 1. 5 plasma volumes.
Aut oant ibodies, immune complexes, t oxins or subst ances bound t o prot eins,
immunoglobulins, t hrombot ic f act ors, coagulat ion f act ors, lipoprot eins, and ot her
immunologic f act ors are removed during plasmapheresis f or a number of
diff erent immunologic, hemat ologic, neurologic, renal, rheumat ologic, and
met abolic disorders. Replacement f luid usually consist s of 60% t o 80% albumin
w it h t he rest of t he solut ion being normal saline. How ever, f resh f rozen plasma
can be used as t he replacement solut ion w hen t here is a concern f or bleeding or
in t he t reat ment of cert ain condit ions such as t hrombot ic t hrombocyt openic
purpura or hemolyt ic uremic syndrome. Ant icoagulat ion is required f or t hese
t reat ment s w it h cit rat e used f or cent rif ugat ion, w hile unf ract ionat ed heparin is
used f or membrane f ilt rat ion.
There is varying evidence f or t he clinical use of plasmapheresis, but according t o
t he American Associat ion of Blood Banks hemapheresis commit t ee, it is st andard
or support ive t herapy f or t he f ollow ing: ant iglomerular basement membrane
disease, rapidly progressive glomerulonephrit is, f amilial hypercholest erolemia,
ABO -incompat ible bone marrow t ransplant , t hrombot ic t hrombocyt openic
purpurahemolyt ic uremic syndrome, post t ransf usion purpura, acut e or chronic
inf lammat ory demyelinat ing polyneuropat hy, G uillain-Barré,
myast henia gravis, cold agglut in disease, cryoglobinemia, rheumat oid art hrit is,
myeloma kidney, Lambert -Eat on syndrome, Sydenham chorea, and pediat ric
aut oimmune disorders associat ed w it h st rep (PANDAS).
What not to Do
The vascular access cat het er used f or plasmapheresis should not be removed
immediat ely af t er plasmapheresis has ended, especially in cases w here albumin
or colloid are used as replacement f luids because of t he deplet ion of coagulat ion
f act ors. The short est half -lif e of t he f act ors is VI I , w hich is bet w een 1. 5 t o 6
hours. Fact ors VI I I and V have a half -lif e of around 12 hours; I X, about 24 hours;
X, 36 hours; XI I , 2 days; and XI , prot hrombin, and f ibrinogen, 3 or more days.
The prot hrombin and act ivat ed part ial t hromboplast in t imes increase af t er a
single plasmapheresis exchange and ret urn t o normal af t er 4 hours. How ever, t he
t ime t o ret urn t o normal levels is dependent on how many plasma volumes are
cycled and how many t reat ment s t he pat ient has undergone. I t is best t o w ait 1
t o 2 days bef ore removing t he cat het er used f or plasmapheresis w henever
possible and t o hold pressure up t o 30 t o 40 minut es, especially in cases w hen a
cat het er needs t o be removed more prompt ly.
Suggested Readings
Clark WF, Rock G A, Buskard N, et al. Therapeut ic plasma exchange: an
updat e f rom t he Canadian Pheresis group. Ann I nt Med 1999; 131: 453.
Hoff man R, ed. Hemat ology Basic Principles and Pract ice. 4t h Ed. New York:
Churchill Livingst one; 2004: 2481–2485.
Kaplan AA, Joy Fridey. Complicat ions of t herapeut ic plasma exchange.

Prescript ion and t echniques of t herapeut ic plasma exchange. Plasmapheresis
w it h hemodialysis equipment . I ndicat ions f or t herapeut ic plasma exchange. Up
To Dat e; O ct ober 2005. ht t p: / / w w w. upt odat e. com/
Smit h JW, Weinst ein R. The AABB Hemapheresis Commit t ee KL. Therapeut ic
apheresis: a summary of current indicat ion cat egories endorsed by t he AABB
and t he American Societ y f or Apheresis. Transf usion 2003; 43: 820.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 62 - Never Thr ead a Tr iple-
Lum en thr ough a C or dis
62
Never Thread a Triple-Lumen through a Cordis
I nt roducer cat het ers (i. e. , Cordis cat het ers) are used f or t he placement of
devices, such as pulmonary art ery cat het ers (PACs) and cert ain cardiac pacing
devices as w ell as w hen ext remely large volume resuscit at ions are needed or
ant icipat ed. Since PACs and pacers may need f requent manipulat ion of t heir
posit ion, t hey are placed t hrough t he hemost at ic valve of t he Cordis, w hich holds
t hem in posit ion and prevent s back leakage of blood. Diff erent devices require
and are mat ched t o diff erent sizes of Cordis cat het ers and it is imperat ive t hat
t he device is appropriat ely mat ched t o t he Cordis diamet er. O ccasionally,
pract it ioners w ill place a st andard cent ral line cat het er t hrough t he Cordis w hen
one of t he af orement ioned f unct ions is not applicable. Despit e it seeming
reasonable t o do so, st andard single-, double-, and t riplelumen cent ral venous
access cat het ers should not be placed t hrough a Cordis f or several reasons.
Watch Out For

First , st andard single-, double-, and t riple-lumen cent ral venous access
cat het ers are not designed t o f it t he cordis and so t here is t he risk of back
leakage of blood. There is also t he risk of air leakage around t he hemost at ic
seal, leading t o air emboli. Addit ionally, pulmonary art ery cat het ers and pacers
have sheat hs t hat are designed t o cover t hem so t hat w hen t heir posit ion is
manipulat ed, t he cat het er remains st erile so as not t o int roduce microorganisms
int o t he bloodst ream. Cent ral line cat het ers are not designed f or use w it h t hese
sheat hs. The port in t he cordis needs t o be kept covered and st erile w it h t he
sheat h or an appropriat e locking cap. Cent ral line cat het ers are not conf igured in
t his manner and should not be jerry-rigged w it h one. Placing a st andard cent ral
line cat het er w it h any number of lumens t hrough a Cordis creat es an inf ect ion
risk.
Thirdly, cent ral line cat het ers are designed f or securing t o t he skin w it h sut ures
using a hub and/ or clamp. Placing t hem t hrough t he Cordis def eat s t his and
leaves t he cent ral line cat het er at risk f or accident al dislodgment . This may
leave t he pat ient in a sit uat ion w here he or she is not receiving a crucial
medicat ion such as an inot rope
or vasopressor and t he hemost at ic valve port is lef t exposed t o t he environment ,

risking inf ect ion.
Last ly, a Cordis should not be lef t in any longer t han necessary. They are
t hrombogenic and lead t o cent ral venous clot t ing, obst ruct ion, and emboli. The
longer it is in and t he larger t he diamet er, t he great er t he risk of t hrombosis.
O nce t he Cordis is no longer needed, it should be rew ired t o a st andard cent ral
line or removed ent irely.
I f it is desirable t o leave t he Cordis in af t er t he PAC or pacer has been
discont inued, or you ant icipat e short -t erm f ut ure need f or t he Cordis, one of t w o
t hings must be done. I t should eit her be capped w it h t he appropriat e locking
cap, or if a second int ravenous access point in addit ion t o t he sideport of t he
Cordis is needed, t here are specially designed single-lumen int ravenous
cat het ers (SLI Cs) t hat go t hrough and lock on t he Cordis. These are t he only
kind of cent ral cat het ers ot her t han pulmonary art ery cat het ers or pacing devices
t hat are accept able t o place t hrough a Cordis. When considering t he use of an
SLI C, one must recognize t hat a Cordis is f elt t o be more t hrombogenic t han a
regular cent ral line w hen t he decision is made t o maint ain t he Cordis f or an
ext ended period of t ime.
Suggested Readings
Hall JB, Schmidt G A, Wood LDH, eds. Principles of Crit ical Care. 2nd Ed.
New York: McG raw -Hill; 1998: 308â 3 22.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 63 - B e Metic ulous in
Tec hnique w hen Ins er ting and C ar ing for C entr al Venous Ac c es s C atheter s in the Intens ive C ar e Unit to
Low er the Inc idenc e of Infec tion
63
Be Meticulous in Technique when Inserting and
Caring for Central Venous Access Catheters in
the Intensive Care Unit to Lower the Incidence of
Infection
Lisa Marcucci MD
Cent ral venous cat het ers are associat ed w it h a signif icant rat e of bloodst ream
inf ect ions and an est imat ed 10% mort alit y in int ensive care pat ient s. Several
st rat egies have been show n t o be eff ect ive in reducing t hese inf ect ions and are
summarized here.
Use Full-barrier Protection When Inserting Central

Venous Catheters
This is t he most bot hersome prot ocol f or most clinicians but should be employed
except in t he most emergent sit uat ions. Full-barrier prot ect ion consist s of t he
surgeon and all assist ant s being f ully gow ned (w it h t he t ies f ast ened as in t he
operat ing room), masked, and gloved w it h st erile gloves and t he pat ient f ully
draped w it h st erile drapes. I f t he procedure is a change of cat het er over a w ire,
t he surgeon should change gloves bet w een removing t he old cat het er and
insert ing t he new cat het er.
Use Chlorhexidine for Skin Prep

Chlorhexidine has a decided advant age over aqueous iodine-based preps in
reducing inf ect ion, and it s use should be adopt ed rout inely w it h ext reme care t o
avoid cont act w it h t he eyes and t he ext ernal ear canal. The prep should be
applied via a concent rically larger circular mot ion f or at least 20 seconds and
should be allow ed t o dry w it hout blot t ing or f anning. Alcohol-based iodine preps
are f elt t o be less eff ect ive.
Use a Sterile Dressing on the Insertion Site
I mmediat ely af t er cat het er insert ion, t he sit e should be prot ect ed w it h a
t ransparent st erile dressing w hile t he insert ion sit e is st ill st erile. The st erile
bandage should not be placed if t he sit e is st ill oozing, and a st erile gauze
should be used t o remove blood (t he ideal cult ure
medium) bef ore t he st erile dressing is placed. Topical ant ibiot ic oint ment has not
been show n t o reduce inf ect ions and should not be used. Ant isept ic impregnat ed
dressings and disks are promising adjunct s t o exit -sit e care.
Use the Catheter w ith the Few est Number of Lumens

Possible and Consider Treated Catheters
The quest ion of w het her t riple-lumen cat het ers have higher inf ect ion rat es
compared w it h single-lumen cat het ers is st ill somew hat uncert ain, w it h t he
majorit y of st udies show ing higher inf ect ion rat es f or t riple-lumen cat het ers.
Cont radict ing t his are a f ew st udies and one recent met a-analysis t hat f ound
t here is no st rong evidence f or an increase in inf ect ions in t riple-lumen cat het ers.
This may be because t he pat ient populat ion st udied in t he lit erat ure is
het erogeneous in t he f ollow ing variables—pat ient locat ion (int ensive care unit
[ I CU] , inpat ient , mixed), pulmonary art ery cat het ers allow ed, blood draw ing
allow ed t hrough cat het ers, t ypes of cat het ers (ant ibiot ic impregnat ed, number of
lumens), sit es of cat het ers (peripherally insert ed, cent rally insert ed), port s
included or not , t unneled and/ or cuff ed cat het ers included or not , and so on.
I n addit ion, cat het ers coat ed w it h a st erilant (e. g. , silver) or an ant ibiot ic (e. g. ,
rif ampin) are also eff ect ive at decreasing inf ect ions; cost -eff ect iveness w ill vary
w it h an inst it ut ion's inf ect ion rat e.
Do Not Use Stopcocks on the Lumen Ports

Use of st opcocks f unct ionally increases t he number of port als in a cent ral venous
cat het er, w it h an at t endant increase in cont aminat ion, and t heref ore should be
avoided. Cat het er hubs t hat incorporat e an ant isept ic barrier should be used
w henever possible.
Suggested Readings
Chaiyakunapruk N, Veenst ra DL, Lipsky BA, et al. Vascular sit e care: t he
clinical and economic benef it s of chlorhexidine gluconat e compared w it h
povidone iodine. Clin I nf ect Dis 2003; 37: 764–771.
Dezf ulian C, Lavelle J, Nallamot hu BK, et al. Rat es of inf ect ion f or single-
lumen versus mult ilumen cent ral venous cat het ers: a met a-analysis. Crit Care
Med 2003; 31: 2385– 2390.
O ’G rady NP, Alexander M, Dellinger EP, et al. G uidelines f or t he prevent ion of

int ravascular cat het er-relat ed inf ect ions. Pediat rics 2002; 110: e51.
Raad I I , Flohn DC, G ilbreat h BJ, et al. Prevent ion of cent ral venous cat het er-
relat ed inf ect ions by using maximal st erile barrier precaut ions during
insert ion. I nf ect Cont rol Hosp Epidemiol 1994; 231–238.
Sit ges-Serra A, Hernandez R, Maest ro S, et al. Prevent ion of cat het er sepsis:
t he hub. Nut rit ion 1997; 13: 30S–35S.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 64 - In a P atient w ith a
P r evious ly P lac ed Vena C ava Filter, do not Us e the J- tip on the Guidew ir e W hen Us ing the S eldinger
Tec hnique to P lac e a C entr al Venous C atheter
64
In a Patient with a Previously Placed Vena Cava
Filter, do not Use the J-tip on the Guidewire
When Using the Seldinger Technique to Place a
Central Venous Catheter
Lisa Marcucci MD
Awori J. Hayanga
Vena cava f ilt ers are placed f or t reat ment of deep vein t hrombosis and/ or
pulmonary embolism. Complicat ions arising f rom t he placement of t hese f ilt ers
include migrat ion, dislodgment , vena cava penet rat ion, and vena cava
t hrombosis. O ne addit ional complicat ion t hat has been increasing in f requency
but is complet ely prevent able is t he ensnarement of t he f ilt er by guidew ires
being used t o insert cent ral venous cat het ers f rom t he subclavian, jugular, and
f emoral approaches.
There are a variet y of vena cava f ilt ers available in t he Unit ed St at es. Some of
t he more commonly used f ilt ers include t he G reen f ield (bot h t it anium and
st ainless st eel models), Simon Nit inol, Bird's Nest , Vena Tech LG M, Vena Tech
TrapEase, and G unt her Tulip f ilt ers. Despit e design diff erences, t hey are all
similar in t hat t hey allow venous f low t hrough t he f ilt er w hile capt uring emboli via
t he radiat ing st rut s of t he f ilt er.
There are various guidew ires t hat are used during t he Seldinger t echnique,
including st raight guidew ires and t he more commonly encount ered 1. 5-, 3-, and
15-mm J-t ip guidew ires packaged in cent ral venous cat het er insert ion kit s (t he
number describes t he radius of t he curve of t he J-t ip). I n vit ro st udies and case
report s show t hat ensnarement of t he guidew ire in t he f ilt er st rut s is possible
w it h all diamet ers of t he J-t ip guidew ire used in placing cent ral venous cat het ers
and does not occur w it h st raight guidew ires. The f ilt ers w it h t he highest
likelihood of ensnaring t he J-t ip guidew ires are t he G reen f ield and Vena Tech
f ilt ers w it h only t he G unt her Tulip report ed as not ent rapping t he J-t ip guidew ire.
Ensnarement of an inf erior vena cava (I VC) f ilt er occurs w hen t he curved end of
a J-t ip guidew ire (see Fig. 64. 1) is pushed t hrough a st rut opening w it h
subsequent hooking of t he w ire ont o t he st rut w hen t he cat het er is pulled back.
This is not ed clinically as t he w ire becoming “st uck” during w ire w it hdraw al. I f
even slight resist ance t o removing t he w ire is not ed, f urt her at t empt s t o remove
t he w ire must be halt ed immediat ely. At t empt ing t o f ree t he guidew ire w it h f orce
can cause
shearing of t he guidew ire or dislodgment of t he f ilt er w it h pot ent ially

cat ast rophic complicat ions including f ilt er caval disrupt ion, perf orat ion,
arrhyt hmias, cardiac t amponade, and deat h. Alt hough some clinicians advocat e
at t empt s t o f ree t he w ire at bedside by pushing t he w ire caudally and applying
t orque t o rot at e t he J-t ip aw ay f rom t he st rut , a more caut ious approach
demands urgent consult at ion w it h int ervent ional radiology. Techniques used t o
remove ensnared guidew ires include using f luoroscopic visualizat ion and
placement of a vascular sheat h and snares t o w ork t he guidew ire f ree.
FIG URE 64. 1. St raight and curved J-t ip double f lexible t ip w ire guide.
What to Do
To help prevent ensnarement of a guidew ire, all pat ient s or pat ient f amilies
should be queried about t he presence of a cava f ilt er bef ore elect ive insert ion of
a cent ral venous cat het er. Wit h an increasing proport ion of f ilt ers placed
percut aneously, f ilt er placement may not be elicit ed as part of a “past surgical
hist ory. ” I n t he event of an elect ive procedure in a pat ient w here t he hist ory
cannot be obt ained, a plain radiograph w ill show t he presence of t his device.
Because pat ient s usually do not know w hat model of vena cava f ilt er t hey have
received, t hey should be specif ically asked about t he placement of a G reen
f ieldd, f ilt er, or birdcage in t he vena cava. ” I f t hey answ er in t he aff irmat ive, a
st raight guidew ire (or t he st raight end of t he J-t ip guidew ire) should be used t o
place t he cent ral venous cat het er. Similarly, in emergent cent ral venous cat het er
insert ions, a st raight guidew ire should be considered t o prevent ensnarement .
Suggested Readings
Dardik A, Campbell KA, Yeo CJ, et al. Vena cava f ilt er ensnarement and
delayed migrat ion: an unusual series of cases. J Vasc Surg 1997; 26: 869–874.
Munir MA, Chien SQ . An in sit u t echnique t o ret rieve an ent rapped J-t ip
guidew ire f rom an inf erior vena cava f ilt er. Anest h Analg 2002; 95: 308–309.
St avropoulus SW, I t kin M, Trerot ola SO . I n vit ro st udy of guide w ire

ent rapment in current ly available inf erior vena cava f ilt ers. J Vasc I nt erv
Radiol 2003; 14: 905–910.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 65 - D o not “ W hip the Tip”
w hen Tes ting a P ulm onar y C atheter befor e Ins er tion
65
Do not “Whip the Tip” when Testing a
Pulmonary Catheter before Insertion
Prior t o placement of a pulmonary art ery (PA) cat het er, t he balloon should be
t est ed, all port s should be f lushed, t he syst em should be primed w it h saline, and
t he equipment calibrat ed or zeroed w it h t he syst em open t o air. Flushing and
priming t he t ubing removes air bubbles t hat can dampen t he w avef orm. Leveling
should be done w it h t he pat ient lying f lat and w it h t he cat het er held at t he level
of t he of t he pat ient 's at rium. The dist al port should be connect ed t o t he
pressure t ransducer, w hich produces t he charact erist ic placement w avef orm on
t he monit or (Fig. 65. 1). Prior t o f loat ing t he cat het er, t he t ransducer connect ions
should be checked by placing a f inger over t he dist al t ip of t he cat het er. This
should cause a rise in pressure of t he PA w avef orm only. Some people shake t he
t ip t o elicit a w avef orm, but t his is t o be avoided. “Whipping t he t ip” w ill cause
all pressures, including t he cent ral venous pressure (CVP), t o rise and w ill not
provide specif ic inf ormat ion about t ransducer connect ions. There is also t he
t heoret ical risk of damage t o t he sensing mechanism and cont aminat ion of t he
cat het er t ip.
FIG URE 65. 1. Pulmonary art ery t racing. (Reused w it h permission f rom Chen
H, Sonnenday CJ, Lillemoe KD, eds. Manual of Common Bedside Surgical
Procedures. 2nd Ed. Philadelphia: Lippincot t Williams & Wilkins; 2000: 100.)
Alt ernat ively, t he pract it ioner may check t he PA cat het er bef ore insert ion by
holding t he cat het er st ill w it h one hand posit ioned at t he 30-cm mark (CVP port ).
I f t he t ip is slow ly raised, t he PA pressure w avef orm only should change if t he
syst em has been assembled correct ly.
Selected Readings
Pulmonary Art ery Cat het er Educat ion Project . ht t p: / / w w w. pacep. org
Summerhill EM, Baram M. Principles of pulmonary art ery cat het erizat ion in t he
crit ically ill. Lung 2005; 183: 209–219.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 66 - R em ove K inked or
C oiled P ulm onar y Ar ter y C atheter s
66
Remove Kinked or Coiled Pulmonary Artery
Catheters
Hari Nathan MD
A post placement chest radiograph should alw ays be obt ained af t er placement of
a pulmonary art ery cat het er (PAC) (Fig. 66. 1). The post placement radiograph
serves several purposes. I t checks f or complicat ions of int roducer cat het er
placement such as pneumot horax or hemot horax. I t conf irms t hat t he PAC has
not been advanced t oo f ar int o t he periphery of t he pulmonary art erial syst em
and it also rules out kinking or coiling of t he PAC.
The risk of kinking t he PAC can be lessened by halt ing advancement of t he
cat het er if any resist ance is f elt upon insert ion. I t should be not ed t hat
appropriat e w avef orm changes, such as f rom right at rial t o right vent ricular or
f rom right vent ricular t o pulmonary art ery, should occur every 15 cm as t he PAC
is advanced (see Figure 65. 1). I f t hese changes are not readily observed, t he
PAC should be w it hdraw n. Minimizing insert ion t ime and using a slow but st eady
insert ion rat e can t ake advant age of f low -direct ed posit ioning of t he PAC (i. e. , a
“f loat ” as opposed t o a “push”) and reduce t he likelihood of t he cat het er kinking
or coiling.
Alt hough a kinked or coiled PAC may give good pressure and w edge t racings and
allow cardiac out put measurement s, it should nevert heless be removed because
of t he risk of t hrombosis and, in ext reme cases, pulmonary inf arct ion. The PAC
should never be w it hdraw n w it h t he balloon inf lat ed t o avoid valvular damage.
Any resist ance during PAC w it hdraw al mandat es t hat t he operat or halt t he
procedure and obt ain a chest radiograph t o ensure t hat t he cat het er has not
knot t ed. I f a kink has developed, int ervent ional radiologist s should be consult ed
t o remove t he PAC under f luoroscopic guidance.
O ne f inal not e is t hat in cardiac surgery pat ient s, a “st uck” PAC must cause
concern t hat t he cat het er w as inadvert ent ly sut ured in place during surgery.
FIG URE 66. 1. Sw an-G anz cat het er looped in inf erior vena cava and
reent ering right at rium. Ant eropost erior close-up view show s t he Sw an-G anz
cat het er t hrough t he superior vena cava (l ong arrow) and right at rium (short
arrow), looping in t he inf erior vena cava (arrowheads) and re-ent ering t he
right at rium (curved arrow). (Reused w it h permission f rom Umali CB. Chest
radiographic examinat ion. I n: I rw in RS, Rippe JM, eds. I rw in and Rippe's
I nt ensive Care Medicine. 5t h Ed. Philadelphia: Lippincot t Williams & Wilkins;
2003: 707. )
Suggested Readings
PACEP Collaborat ive. Pulmonary Art ery Cat het er Educat ion Project . 2006.
Available at ht t p: / / w w w. pacep. org. Accessed March 9, 2006.
Voyce SJ, McCaff ree DR. Pulmonary art ery cat het ers. I n: I rw in RS, Rippe
JM, eds. I rw in and Rippe's I nt ensive Care Medicine. 5t h Ed. Philadelphia:
Lippincot t Williams & Wilkins; 2003: 45–67.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 67 - R em ove a P ulm onar y
Ar ter y C atheter in P atients w ith New - O ns et Ventr ic ular Ar r hythm ia and in m os t Atr ial Ar r hythm ias
67
Remove a Pulmonary Artery Catheter in Patients
with New-Onset Ventricular Arrhythmia and in
most Atrial Arrhythmias
William R. Burns MD
Arrhyt hmias are common in t he crit ically ill pat ient and can result f rom a w ide
variet y of physiologic dist urbances. As t he pot ent ial et iologies are many (int rinsic
heart disease, myocardial ischemia, elect rolyt e abnormalit ies, hypoxemia,
acidosis, volume alt erat ions, endocrine irregularit ies, and ot hers), it is diff icult t o
pinpoint t he underlying pat hology in most cases. How ever, one should cert ainly
not overlook t he role t hat vascular cat het ers play in new -onset arrhyt hmias. I n
part icular, pulmonary art ery cat het ers have a high risk of incit ing bot h at rial and
vent ricular arrhyt hmias, and removal of t he line should be st rongly considered.
The incidence of arrhyt hmias relat ed t o pulmonary art ery cat het erizat ion is best
described during init ial cat het er insert ion. Premat ure at rial and vent ricular
cont ract ions are f requent w hile advancing t he cat het er and have been report ed in
13% t o 87% of pat ient s. More concerning dist urbances, such as nonsust ained
vent ricular t achycardia and right bundle branch blocks, are also common;
how ever, t hese arrhyt hmias are usually w ell t olerat ed. Vent ricular t achycardia,
vent ricular f ibrillat ion, and complet e heart block, on t he ot her hand, require
emergent int ervent ion. Fort unat ely, t hese unst able arrhyt hmias occur in less t han
1% of all cat het erizat ions.
Watch Out For

Cat het er-induced arrhyt hmias are init iat ed by direct physical cont act bet w een
t he cat het er and t he endomyocardial cells. This mechanical st imulat ion can be
suff icient t o t rigger ect opic beat s, as w ell as lif e-t hreat ening arrhyt hmias. I t
should be not ed t hat pat ient s w it h a lef t bundle branch block are at increased
risk of developing complet e heart block during pulmonary art ery cat het erizat ion.
New er cat het ers (w hich are of t en t hinner and more f lexible) may be less likely t o
t rigger t hese rhyt hm dist urbances.
While arrhyt hmias are most of t en not ed during cat het er advancement , t he simple
presence of a f oreign body can be suff icient t o init iat e rhyt hm dist urbances.
Theref ore, t he cont ribut ion of pulmonary art ery
cat het ers t o new -onset arrhyt hmias should alw ays be considered, even in t he
absence of recent cat het er manipulat ion. Acut e myocardial ischemia, recent
myocardial inf arct ion, severe hypoxemia, acidosis, and elect rolyt e abnormalit ies
are key risk f act ors f or arrhyt hmias and also seem t o low er t he t hreshold f or
cat het er-relat ed arrhyt hmias. I n t hese sit uat ions, removal of t he cat het er should
be st rongly considered.
Suggested Readings
Ermakov S, Hoyt JW. Pulmonary art ery cat het erizat ion. Crit Care Clin
1992; 8: 773–806.
Marino PL. The I CU Book. 2nd Ed. Balt imore: Williams & Wilkins; 1997: 158–
159.
Roizen MF, Berger DL, G abel RA, et al. Pract ice guidelines f or pulmonary
art ery cat het erizat ion: an updat ed report by t he American Societ y of
Anest hesiologist s Task Force on Pulmonary Art ery Cat het erizat ion.
Anest hesiology 2003; 99: 988–1014.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 68 - R em em ber that in
Nor m al P hys iology the W edge P r es s ur e is les s than the P ulm onar y Ar ter y D ias tolic P r es s ur e
68
Remember that in Normal Physiology the Wedge
Pressure is less than the Pulmonary Artery
Diastolic Pressure
Jose M. Rodriguez-Paz MD
The pulmonary art ery cat het er (PAC) is a commonly used device in t he crit ically
ill or anest het ized pat ient . I t allow s t he cont inuous monit oring of cardiac
perf ormance under several clinical condit ions. Unf ort unat ely, during t he last f ew
years several report s have show n it s pot ent ial limit at ions, including t he
erroneous int erpret at ion of t he hemodynamic dat a (Fig. 68. 1). The correct use of
t he PAC requires a t horough underst anding of heart physiology and w hat t he
cat het er is measuring.
The PAC is a f low -direct ed cat het er. Dr. Sw an conceived t he idea (by looking at
a sailboat in 1967) of t aking advant age of t he f low generat ed by t he blood going
t hrough t he heart t o guide a cat het er. Essent ially t he PAC uses t he air-f illed cuff
as a sail t o guide t he cat het er t hrough t he diff erent chambers of t he heart . The
main object ive of t he PAC is t o measure t he pressures of t he right heart and “get
t he closest w e can t o t he lef t heart . ” This guided “t our” of t he heart provides
valuable inf ormat ion t hat includes some measured values and a variet y of
calculat ed paramet ers.
I t is basic in t he underst anding of t he healt hy heart t hat t he pulmonary
vasculat ure w orks, under normal condit ions, as a low -pressure syst em. This is
import ant , since blood f low s due t o a pressure gradient (ΔP) bet w een t he sides
of t he heart , as a result of t he pressures generat ed by t he right vent ricle. This
can be expressed by t he equat ion CO = ΔP/ PVR, w here t he f low (CO , cardiac
out put ) depends on t he diff erence bet w een pulmonary art ery pressure (PAP) and
pulmonary art ery w edge pressure (PAWP), and it is inversely proport ional t o t he
pulmonary vascular resist ance (PVR).
Watch Out For

When placing a PAC, several diff erent pressures are observed t hrough t heir
pressure t racings seen on t he monit or. Af t er t he cat het er is int roduced int o t he
superior vena cava (SVC), t he t ypical t racing of t he right at rium (including t he a,
c, and v w aves) is seen. When t he cat het er passes t hrough t he t ricuspid valve,
t here is a subst ant ial change of pressures due t o t he higher pressure syst em
generat ed by t he cont ract ion of t he
right vent ricle (RV), as seen in Figure 68. 1. I t is also apparent t hat t he RV
diast olic pressure is t he same as t he right at rium (RA) pressure, since at end
diast ole t he chambers reach equilibrium. As t he cat het er is advanced, event ually
t he blood f low w ill guide t he t ip of t he PAC t hrough t he pulmonary valve. That
can be easily recognized by t he eff ect t hat t he closure of t he pulmonary valve
has on t he t racing. At t he pulmonary art ery diast olic pressure (PAD), t he
pulmonary valve closes causing t he PAD t o be higher t han t he right vent ricular
diast olic pressure. That is t he dist inct ive f eat ure of advancing t he cat het er int o
t he PA vasculat ure. Also, t he PA pressure t racing show s a charact erist ic dicrot ic
not ch (as w ould be expect ed f rom t he closure of t he pulmonary valve). The PA
pressure is equivalent t o t he syst emic pressures, in t hat PAS represent s t he
pressure generat ed by t he RV and t he PAD by t he recoiling of t he PA vasculat ure
unt il t he pulmonic valve closes.
FIG URE 68. 1. Hemodynamic changes w it h t he advancement of t he pulmonary

art ery cat het er. (Reused w it h permission f rom Marino PL. The pulmonary
art ery cat het er. I n: Marino PL, ed. The I CU Book. 2nd Ed. Balt imore: Williams
& Wilkins; 1998: 157.)
O nce in t he PA vasculat ure, t he f low w ill guide t he cat het er unt il it reaches a
vessel t hat is smaller in diamet er t han t he cat het er w it h t he balloon inf lat ed. At
t hat point t he t racing w ill appear dampened and look similar t o t he t racing
obt ained in t he RA (at rial pressures). That value of pressure is know n as t he PA
w edge pressure (PAWP) or PA occlusion pressure (PAO P). That pressure
represent s t he int errupt ion of f low generat ed by t he RV as a result of t he
inf lat ed balloon obst ruct ing t he capillary bed of t he PA vasculat ure and
corresponds
t o t he pressure t ransmit t ed by t he column of f luid creat ed by t he lef t at rium (LA)

and t ransmit t ed by t he pulmonary veins. Since PAO P is t o t he lef t side of t he
heart as t he CVP is t o t he right side of t he heart , one could make inf erences of
t he st at us of t he f illing of t he lef t vent ricle by assuming t he correlat ion bet w een
lef t vent ricular end-diast olic pressure (LVEDP) and lef t vent ricular end-diast olic
volume (LVEDV). Regret t ably, t his correlat ion is diff icult t o make because t he
PAC and PAO P do not give us inf ormat ion on t he LV compliance
(volume/ pressure) and cont ract ilit y (t he slope of t hat relat ionship).
TABLE 68-1 CARDIAC PRESSURE VALUES
PRESSURE MEAN
CARDIAC
SYSTOLIC/DIASTOLIC PRESSURE
CHAMBER
(MM HG) (MM HG)
Right atrium 0–6 3
Right S: 17–30

ventricle D: 0–6
Pulmonary S: 15–30
10–18
artery D: 5–13
Pulmonary
artery 2–12
wedge
Anot her import ant principle in t he use of t he PAC is t hat if it is correct ly

posit ioned, PAO P is alw ays less t han t he PAS and PAD pressures. This occurs
only if t here is a pat ent column of f luid connect ing t he t ip of t he cat het er w it h t he
pulmonary veins/ LA. Also t his occurs only if t he t ip of t he PAC lies in West zone
I I I (PA >pulmonary venous pressure >alveolar pressure), radiologically locat ed
below t he LA. When t he cat het er is locat ed in eit her of t he ot her West zones,
t he pressures may be diff erent . An easy w ay t o visualize t his is by using Dr.
Sw an's sailboat example. As a sailor, t he goal is t o bring t he boat t o t he area of
maximal w ind (blood f low ) and not t o an area of minimal w ind, w hich w ill make
t he boat st ay st ill or go backw ard, if t he w ind is act ually against t he sail (PAO P
≥PAD).
Suggested Readings
Marino PL. The pulmonary art ery cat het er. I n: Marino PL, ed. The I CU Book.
2rd Ed. Balt imore: Williams & Wilkins; 1998: 154–165.
Summerhill EM, Baram M. Principles of pulmonary art ery cat het erizat ion in t he
crit ically ill. Lung 2005; 183: 209–219.
Voyce SJ, McCaff ree DR. Pulmonary art ery cat het ers. I n: I rw in RS, Rippe
JM, eds. I rw in and Rippe's I nt ensive Care Medicine. Philadelphia: Lippincot t
Williams & Wilkins; 2003: 45–66.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 69 - D o not Us e P ulm onar y
Ar ter y C atheter Meas ur em ents in Tr ic us pid R egur gitation
69
Do not Use Pulmonary Artery Catheter
Measurements in Tricuspid Regurgitation
Dimitris Stefanidis MD, PhD
Ronald F. Sing DO
Most of t he paramet ers obt ained f rom pulmonary art ery (PA) cat het ers are
calculat ed rat her t han direct ly measured; only right at rial and pulmonary art ery
pressures can be direct ly measured. Many PA cat het ers use t he met hod of
t hermodilut ion t o calculat e cardiac out put (CO ) indirect ly. This met hod, how ever,
is subject t o a number of pit f alls t hat can occur during daily int erpret at ion of PA
cat het er readings and invalidat e measurement s. Tricuspid valve regurgit at ion,
low f low st at es, pat ient t emperat ure, presence of int racardiac shunt s, and
volume of inject ed bolus can all int erf ere w it h accurat e measurement s.
Watch Out For

Thermodilut ion is a met hod of measuring blood f low. A cold elect rolyt e solut ion is
inject ed t hrough t he proximal port of t he PA cat het er int o t he right at rium. The
cold f luid mixes w it h blood in t he right heart chambers and t he cooled blood is
eject ed int o t he pulmonary art ery, w here it f low s past a t hermist or locat ed at t he
dist al end of t he PA cat het er. The changes in blood t emperat ure over t ime
recorded by t he t hermist or are used t o creat e a t emperat ure-t ime curve; t he
area under t his curve is equivalent t o t he average cardiac out put in t he absence
of int racardiac shunt s. I n t he case of t ricuspid regurgit at ion, t he cold f luid is
delayed in t he right heart because of back-and-f ort h movement across t he
incompet ent valve, t hus leading t o a prolonged t hermodilut ion curve. This
produces a f alsely low cardiac out put .
A st udy t hat compared cardiac out put measurement s by t hermodilut ion w it h
cont inuous Doppler measurement s by t ransesophageal echocardiography (TEE)
in pat ient s w it h variable degrees of t ricuspid regurgit at ion f ound t hat t he
t hermodilut ion met hod w as associat ed w it h underest imat ion of cardiac out put . I n
part icular, t hey f ound t hat t he higher t he degree of t ricuspid valve regurgit at ion,
t he low er t he CO measurement s; in pat ient s w it h t hird-degree t ricuspid valve
regurgit at ion, t hermodilut ion produced CO measurement s t hat w ere a mean of 2
L/ min low er compared w it h t hose of TEE. Moreover, some degree of t ricuspid
valve regurgit at ion may be common in vent ilat ed pat ient s because of high right -
sided cardiac pressures creat ed
by posit ive-pressure vent ilat ions and may t hus be a common source of CO
measure error w hen t hermodilut ion is used. I n addit ion, anot her st udy has
implied t hat just t he presence of t he PA cat het er may lead t o a low degree of
t ricuspid valve regurgit at ion.
Thus, in pat ient s w it h t ricuspid valve regurgit at ion, especially w hen it is severe,
pulmonary art ery cat het er measurement s t hat are based on t he t hermodilut ion
met hod should be int erpret ed very caref ully, if not avoided complet ely. I n t hese
cases, cont inuous cardiac out put measurement s or Doppler measurement s by
TEE should replace t hermodilut ion-based measurement s.
Suggested Readings
Balik M, Pachl J, Hendl J. Eff ect of t he degree of t ricuspid regurgit at ion on
cardiac out put measurement s. I nt ens Care Med 2002; 28: 1117–1121.
Marino PL. The I CU Book. 2nd Ed. Balt imore: Williams & Wilkins; 1998: 182.
Sherman SV, Wall MH, Kennedy DJ, et al. Do pulmonary art ery cat het ers
cause or increase t ricuspid or pulmonic valvular regurgit at ion? Anest h Analg
2001; 92: 1117–1122.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 70 - Alw ays Tur n on a
P ac ing P ulm onar y Ar ter y C atheter P r ior to Floating The D evic e
70
Always Turn on a Pacing Pulmonary Artery
Catheter Prior to Floating The Device
There are several t emporary t ransvenous pacing devices available f or use in t he
int ensive care unit (I CU). They include pacing w ires, pacing cat het ers, paceport
pulmonary art ery cat het ers (PACs), and pacing PACs. Pacing w ires are devices
t hat use essent ially bare w ires, w hich are placed t hrough an int roducer sheat h
int o t he cent ral circulat ion and advanced int o t he heart unt il pacing is capt ured.
Pacing cat het ers usually have an inf lat able balloon at t heir t ip much like a PAC
t hat allow s blood f low t o help guide t he cat het er f orw ard as it is advanced
t hrough t he int roducer sheat h. Since it depends on blood f low f or eff ect ive
placement , it is not usef ul in emergent episodes of asyst ole and may be diff icult
t o place in t he case of ext reme bradycardia. Paceport and pacing PACs have
similar pot ent ial problems since t hey are f low direct ed f or t heir placement .
I deally, pharmacological t reat ment w it h at ropine, epinephrine, or isoprot eronol
and/ or t ranscut aneous pacing w it h Zoll pads w ill creat e a sit uat ion t hat provides
t ime and adequat e cardiac out put such t hat a t ransvenous syst em can be placed
w hen asyst ole or ext reme bradycardia occurs.
Paceport and pacing PACs diff er in t hat t he paceport PAC, in addit ion t o t he
normal lumens f or measuring cardiac out put , pulmonary art ery pressures, and so
on, has a special lumen t hrough w hich a special w ire mat ched t o t he cat het er
can be int roduced. The w ire exit s t hrough an orif ice so as t o come in cont act
w it h t he right vent ricular w all and eff ect capt ure. The pacing PAC does not have
t his port nor a special w ire but rat her has elect rodes built int o t he w all of t he
cat het er such t hat w hen it is in normal posit ion f or doing cardiac out put
measurement s, t he elect rodes are posit ioned f or aff ect ing pacing capt ure. They
are in a f ixed posit ion on t he device, w hich can make capt ure diff icult if t he
cat het er is not orient ed in opt imal posit ion. While t he orif ice on t he paceport
cat het er is also in a f ixed posit ion, t he w ire exit s t o move somew hat
independent ly so as t o cont act t he vent ricular w all.
Placement of eit her one of t he pacing-capable PACs st art s w it h cent ral venous
cannulat ion, pref erably in t he right int ernal jugular or lef t subclavian posit ion. The
reason f or t his is t hat PACs have a curve molded int o t hem t hat is most
amenable t o placement f rom t hese t w o
posit ions. The lef t int ernal jugular or right subclavian approach may be used but
it is of t en more t echnically challenging t o “f loat ” t he cat het er f rom eit her of t hese
t w o sit es. Femoral placement is also an opt ion but t ends t o have similar
diff icult ies. Even brachial placement has been perf ormed, t hough passing an
int roducer sheat h int o one of t he arm veins is of t en very diff icult . O nce t he
int roducer sheat h is placed and secured, t he PAC is f lushed, covered w it h t he
st erile cover sheat h, connect ed t o t he monit ors, and t hen placed t hrough t he
hemost at ic valve of t he int roducer. O nce it is advanced t o t he 20-cm mark on t he
cat het er, t he balloon is inf lat ed. The dist al elect rode should be connect ed t o lead
V on t he elect rocardiogram (ECG ), w hich should be properly grounded since any
current leakage t o ground above 10 µA can result in vent ricular f ibrillat ion. For
t he pacing PAC, t he pacer pulse generat or is connect ed and t urned on t o 20 mA.
I t is import ant t o use t his higher current f or placement so as t o be sure capt ure
occurs. The elect rocardiogram monit or as w ell as t he dist al port t racing should
be w at ched as t he cat het er is advanced. As t he cat het er ent ers t he at rium a
large and possibly dow n-going P-w ave w ill occur. Whet her t his P-w ave is up or
dow n going or biphasic w ill depend on t he posit ion in t he at rium and t he
relat ionship of t he elect rode t o t he direct ion of at rial depolarizat ion. I deally, you
should have capt ure w hen you achieve a pulmonary art ery t racing t hat also
w edges t he cat het er. Capt ure w ill be ident if ied by a sudden change in t he axis of
t he ECG and w idening of t he Q RS. Premat ure vent ricular cont ract ions (PVCs)
t hat may occur during t he passage of t he PAC t hrough t he vent ricle should not
be mist aken as capt ure. The PVCs are random w hile capt ure should be regular
at t he set rat e (usually 80). O nce you achieve capt ure, t he current is t urned
dow n unt il capt ure is lost and t hen t urned back up unt il capt ure is at t ained again.
Clinicians usually place t he f inal set t ing a couple of milliamps above t he capt ure
point f or a saf e margin. The cat het er may need f requent adjust ment s t o achieve
capt ure at t he low est possible current w hile at t he same t ime t o allow f or
w edging of t he cat het er f or hemodynamic assessment s.
The paceport PAC is placed as w it h ot her PACs unt il a w edge posit ion is
obt ained. O nce t his is done t he w ire is advanced t hrough t he pacing port w it h
t he current set on 20 mA unt il similar changes are seen on t he ECG as described
previously. O nce t his has occurred t he current is reduced unt il capt ure is lost ,
raised again unt il capt ure is obt ained again, and t hen usually set a couple of
milliamperes above t his f or saf e margin.
O n average, most clinicians f ind paceport PACs are easier t o use f or maint aining
capt ure as w ell as proper posit ion f or hemodynamic
measurement , t hough ot hers f avor t he pacing cat het er. I n general, t hese devices
are best chosen w hen it is imperat ive t o have hemodynamic measurement s in
addit ion t o pacing capabilit y. I f only pacing is needed, pacing w ires alone or
pacing cat het ers t hat are not pulmonary art ery cat het ers are pref erred f or
t ransvenous pacing. There are amyriad of product s and conf igurat ions f or t hese
devices and one should become f amiliar w it h t he t ype available at his or her
hospit al so as t o reduce conf usion w hen placement is emergent ly required.
Suggested Readings
Bump TE, Soble JS. Cardiac pacing. I n: Hall JB, Schmidt G A, Wood LDH,
eds. Principles of Crit ical Care. 3rd Ed. New York: McG raw -Hill; 2005: 423–
433.
G ammadge MD. Temporary cardiac pacing. Heart 2000; 83: 715–720.

> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 71 - Make S ur e any C or dis
in P lac e is the C or r ec t S iz e w hen E m er gently Floating a Tr ans venous P ac ing W ir e
71
Make Sure any Cordis in Place is the Correct
Size when Emergently Floating a Transvenous
Pacing Wire
Christina L. Cafeo RN MSN
David G . Hunt RN, BSN
Peter J. Pronovost MD, PhD
O ne of t he most diff icult and f rust rat ing sit uat ions t o occur w hen caring f or a
crit ically ill pat ient is not having all of t he required supplies available, especially
during a crisis sit uat ion. This sit uat ion part icularly seems t o occur w hen cert ain
emergency procedures arise only sporadically. As a result , st aff may be unsure
of bot h t he supplies and t he equipment required t o t reat such emergencies in a
t imely and eff icient manner.
What not to Do
Many clinicians have f aced t he misadvent ures associat ed w it h placing a
t ransvenous pacing w ire. Temporary t ransvenous cardiac pacing is a pot ent ially
lif esaving int ervent ion used primarily t o correct prof ound bradycardia. A
t ransvenous pacing w ire is t ypically placed t hrough a 6 French Cordis cat het er.
I f t he pat ient does not have a Cordis cat het er in place, t he t empt at ion is t o
break open t he closest kit t hat cont ains a Cordis and proceed t o gain cent ral
access. How ever, t his is usually a mist ake; usually, t he closest kit t hat cont ains
a Cordis is one f or insert ing pulmonary art ery cat het ers. I nvariably t hese have
Cordis cat het ers t hat are 7. 5F—t hat is, cat het ers t hat are t oo big t o insert a
t ransvenous pacing w ire t hrough. Similarly, if t he pat ient already has a Cordis in
place, it is almost alw ays a 7. 5F cat het er.
The placement of a t ransvenous w ire has a considerable complicat ion rat e, w it h
a high risk-t o-benef it rat io because of insert ion diff icult ies and pacemaker
malf unct ion. Clinically ident if ied complicat ions include local t rauma,
pneumot horax, hemot horax, arrhyt hmias, and cardiac perf orat ion. I n addit ion,
blood loss, air emboli, and inf ect ion are risks t hat occur w hen Cordis cat het er is
used t hat is t oo large and t hat allow s f or bleeding or an air embolus around t he
pacer w ire, w hich is generally 6 French.
Recommendat ions t o reduce t he risk of not having t he correct -size Cordis
available include ident if icat ion and packaging of all appropriat e supplies f or an
adult pacer w ire insert ion kit (bundle) t oget her. Component s f or t his bundle
should include an adapt er f or t he pacemaker,
an int roducer percut aneous sheat h 6F combo kit , and bipolar t ransvenous pacing
w ire. This is problemat ic as many hospit als purchase t hese component s f rom
diff erent suppliers.
What to Do
There are t w o pot ent ial solut ions t o t his problem. O ne is t o have immediat ely
available a sheat h adapt er t hat can be used w it h most of t he current ly clinically
available 4F t o 8F cordis cat het ers (e. g. , Arrow Sheat h Adapt er w it h Cat h-
G ard). A second solut ion is t o st ore t he pacer w ire insert ion supplies in a
dedicat ed area and aw ay f rom t he rout inely used cordis, cent ral line, and
pulmonary art ery cat het er supplies.
O ne f inal not e is t hat t he designat ion of a unit of French ref ers t o t he size of t he
ext ernal diamet er of a cat het er. O ne unit is equal t o . 033 mm; so a 6F cat het er
has an ext ernal diamet er of approximat ely 0. 198 mm and a 7. 5F cat het er has an
ext ernal diamet er of 0. 2475 mm.
Suggested Readings
Fit zpat rick A, Sut t on R. A guide t o t emporary pacing. BMJ 1992; 304: 365–369.
G ammage MD. Temporary cardiac pacing. Heart 2000; 83: 715–720.

> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 72 - Tr y P lac ing a S kin
Lead to Ac t as a Gr ound W ir e if P ac ing W ir es Ar e not Func tioning C or r ec tly P os toper atively
72
Try Placing a Skin Lead to Act as a Ground Wire
if Pacing Wires Are not Functioning Correctly
Postoperatively
Tammy M. Slater CRNP
Temporary epicardial w ires are rout inely implant ed in pat ient s during cardiac
surgery. Hist orically, f our w ires are placed w it h t he dist al ends of t hese w ires
being held in place on t he epicardium by a sut ure or st aple. The proximal ends
are brought out side t he chest w all t hrough subcut aneous t issue. Tw o at rial w ires
are placed on t he right at rium w it h t he proximal ends of t hose w ires being
brought out t hrough t he chest w all t o t he right of t he subcost al area. Tw o
vent ricular w ires are implant ed on t he ant erior surf ace of t he right vent ricle w it h
t he proximal ends exit ing t he chest w all t o t he lef t of t he subcost al area. The
locat ion of t he at rial w ires on t he right side of t he chest and vent ricle w ires on
t he lef t side is universal. Addit ionally, skin (or ground w ires) may be placed.
These w ires are t he longest of t he w ires and usually exit on t he side of t he at rial
w ires (i. e. , t he right side). I f a skin w ire has not been placed, t hen one of t he
ot her w ires act s as a ground w ire (posit ive pole).
The w ires t hat are implant ed in t he operat ing room w ill dict at e t he t ype of pacing
t hat can be perf ormed. For example, if bot h at rial and vent ricular w ires are
present , t hen a pat ient can be at rially (A) paced, vent ricularly (V) paced, or AV
paced. More recent ly, in uncomplicat ed cardiac surgery cases, it has become
rout ine f or only t w o vent ricular w ires t o be placed. Secondary t o t he at rial w all
being t hinner, t here is a pot ent ial increased risk of bleeding w hen pulling t he
at rial w ires. I f t his is t he sit uat ion, and t here is diff icult y w it h capt ure during
pacing, an ext ra skin or ground lead can be placed at t he bedside t o opt imize
pacing. All skin or ground w ires are placed in t he posit ive pole of t he ext ernal
pacer.
This procedure should be perf ormed asept ically. Temporary pacing leads are in
st erile packaging and should be available in cardiac surgical int ensive care unit s
(I CUs) or in t he cardiac surgery operat ing rooms. The f ollow ing st eps can be
used:
Wash hands t horoughly.
Cleanse t he skin w it h ChloraPrep.
Use st erile gloves.
Not ice t he t emporary pacing lead has a curved needle at one end and a long
st raight needle at t he ot her end. Use t he long st raight needle t o punct ure t he
skin.
Pull t he needle unt il t he w ire has gone t hrough t o t he prot ect ive coat ing.
Take t he w ire w it h t he prot ect ive coat ing and t ie a knot . This w ill secure it t o
t he skin.
Then break t he st raight needle off . You w ill be lef t w it h a st raight long
st ainless st eel shaf t in place int o t he t emporary pacemaker.
At t his t ime, t he curved needle can be removed (t his curved needle is used in
t he operat ing room w hen t hese w ires are init ially placed in t he epicardium).
Suggested Readings
Baumgart ner WA, O w en, SG , Cameron DE, et al. , eds. The Johns Hopkins
Manual of Cardiac Surgical Care. Balt imore: Mosby; 1994: 208–211.
Clochesy JM, Breu C, Cardin S, et al. , eds. Crit ical Care Nursing. 2rd Ed.
Philadelphia: W. B. Saunders; 1996: 182–185.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 73 - K now the B ac kup
Mec hanis m of Ac tion and R equir em ents for E lec tr ic al C ar diover s ion and Antic oagulation for Ventr ic ular
As s is t D evic es
73
Know the Backup Mechanism of Action and
Requirements for Electrical Cardioversion and
Anticoagulation for Ventricular Assist Devices
More t han 2 million Americans have congest ive heart f ailure and of t hese,
approximat ely 60, 000 meet crit eria f or some t ype of cardiac support or
t ransplant . Despit e expansion of crit eria in t he donor pool, t here remains a
limit ed number of donors. This number has essent ially plat eaued over t he last
decade bet w een 2, 000 and 3, 000 annually. Because heart t ransplant at ion is not
a viable opt ion f or many pat ient s, t here has been signif icant eff ort and research
in t he development of alt ernat ive medical and mechanical t herapies. Mechanical
vent ricular assist devices (VADs) are current ly used as a t emporizing measure in
t he set t ing of acut e heart f ailure, a bridge t o t ransplant at ion, or as dest inat ion
t herapy. These devices are report ed t o improve t he qualit y of lif e in t hese
pat ient s w ho are ot herw ise crippled by t heir heart f ailure despit e maximal
medical t herapy.
VADs can be cat egorized in many w ays: short t erm or long t erm; pulsat ile or
nonpulsat ile; int ernal or ext ernal (i. e. , int ra- or ext ra-corporeal). Short -t erm
devices include cardiopulmonary bypass, ext racorporeal membrane oxygenat ion
(ECMO ), and t he Abiomed BVS. Long-t erm devices include t he Novacor, t he
Heart Mat e I and I I , Thorat ec, and t he new er Abiomed AB5000 Vent ricle. The
nonpulsat ile devices are cardiopulmonary bypass and ECMO in addit ion t o t he
Heart Mat e I I . The permanent devices have an int ernal chamber. No device has a
t ot ally implant able energy source.
All VADs except t he Heart Mat e I I , w hich generat es nonpulsat ile f low using a
rot arylike mot or, generat e pulsat ile f low because of t heir design w it h
unidirect ional f low valves. They operat e in a f ill-t o-empt y cycle, similar t o t he
act ion of t he nat ive vent ricle (i. e. , diast ole and syst ole). The det erminant s of t he
st roke volume are t he pre-eject ion chamber volume (preload) and t he out f low
resist ance, eit her pulmonary or syst emic, depending w hich side is mechanically
support ed. Empt ying is init iat ed w hen t he cont rol unit senses t hat t he chamber is
nearly f ull. The rat e of t he eject ion is volume, or preload, dependent . I t is
import ant t o not e t hat t he rat e of out put is independent of t he nat ive cardiac
rhyt hm. The goal is t o maint ain unloading
of t he vent ricle, especially in t hose pat ient s w ho have had t his placed w it h
planned removal af t er recovery of t he myocardium. This t ypically is limit ed t o t he
post cardiac surgery pat ient in w hich a t emporary device is placed.
TABLE 73-1 CLINICAL CHARACTERISTICS OF COM M ON

DEVICES
ABIOMED
DEVICE T HORAT EC NOVACOR
(AB5000)
Arterial tracing Pulsatile Pulsatile Pulsatile
Angioflex
Valves Mechanical Bioprosthet
polyurethane
Para- or
Implantation Paracorporeal intra- Intracorpor
corporeal
Portable,
Not portable, Portable,
Control unit not
large device worn as “be
wearable
Anticoagulation Coumadin Coumadin Coumadin
Maximum
100 mL 65 mL 70 mL
stroke volume
Backup
Backup pump
Pneumatic controller
Backup (also has
handheld needed; no
mechanism pneumatic
pump manual
hand pump)
mechanism
Maximum flow
4–6 L/min 7 L/min 10 L/min
rate
Electrical
No issuea No issue No issue
cardioversion
Cautery No issue No issue No issue
Acetone
Acetone can
can crack
crack the
Precautions the
housing
housing
device
device
a May perform external or internal cardioversion with the dev

b Specific instructions should accompany the device. It inclu
performed in a specific sequence.

Among t he commonly used VADs (Table 73. 1) t here are unique charact erist ics
and specif ic clinical inf ormat ion t hat are import ant t o recognize. These include
w het her t hey require ant icoagulat ion, t he maximum st roke volume, t he
requirement s f or cardioversion and concerns w it h elect rocaut ery, and t he backup
mode in case of emergency. I n addit ion t o t hese import ant management issues,
t he key t o t he management of t he hemodynamics of t hese pat ient s is
underst anding t hat t he pumps are vol ume dependent (as discussed previously)
and t hat many of t hese pat ient s have concomit ant right vent ricular (RV)
dysf unct ion or f ailure and pulmonary hypert ension, w hich may f urt her
compromise RV f unct ion. RV dysf unct ion is very diff icult t o manage and key
hemodynamic goals include t hose t hat opt imize RV f unct ion and prevent ion of
pulmonary hypert ension, w hich w ill result in improved VAD f illing and empt ying.
These include prevent ion of hypoxia and hypercarbia; opt imal RV coronary
perf usion pressure; and maint enance of a normal int rinsic cardiac rhyt hm.
Alt hough t he f illing and empt ying of t he LVAD is not in synchrony w it h t he nat ive
cardiac cycle (rat e), maint enance of sinus or ot her st able rhyt hm w ill pot ent ially
improve RV eject ion and t hus LV f illing.
A key aspect in t he management of such devices is t he need f or ant icoagulat ion,
except in t he case of t he Heart Mat e I . There are clinical sit uat ions t hat may
require discont inuat ion of ant icoagulat ion (e. g. , need f or emergent surgery,
acut e hemorrhagic st roke, and cent ral line placement ). Ant icoagulat ion may be
held f or brief periods of t ime, but it must be reinit iat ed as soon as it is deemed
saf e. These devices are t hrombogenic and pat ient s are at high risk of clot t ing,
w hich is generally and universally cat ast rophic. Even w it h adequat e
ant icoagulat ion, t here is an est imat ed 30% risk of t hrombus f ormat ion.
Thrombosis may occur in t he device it self or at a nat ive valve (e. g. , aort ic).
Changing out t he pumps because of t hrombosis is possible only w it h t he Abiomed
BVS syst em and t he paracorporeal Thorat ec. There are some report s of
changing out t he prost het ic valvular devices in t he lit erat ure as w ell.
Suggested Readings
Deng MC, Naka Y. Mechanical circulat ory support devices—st at e of t he art .
Heart Fail Monit 2002; 2: 120–128.
Mancini D, Burkhoff D. Mechanical device-based met hods of managing and

t reat ing heart f ailure. Circulat ion 2005; 112: 438–448.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 74 - P lac e the D efibr illator
in S ync hr onous Mode w hen C ar diover ting
74
Place the Defibrillator in Synchronous Mode
when Cardioverting
Angela D. Shoher MD
Elect rical cardioversion is used t o manage pat ient s w it h cardiac arrhyt hmias t hat
involve re-ent rant circuit s, such as vent ricular t achycardia or at rial f ibrillat ion. I n
an emergent set t ing any pat ient w it h a vent ricular rat e great er t han 150 w ho is
unst able (i. e. , chest pain, hypot ension, unresponsive) should be t reat ed w it h
synchronized cardioversion. Cardioversion may also be perf ormed on st able
pat ient s w ho have an arrhyt hmia t hat is ref ract ory t o pharmacologic int ervent ion.
Cardioversion t heoret ically st ops t he arrhyt hmia by depolarizing t he re-ent rant
circuit and making it ref ract ory t o propagat ion. Cardioversion delivers energy
t hat is synchronous w it h t he early part of t he Q RS complex. I f t he energy is
delivered in an asynchronous manner, it can induce vent ricular f ibrillat ion. This
occurs w hen t he energy is delivered during t he early phase of repolarizat ion.
When t he cardiovert er is placed in synchronous mode, it aut omat ically
discharges a current t hat coincides w it h a large R or S w ave, t hereby avoiding
t he period of repolarizat ion w hen vent ricular f ibrillat ion can occur. I n synchronous
mode t here is alw ays a delay in energy delivery w hile t he cardiovert er searches
f or t he R or S w ave. I t is import ant t o sw it ch t he mode t o asynchronous delivery
if t he pat ient develops vent ricular f ibrillat ion.
What to Do
When cardiovert ing, t here are t w o possible placement opt ions of t he pads on t he
chest w all. I n t he ant eropost erior posit ion, one pad is placed t o t he right of t he
st ernum ant eriorly and t he second is placed bet w een t he spine and t he t ip of t he
lef t scapula post eriorly. The second opt ion is t o place t he paddles in t he
ant erolat eral posit ion w it h one paddle in t he f ourt h or f if t h int ercost al space in
t he lef t midaxillary line. The second paddle is t hen placed t o t he right of t he
st ernum on t he second or t hird int ercost al space. The ant erior-post erior
approach is opt imal in pat ient s w it h implant able devices t o avoid divert ing
current t o t he device. Pacemakers should be 10 cm f rom direct cont act w it h t he
paddles. The paddles should be placed f irmly against t he skin t o prevent arcing
and skin
burns. Conduct ive gel should be used t o ensure good cont act w it h t he skin.
The amount of energy used t o cardiovert is cont roversial. Prompt cardioversion
prevent s prolonged ischemia and mult iple shocks. Excessive energy, how ever,
can lead t o myocardial damage. The opt imal amount of energy also depends on
t he pat ient 's arrhyt hmia. For a pat ient w ho is relat ively hemodynamically st able,
at rial f ibrillat ion w it h rapid vent ricular rat e (RVR) t hat is ref ract ory t o
pharmacologic t herapy, f or example, 50 joules may be adequat e. For vent rical
t achycardias, how ever, advanced cardiac lif e support guidelines now recommend
st art ing w it h 360 joules (unlike t he increasing amount of joules recommended in
past guidelines). This number also depends on t he t ype of def ibrillat or. I f it
ismonophasic (i. e. , current t ravels in one direct ion bet w een t he paddles) more
energy w ill be required. I f t he machine delivers biphasic current , considerat ion
should be given t o st art ing at 25 joules f or a t rial f ibrillat ion and 150 joules f or
vent ricular t achycardias.
Suggested Readings
Morgan G H, Mikhail MS, Murray MJ, et al. , eds. Clinical Anest hesiology. 3rd
Ed. New York: McG raw -Hill; 2002: 472â 4 73.
Podrid PJ. Basic Principles and Techniques of Cardioversion and

Def ibrillat ion. UpToDat e. ht t p: / / w w w. upt odat e. com
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 75 - Tur n off the
S ync hr oniz ation Mode on the D efibr illator if ther e is no R ec ogniz ed R W ave
75
Turn off the Synchronization Mode on the
Defibrillator if there is no Recognized R Wave
Nirav G . Shah MD
The diff erence bet w een cardioversion and def ibrillat ion is based on w het her t he
synchronizat ion mode on t he cardiovert er-def ibrillat or is t urned on or off . I n
cardioversion, t he elect rical shock is delivered in synchronizat ion (“t urned on”)
w it h t he Q RS complex and can result in t he t erminat ion of a variet y of
arrhyt hmias. At rial f ibrillat ion is one of t he most f requent ly cardiovert ed rhyt hms
and usually requires at least 300 J of energy on t radit ional def ibrillat ors and
bet w een 75 J and 200 J on t he new er biphasic def ibrillat ors. Urgent
cardioversion is undert aken in pat ient s w it h hemodynamic compromise and in
st able pat ient s af t er ant icoagulat ion unless t he arrhyt hmia has last ed less t han
48 hours. Direct -current cardioversion is also used f or at rial f lut t er and most
pract it ioners init ially t ry 50 t o 100 J w it h a good rat e of success.
Supravent ricular t achycardias and vent ricular t achycardias may also be
cardiovert ed but t his is done f ar less f requent ly secondary t o success of
int ravenous ant iarrhyt hmics and unrecognizable R w aves.
Def ibrillat ion, as opposed t o cardioversion, delivers a shock at any t ime during a
cardiac cycle and is used primarily w hen t here is no recognized R w ave, such as
in vent ricular f ibrillat ion, polymorphic vent ricular t achycardia, or rapid vent ricular
t achycardia. The energy requirement is now recommended t o be 360 J f or
successf ul conversion t o sinus rhyt hm. Wit h prompt def ibrillat ion t he success of
t his procedure is usually high. I n pat ient s w it h vent ricular t achycardia t here is a
chance of conversion t o vent ricular f ibrillat ion requiring addit ional shocks f or
rest orat ion of sinus rhyt hm.
For eit her cardioversion or def ibrillat ion several considerat ions are import ant .
First , adequat e sedat ion should be provided if t he pat ient is aw ake and alert .
The pain associat ed w it h elect rical shock is signif icant and ideally t he pat ient
should have adequat e analgesia and amnesia f or t he procedure. Second,
complicat ions f rom t he procedure include myocardial necrosis, myocardial
dysf unct ion, pulmonary or syst emic embolizat ion, skin burns, and det eriorat ion t o
a less st able rhyt hm. Finally, it is possible t hat t he pat ient may progress t o
cardiopulmonary arrest and t he care t eam should be prepared f or t he need t o
t reat t his should it occur.
Suggested Readings
Falk RH. At rial f ibrillat ion. N Engl J Med 2001; 344: 1067–1078.
Tracy CM, Akht ar M, DiMarco JP, et al. American College of

Cardiology/ American Heart Associat ion clinical compet ence st at ement on
invasive elect rophysiology st udies, cat het er ablat ion, and cardioversion: a
report of t he American College of Cardiology/ American Heart
Associat ion/ American College of Physicians-American Societ y of I nt ernal
Medicine t ask f orce on clinical compet ence. Circulat ion 2000; 102: 2309.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 76 - D o not R ule O ut the
P r es enc e of A Myoc ar dial Infar c tion by a Nor m al E lec tr oc ar diogr am
76
Do not Rule Out the Presence of A Myocardial
Infarction by a Normal Electrocardiogram
Jose M. Rodriguez-Paz MD
O ne of t he most challenging clinical scenarios t hat w e encount er in clinical
pract ice and specif ically in t he int ensive care unit (I CU) environment is t he
diagnosis of myocardial ischemia/ inf arct ion (MI ) or acut e coronary syndrome
(ACS). The diagnosis is made by considerat ion of cert ain risk f act ors f or
coronary art ery disease, t he charact erizat ion of clinical sympt oms (especially
pain), and t he result s of diagnost ic st udies (invasive and noninvasive). Moreover,
making t he diagnosis in a t imely f ashion is paramount in order t o reduce t he
morbidit y and mort alit y associat ed w it h t his disease, since t hose pat ient s w it h
ST-elevat ion MI or new lef t bundle branch block require urgent reperf usion
t herapy.
Signs and Symptoms

O ne common clinical f inding in pat ient s present ing w it h MI is chest pain (locat ed
in t he precordium and described as t ight ness or pressure radiat ing t o lef t arm,
jaw, or neck). Commonly t hese pat ient s also present w it h dyspnea, anxiet y,
light headedness and syncope, nausea and vomit ing, and diaphoresis. There are
also many pat ient s w ho present w it h no specif ic sympt oms. I t is accept ed t hat
one in six pat ient s w it h chest pain act ually have an MI , especially in elderly
pat ient s. I t is import ant t o obt ain a good descript ion of t he pat ient 's sympt oms.
I n general, t he likelihood of having an MI w it h chest pain increases if t he pat ient
has had ot her episodes of chest pain f or more t han 1 year, if t he pain is const ant
and described as pressure and radiat es t o t he arms, and if it is associat ed w it h
diaphoresis, nausea, and vomit ing. I n some pat ient s (e. g. , diabet ics) t his
charact erist ic pain may be absent . O t her clinical f indings t hat have an increased
likelihood of being associat ed w it h MI include hypot ension, presence of a t hird
heart sound, and t he presence of pulmonary crackles. Alt hough t hese sympt oms
cannot rule it out , sympt oms t hat low er t he likelihood of ischemia include
reproducibilit y of pain w it h palpat ions and posit ional changes, st abbing pain, and
ot her at ypical present at ions. Nevert heless, t he evidence published does not
support t he sole use of clinical signs and sympt oms w it hout t he support of
elect rocardiography (ECG ), cardiac enzymes, or ot her diagnost ic t ools t o

diagnose ACS or MI , as t hese have low sensit ivit y.
ECG cont inues t o be one of t he most import ant diagnost ic t ools f or ACS. The
current classif icat ion of ACS has moved t ow ard dividing MI int o ST-elevat ion MI
(STEMI ) and non—ST-elevat ion MI (NSTEMI ). I n t he presence of charact erist ic
sympt oms, t he presence of ST-segment elevat ion has high specif icit y f or MI ,
especially f or t ransmural MI (Q -w ave inf arct ). I n general, perioperat ive ischemia
or MI generally is associat ed w it h ST depression rat her t han ST elevat ion.
Unf ort unat ely, ECG s can be normal in half of t he pat ient s w it h ACS. This is
probably even more f requent in pat ient s admit t ed t o t he I CU f or ot her causes or
in t he case of int raoperat ive MI , in w hich t he lack of t ypical sympt oms or
at ypical ECEs changes may make t he diagnosis of MI very diff icult . I t is
import ant t o not e t hat ECG w ave-f orms are det ermined by not only t he
pat hophysiology of t he pat ient but also by t echnical f act ors, t hus det ermining t he
out put ECG . The elect rical changes regist ered by t he ECG are relat ed t o t he
mass of myocardium damaged and t hus t he bigger t he area, t he more evident
t hat t he changes and t he t errit ory are aff ect ed by t he MI , alt hough t his not
alw ays t rue. For example, in t he case of concomit ant right vent ricular (ST
elevat ion in V1 and V2) and post erior MI (ST depression in V1—V3), bot h in t he
t errit ory of t he right coronary art ery, t he elect rical f orces t hat result af t er MI
neut ralize each ot her and ST changes may not be seen. This is very import ant
because t he pat hophysiology and t reat ment of t he right vent ricle MI is very
diff erent t han t he MI in t he lef t heart , requiring volume loading and early use of
inot ropes t o maint ain preload t o t he lef t vent ricle.
TABLE 76-1 RELATIONSHIP OF CORONARY ANATOM Y

AND ECG LEAD FINDINGS
CORONARY
REGION ECG LEADS
ART ERY
Inferior wall II, III, a VF RCA
Large R wave (R/S ratio

Posterior >1) in V1, V2, or V3 with
ST depression, T-wave RCA
wall
changes in V1
Lateral wall I, a VL, V4–V6 Circumflex
Anterolateral I, a, V1–V6 LAD
Anteroseptal V1–V3 LAD
Anterior V1–V4 LAD
V3R–V5R (transient
Right changes), V4R the
RCA
ventricle highest specificity and
sensitivity
LAD, left anterior descending; RCA, right coronary

artery; V3R, right precordial lead.
The ECG f indings t hat charact erist ically are f ound in MI are t he presence of Q
w aves, ST-segment elevat ion (≥1 mm) or depression, new conduct ion def ect s
and T-w ave depression in specif ic leads, depending upon t he t errit ory aff ect ed
(Table 76. 1). The change w it h t he highest probabilit y of being f ound in MI is t he
ST-segment elevat ion, especially in t he presence of reciprocal changes. At t he
same t ime, and despit e t he number of cases of MI w it h normal ECG , t he
presence of a normal ECG decreases t he probabilit y of MI but t his does not rule
it out .
Suggested Readings
Birnbaum Y, Drew BJ. The elect rocardiogram in ST elevat ion acut e myocardial
inf arct ion; correlat ion w it h coronary art ery anat omy and prognosis. Post grad
Med J 2003; 79: 490–504.
Hurf ord WE. Crit ical Care Handbook of t he Massachuset t s G eneral Hospit al.
Panju AA, Hemmelgarn BR, G uyat t t G H, et al. The rat ional clinical
examinat ion: is t his pat ient having a myocardial inf arct ion. JAMA
1998; 280: 1256–1263.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 77 - D o not us e the 3 or 5
Lead E C G Monitor as a 12- lead E lec tr oc ar diogr am
77
Do not use the 3 or 5 Lead ECG Monitor as a 12-
lead Electrocardiogram
Tuhin K. Roy MD, PHD
Cont inuous elect rocardiogram (ECG ) monit oring is a st andard monit or ut ilized f or
t he det ect ion of arrhyt hmias and ischemia. Many common monit oring syst ems
ut ilize 3 or 5 leads inst ead of t he 12 leads capt ured by t he st andard ECG . Even
t hough bot h syst ems can pot ent ially be used f or det ect ion of arrhyt hmias as w ell
as ischemia, t he limit at ions of t he 3- or 5-lead syst em monit ors should be
underst ood so t hat t he care of t he pat ient may be opt imized by obt aining f urt her
st udies as necessary. The st andard 12-lead ECG provides a 10-second snapshot
of t hree bipolar leads (I , I I , I I I ), t hree augment ed unipolar leads (a VR, a VL, a
VF), and six precordial leads (V1 t o V6); t he st andard 5-lead syst em allow s t he
monit oring of seven diff erent leads (I , I I , I I I ; a VR, a VL, a VF; and V5). I t is
import ant t o remember t hat t he ECG ref lect s only elect rical act ivit y and not
heart f unct ion (e. g. , pulseless elect rical act ivit y) and can be alt ered by
physiologic f act ors, pat hophysiological f act ors (e. g. , elect rolyt e imbalances,
pacemakers, pericardit is, hypot hermia, subarachnoid hemorrhage), and t echnical
f act ors (e. g. , calibrat ion, lead posit ion, and body posit ion).
Arrhythmia Detection
Accurat e det ect ion and ident if icat ion of arrhyt hmias is crucial since inadequat e
or inappropriat e t reat ment can be harmf ul or f at al. Alt hough lead I I is commonly
used f or t he det ect ion of cardiac dysrhyt hmias since it s direct ion parallels at rial
depolarizat ion and result s in t he maximum P-w ave amplit ude, lead V1 is more
usef ul f or diagnosing bundle branch blocks and discriminat ing bet w een
vent ricular t achycardia (VT) and supravent ricular t achycardia (SVT) w it h
aberrant conduct ion. Bot h of t hese leads should be monit ored if possible so t hat
t he relat ionship bet w een at rial and vent ricular depolarizat ions can be
det ermined.
Premat ure at rial cont ract ions t ypically result in abnormal P w aves f rom an
ect opic at rial f ocus w it h a variable PR int erval and no compensat ory pause. I f
part of t he vent ricular conduct ion syst em is ref ract ory, t he result ing aberrant
conduct ion leads t o an abnormal Q RS complex and can be conf used w it h a
premat ure vent ricular cont ract ion. Look f or a preceding (likely abnormal) P
w ave, a Q RS complex suggest ive
of a right bundle branch block, an rSR’ in V1, and concordance of t he init ial
def lect ion w it h t he preceding beat t o dist inguish t hese f rom premat ure vent ricular
cont ract ions.
Paroxysmal SVT result s f rom t ransient act ivit y of an ect opic at rial pacemaker
and is seen as a rapid regular rhyt hm w it h abnormal P w aves and normal Q RS
complexes. SVT w it h aberrant conduct ion (as ment ioned earlier) can be conf used
w it h vent ricular t achycardia. At rial f lut t er is charact erized by saw t oot h f lut t er (F)
w aves best seen in leads I I and V1 w it h normal Q RS complexes t ypically
conduct ed at 2: 1, 3: 1, or 4: 1. I n at rial f ibrillat ion, P w aves cannot be ident if ied
and t he Q RS complexes are narrow w it h irregular int erval.
Junct ional rhyt hms such as at riovent ricular (AV) nodal re-ent ry t achycardia can
result f rom ect opic act ivit y near t he AV node t hat is conduct ed ret rograde as
w ell as ant erograde, leading t o invert ed P w aves in leads I I , I I I , and a VF and
posit ive P w aves in a VR. High nodal rhyt hms produce short ened PR int ervals,
w hereas low nodal rhyt hms produce P w aves t hat f ollow a normal Q RS. I f t he
rhyt hm is midnodal, t he P w ave can dist ort t he Q RS complex, causing a pseudo
S w ave in t he inf erior leads and a pseudo R w ave in V1.
AV re-ent rant t achycardias are due t o accessory conduct ion pat hw ays t hat
permit vent ricular pre-excit at ion. Wolff -Parkinson-Whit e syndrome (WPW)
generat es a regular rhyt hm w it h a delt a w ave and a slurred Q RS complex
result ing f rom early vent ricular depolarizat ion. Type A WPW is charact erized by
an upright delt a w ave and Q RS complex in t he precordial leads w it h an upright
dominant R w ave in V1; t his can be conf used w it h a right bundle branch block
(RBBB). Type B WPW has a dow nw ard delt a w ave and Q RS complex in leads V1
and V2 w it h upw ard def lect ions in t he ot her precordial leads similar t o a lef t
bundle branch block (LBBB). The majorit y of t achycardias in WPW are classif ied
as ort hodromic and result f rom ant egrade conduct ion of impulses t hrough t he AV
node t hat ret urn t o t he at ria via t he accessory pat hw ay; P w aves f ollow t he
normal-w idt h Q RS complexes and no delt a w aves are seen. Ant idromic AV re-
ent rant t achycardia result s f rom ant egrade conduct ion via t he accessory
pat hw ay (causing w ide Q RS complexes) w it h ret rograde conduct ion via t he AV
node. At rial f ibrillat ion in WPW result s in an irregular w ide complex t achycardia.
Premat ure vent ricular cont ract ions result in a prolonged abnormal Q RS complex
since t he ect opic vent ricular pacemaker impulses are conduct ed via vent ricular
muscle rat her t han t he Purkinje f ibers; t ypically, t he Q RS volt age is increased
and t he T w ave appears invert ed. Ret rograde or blocked at rial depolarizat ion
may lead t o t he
presence of P w aves on t he t racing. A compensat ory pause occurs since a
subsequent depolarizat ion f rom t he sinoat rial (S-A) node is blocked. An R-on-T
phenomenon may happen if t his occurs near t he vulnerable period of t he
preceding depolarizat ion, leading t o vent ricular f ibrillat ion. The Q RS complex
may be of t he RBBB variant w it h a prominent R w ave in V1 or may be of t he
LBBB variant w it h not ching of t he S w ave w it h less acut e dow nsloping.
Vent ricular t achycardia appears as mult iple, rapid, regular premat ure vent ricular
cont ract ions. The Q RS complexes are w ide and must be dist inguished f rom SVT
w it h aberrancy. I f t he Q RS complexes appear t o be posit ive in lead V1, a
vent ricular origin is suggest ed by posit ive def lect ions in all precordial leads, a
durat ion >0. 14 seconds, and charact erist ic R-w ave morphologies in lead V1. I f
t he Q RS complexes are negat ive in lead V1, a vent ricular origin is suggest ed by
negat ive def lect ions in all precordial leads and a durat ion >0. 16 seconds. Axis
deviat ion >40 degrees, a Q S or negat ive complex in lead V6, and negat ive P
w aves in a VR are also suggest ive of a vent ricular origin. This rhyt hm is
charact erized by f usion/ capt ure beat s as w ell as AV dissociat ion. Vent ricular
f ibrillat ion appears as an irregular w avy line. Torsades de point es, associat ed
w it h alt ered repolarizat ion and Q T prolongat ion, may have a similar appearance
t o polymorphic VT or vent ricular f ibrillat ion (VF).
O t her causes of Q RS prolongat ion include lef t vent ricular hypert rophy (LVH).
Conduction Defects
Conduct ion def ect s include SA node block (no at rial excit at ion and no P w ave),
at riovent ricular block (w it h alt ered conduct ion of t he at rial impulse t hrough t he
AV node), and int ravent ricular def ect s including RBBB and LBBB.
Alt hough t he Q RS complexes are narrow in f irst -degree at riovent ricular block
(PR int erval >0. 21 seconds) and Type I (Wenckebach) second-degree block, t he
durat ion of t he Q RS complex in Type I I second-degree block depends on it s
locat ion w it hin t he conduct ion syst em. Complet e at riovent ricular dissociat ion
(t hird-degree block) can result in narrow or w ide Q RS complexes depending on
t he locat ion of t he ect opic pacemaker causing vent ricular depolarizat ion. I f it is
near t he AV node, t he complexes appear normal w it h a vent ricular rat e of 40 t o
60 per minut e; if it is inf ranodal, t he rat e is usually <40 per minut e and t he
complexes appear w ide.
Leads I , V1, and V6 are especially import ant f or diagnosing int ravent ricular
conduct ion def ect s. I n complet e LBBB, all impulses are conduct ed via t he right
bundle, causing a w idened Q RS w it h a negat ive rS complex in V1 and a w ide
not ched R w ave in leads I , a VL, and V6. The Q RS complex is not w idened in lef t
ant erior or post erior hemiblock. Lef t ant erior hemiblock leads t o lef t -axis
deviat ion w it h a small Q in I and a and a small R in I I , I I I , and a VF; lef t
post erior hemiblock leads t o right -axis deviat ion w it h a small R in I and a and a
small Q in I I , I I I , and a. RBBB causes a w ide Q RS and can generat e rSR’
complexes in V1 t o V3 w it h w ide S w aves in leads I and V6.
I n summary, w idening of t he Q RS complex can be associat ed w it h right and lef t
bundle branch blocks as w ell as ect opy arising f rom t he right and lef t vent ricles,
w hich may not be det ect ed if lead I I is monit ored exclusively. This highlight s t he
import ance of lead V1 t o assist w it h t he accurat e det ect ion and diagnosis of
conduct ion def ect s and arrhyt hmias during monit oring. O bt aining a hard-copy 12-
lead ECG should be st rongly considered w hen aberrant conduct ion or conduct ion
def ect s are suspect ed.
TABLE 77-1 RELATIONSHIP OF CORONARY

DISTRIBUTION TO POSSIBLE ECG CHANGES
CORONARY
LEADS MYOCARDIUM
DIST RIBUT ION
Inferior LV wall RA, RV,

II, III,
RCA interatrial septum SA node,
a
AV node
Lateral LV wall SA node, AV

I, a LCX
node
V3–V6 LAD Anterolateral LV wall
LAD, left anterior descending coronary artery; LCX, left

circumflex artery; RCA, right coronary artery.
Ischemia Detection
The opt imal leads f or det ect ion of ischemia are diff erent f rom t he opt imal leads
f or det ect ion of arrhyt hmias. Common indicat ors of ischemia are ST-segment
changes, T-w ave inversions, or t he development of Q w aves in leads
corresponding t o diff erent areas of myocardium (Table 77. 1); not e, how ever, t hat
ischemia may be present even in t he absence of ECG changes.
ST-segment changes are t ypically measured relat ive t o t he PQ baseline. A
generally accept ed def init ion of ischemia is horizont al or
dow nsloping ST-segment depression >1 mm f rom baseline or ST-segment

elevat ion >1 mm in any lead 60 or 80 milliseconds af t er t he J point . Leads I I and
V5 (ref lect ing t he inf erior and ant erolat eral dist ribut ions) have t radit ionally been
used f or ischemia det ect ion. A st udy by London et al. assessed t he ut ilit y of
various leads in det ect ing ischemia in pat ient s w it h know n coronary disease
undergoing surgery. Lead V5 w as f ound t o be t he most sensit ive (75%); adding
lead I I (of t en used f or arrhyt hmia det ect ion) improved t his t o 80%, but adding
leads I I and V4 improved t his f urt her t o 96%. A more recent st udy by
Landesberg et al. in pat ient s undergoing vascular surgery f ound t hat V4 w as t he
most sensit ive f or ischemia among pat ient s w ho had a myocardial inf arct ion
(83%) and t hat using 2 t o 3 precordial leads det ect s more t han 92% of all
ischemia t hat w ould be picked up by using all 12 leads. Not e t hat simult aneous
recording of mult iple precordial leads may not be available on all monit ors.
Alt hough t he 5-lead ECG may a usef ul t ool f or det ect ion of ischemia, as w it h a
12-lead, t he w avef orms obt ained are sensit ive t o lead placement , especially w it h
regard t o t he precordial elect rodes. O f t en during surgical procedures or in
pat ient s w it h dressings in place, t he placement of t he leads is not ideal.
Accurat e ischemia det ect ion also relies on adequat e skin cont act of t he leads.
To reduce elect rical and caut ery int erf erence, monit ors rely on f ilt ers t hat reduce
art if act and drif t but t hat may prevent accurat e ST-segment analysis. Monit ors
should be sw it ched t o diagnost ic mode t o allow t he appropriat e bandw idt h f or
det ect ion of ischemia.
Cont inuous ST-segment t racking is available t o monit or changes but is subject t o
limit at ions and requires vigilance w it h regard t o drif t ing set point s t hat need t o
be checked and adjust ed on a regular basis. Also, ischemia may be t ransient and
not ref lect ed on serial ECG s, providing t he rat ionale f or cont inuous 12-lead ECG
monit oring w it h simplif ied lead syst ems t hat have recent ly become available.
Finally, int raoperat ive ECG monit oring may not det ect a subst ant ial f ract ion of
int raoperat ive ischemic event s as compared w it h t he use of echocardiography t o
det ect regional w all mot ion abnormalit ies. I f ischemia is suspect ed based on a 3-
or 5-lead ECG , a 12-lead ECG should be obt ained, if possible. I n addit ion, t he
use of ot her modalit ies including echocardiography should also be considered if
ischemia is suspect ed in t he absence of ECG changes.
Suggested Readings
Edhouse J, Morris F. Broad complex t achycardia: part I . BMJ 2002; 324: 719–
722.
Edhouse J, Morris F. Broad complex t achycardia: part I I . BMJ 2002; 324: 776–
779.
Landesberg G , Mosseri M, Wolf Y, et al. Perioperat ive myocardial ischemia
and inf arct ion. Anest hesiology 2002; 96: 264–270.
London MJ, Hollenberg M, Wong MG . I nt raoperat ive myocardial ischemia:

localizat ion by cont inuous 12-lead elect rocardiography. Anest hesiology
1988; 69: 232–241.
Miller RD, ed. Anest hesia. 5t h Ed. New York: Churchill Livingst one;
2000: 1231–1254.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 78 - P ut a B oar d under the
P atient w hen doing C hes t C om pr es s ions
78
Put a Board under the Patient when doing Chest
Compressions
Elizabeth A. Hunt MD, MPH
Cameron Dezfulian MD
Chest compressions have long been considered one of t he crit ical element s of
basic and advanced lif e support , as t hey provide vit al subst rat e t o t he brain and
myocardium during cardiac arrest . The qualit y of compressions delivered during
cardiac arrest aff ect s t he success of t he pat ient 's out come.
What to Do
How ever, it has become increasingly clear t hat healt h care providers do not
alw ays perf orm high-qualit y cardiopulmonary resuscit at ion (CPR). To address
t his issue, t he 2005 American Heart Associat ion G uidelines f or CPR and
Emergency Cardiovascular Care provides a number of evidence-based
recommendat ions det ailing how t o deliver eff ect ive chest compressions. For
example, t hese guidelines and t he evidence st ress t he import ance of providing
chest compressions as early as possible af t er cardiovascular collapse,
minimizing int errupt ions t o compressions, and permit t ing chest -w all recoil. The
guidelines also give specif ic recommendat ions on how t o perf orm chest
compressions on adult s, including a rat e of 100 compressions per minut e, a
compression t o vent ilat ion rat io of 30: 2 f or t he unint ubat ed pat ient , and a
compression dept h of 1. 5 t o 2 in (4 t o 5 cm).
I n addit ion t o t hese recommendat ions, t he guidelines also specif ically st at e, “t o
maximize t he eff ect iveness of compressions, t he vict im should lie supine on a
hard surf ace (e. g. , backboard or f loor). ” I n addit ion, t he guidelines st at e, “‘CPR-
f riendly’ def lat able mat t resses have been st udied, and t hey do not provide an
adequat e surf ace on w hich t o perf orm chest compressions. ” I n animal and
mannequin st udies, compressions delivered on a sof t surf ace (e. g. , f oam
mat t ress or f ully inf lat ed or def lat ed air mat t ress) result ed in signif icant
dissipat ion of t he dow nw ard f orce of compression and inadequat e dept h of
compression w hen compared w it h placement of t he mannequin on t he f loor.
Furt hermore, resuscit at ors f elt t heir compressions w ere more unst able and less
eff ect ive w hen perf ormed on any of t he mat t resses w hen compared w it h t he
f loor.
The af orement ioned st udies also show ed t hat t he f orce of compressions
decreased w it h t he increasing height of t he bed. I t is possible
t hat t he f loor is such an eff ect ive surf ace f or chest compressions because of t he
int eract ion bet w een t he hard surf ace and t he posit ion of t he rescuer over t he
pat ient 's chest . Wit h t his in mind, if a pat ient suff ers a cardiac arrest w hile in a
hospit al bed, in addit ion t o using a backboard, rescuers should also low er t he
bed and/ or use a st ool on w hich t o st and in an eff ort t o ensure t he pat ient 's
chest (and not just t he mat t ress) is compressed t o t he suggest ed dept h of 4 t o 5
cm.
Suggested Readings
2005 American Heart Associat ion guidelines f or cardiopulmonary resuscit at ion
and emergency cardiovascular care, part 4: adult basic lif e support .
Circulat ion 2005; 112: 19–34.
Perkins G D, Benny R, G iles S, et al. Do diff erent mat t resses aff ect t he
qualit y of cardiopulmonary resuscit at ion? I nt ensive Care Med 2003; 29: 2330–
2335.
Boe JM, Babbs CF. Mechanics of cardiopulmonary resuscit at ion perf ormed
w it h t he pat ient on a sof t bed vs a hard surf ace. Acad Emerg Med
1999: 6: 754–757.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 79 - K now the P ac em aker
Alphabet S oup
79
Know the Pacemaker Alphabet Soup
Crit ically ill pat ient s may require urgent or emergent pacing at any t ime during
t heir hospit al course. This may occur more f requent ly in medical or surgical
cardiac int ensive care unit s (I CU), but a w orking know ledge of t he st andard
nomenclat ure is import ant f or all I CU caregivers.
Many w ill care f or pat ient s w it h implant ed pacemakers f or longst anding
condit ions. An increasing number of pat ient s w ill have a pacemaker plus an
Aut omat ic I mplant able Cardiovert er Def ibrillat or (AI CD) or bivent ricular pacing
w ires, as t his becomes more w idespread in t he management of pat ient s w it h
heart f ailure. Many of t hese implant ed pacemakers have sophist icat ed set t ings.
This is in cont rast t o t he much more simplif ied scenario used in an urgent
sit uat ion requiring t ransvenous pacing or epicardial pacing in post –cardiac
surgery pat ient s. How ever, despit e being more simplif ied, t he nomenclat ure is
t he same in bot h implant ed and t emporary pacing.
Watch Out For

The st andard nomenclat ure f or pacing is a f ive-let t er designat ion f or t he mode of
pacing and sensing of int rinsic elect rical (cardiac) act ivit y. I n t he set t ing of
urgent or emergent pacing, only t he f irst t hree let t ers of t he I nt er-Societ y
Commission f or Heart Disease Resources (I CHD) are t ypically used (Table 79. 1).
The f irst let t er ref ers t o t he cardiac chamber paced, t he second let t er
represent s t he chamber sensed, and t he t hird let t er represent s t he mode, or
w hat happens in response t o t he sensing. The lat t er t w o-let t er designat ions are
generally used in int ernal pacemakers and ref lect more sophist icat ed
programming t hat is beyond t he scope of t his chapt er.
The met hod used t o pace pat ient s depends on t he w ires t hat are used (i. e. ,
w het her t here is capabilit y t o pace t he at ria, vent ricles, or bot h) and t he mode
designat ed on t he pacer box. The capabilit y t o pace one or bot h chambers
depends on t he t ype (or number) of w ires placed or t he t ype of t ransvenous
device t hat has been placed. I n t he set t ing of post –cardiac surgery pat ient s, t he
ground w ire is placed in t he posit ive pole of t he pacer box. By convent ion, t he
ground w ire is t he longer of t he t w o vent ricular w ires or a new ly placed skin
lead. For at rial epicardial w ires, eit her can be used as t he grounding w ire. I n t he
set t ing of t ransvenous pacing, t he posit ive lead on t he w ire is placed in t he

posit ive pole of t he pacer box.
TABLE 79-1 PACING NOM ENCLATURE AND M
Cardiac
Cardiac Response
Cham ber
Cham ber Sensed
ICHD Sensed
Paced (First Events
DESIGNAT ION a (Second
Letter (T hird Lett
Letter
Designation) Designatio
Designation)
AAI Atrial Atrial Inhibited
n/a (no
AOO Atrial None
inhibition)
VVI Ventricular Ventricular Inhibited
n/a (no
VOO Ventricular None
inhibition)
V pacing
DVI Dual b Ventricular inhibited by
V sensing
DDI Dual b Dual b Inhibited

n/a (no
DOO Dual b None
inhibition)
a Inter-Society Commission for Heart Disease Resources co

b Dual is both atrial and ventricular chambers.
There are t w o general modes of pacing, asynchronous and synchronous.

Asynchronous pacing w ill result in elect rical out put t o t he myocardium at a f ixed
or set rat e, regardless of w hat t he underlying rat e and rhyt hm are. There is no
f eedback f rom t he myocardium t o t he ext ernal pacing device. This mode is
t ypically used f or at rial pacing because t he at ria are diff icult t o sense since
t hese chambers have less mass. Asynchronous pacing is generally not used f or
vent ricular pacing because it increases t he risk of t he R-on-T phenomenon. I f
t his occurs and t he vent ricle is t riggered w hen it is in a vulnerable phase of t he
cardiac cycle, vent ricular f ibrillat ion can result .
Synchronous or demand pacing is used t o describe modes t hat are responsive t o
t he nat ive rhyt hm. That is, t he nat ive rhyt hm is sensed by t he pacing w ires and
give f eedback t o t he pacer box such t hat it w ill f ire an elect rical current only if
t he nat ive rat e is below t he predet ermined rat e (or rat e set on t he pacer box).
Suggested Readings
Art usio JF, Yao F-SF. Anest hesiology: Problem-orient ed Pat ient Management .
3rd Ed. Philadelphia: Lippincot t Williams & Wilkins; 1993: 199–213.
Baumgart ner WA. The Johns Hopkins Manual of Cardiac Surgical Care. St .
Louis: Mosby; 1994: 208–212.
Fust er V, Alexander RW, Schlant RC. Hurst 's t he Heart , Art eries and Veins.
9t h Ed. New York: McG raw -Hill; 1998: 1023–1052.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 80 - D o not us e C ooling
B lankets to C ool
80
Do not use Cooling Blankets to Cool
A pat ient t hat “needs t o be cooled” is not an inf requent occurrence in t he
int ensive care unit (I CU). I n order t o eff ect ively low er a pat ient 's core body
t emperat ure, an underst anding of t emperat ure physiology is required.
The human body can be roughly divided int o t w o t hermal compart ment s: a core
compart ment , consist ing of t he t runk and head, excluding t he skin, and a
peripheral compart ment , consist ing of t he skin and ext remit ies. The core
t emperat ure is regulat ed by limit ing or increasing heat t ransf er t o t he periphery
t hrough vasoconst rict ion and vasodilat at ion, respect ively. Heat loss f rom t he
peripheral compart ment is regulat ed t hrough changes in skin perf usion (again
t hrough vasodilat at ion or vasoconst rict ion) and by increasing or decreasing t he
product ion of sw eat .
When cooling blanket s are placed over or under a pat ient , t hey increase
sympat het ic t one (w it h concomit ant increase in syst emic vascular resist ance
index) and induce vasoconst rict ion in t he skin. This act ually prevent s heat
dissipat ion and causes an increase in core body t emperat ure, not a decrease.
This physiological response complicat es at t empt s t o induce t herapeut ic
hypot hermia or normot hermia by ot her means of ext ernal cooling. I n addit ion,
heat product ion w ill be increased t hrough t he shivering response and, in lat er
phases, t hrough t he increased met abolism of f at s, carbohydrat es, and prot eins.
As a result t here can be an increase (not a decrease) in body t emperat ure.
TABLE 80-1 M ETHODS OF PERIPHERAL COOLING
HEAT
MET HOD COMMENT S
T RANSFER
Fans Convection Increased
infection risk
Ice packs Conduction Neck, axilla, groin
Cold-water
Evaporation —
sponge bath
There are more eff ect ive w ays t o cool t he core body t emperat ure t hat use t he
w ell-described f our mechanisms of heat t ransf er: convect ion, conduct ion,
evaporat ion, and radiat ion. Some of t he more common w ays are summarized in
Tables 80. 1 and 80. 2.
TABLE 80-2 M ETHODS OF CORE BODY COOLING
HEAT
MET HOD COMMENT S
T RANSFER
Antipyretic Acetaminophen,
—
agents dantrolene
Cooled
intravenous Conduction Neck, axilla, groin
fluids
Cold-water Gastric, bladder,

Conduction
lavage peritoneal
Antishivering Conduction,
Paralysis
maneuvers radiation
Suggested Readings
Creechan T, Vollman K, Kravut ske ME. Cooling by convect ion vs cooling by
conduct ion f or t reat ment of f ever in crit ically ill adult s. Am J Crit Care
2001; 10: 52–59.
Henker R, Rogers S, Kramer DJ, et al. Comparison of f ever t reat ment s in t he

crit ically ill: a pilot st udy. Am J Crit Care 2001; 10: 276–280.
O ’Donnell J, Axelrod P, Fisher C, et al. Use and eff ect iveness of hypot hermia
blanket s f or f ebrile pat ient s in t he int ensive care unit . Clin I nf ect Dis
1997; 24: 1208–1213.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 81 - D o not P lac e Fem or al
Ar ter ial Lines or Venous C atheter s Above the Inguinal Ligam ent or B elow the Thigh C r eas e
81
Do not Place Femoral Arterial Lines or Venous
Catheters Above the Inguinal Ligament or Below
the Thigh Crease
Susanna L. Matsen MD
The majorit y of int ensive care pat ient s require art erial and cent ral venous lines
f or purposes of cont inuous blood pressure monit oring, blood gas sampling, and
t he administ rat ion of large volumes of f luid and vasoact ive subst ances. The
f emoral areas provide an access locat ion f or rapid art erial or venous cannulat ion
in an emergency sit uat ion, or alt ernat ively w hen ot her locat ions are not
available. Despit e t he relat ive accessibilit y and superf icial nat ure of t hese
vessels, t heir cannulat ion can present risks of w hich t he prudent clinician should
be aw are.
Watch Out For

Punct ure sit es f or f emoral lines should lie bet w een t he groin crease and t he
inguinal ligament ; punct ures out side of t hese zones risk complicat ions. The
inguinal ligament st ret ches f rom t he ant erior superior iliac spine dow n t o t he
pubic symphysis and marks t he inf erior border of t he perit oneum. Hence,
punct ures above t his point w ill likely ent er t he perit oneal cavit y, w it h pot ent ial
visceral injury and bact erial seeding. Aspirat ion of ent eric cont ent s or ascit ic
f luid should immediat ely alert t he operat or of t his event ualit y, and a general
surgical t eam should be alert ed.
Conversely, at t empt ing cannulat ion t oo inf eriorly (below t he groin crease) risks
lacerat ion of branch vessels. The common f emoral art ery branches int o t he
superf icial f emoral and deep f emoral art eries at t his point . Similarly, in t he
venous syst em, t he great er saphenous vein joins t he deep f emoral vein roughly 3
cm inf erior t o t he inguinal ligament , f orming t he f emoral vein. Alt hough t he
operat or may obt ain a “f lash” of blood int o t he syringe at a dist al sit e, t here may
be diff icult y t hreading t he w ire t hrough t he junct ion of t he superf icial t o t he
common f emoral art ery. More signif icant ly, t he needle may lacerat e t he f emoral
art ery “t hrough and t hrough” and also injure t he vein, event ually leading t o an
art eriovenous f ist ula (usually >2 days af t er t he injury). A pat ient w it h diminished
dist al pulses, a groin bruit , pulsat ile mass, and pot ent ially a drop in blood
pressure af t er at t empt s at f emoral cannulat ion should be suspect ed of an
art eriovenous (AV) f ist ula, and a high-qualit y vascular duplex st udy should be
ordered. Evidence of
f low bet w een t he art ery and vein should prompt an immediat e vascular surgery
consult .
O t her iat rogenic et iologies of pulsat ile masses in t he groin include f emoral art ery
pseudoaneurysms. These result f rom dissect ion of t he art erial w all by t he needle
or w ire during line placement , leading t o w eakening of t he w all and aneurysmal
dilat at ion. G iven t he risk of grow t h of t he pseudoaneurysm and pot ent ial rupt ure,
t hese should also be ref erred t o vascular surgery.
Undet ect ed bleeds f rom t he f emoral art ery af t er at t empt s at cannulat ion may
lead t o ret roperit oneal hemat omas. Such a bleeding source should alw ays
remain in t he back of t he mind of an int ensive care t eam w it h a pat ient w ho
drops his hemat ocrit w it hout any apparent bleeding source. The blood remains
ext raperit oneal, dissect ing t he pot ent ial space bet w een t he perit oneal sac and
t he ret roperit oneum. Pat ient s may present w it h leg paresis (secondary t o
irrit at ion of t he nerves coursing t hrough t his area), abdominal or groin pain, or
simply a f lank bruise. Alt hough t he blood usually courses post eriorly, it may
dissect ant eriorly int o t he space of Ret zius. A cont inuous drop in hemat ocrit or a
sympt omat ic bleed w it h a comput ed t omography (CT) scan demonst rat ing
ret roperit oneal blood should prompt surgical repair of t he vessel.
O ne f inal not e is t hat t he operat or must be f amiliar w it h t he medial-lat eral
anat omy of t he f emoral vessels, commonly recalled by t he mnemonic NAVEL
w hen moving lat eral t o medial. NAVEL st ands f or nerve most lat erally, t hen
art ery next most medial, t hen vein, empt y space, and l ymphat ics most medially.
Selected Readings
75.
Moore KL, Dalley AF. Clinically O rient ed Anat omy. 4t h Ed. Philadelphia:
Lippincot t Williams & Wilkins; 1999: 541–549.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 82 - C ons ider R es pir ator y
Var iation on the Ar ter ial- Line Monitor Tr ac ing as a S ign of Hypovolem ia
82
Consider Respiratory Variation on the Arterial-
Line Monitor Tracing as a Sign of Hypovolemia
James F. Weller MD
During t he inspirat ory phase of spont aneous vent ilat ion, negat ive int rat horacic
pressure leads t o augment at ion of right vent ricular (RV) f illing. This leads t o a
lef t w ard shif t in t he int ervent ricular sept um, decreasing t he size and compliance
of t he lef t vent ricle. I n addit ion, t he augment ed right vent ricular st roke volume
pools in t he pulmonary circulat ion, causing a f urt her decrease in lef t vent ricular
(LV) f illing volume. Decreased LV volume is associat ed w it h a decrease in LV
st roke volume and syst emic blood pressure. Wit h exhalat ion, blood pooled in t he
right vent ricle and pulmonary circulat ion ret urns t o t he lef t vent ricle; st roke
volume and syst emic blood pressure increase. Under normal condit ions, t he
inspirat ory decrease in art erial blood pressure does not exceed 5 t o 10 mm Hg.
A reversed pulsus paradoxus, or syst olic pressure variat ion, has been described
during mechanical vent ilat ion. The posit ive int rat horacic pressure of mechanical
inspirat ion decreases venous ret urn t o t he right heart and increases RV
af t erload. I n addit ion, increased alveolar pressure “squeezes” blood f rom t he
pulmonary circulat ion int o t he lef t vent ricle. The result is an inspirat ory decrease
in RV st roke volume and an inspirat ory increase in LV st roke volume and
syst emic art erial blood pressure. G iven st at ic art erial compliance, syst emic
art erial pulse pressure (t he diff erence bet w een syst olic and diast olic blood
pressure) is direct ly proport ional t o st roke volume. Thus, any changes in venous
ret urn or pulmonary blood f low t hat lead t o an increase in LV st roke volume w ill
also lead t o an increase in pulse pressure.
Under hypovolemic condit ions, mechanical respirat ory eff ect s on venous ret urn,
st roke volume, and art erial blood pressure are accent uat ed f or at least f our
reasons: (i) t he vena cava is more collapsible; (ii) an underf illed right at rium is
more compliant and, t heref ore, more suscept ible t o changes in pleural pressure;
(iii) more of t he lung is in West zone I (Palveolar >P ar ter ial >P venous ), w here
respirat ory eff ect s on pulmonary blood f low are most pronounced; and (iv) bot h
vent ricles are more sensit ive t o changes in preload w hen operat ing on t he st eep
port ion of t he Frank-St arling curve. All f our of t hese mechanisms accent uat e t he
mechanical inspirat ory decrease in RV st roke volume, suggest ing t hat it is t he
subsequent expirat ory decrease in LV out put
t hat is most ly responsible f or t he prof ound respirat ory variat ion in blood
pressure encount ered in t he set t ing of hypovolemia. I f t he diff erence is great er
t han 10 mm Hg, hypovolemia must be considered as t he most common cause.
Suggested Readings
Michard F. Changes in art erial pressure during mechanical vent ilat ion.
Thomas S, Kramer J, eds. Manual of Cardiac Anest hesia. 2nd Ed. New York:
Churchill Livingst one; 1993: 134–135.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 83 - D o Not Us e Low -
Molec ular - W eight Hepar in in P atients w ith E pidur als
83
Do Not Use Low-Molecular-Weight Heparin in
Patients with Epidurals
Awori J. Hayanga MD
Elliott R. Haut MD
Neuraxial anest hesia and analgesia (eit her by perioperat ive spinal or
post operat ive epidural cat het er) has been show n t o signif icant ly improve pain
cont rol and ot her out comes af t er major surgery as w ell as f or t rauma pat ient s
w it h mult iple rib f ract ures. Simult aneously, dat a have show n low er rat es of
venous t hromboembolic event s in pat ient s receiving chemical venous
t hromboembolic event prophylaxis w it h some f orm of ant icoagulant (i. e. ,
unf ract ionat ed or low -molecular-w eight heparin [ LMWH] ). How ever, t he
concomit ant use of ant icoagulant s and spinal or epidural anest hesia can result in
t he rare, but pot ent ially devast at ing, spinal hemat oma.
Pat ient s w it h signif icant uncorrect ed bleeding disorders and t hose receiving
ant icoagulat ion are at signif icant ly elevat ed risk of developing spinal hemat oma
in t he set t ing of neuraxial anest hesia. The diagnosis of spinal hemat oma is
complicat ed by t he concealed nat ure of t he bleeding. A high index of suspicion
must be maint ained. Pat ient s w ho have neurologic f indings (e. g. , w eakness,
decreased sensat ion, or incont inence) af t er undergoing a lumbar punct ure or
receiving spinal or epidural analgesia require emergent evaluat ion f or possible
spinal hemat oma. Back pain is a common f inding but is not present in all cases.
Det ailed physical examinat ion should also include a rect al exam t o evaluat e t one.
These physical examinat ion f indings should not be simply at t ribut ed t o t he
epidural anest het ic. A comput ed t omography (CT) scan can of t en be obt ained
rapidly and may yield t he diagnosis. How ever, magnet ic resonance imaging
(MRI ) or myelography may be necessary f or def init ive diagnosis.
The t reat ment f or a sympt omat ic pat ient is urgent surgical int ervent ion, usually a
laminect omy, and evacuat ion of t he blood. Expedient decompression of t he spinal
hemat oma (w it hin 8 t o 12 hours of onset of sympt oms) is essent ial t o avoid
permanent loss of neurologic f unct ion.
Current guidelines f rom t he Second American Societ y of Regional Anest hesia
(ASRA) Consensus Conf erence on Neuraxial Anest hesia and Ant icoagulat ion
include t he f ollow ing suggest ions relat ed t o t he use of LMWH:
1. I n pat ient s receiving preoperat ive prophylact ic LMWH, insert ion of t he spinal
needle should be delayed f or at least 10 t o 12 hours af t er t he most recent
LMWH prophylaxis dose. For pat ient s on t reat ment -dose LMWH, t he delay
should be 24 hours f rom t he last dose.
2. Single-dose anest het ic is pref erable t o cont inuous epidural anest hesia.
3. LMWH dosing should be delayed f or at least 2 hours af t er spinal needle
placement . I f a hemorrhagic aspirat e occurs during t he init ial spinal needle
placement , LMWH should not be given f or 24 hours.
4. Single-daily LMWH dosing may be saf ely used w it h indw elling cat het er-based
cont inuous epidural anest hesia as long as t he f irst post operat ive dose is
given at least 6 t o 8 hours af t er surgery. The second dose should be at least
24 hours lat er. Cat het er removal should be a minimum of 10 t o 12 hours
af t er t he last dose.
5. Tw ice-daily LMWH dosing should not be used in pat ient s w it h indw elling
cat het ers because of t he pot ent ial associat ion w it h increased risk of spinal
hemat oma.
Any decision t o implement LMWH prophylaxis in t he presence of an indw elling

cat het er must be made w it h ext reme care, and ext reme vigilance of t he pat ient 's
neurologic st at us is w arrant ed.
Suggested Readings
Douket is JD, Kinnon K, Crow t her MA. Ant icoagulant eff ect at t he t ime of
epidural cat het er removal in pat ient s receiving t w ice-daily or once-daily low -
molecular-w eight heparin and cont inuous epidural analgesia af t er ort hopedic
surgery. Thromb Haemost 2002; 88: 37–40.
Horlocker TT, Wedel DJ, Benzon H, et al. Regional anest hesia in t he

ant icoagulat ed pat ient : def ining t he risks (t he Second ASRA Consensus
Conf erence on Neuraxial Anest hesia and Ant icoagulat ion). Reg Anest h Pain
Med 2003; 28: 172–197.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 84 - Adm inis ter E pidur al
tes t D os es
84
Administer Epidural test Doses
Rahul G . Baijal MD
The spinal canal cont ains t he spinal cord and it s coverings (or meninges), f at t y
t issue, and t he venous plexus. I t ext ends f rom t he f oramen magnum t o t he sacral
hiat us. The meninges are composed of t hree layers: t he pia mat er, t he arachnoid
mat er, and t he dura mat er. The pia mat er is adherent t o t he spinal cord, w hereas
t he arachnoid mat er is adherent t o t he t ough, f ibroelast ic dura mat er. The
subarachnoid space, locat ed bet w een t he pia mat er and t he arachnoid mat er,
cont ains cerebrospinal f luid. The epidural space is a pot ent ial space, cont aining
f at , alveolar t issue, nerve root s, and a net w ork of art eries and venous plexus.
No f luid is present in t he epidural space. The epidural space ext ends cephalad
f rom t he f oramen magnum and caudad t o t he sacral hiat us. The space is bound
ant eriorly by t he post erior longit udinal ligament , lat erally by t he int ervert erbral
f oramen, and post eriorly by t he ligament um f lavum. The spinal cord ext ends f rom
t he f oramen magnum t o L1 in adult s and L3 in children. The ant erior and
post erior nerve root s at each spinal level exit t he int ervert ebral f oramina f orming
spinal nerves f rom C1 t o S5.
Epidural anest hesia is produced by inject ion of a local anest het ic solut ion int o
t he epidural space. Cont inuous epidural anest hesia is achieved by placement of
a cat het er int o t he epidural space. The epidural space is ident if ied by t he “loss
of resist ance” t echnique, ref lect ing passage f rom an area of high resist ance in
t he ligament um f lavum t o an area of low resist ance in t he epidural space. Af t er
t he epidural needle is posit ioned in t he ligament um f lavum, a glass syringe w it h a
f reely movable plunger is at t ached t o t he needle and cont inuous pressure is
applied t o t he plunger. I f t he needle is posit ioned correct ly in t he ligament um
f lavum, t he plunger should not inject w hen pressure is applied. When t he needle
passes int o t he epidural space, a loss of resist ance is achieved, and t he plunger
w ill easily inject .
Watch Out For

Follow ing successf ul placement of t he needle in t he epidural space, a cat het er is
placed t hrough t he needle and advanced 3 t o 5 cm int o t he epidural space t o
allow repeat ed and increment al inject ions of local anest het ic solut ion. The
short er dist ance t he cat het er is advanced, t he more likely t he cat het er may
become dislodged. Conversely, t he f art her t he
cat het er is advanced, t he more likely t he cat het er may ent er an epidural vein,
punct ure t he dura mat er and ent er t he subarachnoid space, exit t he
int ervert ebral f oramen, or enclose a nerve root . Even a cat het er successf ully
placed is subject t o migrat ion. Movement of a pat ient w it h an epidural cat het er
may dislodge t he cat het er f rom it s init ial appropriat e posit ion.
Epidural analgesia has t he pot ent ial t o produce local anest het icinduced syst emic
t oxicit y t hrough unint ent ional administ rat ion int o an epidural vein. Addit ionally,
unint ent ional administ rat ion of an epidural dose of local anest het ic in t he
subarachnoid space w ill produce a rapid spinal, or subarachnoid, analgesia. A
t est dose f ollow ing placement of t he epidural cat het er is designed t o det ect bot h
subarachnoid and int ravascular inject ion. The classic t est dose combines 3 mL of
local anest het ic solut ion and epinephrine, t ypically 3 mL of 1. 5% lidocaine
(secondary t o rapid onset ) w it h 1: 200, 000 epinephrine. The lidocaine, if inject ed
int o t he subarachnoid space, w ill produce a rapid spinal analgesia t hat may be
manif est as analgesia t o t he saddle area and cardiopulmonary depression
secondary t o sympat het ic blockade. The lidocaine, if inject ed int ravascularly, w ill
aff ect primarily t he cent ral nervous and cardiovascular syst ems, w it h t he cent ral
nervous syst em aff ect ed at low er levels. Four t o seven t imes t he dose of local
anest het ic necessary t o produce convulsions is required t o produce
cardiovascular depression. The epinephrine, if inject ed int ravascularly, w ill
produce a 20% or more rise in heart rat e. Reapplicat ion of t he t est dose t hrough
t he cat het er is paramount t o det ect unint ent ional inject ion int ravascularly or in
t he subarachnoid space.
Suggested Readings
Miller RD, ed. Miller's Anest hesia. 6t h Ed. Philadelphia: Elsevier; 2005: 1677–
1678.
Morgan EG , Maged SM, eds. Clinical Anest hesiology. New York: McG raw -Hill;
2002: 270–271.
St oelt ing RK, Miller RD, eds. Basics of Anest hesia. Philadelphia: Elsevier;
2002: 179–181.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 85 - C ons ider an
Intr avenous Naloxone D r ip for Tr eatm ent of P r ur itus As s oc iated w ith E pidur al Analges ia
85
Consider an Intravenous Naloxone Drip for
Treatment of Pruritus Associated with Epidural
Analgesia
Cont inuous epidural analgesia can be benef icial f or post operat ive pain cont rol.
Perioperat ively, it may at t enuat e t he physiologic response t o surgery. This
pot ent ially decreases morbidit y and mort alit y w hen compared w it h syst emic
opioids by improving gast roint est inal f unct ion, decreasing pulmonary
complicat ions, decreasing myocardial inf arct ions, improving ment al st at us, and
decreasing t hrombot ic event s. I t is indicat ed f or pat ient s undergoing almost all
surgeries ext ending f rom t he t horax t o t he low er ext remit ies. There are f ew
cont raindicat ions; t hese include pat ient ref usal, coagulopat hy, and local inf ect ion
at t he insert ion sit e. Alt hough cont inuous epidural analgesia is quit e eff ect ive
w hen properly managed, care must be t aken t o balance t he benef icial eff ect s
w it h t he many possible side eff ect s.
There are several choices of medicat ions and medicat ion combinat ions t hat can
be used f or epidural inf usion. Local anest het ics can produce an excellent
sensory block but may also cause a mot or block, hypot ension, and/ or
bradycardia. O pioids produce good analgesia but may be associat ed w it h
nausea, vomit ing, urinary ret ent ion, and prurit us. Theref ore, f or post operat ive
pain cont rol, t he combinat ion of a local anest het ic and an opioid is of t en t he best
choice. There are many diff erent local anest het ics and opioids t hat are available
and t he choice of w hat t o use should be t ailored t o t he pat ient and sit uat ion.
Adjuvant medicat ions include clonidine and epinephrine, but t heir ut ilit y has not
been proven t o be eff ect ive.
Signs and Symptoms

The most f requent side eff ect of epidural opioid use is prurit us. The prevalence
in t he lit erat ure has been report ed t o be 0% t o 100%, but experienced providers
anecdot ally report it s occurrence in about 60% of pat ient s. The cause is
unknow n, but it is not relat ed t o hist amine release. There may be cent ral
act ivat ion of an it ch cent er in t he low er medulla involving t he t rigeminal nucleus.
Prurit us may occur at any dose, but severit y seems t o be dose-relat ed.
Prurit us is an undesired eff ect of cont inuous epidural opioid use but is not
dangerous t o t he pat ient and should not prevent t he use
of perioperat ive epidurals. I t can be cont rolled w it h an int ravenous inf usion of an
opioid ant agonist such as naloxone (Narcan). The diff icult y in t his st rat egy is
f inding t he dose t hat decreases t he prurit us w it hout compromising t he analgesia.
A good st art ing dose is 1 t o 5 µg/ kg/ hr, w hich can of t en accomplish t his goal.
There is also some evidence t hat using an opioid agonist -ant agonist (e. g. ,
nalbuphine, but orphanol) may cont rol prurit us w it h bet t er analgesia preservat ion.
An addit ional benef it of t hese medicat ions is t hat t hey may decrease t he
incidence of respirat ory depression associat ed w it h epidural opioids t hat can be
f at al if unnot iced.
Suggested Readings
Caldw ell M, Wu C. Eff ect of post -operat ive analgesia on pat ient morbidit y.
Best Pract Res Clin Anest hesiol 2002; 16: 549–563.
Miller RD, ed. Miller's Anest hesia. 6t h Ed. Philadelphia: Elsevier; 2000: 2737–
2744.
Raj PP, ed. Pract ical Management of Pain. 3rd Ed. New York: Mosby;
2000: 689–707.
Richman J, Wu C. Epidural analgesia f or post operat ive pain. Anest hesiol Clin
Nort h America 2005; 23: 125–140.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 86 - B e Metic ulous w hen
D os ing B upivac aine in P atients w ith B oth E pidur al and P leur al C atheter s
86
Be Meticulous when Dosing Bupivacaine in
Patients with Both Epidural and Pleural
Catheters
Many pat ient s being admit t ed t o t he int ensive care unit (I CU) af t er t horacic
surgery w ill have bot h epidural and pleural cat het ers f or post operat ive pain
cont rol. The epidural cat het er w ill usually be at t ached t o a pat ient -cont rolled
analgesia (PCA) device w it h bot h a cont inuous and a demand dose; pain
medicat ion administ ered via a pleural cat het er is usually given as a cont inuous
inf usion.
This can be an area of pot ent ial conf usion f or I CU st aff because epidural and
pleural cat het ers of t en look ident ical; a pleural cat het er is usually a cat het er
t aken f rom an epidural kit and insert ed int o t he pleural space. I n addit ion t o t he
ident ical appearance, t he epidural and pleural cat het ers may be posit ioned next
t o each ot her and covered w it h dressings, making it ext remely easy f or t he
nursing st aff t o mist ake one cat het er f or t he ot her. To avoid t his conf usion, t he
epidural and pleural cat het ers should be clearly marked near t he sit e of
inject ion.
G iven t he mist akes t hat have occurred w it h epidural and int ravenous (I V)
cat het ers, such as inadvert ent I V administ rat ion of bupivacaine int ended f or an
epidural cat het er or t he epidural administ rat ion of pot assium or lipid emulsion, it
is mandat ory t hat met iculous care should also be used w hen dosing epidural and
pleural cat het ers. O ne drug t hat is part icularly problemat ic if it is administ ered in
t he w rong cat het er in t he w rong dose is bupivacaine, w hich is a local anest het ic
used in bot h epidural and pleural cat het ers f or post operat ive pain cont rol.
Bupivacaine w orks by binding t o t he sodium channel of a nerve cell and
prevent ing propagat ion of t he nerve impulse along t he cell membrane. I t t hen
dissociat es very slow ly f rom t he sodium channel.
The usual scenario w here bupivacaine causes t rouble is w hen t he epidural dose
is inject ed int o t he pleural cat het er. This can cause a severe t oxicit y. Syst emic
absorpt ion of bupivacaine f rom a pleural cat het er (virt ually 100%) is great er t han
f rom an epidural cat het er because of t he increased vascularit y of t he pleural
space as compared w it h t he epidural space. Signs of t oxicit y are usually
manif est ed by t he cent ral nervous syst em (CNS) bef ore t he cardiovascular
syst em. CNS manif est at ions include light headedness, t innit us, perioral
numbness, disorient at ion, muscle t w it ching, t onic-clonic seizures,
unconsciousness, and respirat ory arrest . Cardiovascular manif est at ions
include hypot ension, sinus bradycardia, sinus node arrest , vent ricular
arrhyt hmias, and circulat ory arrest . Because of bupivicaine's slow dissociat ion
f rom t he sodium channel, resuscit at ion af t er cardiovascular collapse due t o
bupivacaine t oxicit y can be diff icult , at t imes requiring cardiopulmonary bypass
unt il bupivacaine dissociat es f rom cardiac sodium channels.
O ne f inal not e is t hat bupivacaine t oxicit y can f ollow inadvert ent int ravascular
inject ion or syst emic absorpt ion f rom t he epidural sit e w it h migrat ion of t he
cat het er.
Suggested Readings
Miller R, ed. Miller's Anest hesia. 6t h Ed. Philadelphia: Churchill Livingst one;
2005: 1648.
Sculz-St ubner S. Regional analgesia in t he crit ically ill. Crit Care Med
2005; 33: 1400â 1 407.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 87 - D o not D is m is s R ib
Fr ac tur es as Tr ivial and C ons ider an E pidur al C atheter for P ain C ontr ol in Multiple R ib Fr ac tur es
87
Do not Dismiss Rib Fractures as Trivial and
Consider an Epidural Catheter for Pain Control
in Multiple Rib Fractures
Bryan A. Cotton MD
Rib f ract ures are t he most common injury associat ed w it h blunt chest t rauma,
w it h a report ed incidence of at least 10% t o 25%. Mult iple rib f ract ures are
associat ed w it h high-energy mechanisms such as mot or vehicle crashes and
pedest rians st ruck by aut omobiles. I n t he elderly, how ever, t hey may be
associat ed w it h mechanisms such as low -level f alls. These injuries are
f requent ly, and inappropriat ely, dismissed as t rivial w it h inadequat e at t ent ion
given t o t he cont rol of pain in t hese pat ient s. Rib f ract ures are likely t o be
associat ed w it h concomit ant injuries such as pulmonary cont usions and hemo-
pneumot horaces (f requent ly requiring t ube t horacost omy). I n addit ion, rib
f ract ures may also serve as a marker f or signif icant solid-organ injury. Lef t -sided
rib f ract ures are associat ed w it h a 15% t o 20% risk of spleen injuries, w hile
right -sided rib f ract ures are associat ed w it h a 10% t o 15% risk of hepat ic injury.
Watch Out For

Mult iple rib f ract ures usually cause signif icant pain and limit ed chest -w all
excursion and t hus rest rict t he pat ient 's abilit y t o cough and breat he deeply. This
result s in poor clearance of secret ions, at elect asis, and decreased f unct ional
residual capacit y w it h t he clinical correlat es of t his being hypoxia, vent ilat ion-
perf usion mismat ches, and respirat ory f ailure. I n t he elderly in part icular, t his
result s in increased lengt h of vent ilat or days, int ensive care unit (I CU) lengt h of
st ay, and pneumonia rat es. As f ew as t hree rib f ract ures have been associat ed
w it h an approximat e 20% mort alit y and a 30% risk of developing pneumonia in
t hose more t han 65 years of age. I n f act , t he risk of mort alit y increases t o
almost 35% and t he pneumonia rat e t o great er t han 50% w hen t hese pat ient s
sust ain six or more rib f ract ures. How ever, pat ient s age 45 and older w it h more
t han f our rib f ract ures are also at a dramat ically increased risk of pulmonary
complicat ions and deat h. Theref ore, aggressive pulmonary t oilet and adequat e
pain cont rol (as evidenced by t he abilit y t o cough, breat he deeply, and
reproducibly perf orm large inspirat ory volumes on spiromet ry) should not be
limit ed t o t he elderly t rauma pat ient w it h rib f ract ures.
What to Do
Several opt ions exist t o achieve adequat e pain cont rol in t hese pat ient s,
including t he use of int ravenous opiat es w it h or w it hout nonst eroidal ant i-
inf lammat ory drug (NSAI D) supplement at ion, int ercost al blocks, paravert ebral
blocks or inf usion cat het ers, and epidural analgesia. Compared w it h ot her f orms
of analgesia, numerous st udies have not ed superior pain cont rol w hen pat ient s
w it h rib f ract ures receive t imely placement of an epidural cat het er. Even w hen
compared t o pat ient s w it h low er chest -injury scores and no epidural cat het er,
pat ient s w it h higher injury scores receiving an epidural have signif icant ly low er
morbidit y and mort alit y. Specif ically, t he use of epidural analgesia f or pain
cont rol is associat ed w it h a decreased risk of nosocomial pneumonia and short er
durat ion of mechanical vent ilat ion. The great est impact of epidural analgesia (as
compared w it h int ravenous opiat es) appears t o be among t hose w it h f ive or more
f ract ures and in t he elderly populat ion. Despit e t he overw helming evidence
support ing t he use of epidural analgesia, a recent evaluat ion of t he Nat ional
Trauma Dat a Bank (NTDB) not ed t his modalit y is signif icant ly underut ilized. I n
some cent ers, t he lack of available experienced (or commit t ed) anest hesia
personnel is t he single great est limit ing f act or t o obt aining an epidural f or pain
cont rol.
Epidural cat het er placement and analgesia are associat ed w it h t he risk of
hypot ension, dural punct ure, spinal cord injury, and urinary ret ent ion. The abilit y
t o ut ilize epidural analgesia is of t en limit ed by concomit ant injuries and pre-
exist ing illnesses. These include acut e spine f ract ure, previous spine def ormit y,
impaired ment al st at us, hemodynamic inst abilit y, coagulopat hy, and chest -w all
inf ect ion and/ or sof t -t issue injury. Enoxaparin and ot her low -molecular-w eight
heparins should be held 24 hours prior t o placement of any f orm of neuraxial
anest hesia (epidural) and not rest art ed unt il 24 hours af t er removal. During t his
int erval, subcut aneous heparin (5, 000 unit s every eight hours) should be ut ilized.
I n addit ion t o convent ional int ravenous opiat es or pat ient -cont rolled analgesia
(PCA), ot her analgesia met hods such as int ercost al nerve (rib) blocks,
cont inuous paravert ebral blocks, and int rapleural blocks exist and are quit e
f easible. I nt ercost al nerve blocks are usually perf ormed w it h a local anest het ic
such as bupivacaine, as eit her a bolus dose t hat may be repeat ed or a
cont inuous inf usion via cat het er. Pain cont rol w it h bolus inject ion is usually
eff ect ive f or 8 t o 24 hours w it h each inject ion. St udies have demonst rat ed t hat
t his approach, w het her administ ered as a single inject ion or via cont inuous
cat het er, provide suff icient pain cont rol in pat ient s w it h mult iple rib f ract ures.
This approach carries a small risk of pneumot horax and is not
suit able f or post erior rib f ract ures. Thoracic paravert ebral inf usions have been
show n t o be an eff ect ive met hod of providing pain cont rol in pat ient s w it h bot h
unilat eral, as w ell as bilat eral, mult iple rib f ract ures. This is support ed by t he
observat ion of a sust ained improvement in pulmonary f unct ion t est ing and
oxygenat ion. Paravert ebral blocks are saf er and easier t o perf orm t han t horacic
epidurals and are not associat ed w it h hemodynamic inst abilit y or urinary
ret ent ion. This approach also carries a small risk of pneumot horax. Pleural-
based anest het ic administ rat ion lacks t he pot ent ial f or cent ral nervous syst em
depression as w ell as t he need f or repeat ed inject ions. This met hod, how ever,
has decreased eff icacy in t he presence of hemot horaces or pleural eff usions and
requires int errupt ion of chest -t ube drainage.
O ne f inal not e is t hat regardless of t he modalit y used, adjunct s such as
int ermit t ent posit ive pressure breat hing (I PPB) or int rapulmonary percussive
vent ilat ion (I PV) should be ut ilized early in t he care of t hese pat ient s.
Suggested Readings
Bulger EM, Arneson MA, Mock CN, et al. Rib f ract ures in t he elderly. J
Trauma 2000; 48: 1040–1046.
Flagel BT, Luchet t e FA, Reed RL, et al. Half -a-dozen ribs: t he breakpoint f or
mort alit y. Surgery 2005; 138: 717–723.
Holcomb JB, McMullin NR, Kozar RA, et al. Morbidit y f rom rib f ract ures
increases af t er age 45. J Am Coll Surg 2003; 196: 549–555.
Karmakar MK, Ho AM. Acut e pain management of pat ient s w it h mult iple
f ract ured ribs. J Trauma 2003; 54: 615–625.
Pain management in blunt t horacic t rauma: an evidence-based out come

evaluat ion. East ern Associat ion f or t he Surgery of Trauma; 2004. Available at
ht t p: / / east . org/ t pg/ painchest . pdf
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 88 - Us e a Tw o- S tep
Tec hnique w ith R adiogr aphic C onfir m ation w hen P lac ing a Feeding Tube
88
Use a Two-Step Technique with Radiographic
Confirmation when Placing a Feeding Tube
Placement of a f eeding t ube is not w it hout peril t o t he pat ient and requires a
high level of concern on t he part of t he pract it ioner, as illust rat ed in t he f ollow ing
scenario.
Case
Mr. JCP w as a 68-year-old male w ho w as post operat ive day f our f rom an aort ic-
valve replacement complicat ed by a st roke. Despit e having no previous hist ory of
pulmonary disease, his airw ay pressures on t he vent ilat or w ere persist ent ly high,
as not ed on morning rounds. The nurse f rom t he night shif t not ed a signif icant
amount of residual f rom his f eeding t ube. She w as inst ruct ed t o cont inue t he
f eeds because t he st omach “makes a lot of secret ions. ” O n radiology rounds, a
signif icant lef t -sided pneumot horax w as not ed w it h t he f eeding t ube posit ioned in
t he lef t pleural space.
Discussion
Feeding t ubes are commonly placed in t he int ensive care unit (I CU) f or ent eral
f eeding in t he pat ient w ho is not able t o orally t ake nut rit ion. Many t ypes of
t ubes are available t o accomplish t his. O ne t ype of t ube t hat is commonly used
is t he f lexible small bore/ st ylet -guided t ube (e. g. , Dobhoff ). These t ubes have
t he advant age of being small and comf ort able. They are more suit able f or
medium-t erm f eeding. Because t hese t ubes are small and f lexible, t hey require a
met al st ylet t o give t hem enough rigidit y f or placement . Unf ort unat ely, t his
rigidit y and small size make t hem ideal f or inadvert ent bronchial placement . This
occurs in about 4. 4% of f eeding-t ube placement s. A correct ly posit ioned
endot racheal t ube w it h t he cuff inf lat ed does not preclude placement int o t he
main-st em bronchus. I t has been post ulat ed t hat t he endot racheal t ube may act
as a guide f or t he f eeding t ube t o f ollow. Bot h recognized and unrecognized, t he
result s can be devast at ing and complicat ions include pneumot horax,
“isocalot horax, ” pneumonia, empyema, sepsis, and deat h.
What to Do
To minimize t hese pot ent ially f at al complicat ions, a t w o-st ep radiographic
met hod is commonly used t o det ect bronchial-t ube placement . The t w o-st ep
t echnique involves t aking t w o radiographs, one af t er part ial insert ion of t he t ube
and t he ot her af t er f ull insert ion of t he t ube. The purpose of t his process is t o
conf irm t hat t he f eeding t ube is in t he esophagus bef ore f ully advancing it t o
avoid punct uring lung t issue and causing a pneumot horax. I n t he f irst st ep, t he
f eeding t ube is advanced t o a point past t he carina, but bef ore ent ering t he
bronchials, and a radiograph is t aken. I f t he f eeding t ube is in t he bronchus, it
w ill be “angled” t ow ard one of t he main-st em bronchi. I f it is in t he esophagus, it
w ill be midline. Af t er t he f eeding t ube is conf irmed t o be midline and below t he
carina, a second radiograph is t aken t o conf irm t hat it is in t he st omach (as
opposed t o curled back int o t he esophagus). I mplement at ion of t his prot ocol
signif icant ly reduces risk of pneumot horax (0. 38% vs. 0. 09%).
To summarize:
1. Prior t o placement , t he t ube should be measured t o be just below t he carina

(t ypically 30 cm).
2. A radiograph is obt ained at t he measured lengt h (est imat ed carina) t o
conf irm t hat t he st ylet ed t ube is in t he midline posit ion and below t he carina
(i. e. in t he esophagus and not t he airw ay).
3. When t here is uncert aint y concerning t he t ube posit ion, radiology
consult at ion should be obt ained.
4. The t ube can t hen be advanced w it h t he st ylet unt il in appropriat e posit ion.
5. A second x-ray should be perf ormed t o conf irm t hat t he st ylet ed t ube is in
t he st omach/ duodenum.
6. O nce in appropriat e posit ion, t he st ylet should be removed and cannot be
reinsert ed f or any reason.
7. Feedings can be st art ed if t he t ube is in t he st omach/ duodenum at t he
discret ion of t he ordering physician and at t ending.
Suggested Readings
Mardenst ein EL, Simmons RL, O choa JB. Pat ient saf et y: eff ect of inst it ut ional
prot ocols on adverse event s relat ing t o f eeding t ube placement in t he
crit ically ill. J Am Coll Surg 2004; 199: 39–47.
Roubenoff R, Ravich WJ. Pneumot horax due t o nasogast ric f eeding t ubes.
Report of f our cases, review of t he lit erat ure, and recommendat ions f or
prevent ion. Arch I nt ern Med 1989; 149: 184–188.
Torringt on KG , Bow man MA. Fat al hydrot horax and empyema complicat ing a
malposit ioned nasogast ric t ube. Chest 1981; 79: 240–242.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 89 - P er for m D oppler
Ultr as ound befor e P lac ing S equential C om pr es s ion D evic es
89
Perform Doppler Ultrasound before Placing
Sequential Compression Devices
Deep vein t hromboses (DVTs) are ext remely common in many groups of
hospit alized pat ient s and cont ribut e t o pat ient morbidit y and mort alit y by t heir
propensit y t o embolize t o t he lung, w here t hey can lead t o severe hypoxemia,
increased dead space, respirat ory insuff iciency, and right -sided heart f ailure w it h
progression t o cardiorespirat ory arrest . Many DVTs develop in t he deep pelvic
veins t hough a subst ant ial number originat e in t he deep veins of t he t high. DVTs
of t he low er leg below t he knee are common also but rarely embolize.
Several f act ors promot e t he t endency t o develop DVTs. I ll pat ient s t end t o be
hypercoagulable, especially t hose w it h cert ain diseases such as cancer and
pancreat it is. This combined w it h immobilit y and venous st asis great ly enhances
t he likelihood of developing clot . Development of clot is a surprisingly rapid
event . Pat ient s undergoing surgery have been show n t o develop DVTs on t he
operat ing t able short ly af t er induct ion of anest hesia.
Prevent ion of DVTs or pulmonary embolisms (PEs) is a major concern f or
hospit alized pat ient s. There are several modalit ies t hat are used t o prevent t his
complicat ion; recent evidence suggest s t hat t he ones chosen f or an individual
pat ient should be based on risk-st rat if icat ion prof iles. These modalit ies include
subcut aneously inject ed heparin, heparinoids such as enoxaparin,
t hromboembolic det errent (TED) st ockings, and sequent ial compression devices
(SCDs). While heparin and heparinlike compounds w ork pharmacologically, TED
st ockings and SCDs w ork mechanically by compressing t he venous st ruct ures in
t he ext remit ies. The primary mechanism of act ion f or reducing t he risk of clot
f ormat ion is t he release of t issue-t ype plasminogen act ivat or (t PA) f rom t he
endot helium in response t o t he mechanical compression of t he veins. The
f ormat ion of plasmin is enhanced, w hich presumably promot es t he lysis of any
clot t hat is init iat ed. This is w hy st udies have show n t hat placing TED st ockings
or SCDs on t he arms of a pat ient is just as eff ect ive as placing t hem on t he legs.
Secondarily, t he compression of t he veins helps t o prevent st asis, w hich is
t hought t o promot e clot t ing.
Watch Out For

This mechanical eff ect of compressing and propelling blood by TEDs/ SCDs may
have delet erious eff ect s as w ell. I f t he pat ient already has a clot burden in t he
legs, t he main concern is t hat t he mechanical compression of t he deep vein by
placement of t he TEDs/ SCDs may dislodge t he clot leading t o a PE. For t his
reason, t he package insert on t he most commonly used brands of SCDs direct
t he clinician t o remove t he device f rom t he ext remit y w here t here is a know n
DVT. Also, it is import ant t o assess w het her any clot burden already exist s prior
t o placing t he compression devices. For ambulat ory pat ient s ent ering t he
hospit al, t his is usually not an issue; ideally, t he SCDs w ould have been placed
and maint ained f rom t he out set of admission. How ever, t his may not alw ays
happen and w hen t hey are t aken off f or procedures or t est s, it is not unusual f or
t heir replacement t o be f orgot t en. During t his unprot ect ed t ime period, clot
burden may already have st art ed t o develop.
What to Do
Thus w hen init ial placement of TEDs or SCDs w as not perf ormed, t hey have not
been replaced w it hin approximat ely 6 hours af t er t heir removal, or a pat ient is
received in t ransf er w it h an uncert ain hist ory of clot burden, many experienced
clinicians w ill obt ain a venous Doppler st udy t o rule out t he presence of DVT
prior t o applicat ion or reapplicat ion of t he devices. As none of t he modalit ies f or
prevent ing DVTs are 100% eff ect ive, it w ould seem prudent t o also perf orm
Doppler st udies on pat ient s receiving pharmacological DVT prophylaxis w hen
mechanical devices are t o be added t o t he prophylact ic regimen as w ell as af t er
a period of immobilit y or in pat ient s judged t o be at high risk of DVT.
Suggested Readings
Caprini JA. Thrombosis risk assessment as a guide t o qualit y pat ient care.
Dis Mon 2005; 51: 70–78.
G oldhaber SZ, Turpie AG . Prevent ion of venous t hromboembolism among

hospit alized medical pat ient s. Circ 2005; 111: e1-e3.
Turpie AG , Chin BS, Lip G Y. Venous t hromboembolism: pat hophysiology,

clinical f eat ures and prevent ion. BMJ 2002; 325: 887–890.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 90 - C ons ider C hanging the
Foley C atheter w hen P atient has a Ur inar y Tr ac t Infec tion
90
Consider Changing the Foley Catheter when
Patient has a Urinary Tract Infection
Melissa S. Camp MD
Urinary t ract inf ect ions (UTI s) are one of t he most common nosocomial inf ect ions
in int ensive care unit (I CU) pat ient s. They almost alw ays occur in t he set t ing of a
Foley cat het er and t he risk of developing a UTI increases w it h t he lengt h of t ime
t he Foley cat het er is lef t in place. The risk of UTI is approximat ely 5% per day
w it h a Foley cat het er.
The diagnosis of a urinary t ract inf ect ion is made based on urinalysis and urine
cult ure. Findings on t he urinalysis t hat are suggest ive of a UTI include pyuria and
bact eriuria. A urine cult ure w it h great er t han 105 bact eria per mL is consist ent
w it h a diagnosis of a UTI . The most common organism responsible f or UTI s is
Escheri chi a col i. O t her common organisms include Proteus, Kl ebsi el l a,
Pseudomonas, Enterococcus, Staphyl ococcus epi dermi di s, Staphyl ococcus
aureus, and Candi da. Ant ibiot ic t reat ment f or a UTI should be based on t he
sensit ivit ies f rom cult ure result s, and recommended durat ion of t herapy is 7 t o
14 days depending on t he severit y of t he inf ect ion. Urosepsis, f or example,
should be aggressively t reat ed w it h a 14-day course of ant ibiot ics.
What to Do
To reduce t he risk of developing urinary t ract inf ect ions, Foley cat het ers should
be removed as soon as possible. Foley cat het ers become coat ed w it h a
bact erial biof ilm t hat encourages adherence of bact eria. The biof ilm prot ect s
bact eria f rom t he host immune response as w ell as f rom eradicat ion w it h
appropriat e ant ibiot ic t herapy. This can make it diff icult t o clear t he source of
inf ect ion w it hout removing or changing t he Foley cat het er. I f a Foley cat het er is
st ill clinically indicat ed and a pat ient develops a UTI , t he Foley should be
changed in addit ion t o inst it ut ing ant ibiot ics. Cont raindicat ions t o changing t he
Foley cat het er include severe scrot al edema and recent urologic or bladder
surgery, w hich may make it t echnically diff icult or pot ent ially harmf ul t o remove
and reinsert t he Foley.
Candiduria may be due t o asympt omat ic colonizat ion or a UTI . I f a UTI is
suspect ed, t he Foley cat het er should be removed or changed and ant if ungal
t herapy should be considered. Treat ment of candiduria
should be inst it ut ed in t ransplant pat ient s and crit ically ill pat ient s w it h mult iple
sit es of f ungal colonizat ion as t hey are at high risk f or disseminat ed f ungal
inf ect ion.
I t should be remembered t hat pat ient s w it h chronic indw elling Foley cat het ers
w ill of t en have chronic bact erial colonizat ion. Ant ibiot ic t reat ment and/ or
changing t he Foley cat het er in t hese sit uat ions is not indicat ed unless t he pat ient
develops ot her sympt oms consist ent w it h syst emic inf ect ion such as f ever,
hypot ension, or changes in ment al st at us.
Suggested Readings
Ferri F, ed. The Care of t he Medical Pat ient . St . Louis: Mosby; 2001: 549–
554.
Lanken P, ed. The I nt ensive Care Unit Manual. Philadelphia: Saunders;

2001: 137–139.
Nicolle L. Cat het er relat ed urinary t ract inf ect ion. Drugs Aging 2005; 22: 627–
639.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 91 - D o Not Flus h Ur eter al
S tents if a Ur ologic al C ons ultation is Available
91
Do Not Flush Ureteral Stents if a Urological
Consultation is Available
Post operat ive urology, gynecology, and ot her abdominal surgical pat ient s of t en
come t o t he int ensive care unit (I CU) w it h ext ernally draining uret eral st ent s;
some of t hese pat ient s may have undergone reimplant at ion of t he uret ers during
urinary diversion. I nvariably, in t he middle of t he night t he likely bloody urine
out put w ill decrease and t he house off icer may be present ed w it h t he opport unit y
or request t o f lush t hese st ent s. How ever, bef ore doing so t here are several
import ant f act ors t o consider.
The prime considerat ion is w het her t here exist s any ot her reason t he pat ient may
have decreased urine out put t hrough t he st ent s. The ast ut e clinician w ill evaluat e
t he post operat ive f luid resuscit at ion, cardiac st at us, and inf ect ious st at us of t he
pat ient . An exam should be perf ormed t o make sure t he pat ient does not have
any ot her cat het ers in place draining t he bladder or kidneys. An assessment
must include t he amount of f luid draining around t he st ent s in a capillary-t ype
manner. I mmediat ely post operat ively, uret eral edema may cause uret eral
obst ruct ion, and t heref ore t he majorit y of urine out put t ypically comes f rom t he
st ent s in t he early post operat ive period. As edema decreases, t he uret ers may
drain urine around t he st ent s, decreasing t he out put f rom t he st ent s but
maint aining good overall urine out put .
I f all ot her causes of decreased urine out put are ruled out and t he uret eral
st ent s appear t o be clogged, a urologist should be called t o assess t he st ent s. I t
is recommended t hat you do not manipulat e t he st ent as t he st ent may become
dislodged or penet rat e t he renal parenchyma, pelvis, uret er, or anast omosis. I f a
consult at ion is not possible, you should draw back on t he st ent w it h a 3- t o 5-
cm 3 syringe. I f t here is no ret urn, you may gent ly f lush t he st ent w it h 3 t o 5 cm3
of normal saline over 5 t o 10 seconds. The st ent should t hen be ret urned t o
st raight drain. The slow int roduct ion of a small volume helps prevent t he abrupt
rise of renal pelvic pressure, w hich can cause pyelovenous backf low of urine int o
t he blood st ream and f orniceal rupt ure.
Suggested Readings
Walsh PC, ed. Campbell's Urology. 8t h Ed. Philadelphia: WB Saunders;
2002: 3745â 3784. 212
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 92 - Make S ur e the C uff
( P ilot B alloon) is C om pletely D eflated on a C uffed Tr ac heos tom y Tube B efor e a P as s y- Muir S peaking
Valve is P lac ed
92
Make Sure the Cuff (Pilot Balloon) is Completely
Deflated on a Cuffed Tracheostomy Tube Before
a Passy-Muir Speaking Valve is Placed
Molly B. Campion MS
A speaking valve is a one-w ay valve t hat is used w it h t racheost omy pat ient s t o
redirect airf low, allow ing t hem t o use t heir voice and speak. Some speaking
valves have also been show n t o assist in reducing aspirat ion risk in t he
t racheost omy pat ient . The Passy-Muir valve is one t ype of speaking valve
commonly used and market ed f or it s “201C; no-leak” closed syst em design.
O ne-w ay speaking valves allow airf low int o t he t racheost omy t ube on inhalat ion
but close on exhalat ion, t hus f orcing air t o pass bet w een t he t racheost omy t ube
and t he w alls of t he t rachea and ult imat ely t hrough t he vocal cords and upper
airw ay (Fig. 92. 1). This helps t o est ablish a more “201C; normal” airf low pat t ern
f or t he pat ient in addit ion t o rest oring verbal communicat ion.
FIG URE 92. 1. Uncuff ed Tracheost omy Tube w it h Passy-Muir Closed-posit ion
“201C; No-leak” Speaking Valve.
I n a pat ient w it h a cuff ed t racheost omy t ube, t he cuff (pilot balloon) must be
complet ely def lat ed prior t o placement of t he speaking valve in order f or t he
exhaled air t o move around t he t racheost omy t ube and airw ay w alls. I f t he cuff
remains inf lat ed and a speaking valve is placed, t he cuff w ill obst ruct exhaled
airf low and t he pat ient w ill be unable t o exhale or breat he (Fig. 92. 2). This
scenario can quickly lead t o a respirat ory arrest .
FIG URE 92. 2. Cuff ed Tracheost omy Tube.
Suggested Readings
Suit er DM, McCullough G H, Pow el PW. Eff ect s of cuff def lat ion and one-w ay
t racheost omy speaking valve placement on sw allow physiology. Dysphagia
2003; 18: 284–292.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 93 - K now The C linic ally
Im por tant Is s ues W ith Us ing an Intr a- Aor tic B alloon P um p
93
Know The Clinically Important Issues With Using
an Intra-Aortic Balloon Pump
Low cardiac out put syndrome is def ined as inadequat e perf usion as a result of
impaired pumping f unct ion and a cardiac index of <2. 2 L/ min/ m2 . Clinical
manif est at ions of hypoperf usion are list ed in Table 93. 1. Management of low
cardiac out put is charact erized by opt imizat ion of hemodynamic paramet ers,
w hich include heart rat e, rhyt hm, preload, af t erload, and cont ract ilit y. All of
t hese paramet ers play a signif icant role in t he balance of oxygen supply and
demand of t he myocardium. Mechanical circulat ory support must be considered
t o opt imize t hese paramet ers, especially w hen t here is ongoing ischemia.
The int ra-aort ic balloon pump (I ABP) is a device t hat can support t hese
hemodynamic goals. I t is designed t o augment coronary blood f low and can
reduce t he vascular impedance seen by eject ing myocardium during syst ole. I n
it s capacit y t o opt imize myocardial oxygen supply and demand and t o improve
cardiac out put , dist al perf usion generally improves as w ell.
The indicat ions f or placement of an I ABP include complicat ions of a myocardial
inf arct ion (including mit ral regurgit at ion); ongoing ischemia or chest pain despit e
maximal medical t herapy f or myocardial ischemia; prophylact ic support in
pat ient s scheduled f or cardiac surgery w it h a high-grade lef t main coronary
art ery st enosis; and post – cardiac surgery in a low cardiac out put st at e. The
cont raindicat ions include aort ic insuff iciency, aort oiliac disease, and
irreversiblemyocardial disease. The f irst , aort ic insuff iciency, is an absolut e
cont raindicat ion because inf lat ion of t he balloon during diast ole w ill cause an
increase in regurgit ant f low, adding signif icant ly t o preload and increasing
myocardial oxygen demand and risk of injury t o t he aort ic valve. The ot her
cont raindicat ions are relat ive. Many pat ient s have some degree of aort oiliac
disease and if t he pat ient is suff ering an acut e decompensat ion t hat is
reversible, t hen t here may be a role f or t he I ABP, even if he or she is not a
candidat e f or revascularizat ion, vent ricular assist placement , or t ransplant .
The cat het er part of t he I ABP is made up of a nont hrombogenic balloon t hat
inf lat es and def lat es in synchrony w it h t he cardiac cycle. The I ABP produces it s
desired eff ect t hrough count erpulsat ion; t he
balloon inf lat es during diast ole and def lat es during syst ole, or count er t o w hat
t he int ra-aort ic blood volume and pressures are. The inf lat ion during diast ole
increases t he int ra-aort ic pressure during t he port ion of t he cardiac cycle in
w hich t he coronary art eries are perf used, t hus augment ing f low, especially
t hrough ischemic vessels t hat may have lost t heir aut oregulat ion. The I ABP
def lat es prior t o opening of t he aort ic valve t o decrease af t erload and opt imize
eject ion.
TABLE 93-1 CLINICAL M ANIFESTATIONS OF

SYSTEM IC HYPOPERFUSION
Tachycardia
Hypotension
Oliguria
Dry mucous membranes
Confusion
Tachypnea
Cold and clammy or mottled extremities
Pulmonary edema
Low cardiac index
Increased calculated systemic vascular resistance

Progressive acidosis
I ABPs are of t en placed in cardiac cat het erizat ion labs by cardiologist s under
f luoroscopy or in t he cardiac surgery operat ing rooms eit her w it h or w it hout
visualizat ion on t ransesophageal echo (TEE). The correct placement of t he I ABP
is ext remely import ant because malposit ion can lead t o serious complicat ions
including occlusion of t he renal or mesent eric vessels and ischemia of t hese end
organs. I f it is t oo cephalad, it can obst ruct lef t vent ricular out f low by
obst ruct ing t he valve or even perf orat ing t he aort a and f low t o t he proximal
major vessels. The most common sit e f or placement is a f emoral art ery, but it
may also be placed direct ly int o t he aort a in t he chest at t he t ime of surgery if
access dist ally is not f easible. Because of t he pot ent ial complicat ions, t he
posit ion of all I ABPs should be conf irmed radio-graphically upon admission t o t he
I CU af t er placement , regardless of how it w as placed (f luoroscopy, TEE). The
opt imal placement is w it h t he t ip in t he t horacic aort a, just dist al t o t he lef t
subclavian art ery. The I ABP device has a radiopaque t ip t hat can be used t o
conf irm appropriat e posit ioning of t he I ABP and must be visualized (Table 93. 1).
I f it is not visualized, t he chest radiograph should be repeat ed eit her more
cephalad or more caudally. I f t he I ABP needs t o be reposit ioned, t he I ABP must
be paused and t he cat het er eit her advanced or w it hdraw n w it hin t he st erile
sheat h.
I n order f or t he I ABP t o be properly synchronized w it h t he cardiac cycle, t he

inf lat ion of t he balloon can be t riggered by t he elect rocardiogram (ECG ) t racing,
t he art erial t racing, det ect ion of pacing, or an int ernal t rigger. I n t he current
generat ion of I ABP devices, t he ECG is t he def ault mode. I n t he presence of
arrhyt hmias, st rong considerat ion should be given t o using t he art erial t racing t o
obt ain more consist ent t riggering. The pacing modalit y is commonly used f or
individuals w ho are being paced but is not required w it h t he new er generat ion of
t he devices. The int ernal t rigger is reserved f or nonpulsat ile f low, f or example,
w hile on bypass. When evaluat ing f or appropriat e t iming, t he I ABP should inf lat e
at t he dicrot ic not ch on t he art erial t racing, ref lect ing inf lat ion af t er t he aort ic
valve has closed.
Weaning of t he I ABP can be accomplished by decreasing t he f requency of
inf lat ion t o def lat ion in relat ion t o t he cardiac cycle, t hat is, f rom a 1: 1 rat io w it h
every heart beat t o a 1: 3 rat io w it h every heart beat . The degree of balloon f ill is
less of t en adjust ed. When evaluat ing t olerance of w eaning, one should monit or
t he ECG and dist al perf usion. A common w eaning st rat egy is t o w ean f rom 1: 1 t o
1: 2 t o 1: 3 w it h int erval ECG and cardiac out put evaluat ions, if available. The
goal should be t o remove t he I ABP as soon as clinically appropriat e t o decrease
t he risk of complicat ions, w hich include inf ect ion, bleeding, t hrombosis and
compromise of dist al perf usion, and visceral ischemia. Aort ic dissect ion is also a
pot ent ial complicat ion, but it generally occurs at t he t ime of placement .
Suggested Readings
Basket t RJ, G hali WA, Mait land A, et al. The int raaort ic balloon pump in
cardiac surgery. Ann Thorac Surg 2002; 74: 1276–1287.
Hanlon-Pena PM, Ziegler JC, St ew art R. Management of t he int raaort ic

balloon pump pat ient : pharmacologic considerat ions. Crit Care Nurs Clin Nort h
Am 1996; 8: 389–408.
Papaioannou TG , St ef anadis C. Basic principles of t he int raaort ic balloon

pump and mechanisms aff ect ing it s perf ormance. ASAI O J 2005; 51: 296–300.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 94 - R em em ber that R ight
Hear t Failur e is a C om m on And Im por tant C om plic ation/Managem ent C hallenge Follow ing P lac em ent of
a Left Ventr ic ular As s is t D evic e
94
Remember that Right Heart Failure is a Common
And Important Complication/Management
Challenge Following Placement of a Left
Ventricular Assist Device
Lef t vent ricular assist devices (LVADs) including int ra-aort ic balloon pumps,
percut aneous Tandem Heart devices, ext racorporeal, and implant ed vent ricular
assist devices have signif icant ly improved t he management and out come f or
pat ient s w it h t he most severe f orms of heart f ailure. Acut e or chronic right
vent ricular f ailure, how ever, remains one of t he most serious and diff icult
condit ions t o manage f or pat ient s w it h heart f ailure, especially f ollow ing LVAD
placement .
Watch Out For

Causes of right vent ricular dysf unct ion include pre-exist ing myocardial disease,
increased preload and/ or right vent ricular dilat at ion, ischemia, arrhyt hmias,
increased int rat horacic pressures f rom mechanical vent ilat ion, and especially
acut e or chronic f orms of pulmonary hypert ension. Frequent ly a posit ive
f eedback cycle exist s, init iat ed by somet hing t hat causes right vent ricular
pressure overload, int ervent ricular sept um displacement , increased w all t ension
w it h or w it hout reduced right vent ricular perf usion of decreased right vent ricular
cardiac out put . Experienced implant programs evaluat e vent ricular device
candidat es f or t he presence of severe right vent ricular dysf unct ion pri or t o
LVAD/ RVAD placement using some combinat ion of echocardiography and
measures of right vent ricular f unct ion including cent ral venous pressures and t he
presence of upst ream organ f ailure due t o right vent ricular dysf unct ion including
renal insuff iciency and hepat ic synt het ic f ailure w it h elevat ed prot hrombin t imes,
aminot ransf erase levels, and hyperbilirubinemia.
I n addit ion t o t he older modalit ies, new er measures of right vent ricular
dysf unct ion include right vent ricular st roke w ork index: (mean pulmonary art ery
pressure — right at rial pressure [ or cent ral venous pressure] / (cardiac index ×
heart rat e) RVSWI = [ (PAPm - RAP) × CI ] / HR = (mm Hg·mL)/m 2. Several st udies
have suggest ed t hat RVSWI values less t han 300 (mm Hg·mL)/m 2 are
independent predict ors of t he need f or an RVAD. A more recent measure of right
vent ricular f ailure is t he t ricuspid annular plane syst olic excursion (TAPSE) of t he
lat eral edge of t he t ricuspid annulus on a st andard f our-chamber echo view of
t he heart . A normal TAPSE is 20 t o 25 mm and is usually easy t o obt ain.
Signif icant decreases in t he TAPSE suggest t he presence of severe right
vent ricular dysf unct ion. I n most cases, pat ient s w it h pre-exist ing bivent ricular
dysf unct ion w ill also have an RVAD insert ed at t he t ime of LVAD placement . For
t hose pat ient s w ho do not , recognizing and t reat ing new right vent ricular
dysf unct ion is among t he most import ant perioperat ive management challenges in
t he int ensive care unit (I CU).
Signs and Symptoms

Right vent ricular dysf unct ion usually manif est s as some combinat ion of low
cardiac out put , hypot ension, and in t he presence of an LVAD, decreased f low s
f rom t he device. Usually, a cent ral venous pressure measurement is available
and is of t en elevat ed as t he right vent ricle is dilat ed, poorly compliant , or volume
overloaded. An emergent echocardiogram is crucial t o obt ain and usually
conf irms t he diagnosis especially if pre-exist ing echocardiography did not show
right vent ricular f ailure prior t o LVAD placement . O t her condit ions such as
t amponade, obst ruct ions t o LVAD inf low, excessive LVAD f low s, excessive
syst emic vasopressor use, and increases in pulmonary vascular resist ance can
mimic, cause, or w orsen right vent ricular dysf unct ion. Rising cent ral venous
pressures w it h const ant or decreasing LVAD f low s is probably t he most common
present at ion of RV dysf unct ion.
What to Do
Treat ment f or right vent ricular dysf unct ion f ocuses on correct ing t he underlying
cause (i. e. , draining blood in t amponade, opt imizing device t iming and rat es).
Probably t he most common t reat ment is t o relieve right vent ricular af t er load by
avoiding or t reat ing any cause of respirat ory or met abolic acidosis and
hypoxemia, as w ell as using direct pulmonary vasodilat ors such as nit ric oxide,
inhaled prost acyclines, or, rarely, ot her nit ric oxide donors such as
nit roglycerine. I f nit ric oxide is not already in use, st art ing doses of 80 ppm
dow n t o 20 ppm is commonly used. Right vent ricular dysf unct ion f requent ly
manif est s more slow ly in pat ient s w it h bivent ricular f ailure w hen nit ric oxide is
being w eaned, especially at 5 ppm or less, and is of t en t he reason nit ric oxide
may need very slow w eaning. I not ropes are also commonly used t o avoid or
t reat right vent ricular f ailure. Phosphodiest erase I I I inhibit ors such as milrinone
(0. 375 t o 0. 5 µg/ kg/ min) are most of t en used because t hey
may have some eff ect on reducing pulmonary vascular t one. Dobut amine,
isoprot erinol, and epinephrine may also improve right vent ricular dysf unct ion but
may also cause more arrhyt hmias. A new er class of inot ropes t hat act by
increasing myocardial t roponin sensit izat ion t o int racellular Ca+ may soon be
used. Even if hypot ension is present , very caref ul diuresis or f luid removal w it h
cont inuous renal replacement devices may also improve right vent ricular f unct ion
in t he set t ing of severe volume overload w it h elevat ed cent ral venous pressures.
Finally, placement of a t emporary RVAD (Abiomed or Thorat ec), cardiac
ext racorporeal membrane oxygenat ion (ECMO ), or emergent heart t ransplant
may be required f or w orsening or ref ract ory right vent ricular f ailure.
Suggested Readings
Furukaw a K, Mot omura T, Nose Y. Right vent ricular f ailure af t er lef t
vent ricular assist device implant at ion: t he need f or an implant able right
vent ricular assist device. Art if O rgans 2004; 29: 369–377.
G hio S, Recusani F, Klersy C, et al. Prognost ic usef ulness of t he t ricuspid

annular plane syst olic excursion in pat ient s w it h congest ive heart f ailure
secondary t o idiopat hic or ischemic dilat ed cardiomyopat hy. Am J Cardiol
2000; 85: 837–842.
Kavarana MN, Pessin-Minsley MS, Urt echo J, et al. Right vent ricular
dysf unct ion and organ f ailure in lef t vent ricular assist device recipient s: a
cont inuing problem. Ann Thorac Surg 2002; 73: 745–750.
O chiai Y, McCart hy PM, Smedira NG , et al. Predict ors of severe right

vent ricular f ailure af t er implant able lef t vent ricular assist device insert ion:
analysis of 245 pat ient s. Ci rcul ati on 2002; 106: 198–202.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 95 - K now the C om m on
P r oblem s As s oc iated W ith C ar diopulm onar y S uppor t Us ing E xtr ac or por eal Mem br ane O xygenation
95
Know the Common Problems Associated With
Cardiopulmonary Support Using Extracorporeal
Membrane Oxygenation
Venoart erial ext racorporeal lif e support (VA ECMO ) is among t he more advanced
met hods t o t emporarily support pat ient s w it h eit her cardiac or cardiac and
pulmonary f ailure unt il pat ient recovery, assist device implant at ion, or t ransplant .
The indicat ions and cont raindicat ions f or t his t ype of support are list ed in Table
95. 1. Managing t hese pat ient s is challenging, and sophist icat ed t eams of
perf usionist s, int ensivist s, and surgeons are required. The advant ages of using
ECMO inst ead of a vent ricular assist device include
t he abilit y t o cannulat e pat ient s at a bedside

immediat e bivent ricular support
bot h cardiac and pulmonary support
abilit y t o assess pat ient s off ECMO prior t o decannulat ion
abilit y t o change t o venovenous cannulat ion if hemodynamic f ailure improves
w hile pulmonary support is st ill required
TABLE 95-1 INDICATIONS AND CONTRAINDICATIONS

OF ADULT ECM O
INDICAT IONS CONT RAINDICAT IONS
Cardiogenic shock Unwitnessed cardiac arrest

Cardiac trauma Aortic insufficiency
Respiratory
Aortic dissection
insufficiency
Status asthmaticus Cardiac arrest >30 minutes
No correctable anatomic
Smoke inhalation
defect
Massive pulmonary
Terminal illness
embolism
Severe diabetes mellitus

Severe peripheral vascular
Donor heart
disease
preservation
Recent cerebrovascular
accident
The disadvant ages include
more bleeding, especially af t er cardiot omy or heart or lung t ransplant at ion

increased lef t vent ricular af t erload
limb ischemia
t hromboembolic injury t o lung and/ or brain
incomplet e lef t vent ricular decompression w it h pulmonary edema

uneven oxygenat ion t o upper and low er body
hemolysis
increased inf ect ions
signif icant expense and resource needs
Experienced ECMO providers have developed management st rat egies f or some
of t hese problems.
Postoperative Bleeding
Bleeding is t he most common complicat ion of VA ECMO . Typically, ECMO
programs have prot ocols f or t reat ing bleeding w it h appropriat e blood product
replacement . Met hods t o reduce syst emic ant icoagulat ion involve some
combinat ion of decreasing doses of heparin t o allow t he kaolin-act ivat ed clot t ing
t ime t o approach very low levels such as 160 t o 180 seconds f or several hours
or in ext remely rare sit uat ions f or several days, t hough t his is a very risky
procedure and requires ext remely close communicat ion among t he ECMO t eam
members. I n any bleeding sit uat ion and especially w hen a low -dose or no-
heparin condit ion exist s, it is mandat ory t o have a redundant ECMO circuit
primed and ready at t he bedside in case t he act ive circuit develops signif icant
clot . O t her st rat egies t o reduce bleeding w hile on ECMO include aprot inin
inf usions; using heparin-bonded circuit t ubing; increasing f low rat es t o reduce
st agnat ion; and f requent surgical explorat ions t o remove clot (surgical w ounds
are usually lef t open ot her t han a st erile occlusive pat ch and elect rocaut ery is
generously applied t o any bleeding point s).
Low Flow States

Low f low st at es are usually due t o reduced int ravascular volume but can also be
due t o compression of venous inf low cannula f rom body posit ion or low er-
ext remit y edema. I ncreasing t he int ravascular volume and/ or reposit ioning t he
pat ient f requent ly improves t his problem. O t her causes of blood-f low impedance
around t he heart include t amponade, t ension pneumot horax, and clot in t he
circuit . I f periodic pulsat ion is not det ect ed by art erial line, echocardiography is
used t o det ermine if an at rial sept ost omy is needed in order t o vent a
noncont ract ing chamber, dilat ed a lef t vent ricle, as w ell as look f or t he presence
of lef t vent ricular clot .
Regional Ischemia
Because most VA ECMO is perf ormed using t he f emoral vein and art ery,
perf usion t o t he ipsilat eral leg is of t en impaired. The usual met hod t o reduce or
t reat t his is t o insert a small, 14-gauge perf usion cannula in t he post erior t ibial
art ery and t o inf use 100 t o 200 mL/ min of f low t hrough t his line. Anot her
complicat ion of perf using oxygenat ed blood int o t he art erial syst em using t he
f emoral art ery is inadequat e oxygenat ion of t he upper ext remit ies and head. This
sit uat ion manif est s it self by t he low er half of t he body looking pink w hile t he
upper half is cyanot ic. The usual t reat ment is t o insert and perf use oxygenat ed
blood t hrough a venous jugular line; t his creat es, in eff ect , a VA-V ECMO circuit .
Hypotension
Blood pressure f requent ly requires some combinat ion of vasopressor and
inot rope support (especially w hen w eaning pat ient s). Norepinephrine 0. 08 t o 0. 2
µg/ kg/ min and milrinone 0. 375 t o 0. 5 µg/ kg/ min or dobut amine 10 t o 20
µg/ kg/ min are of t en used f or t hese purposes.
Neurological Injury
Periodic sedat ion holidays are import ant t o det ect new onset or presence of
serious neurologic morbidit y in t hese pat ient s.
Selected Readings
Meurs KV, Lally KP, Peek G , et al. , eds. Ext racorporeal Cardiopulmonary
Support in Crit ical Care. 3rd Ed. Ann Arbor, MI : ECLS O rganizat ion; 2005.
Mielck F, Q uint el MI . Ext racorporeal membrane oxygenat ion. Curr O pin Crit
Care 2005; 11: 87–93.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 96 - Tr eat “ 201C ;w r inkling”
in the Abiom ed D iaphr agm as a P os s ible S ign of Hypovolem ia
96
Treat “201C;wrinkling” in the Abiomed
Diaphragm as a Possible Sign of Hypovolemia
Frank Rosemeier MD
Vent ricular assist devices (VADs) are indicat ed f or pat ient s w it h vent ricular
f ailure on maximum pharmacological support in t he post -cardiopulmonary bypass
period, f ollow ing an acut e myocardial inf arct ion, as bridge t o t ransplant at ion, or
dest inat ion t herapy f or permanent use. Most of t hese devices, w hich can be
used in eit her an int racorporeal or ext racorporeal posit ion, are displacement
pumps and assist t he lef t vent ricle. I n a lef t VAD (LVAD) conf igurat ion, a device
inf low cannula connect s t he lef t at rium or lef t vent ricular apical region t o t he
pump chamber. Blood ret urns t o t he ascending aort a via a device out f low
cannula. For a right -sided VAD (RVAD), t he inf low cannula is sit uat ed in t he right
at rium or right vent ricle w it h t he out f low cannula connect ed t o t he main
pulmonary art ery. These t ubular cannulae are t ypically radiolucent and are not
easily ident if iable on a plain radio-graph.
The Abiomed assist device is most f requent ly used in an emergency f or it s ease
of implant at ion, robust ness of operat ion, and versat ilit y as it can be conf igured
as an LVAD, an RVAD, or BiVAD. I t is simple t o operat e, inexpensive, and robust
f or t ransport purposes. I t has t he added advant age in t hat t he pump sect ion can
be changed at t he bedside in 1 minut e. Draw backs include a high re-explorat ion
rat e f or hemorrhage. A clear plast ic housing device t hat is placed below t he level
of t he heart cont ains a pulsat ile dual-chamber assembly in series. Filling of t he
upper chamber is via passive gravit y drainage f rom t he vent ricle or at rium.
Adjust ing t he height of t he device relat ive t o t he pat ient 's heart level w ill cont rol
preload condit ions. As t he device is low ered, t he f low int o t he upper chamber
w ill increase, short ening diast olic f illing t ime and increasing device out put . The
low er chamber perf orms t he pumping act ion at a const ant st roke volume of
around 88 mL w it h an out put limit at ion of 4 t o 5 L/ min. High syst emic vascular
resist ance may increase t he durat ion of t he empt ying of t he low er chamber,
prolonging syst ole w it h a possible decrease in out put .
Watch Out For
For t he Abiomed device, it is import ant t o recognize t hat a reduced rat e of pump
cycles may serve as a clinical indicat or of inadequat e preload
or excessive af t erload. Typically, in hypovolemia t he chambers might cycle

slow ly, causing w rinkling in t he diaphragm, w hich may be t he f irst sign of a
problem. Volume loading w ill help rest ore device out put , as can t he
Trendelenburg posit ion. I n addit ion, manipulat ion of t he ext ernal housing device
may help t emporize t he clinical hypovolemia. This is accomplished by increasing
t he height diff erent ial by low ering t he device, w hich may improve t he pressure
gradient and t heref ore f low t o t he upper chamber t emporarily. I n addit ion, t his
maneuver may be helpf ul as a diagnost ic procedure. Caut ion is advised w hen
raising t he head of t he pat ient in an eff ort t o increase t he pressure diff erent ial
as venous ret urn t o t he pat ient 's right heart may f urt her diminish.
Alt hough f illing and out put occur independent ly of t he nat ive cardiac rhyt hm,
pat ient s w it h assist devices generally do not t olerat e vent ricular f ibrillat ion,
t achycardia, or asyst ole w ell. The f illing of t he chamber requires a pressure
gradient all t he w ay f rom t he right vent ricle (RV) t hrough t he pulmonary
vasculat ure bed t o t he lef t at rium/ vent ricle. Loss of RV out put can result in a
decrease of LVAD preload. Low out put of an LVAD can also be relat ed t o a
f ailing right heart , increased pulmonary resist ance, or f low obst ruct ion f rom t ube
kinking or clot f ormat ion. A caref ul inspect ion of t he t ubing and chambers is
mandat ory w hen unexpect ed clinical hypovolemia or low LVAD out put is
observed.
Suggested Readings
G oldst ein DJ, O z MC, Rose EA. I mplant able lef t vent ricular assist devices. N
Engl J Med 1998; 339: 1522–1533.
Hensley FA, Mart in DE, G ravlee G P, eds. A Pract ical Approach t o Cardiac
Anest hesia. 3rd Ed. Philadelphia: Lippincot t Williams & Wilkins; 2003: 557–
573.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 97 - Adm inis ter Antibiotic s
B efor e P lac ing a C hes t Tube in Tr aum a P atients
97
Administer Antibiotics Before Placing a Chest
Tube in Trauma Patients
Peter G . T homas DO
Patrick K. Kim MD
The role of ant ibiot ics during t he placement of t ube t horacost omy f or hemot horax
or pneumot horax in t he t rauma pat ient remains cont roversial. Every year, more
t han 100, 000 chest t ubes are placed in t rauma pat ient s f or hemopneumot horax.
Mult iple st udies have been perf ormed over t he last f our decades t o det ermine
w het her or not ant ibiot ics decrease t he rat e of inf ect ion af t er chest -t ube
insert ion.
I n 2000, t he East ern Associat ion f or t he Surgery of Trauma (EAST) Pract ice
Management G uidelines Work G roup review ed all of t he prospect ive art icles and
t w o met a-analyses on t he t opic published bet w een 1977 and 1997. The group
ident if ied class I (prospect ive randomized cont rolled t rials) and class I I (clinical
st udies in w hich t he dat a w ere collect ed prospect ively, and ret rospect ive
analyses t hat w ere based on clearly reliable dat a) st udies upon w hich t o base
t heir recommendat ions. The aut hors concluded t hat t here is suff icient dat a t o
make a level I I I recommendat ion f or prophylact ic ant ibiot ic use in pat ient s
undergoing t ube t horacost omy f or t rauma. There is no apparent eff ect on
empyema rat e, but ant ibiot ics may reduce t he incidence of pneumonia. The
guidelines suggest t hat ant ibiot ics should not be given f or more t han 24 hours.
The aut hors w ere able t o make only level I I I recommendat ions (def ined as
“201C; support ed by available dat a but adequat e scient if ic evidence is lacking”)
because t he st udies had markedly diff erent st udy designs and t reat ment crit eria,
and as such, consensus conclusions could not be made based by combining t he
dat a.
Wilson and Nichols review ed t he same dat a and published an edit orial t o t he
EAST guidelines in 2000. Their int erpret at ion of t he dat a re-emphasized t he
correlat ion bet w een t he use of ant ibiot ics during chest -t ube placement and t he
decrease in t he rat e of pneumonia. When t he t hree review ed st udies t hat had
Cent ers f or Disease Cont rol and Prevent ion (CDC)–conf orming crit eria f or
pneumonia w ere combined, t here w as a 9. 4% incidence of pneumonia f or t he
placebo group and a 0. 8% (p = 0. 003) incidence f or t he ant ibiot ic group. When
t he eight papers t hat compared ant ibiot ics versus no ant ibiot ics w ere combined,
t he pneumonia rat e w as 14. 8% f or t he placebo group and 4. 1% (p = 0. 001) f or
t he ant ibiot ic group.
The lit erat ure published af t er t he EAST guidelines st ill f ails t o end t his
cont roversy. A randomized, prospect ive st udy f rom 1998 show ed similar benef it s
f or pat ient s receiving ant ibiot ics. Sevent y-one pat ient s t reat ed w it h ant ibiot ics
w ere compared w it h 68 t hat received placebo. The t reat ed group had no
inf ect ions w hile t he placebo group had f our (p = 0. 05).
The most recent st udy w as a w ell-designed prospect ive, randomized, double-
blind mult icent er t rial. I t compared t hree groups of pat ient s: t hose w ho received
ant ibiot ics f or t he ent ire lengt h of chest t ube placement ; 24 hours of ant ibiot ics;
and placebo. The aut hors concluded t hat t here w as a slight decrease in t he
report ed empyema rat e, alt hough it w as not st at ist ically signif icant . They also
f ound no decrease in t he rat e of pneumonia. The limit at ion of t his st udy w as t hat
it w as severely underpow ered w it h an accrual rat e of less t han 20% of t he
predict ed needed number of pat ient s t o prove st at ist ical signif icance.
What to Do
Alt hough it is apparent t hat t here is no def init ive answ er t o t he quest ion of
prophylact ic ant ibiot ics w hen placing chest t ubes f or t rauma, some
recommendat ions can be made based on t he exist ing lit erat ure. First , ant ibiot ics
do appear t o decrease t he rat e of pneumonia in pat ient s w ho undergo chest -t ube
placement . Second, t he ant ibiot ic used should primarily cover Staphyl ococcus
aureus and Staphyl ococcus epi dermi di s, as t hese are t he most common
organisms recovered. Third, it appears t hat a single preprocedure dose of
ant ibiot ics is as eff icacious as 24 hours of ant ibiot ics. Judicious use of
ant ibiot ics should be pract iced in order t o limit ant ibiot ic-resist ant st rains of
bact eria.
Suggested Readings
G onzalez RP, Holevar MR. Role of prophylact ic ant ibiot ics f or t ube
t horacost omy in chest t rauma. Am Surg 1998; 64: 617–620; discussion 620–
621.
Luchet t e FA, Barrie PS, O sw anski MF, et al. Pract ice management guidelines
f or prophylact ic ant ibiot ic use in t ube t horacost omy f or t raumat ic
hemopneumot horax: t he EAST Pract ice Management G uidelines Work G roup.
East ern Associat ion f or Trauma. J Trauma 2000; 48: 753–757. Also available at
ht t p: / / w w w. east . org/ t pg. ht ml
Maxw ell RA, Campbell DJ, Fabian TC, et al. Use of presumpt ive ant ibiot ics
f ollow ing t ube t horacost omy f or t raumat ic hemopneumot horax in t he
prevent ion of empyema and pneumonia: a mult i-cent er t rial. J Trauma
2004; 57: 742–749.
Wilson RF, Nichols RL. The EAST pract ice management guidelines f or
prophylact ic ant ibiot ic use in t ube t horacost omy f or t raumat ic
hemopneumot horax: a comment ary. J Trauma 2000; 48: 758–759.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 98 - B e S us pic ious of a
Lar ge P er s is tent Air Leak in a C hes t Tube
98
Be Suspicious of a Large Persistent Air Leak in a
Chest Tube
David T. Efron MD
To underst and how t o t roubleshoot a persist ent air leak f rom a chest t ube, one
must t horoughly underst and t he ent ire algorit hm describing t he creat ion and
subsequent evacuat ion and t reat ment of a pneumot horax.
The chest cavit y is a rigid st ruct ure made so by t he bony scaff old of t he rib
cage, st ernum, and t horacic vert ebrae. The lef t and right hemit horaces are
usually anat omically dist inct cavit ies under normal circumst ances. How ever,
physiologically t hey are int erdependent due t o a sof t , relat ively compliant
mediast inum. Normally, each hemit horax is composed of a rigid out er shell
(musculoskelet al) lined on t he inside w it h t he pariet al pleura. The lung is an
aerat ed organ, sof t and compliant , t he out side of w hich is lined w it h t he visceral
pleura. The space bet w een t he pleurae is maint ained at a pressure of negat ive
5cm H2 O . This pressure, in addit ion t o int ra-alveolar lubricat ion and surf act ant ,
allow s t he lung t o remain open and aerat ed and prevent s collapse on it self .
Thus, t he movement of t he diaphragm in a cephalad-caudad direct ion aided by
much smaller, but not insignif icant movement of t he int ercost als muscle brings air
int o and out of t he lungs via t he t racheobronchial t ree. I t is a syst em t hat is open
t o t he out side environment but closed t o t he chest cavit y. Violat ion of any of t he
barriers of t he closed chest cavit y result s in release of t he negat ive pressure in
t he pleural space and collapse of t he lung. O ngoing leakage of air int o t he
pleural space result s in t ension pneumot horax w hen t hat air is not able t o
escape. A chest t ube t hat is placed t o alleviat e t his w ill demonst rat e an ongoing
air leak in t he w at er-seal chamber.
Air leaks result f rom disrupt ion of any one part of t he sealed chest cavit y.
Diagnosis of t he source of an air leak is simple in t he majorit y of cases and vit al
t o undert aking appropriat e act ion. The best w ay t o t hink of a chest -t ube circuit is
t o consider everyt hing f rom t he t ube t o t he w at er seal (t he second chamber) an
ext ension of t he pleural cavit y. A persist ent air leak can be caused only by one
of t hree basic mechanisms: pat ient pat hology; t ube pat hology; or Pleuravac
pat hology.
Patient Pathology
Tracheobronchial disrupt ion
Parenchymal injury or t ear
Esophageal perf orat ion
Unsealed chest -w all disrupt ion (i. e. , open pneumot horax or “201C; sucking
chest w ound”)
Tube Pathology
Leak at insert ion sit e (i. e. , hole in chest w all t oo big, poor seal around hole)
Proximal hole in t ube out side of chest w all
Perf orat ion of chest t ube
Poor seal at connect ion of t ube t o Pleuravac t ube
Pleuravac Pathology
I mproperly prepared apparat us
Broken apparat us
The f ollow ing is a suggest ed algorit hm of how t o diagnose t he source and

pot ent ially f ix t he air leak:
1. I nspect t he pleuravac f or obvious def ect s in set up or component s. Correct

any f ault y set ups or, if damaged, replace t he ent ire syst em.
2. I nspect all connect ions w it hin t he circuit and bet w een t he t ubes and
Pleuravacs. Remove all t ape t hat obscures your view of t hese joint s. Ensure
t hat t hese connect ions are snug and airt ight .
3. I nspect t he chest -t ube insert ion sit e. The dressing must be t aken dow n
complet ely t o do t his properly. You need t o ensure t hat t he chest t ube it self
is int act and no side holes are visible out side t he chest w all. A chest
radiograph is also usef ul in obese pat ient s w here t he chest w all is t oo t hick
t o assess t he relat ionship of t he last hole t o t he chest cavit y. I f no obvious
cause is f ound, it is best t o replace an occlusive dressing at t his sit e bef ore
proceeding.
4. Serial clamping of t he t ubes in t he drainage syst em is t hen perf ormed.
St art ing just above t he Pleuravac, t he t ube is clamped and t he leak assessed
w hile t he pleuravac remains on suct ion. I f a leak persist s at t he Pleuravac
w hen a clamp is applied t o t he t ube, t his means t here must be a leak in t he
circuit dist al t o t he clamp and a search should be direct ed t o t his area. The
clamp is t hen serially
moved up t he t ubing t ow ard t he pat ient . I f t he leak disappears w hen t he t ube

is clamped, t hen t he leak is proximal (i. e. , on t he pat ient side) t o t he clamp.
O ne f inal not e is t hat experienced clinicians usually report t hat t he most f requent
source of persist ent air leak (and t he easiest t o f ix) is t he chest -t ube circuit .
Suggested Readings
Baumann MH. What size chest t ube? What drainage syst em is ideal? And
ot her chest t ube management quest ions. Curr O pin Pulm Med 2003; 9: 276–
281.
Cerf olio RJ. Recent advances in t he t reat ment of air leaks. Curr O pin Pulm
Med 2005; 11: 319–323.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 99 - Tr eat Any Milky Fluid
C om ing Fr om the C hes t or Abdom en as C hylous Until P r oven O ther w is e
99
Treat Any Milky Fluid Coming From the Chest or
Abdomen as Chylous Until Proven Otherwise
Rajan G upta MD
The anat omy of t he t horacic duct can be variable; how ever, it usually originat es
at t he conf luence of t he cist erna chili on t he right side of t he aort a at t he L1-L2
level. I t courses cephalad t hrough t he aort ic hiat us and crosses t he midline
behind t he aort ic arch at t he level of T4-T5. The duct t hen t ravels along t he lef t
side of t he esophagus int o t he neck t o drain int o t he venous syst em at t he
conf luence of t he lef t subclavian and jugular veins. The t horacic duct t ransport s
up t o 4 L/ day of chyle, depending on diet , drug int ake, and int est inal f unct ion.
Chyle is primarily lymphat ic f luid f rom t he gast roint est inal t ract , w it h a smaller
cont ribut ion f rom lymphat ic drainage of t he chest . Since 60% t o 75% of
absorbed diet ary f at passes t hrough t his syst em mainly as chylomicrons, chyle is
primarily composed of t riglycerides, cholest erol, and f at -soluble vit amins. These
component s give chyle it s charact erist ic milky w hit e appearance. The t horacic
duct is also t he main conduit f or t he ret urn of ext ravasat ed prot eins t o t he
circulat ion. The cellular component is predominant ly lymphocyt es.
Chyle leaks are most commonly secondary t o t rauma or malignancy (e. g. ,
lymphoma, chronic lymphocyt ic leukemia, met ast at ic disease); how ever, t hey can
also be congenit al or secondary t o sarcoidosis or inf ect ious et iologies (e. g. ,
hist oplasmosis, t uberculosis). Traumat ic chyle leaks are of t en iat rogenic, usually
as a complicat ion of t horacic procedures. Esophagect omy and surgical
correct ion of congenit al heart disease are surgical procedures w it h part icularly
high risk. Abdominal procedures such as open aort ic aneurysm repair,
ret roperit oneal lymph node dissect ion, Nissen f undoplicat ion, liver
t ransplant at ion, and perit oneal dialysis cat het er placement have also been
implicat ed in chylous ascit es. Approximat ely 20% of t raumat ic chyle leaks are
secondary t o penet rat ing or blunt t rauma.
Alt hough chyle is of t en milky w hit e in appearance, it may also be clear,
especially if ent eral nut rit ion has not been recent ly administ ered. Thus, a chyle
leak should not necessarily be excluded w hen clear pleural or perit oneal f luid is
drained. Conversely, a chronic pleural eff usion can have a milky charact er due t o
lysis of red blood cells and neut rophils in t he f luid. These t ypes of eff usions are
of t en secondary
t o t uberculosis or rheumat oid disease. Purulent ascit es due t o bact erial

perit onit is can also have a similar appearance. Theref ore, if a chyle leak is
suspect ed or a milky or clear f luid of unknow n et iology occurs, t he f luid should
be analyzed f or t riglyceride concent rat ion and lymphocyt e levels. I f t he
t riglyceride concent rat ion is above 110 mg/ dL, t he f luid is chyle in 99% of t he
cases. I f t he t riglyceride concent rat ion is below 50 mg/ dL, t he f luid is chyle in
only 5% of cases. Alt ernat ively, Sudan I I I st ain can be used t o det ermine t he
presence of f at globules. O f t en, t he diagnosis can be made clinically by
administ ering cream orally and examining t he change in appearance of t he f luid.
What to Do
The approach t o management of a chylot horax depends on t he et iology.
G enerally, a graded approach should be considered, st art ing w it h t he least
invasive management st rat egy. A regimen of percut aneous drainage of t he f luid,
parent eral nut rit ion w it h volume replacement , and a f at -f ree diet supplement ed
w it h medium-chain t riglycerides w ill enable 80% of nont raumat ic cases t o resolve
w it hin 2 t o 3 w eeks. Somat ost at in is an addit ional adjunct t hat may be ut ilized.
Chyle leaks secondary t o malignancy usually require t reat ment of t he malignancy
it self , such as radiat ion f or a lymphoma. Persist ent chylous drainage beyond 2
w eeks can begin t o compromise nat ural immune response and more invasive
procedures should be considered at t his t ime. I n addit ion, a chyle leak
secondary t o surgical or t raumat ic disrupt ion is less likely t o heal w it h
conservat ive measures and w ill of t en require more invasive int ervent ion.
Percut aneous t ransabdominal cat het er embolizat ion or needle disrupt ion of
ret roperit oneal lymphat ic vessels has been eff ect ive in managing t hese leaks in
70% of cases. The advent of minimally invasive t echniques such as video-
assist ed t horacic surgery (VATS) has led t o earlier int ervent ion in select cases.
Surgical opt ions f or chylot horax include ligat ion of t he duct , pleurodesis, and
pleuroperit oneal and pleurovenous shunt ing.
Suggested Readings
Cope C, Kaiser LR. Management of unremit t ing chylot horax by percut aneous
embolizat ion and blockage of ret roperit oneal lymphat ic vessels in 42 pat ient s.
J Vasc I nt erv Radiol 2002; 13: 1139–1148.
Merrigan BA, Wint er DC, O ’Sullivan G C. Chylot horax. Br J Surg 1997: 84: 15–
20.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 100 - K now the C onditions
that C aus e an Inac c ur ate P uls e O xim etr y R eading
100
Know the Conditions that Cause an Inaccurate
Pulse Oximetry Reading
Eliahu S. Feen MD
Jose I. Suarez MD
Basic Principles
Pulse oximet ry is t he t echnique of measuring art erial blood oxygenat ion t hrough
ident if icat ion of pulsat ile blood. The f undament al physiological principle t hat t he
t echnique uses is spect rophot omet ry, or t he measurement of t he absorpt ion of
light by blood and t issues. O ne simple w ay t o underst and t he basic principle is
t o imagine a t est t ube f illed w it h some subst ance and shining light of a cert ain
w avelengt h upon it (incident light ). Some of t he light is t ransmit t ed (“t ransient ”
light ) t hrough t he subst ance and some is absorbed. The amount of light
t ransmit t ed depends upon t he absorpt ive (and ref lect ive) nat ure of t he part icular
subst ance. The Beer-Lambert law relat es t his mat hemat ically:
C = (1/d × α ) loge ( I i / I t )
w here C is t he concent rat ion of t he part icular subst ance in t he t est t ube, d is t he
pat h lengt h of t he light t hrough t he t est t ube, α is an absorpt ion coeff icient f or
t he subst ance in quest ion at a part icular w avelengt h of light , I i is t he int ensit y of
t he incident light , and I t is t he int ensit y of t he t ransmit t ed light . When t his
principle is applied t o blood oxygenat ion, t hen t he quest ion becomes one of
det ermining t he absorpt ion of light by hemoglobin.
I n adult s, t here are f our f orms of hemoglobin, oxyhemoglobin (HbO2 ),
met hemoglobin (met Hb), carboxyhemoglobin (HbCO ), and “reduced” hemoglobin
(Hb). Each f orm of hemoglobin absorbs a diff erent amount of light at a given
w avelengt h. Using f our diff erent w avelengt hs of light , t he Beer-Lambert law can
be solved f or each of t he f our f orms of hemoglobin t o give a concent rat ion in a
sample of blood. Wit h t hese concent rat ions, t he f ract ional hemoglobin sat urat ion
(percent age of oxyhemoglobin in blood) can be calculat ed as f ollow s:
Except in pat hological st at es, t he concent rat ions of met Hb and HbCO are low.
These f orms of hemoglobin do not cont ribut e import ant ly
t o t issue oxygen delivery. Physiologically, w hat is import ant is oxygen delivery. A

f unct ional oxygen sat urat ion has been def ined as f ollow s:
This equat ion requires only t w o w avelengt hs of light in order t o solve f or

concent rat ion t hrough t he Beer-Lambert law.
I n most clinical set t ings, t he t w o w avelengt hs are a red light w avelengt h and an
inf rared w avelengt h. Thus, w e have oximet ry t hrough t he measurement of light
absorpt ion by blood. The problem t hen becomes det ermining t he light absorpt ion
of art erial blood as opposed t o venous and capillary blood and t issue absorpt ion
of light . Pul se oximet ry solves t his problem by measuring t he diff erence in t ot al
absorpt ion of light at t he great est dist ance and t he least dist ance over a f ixed
span of t issue cont aining art erioles. The least dist ance is t he dist ance of t issue
plus venous and capillary blood plus a dist ance of art eriolar blood at diast ole.
The great est dist ance is t he least dist ance plus t he dist ance creat ed by t he
ext ra volume of blood cont ained in t he art erioles at syst ole (t he “pulse”). The
diff erence bet w een t hese dist ances is creat ed by art erial blood alone, so light
absorpt ion diff erences represent art erial blood. This f igures int o t he pat h lengt h
of t he Beer-Lambert law (d in t he equat ion above). Ult imat ely, using t his
inf ormat ion, t he oxygen sat urat ion of art erial blood can be comput ed.
Inaccurate Readings
Based upon t hese principles, t he reasons f or spurious pulse oximet er readings
can be underst ood. Any process t hat limit s t he art eriolar blood f low int o
peripheral sit es w here pulse oximet er readings are perf ormed w ill eit her f alsely
low er t he readings or preclude t he abilit y t o get readings. The simple example of
placing a pulse oximet er probe on a f inger of an arm compressed by a
sphygmomanomet er w ill demonst rat e t his idea. When t he cuff pressure exceeds
syst olic blood pressure, art erial blood f low t apers off , and t he pulse oximet er
reading drops off as w ell. Similarly, peripheral vasoconst rict ion due t o any cause
(e. g. , shock) w ill cause f alsely alt ered readings. Addit ionally, any ambient
int erf erence in light t ransmission (or absorpt ion) w ill alt er t he readings.
Examples include nail polish on t he f ingers upon w hich t he pulse oximet er probe
is placed and f luorescent light in t he room. Many manuf act urers develop signal-
processing t echniques t o account f or ambient light around t he pat ient . Some
int ravenous dyes (e. g. , met hylene blue) can alt er t he light absorpt ion w it hin
blood and t hereby spuriously
alt er t he pulse oximet ry result . As ment ioned bef ore, in some pat hophysiologic
st at es, t here is increased concent rat ion of met hemoglobin and
carboxyhemoglobin, w hich w ere not f igured int o t he f unct ional oxygen sat urat ion
equat ion. I n such st at es (e. g. , carbon monoxide poisoning) t he pulse oximet er
reading can be f alse. Carboxyhemoglobin is int erpret ed as oxyhemoglobin by
st andard pulse oximet ers, so a f alsely elevat ed reading w ill result .
Met hemoglobin (w hich has t he visual appearance of a chocolat e brow n color) w ill
f alsely low er t he pulse oximet er reading. O t her met abolic st at es have been
report ed t o cause changes. Hyperbilirubinemia, on t he ot her hand, does not alt er
t he act ual pulse oximet er reading because of t his condit ion it self . How ever, many
ict eric pat ient s w ill have elevat ed serum levels of carboxyhemoglobin because of
t he cat abolism of hemoglobin, and thi s may inf luence t he real oxygen sat urat ion.
Suggested Readings
Barker SJ, Tremper KK, Hyat t J. Eff ect s of met hemoglobinemia on pulse
oximet ry and mixed venous oximet ry. Anest hesiology 1989; 70: 112–117.
Miller RD, ed. Anest hesia. 5t h Ed. Philadelphia: Churchill Livingst one; 2000.
Tremper KK, Barker SJ. Pulse oximet ry. Anest hesiology 1989; 70: 98–108.
Veyckermans F, Baele P, G uillaume JE, et al. Hyperbilirubinemia does not

int erf ere w it h hemoglobin sat urat ion measured by pulse oximet ry.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 101 - R em em ber that
P uls e O xim etr y is Inac c ur ate at Low er S atur ation Levels
101
Remember that Pulse Oximetry is Inaccurate at
Lower Saturation Levels
Nirav G . Shah MD
Pulse oximet ry is a noninvasive met hod t o measure art erial hemoglobin
sat urat ion. The probe f or t he pulse oximet er consist s of t w o light -emit t ing probes
and a phot odet ect or. Pulse oximet ers are able t o dist inguish bet w een
oxygenat ed and deoxygenat ed hemoglobin on t he basis of t heir diff erent ial
absorpt ion of t w o w avelengt hs of light . The phot odiodes t urn on and off several
hundred t imes per second, allow ing t he absorpt ion of light by oxyhemoglobin and
deoxyhemoglobin t o be cont inuously recorded. I t is t he absorpt ion during
pulsat ile f low t hat is a measure of art erial hemoglobin sat urat ion, w hile t he
absorpt ion during nonpulsat ile f low is a f unct ion of t he surrounding t issue and
venous blood.
Watch Out For

Alt hough pulse oximet ry is universally used in t he int ensive care unit (I CU), t he
experienced pract it ioner w ill underst and t he limit at ions and inaccuracies inherent
in using it . First , microprocessors in pulse oximet ers are calibrat ed using
ref erence t ables compiled by exposing healt hy volunt eers t o decreasing FI O2
(f ract ion of inspired oxygen) t o produce SaO2 (oxygen sat urat ion) ranging f rom
75% t o 100% by co-oximet ry. Thus, any reading below 75% is a calculat ed value
ext rapolat ed f rom healt hy volunt eers. At 75% t o 83% t he bias on pulse oximet ry
is 8% f rom co-oximet ry and t his value does not reach an accept able 3% bias
unt il t he sat urat ion is >83% on pulse oximet ry.
A second limit at ion of pulse oximet ry is t he insensit ivit y of t he measured values
t o changes in t he part ial pressure of oxygen because of t he f lat shape of t he
oxyhemoglobin dissociat ion curve at t he upper end of t he spect rum. For example,
in a pat ient receiving supplement al oxygen t herapy, a f all in PaO2 (part ial
pressure of oxygen in art erial blood) of 140 t o 65 mm Hg w ould be required
bef ore a signif icant decrease in oxygen sat urat ion is picked up by t he pulse
oximet er.
A t hird limit at ion is t he inabilit y of t he phot odet ect or t o pick up adequat e signals
t o generat e a value. This can occur w it h poor t issue perf usion (e. g. , peripheral
vasoconst rict ion, pressor administ rat ion w here t he probe can cause t issue injury,
or shock), excessive pat ient
movement (e. g. , seizures or shivering), excessive ambient light , severe anemia,

and hypot hermia.
O ne f inal not e is t hat it is w ell know n t hat t he pulse oximet er cannot diff erent iat e
bet w een oxyhemoglobin and met hemoglobin or carboxyhemoglobin, and t he
presence of eit her of t he lat t er t w o in t he vascular syst em w ill provide an
inaccurat e assessment of t he amount of oxygenat ed hemoglobin.
Suggested Readings
Schnapp LM, Cohen NH. Pulse oximet ry: uses and abuses. Chest
1990; 98: 1244.
Weinberger SE. Recent advances in pulmonary medicine. N Engl J Med

1993; 328: 1389â 1 397.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 102 - C ons ider P lac ing an
Intr ac r anial P r es s ur e Monitor in P atients w ith Glas gow C om a S c ale ≤8
102
Consider Placing an Intracranial Pressure
Monitor in Patients with Glasgow Coma Scale ≤8
Jose I. Suarez MD
Neuronal injury is usually classif ied as primary or secondary. Primary injury
ref ers t o t he int racranial processes t hemselves (i. e. , ischemic or hemorrhagic
st rokes, t rauma, neoplasms) w hereas secondary injury ref ers t o syst emic
condit ions t hat may w orsen t he primary injury. Common processes t hat lead t o
secondary injury include f ever, seizure, hypot ension, and elevat ed int racranial
pressure (I CP). Uncont rolled elevat ed I CP w ill lead t o cerebral ischemia and
most experienced clinicians init iat e t reat ment at an upper t hreshold of 20 t o 25
mmHg. Such t hreshold is based on prospect ive analysis of pat ient s w it h
t raumat ic brain injury (TBI ) and out come relat ed t o I CP elevat ions. The opt imal
predict ive value of out come w as 20 mmHg alt hough dat a are lacking f rom
randomized, cont rolled t rials invest igat ing various I CP t reat ment t hresholds.
ICP M onitoring
The best w ay t o det ermine I CP is t hrough cont inuous monit oring. The
neurological examinat ion may not reliably correlat e w it h I CP elevat ions,
especially in pat ient s in coma. Most of t he dat a available regarding select ion of
pat ient s f or I CP monit oring come f rom TBI pat ient s and most pract it ioners have
ext rapolat ed such crit eria f or pat ient s w it h various int racranial abnormalit ies.
Current recommendat ions indicat e t hat I CP monit oring is appropriat e in pat ient s
w it h severe TBI (G lasgow Coma Scale [ G CS] of 3 t o 8 af t er cardiopulmonary
resuscit at ion) w it h an abnormal admission head comput ed t omography (CT)
scan. An abnormal head CT scan is def ined as one t hat demonst rat es
hemat omas, cont usions, edema, or compressed basal cist erns. I CP monit oring is
also appropriat e in pat ient s w it h severe TBI w it h a normal head CT scan but w ho
present w it h t w o or more of t he f ollow ing: age >40 years, unilat eral or bilat eral
mot or post uring, and syst olic blood pressure <90 mm Hg. Essent ially, any
pat ient w it h G CS ≤8 should be considered f or I CP monit oring regardless of t he
underlying int racranial or syst emic condit ion.
Several t ypes of I CP-monit oring devices are commercially available t oday. These
include int ravent ricular cat het ers (I VCs),
parenchymal cat het ers, subarachnoid bolt s, epidural cat het ers, and lumbar
drains (t o measure t he cerebrospinal f luid [ CSF] pressure in t he lumbar spinal
space). When I VCs are insert ed, a pressure t ransducer can be connect ed t o
measure t he f luid pressure of t he CSF t hat drains. I VCs are considered t he gold
st andard of direct I CP monit oring, because f luid conveys a pressure w ave
reliably. They also have t he added benef it of t reat ing I CP elevat ions by draining
CSF. I VCs are associat ed w it h inf ect ions t hat may range f rom 5% t o 20% and
are more common af t er day f ive of insert ion. Accordingly, some pract it ioners
have recommended replacing cat het ers t hat are more t han 5 days old. O t her
clinicians believe changing t he cat het er should be done only if t here is evidence
of an inf ect ion t hat has not been localized t o a source ot her t han t he cent ral
nervous syst em (CNS). O t her devices include parenchymal cat het ers, such as
t he Camino and Codman MicroSensor cat het ers, w hich are insert ed t hrough a
burr hole in t he skull w it h t he cat het ers being insert ed t hrough t he dura int o brain
parenchyma (or t he vent ricles, if so desired). Hemorrhagic and inf ect ious
complicat ions appear t o be about t he same as w it h an I VC. Subarachnoid bolt s
and epidural cat het ers are less commonly used.
The Glasgow Coma Scale

The best w ay t o det ermine level of consciousness in t he int ensive care unit (I CU)
is by calculat ing t he G CS, w hich is a reproducible 15-point scale in w hich t he
maximum score is 15 and t he minimum score is 3. The scale describes eye
opening, verbal behavior, and mot or responsiveness. The scale has a high
int erobserver reliabilit y among physicians of mult iple disciplines and nursing
st aff . The scale has a good sensit ivit y and reliabilit y (int raclass correlat ion
coeff icient 0. 8 t o 1 f or t rained users). G CS scores bet w een 3 and 8 correspond
t o unresponsive pat ient s. Alt hough t he calculat ion of t he G CS sum score in
crit ically ill subject s has mult if acet ed applicat ions, t he presence of unt est able
component s of t he G CS limit s it s usef ulness. The most common unt est able
f eat ure of t he G CS is t he verbal score, w hich is usually due t o t he presence of
endot racheal int ubat ion. A predict ed verbal score f or subject s w ho are int ubat ed
has been described and validat ed (Tabl e 102. 1), w it h t he t ot al maximum score of
t he G CS in int ubat ed pat ient s st ill 15. The accuracy of t his model w as conf irmed
by comparing t he predict ed verbal score w it h t he act ual verbal score in t he t est
dat a set (n = 736, r = 0. 92, R 2 = 0. 85, p = 0. 0001). The derivat ion of t he verbal
score based on t he eye and mot or component s
w as f urt her validat ed in a larger dat a set (n = 24, 085, r = 0. 97, p = 0. 0001).
TABLE 102-1 CALCULATED PREDICTED GCS VERBAL
SCORE FROM EYE AND M OTOR SCORES USING
LINEAR REGRESSION
GCS EYE SCORE
GCS MOTOR SCORE 1 2 3 4
1 1 1 1 2
2 1 2 2 2
3 2 2 3 3
4 2 3 3 4
5 3 3 4 4
6 3 4 4 5
Reused with permission from Rutledge R, Lentz CW,

Fakhry S, et al. Appropriate use of the Glasgow Coma
Scale in intubated patients: a linear regression
prediction of the Glasgow verbal score from the
Glasgow eye and motor scores. J Trauma 1996;41:514–
522.
Suggested Readings
Meredit h W, Rut ledge R, Fakhry SM, et al. The conundrum of t he G lasgow
Coma Scale in int ubat ed pat ient s: a linear regression predict ion of t he
G lasgow verbal score f rom t he G lasgow eye and mot or scores. J Trauma
1998; 44: 839–844.
Rut ledge R, Lent z CW, Fakhry S, et al. Appropriat e use of t he G lasgow Coma
Scale in int ubat ed pat ient s: a linear regression predict ion of t he G lasgow
verbal score f rom t he G lasgow eye and mot or scores. J Trauma
1996; 41: 514–522.
Teasdale G , Jennet t B. Assessment of coma and impaired consciousness: a

pract ical scale. Lancet 1977; 2: 81–84.
Teasdale G , Knill-Jones R, van der Sande J. O bserver variabilit y in assessing

impaired consciousness and coma. J Neurol Neurosurg Psychiat ry
1978; 41: 603–610.
The Brain Trauma Foundat ion. The American Associat ion of Neurological
Surgeons. The Joint Sect ion on Neurot rauma and Crit ical Care. I ndicat ions f or
I nt racranial Pressure Monit oring. J Neurot rauma 2000; 17: 479–491.
Surgeons. The Joint Sect ion on Neurot rauma and Crit ical Care. I nt racranial
Pressure Treat ment Threshold. J Neurot rauma 2000; 17: 493–495.
> Table of C ontents > D evic es /Lines /Tubes /C atheter s /D r ains /P r oc edur es > 103 - K now How to Us e the
Lic ox S ys tem to Meas ur e B r ain Tis s ue O xygenation
103
Know How to Use the Licox System to Measure
Brain Tissue Oxygenation
Aaron Bransky MD
Heidi L. Frankel MD
Traumat ic brain injury is a leading cause of deat h and disabilit y in young people.
I n an at t empt t o direct t herapy, int racranial pressure (I CP) monit oring and
calculat ed cerebral perf usion pressure (CPP) have been used t o make bedside
management decisions. How ever, several st udies have show n t hat cerebral
inf arct ion can occur even w it h normal I CP and CPP.
G iven t his inf ormat ion, brain t issue oxygen monit oring has been invest igat ed t o
give an earlier marker of brain ischemia. O ne new t echnology has been t he use
of a Clark-t ype cat het er t o measure part ial oxygenat ion of brain t issue (PbO 2 ).
These probes can be placed in t he brain cort ex during craniot omy or in t he
int ensive care unit (I CU) by placing a burr hole. There are diff ering opinions as
t o w het her t he probe should be placed in healt hy brain t issue or t he penumbra of
injury.
How ever, it is generally agreed t hat PbO2 levels should be maint ained great er
t han 15 t o 20 mmHg. I f an acut e decrease is discovered, it may represent an
earlier sign of brain ischemia t han eit her elevat ed I CP or CPP. Cont radict ory t o
t herapy f or elevat ed I CP, if a pat ient develops low PbO2 , t he physician should
low er t he head of t he bed and consider permissive hypercapnia f or cerebral
vasodilat at ion. Addit ionally, raising global oxygen delivery by means such as
increasing FI O2 (f ract ion of inspired oxygen) and t ransf using packed red blood
cells, or raising cardiac out put has show n some success in raising PbO2 . The
manuf act urer (Licox) recommends use f or up t o 5 t o 7 days.
Suggested Readings
St ief el MF, Spiot t a A, G racias VH, et al. Reduced mort alit y rat e in pat ient s
w it h severe t raumat ic brain injury t reat ed w it h brain t issue oxygen monit oring.
J Neurosurg 2005; 103: 805–811.
Valadka AB, G opinat h SP, Cont ant CF, et al. Relat ionship of brain t issue PO 2
t o out come af t er severe head injury. Crit Care Med 1998; 26: 1576–1581.
van den Brink WA, van Sant brink H, St eyerberg EW, et al. Brain oxygen
t ension in severe head injury. Neurosurgery 2000; 46: 868–876; discussion
876–878.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 104 - P r eoxygenate P atients B efor e
Intubation
104
Preoxygenate Patients Before Intubation
Rahul G . Baijal MD
The purpose of preoxygenat ing a pat ient bef ore induct ion of general anest hesia
and paralysis is t o provide maximum t ime t hat a pat ient can t olerat e apnea.
Maximum preoxygenat ion is achieved w hen t he alveolar, art erial, t issue, and
venous compart ment s are f illed w it h oxygen. Pat ient s in w hom oxygen ext ract ion
is increased (e. g. , hypert hermia, acidosis, hypercarbia) or oxygen loading is
decreased (e. g. , decreased f unct ional residual capacit y, hemoglobin
concent rat ion, alveolar vent ilat ion, cardiac out put ) desat urat e f ast er during
apnea t han a healt hy pat ient and consequent ly require maximum preoxygenat ion.
Various f act ors may necessit at e preoxygenat ion w hen mask vent ilat ion is not
possible, including diff icult y maint aining airw ay pat ency; a f ull st omach w here
pressure of t he upper abdomen may induce regurgit at ion; ant icipat ed diff icult
airw ay requiring increased apneic t ime; morbid obesit y w here high pressures are
required t o vent ilat e t he lung; and pregnancy w here increased abdominal
pressure may also induce regurgit at ion.
What to Do
Funct ional residual capacit y (FRC) is t he volume remaining in t he lungs at
exhalat ion f ollow ing normal t idal volume breat hing. FRC is approximat ely 2, 500
mL in a healt hy adult and is reduced as a pat ient is moved f rom an upright t o a
supine or prone posit ion. FRC is addit ionally reduced by 15% t o 20% f ollow ing
induct ion of anest hesia. During preoxygenat ion, t he pat ient inspires 100% oxygen
via a f ace mask bef ore induct ion, replacing nit rogen w it h oxygen in t he pat ient 's
FRC. Normal oxygen consumpt ion in a healt hy adult is approximat ely 250
mL/ min. O xygen desat urat ion may occur as rapidly as 30 t o 60 seconds in a
healt hy adult w it h an FRC of 21% oxygen f ollow ing induct ion of anest hesia and
subsequent apnea, despit e normal init ial oxygen sat urat ion. Denit rogenat ion
during spont aneous breat hing is 95% complet e w it hin 3 minut es w hen a pat ient is
breat hing a normal t idal volume of 100% oxygen. This increases t he margin of
saf et y t o approximat ely 4 t o 6 t imes during periods of apnea f ollow ing induct ion
of anest hesia. Preoxygenat ion w it h eight maximum deep breat hs over 60
seconds result s in art erial oxygenat ion t hat is not diff erent f rom t idal volume
breat hing f or 3 minut es. This t echnique increases minut e vent ilat ion above
FRC and minimizes nit rogen rebreat hing, ensuring w ashout of FRC; addit ionally,
t aking eight deep breat hs may open collapsed airw ays, increasing FRC oxygen
st ore. Four maximum breat hs over 30 seconds also increases art erial
oxygenat ion, but t he t ime f or hemoglobin desat urat ion is short ened compared
w it h pat ient s breat hing normal t idal volume f or 3 minut es or t aking eight
maximum breat hs over 60 seconds.
Prevent able reasons t hat maximum preoxygenat ion may not be achieved include
a leak under t he mask, allow ing inspirat ory ent rainment of room air, and
insuff icient preoxygenat ion t ime. Avoiding a leak bet w een t he mask and t he f ace
is an import ant f act or t hat cannot be compensat ed f or by increased durat ion of
preoxygenat ion.
Suggested Readings
Baraka AS, et al. Preoxygenat ion: comparison of maximal breat hing and t idal
volume breat hing t echniques. Anest hesiology 1999; 91: 612–616.
Miller RD. Miller's Anest hesia. 6t h Ed. Philadelphia: Elsevier; 2005: 617–719.
East w ood WD. A st udy of dinit rogenat ion w it h some inhalat ion and anest het ic
syst ems. Anest hesiology 1955; 16: 861–867.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 105 - Us e C r ic oid P r es s ur e w hen
P er for m ing R apid S equenc e Intubation or B ag- Mas k Ventilation
105
Use Cricoid Pressure when Performing Rapid
Sequence Intubation or Bag-Mask Ventilation
Rahul G . Baijal MD
Cricoid pressure, f irst described by Sellick in 1961, is used t o occlude t he upper
esophagus t o reduce t he aspirat ion of gast ric cont ent s during rapid-sequence
int ubat ion. Pulmonary aspirat ion of gast ric cont ent s occurs in t hree st ages. First ,
gast ric cont ent s ref lux int o t he esophagus w hen t he low er esophageal sphinct er
relaxes. Esophageal cont ent s t hen ref lux int o t he pharynx w hen t he upper
esophageal sphinct er relaxes. Finally, pulmonary aspirat ion of pharyngeal
cont ent s occurs af t er loss of laryngeal ref lexes.
To brief ly review, t he cricoid cart ilage is t he only upper airw ay cart ilaginous
st ruct ure t hat is a complet e ring. The esophagus begins at t he low er border of
t he cricoid cart ilage. Cricoid pressure replaces t he f unct ion of t he upper
esophageal sphinct er by compressing t he lumen of t he upper esophagus bet w een
lamina of t he cricoid cart ilage and t he body of t he sixt h cervical vert ebrae,
prevent ing regurgit at ion of esophageal cont ent s int o t he pharynx (t he second
phase of aspirat ion). The upper esophageal sphinct er is f ormed ant eriorly by t he
lamina of t he cricoid cart ilage and post eriorly by t he cricopharyngeus muscle,
w hich is at t ached t o t he lat eral aspect s of t he cricoid cart ilage. Upper
esophageal sphinct er t one is 40 mm Hg in aw ake pat ient s and decreases t o less
t han 10 mm Hg during induct ion of anest hesia. Regurgit at ion of esophageal
cont ent s int o t he pharynx occurs w hen t he upper esophageal sphinct er pressure
is less t han 25 mm Hg. Theref ore, induct ion of anest hesia can decrease upper
esophageal sphinct er pressure suff icient ly t o allow regurgit at ion of esophageal
cont ent s int o t he pharynx. Cricoid pressure count eract s t his reduct ion in upper
esophageal pressure. I t is import ant t o not e t hat ot her cart ilaginous st ruct ures in
t he upper airw ay are u-shaped; mist akenly exert ing pressure on t hese st ruct ures
w ill be ineff ect ive and possibly lead t o airw ay damage or dist ort ion during
int ubat ion.
To avoid t his, t he cricoid cart ilage should f irst be ident if ied and palpat ed bef ore
induct ion of anest hesia. Wit h single-handed cricoid pressure, t he t humb and
middle f inger can be placed on eit her side of t he cricoid cart ilage w it h t he index
f inger caudal. The t humb and middle f inger prevent lat eral movement of t he
cricoid cart ilage and t he
index f inger provides ant erior-post erior pressure. The disadvant age of t his
t echnique is t hat t he ext ended neck w ill t end t o collapse t he arch and reduce t he
glot t ic view. Anot her single-handed t echnique is t o place t he palm of t he hand on
t he st ernum and apply pressure t o t he cricoid cart ilage w it h only t he index and
middle f ingers. I n t he t w o-handed, or bimanual, t echnique, cricoid pressure is
perf ormed as w it h t he single-handed t echnique except t hat t he assist ant 's ot her
hand provides count erpressure beneat h t he cervical vert ebrae, support ing t he
neck.
Cricoid pressure can w it hst and esophageal pressures of 50 cm H2 O (1 cm H2 O =
0. 73 mm Hg). Cricoid pressure can overcome gast ric pressures associat ed w it h
f ast ing (<18 cm H2 O ), delayed gast ric empt ying (<50 cm H2 O ), eruct at ion (<20
cm H2 O ), pregnant supine pat ient s (<35 cm H2 O ), or f asiculat ions associat ed
w it h succinylcholine administ rat ion (<40 cm H2 O ). How ever, vomit ing, w hich
creat es esophageal pressures >60 cm H2 O , can overcome cricoid pressure,
leading t o regurgit at ion and pulmonary aspirat ion. Cricoid pressure should begin
bef ore t he pat ient is f ully asleep, w it h 10 N of pressure, increasing t o 30 N of
pressure w hen loss of consciousness is est ablished. Recommended cricoid
pressure t o prevent pulmonary aspirat ion is bet w een 30 and 40 N (10 N ~ 1 kg),
but pressure >20 N can cause pain and ret ching in aw ake pat ient s, and pressure
>40 N can dist ort t he t rachea, making int ubat ion diff icult . Ret ching causes a
ref lex relaxat ion of t he low er esophageal sphinct er and rapid dist ent ion of t he
esophagus, relaxing t he upper esophageal sphinct er and allow ing expulsion of
vomit . Cricoid pressure maint ained during ret ching allow s esophageal pressure
t o rise, increasing t he risk of esophageal rupt ure. Cricoid pressure is addit ionally
cont raindicat ed in pat ient s w it h cricot racheal injury, act ive emesis, or unst able
cervical spine injuries.
Finally, bag-mask vent ilat ion provides posit ive pressure vent ilat ion w it h delivery
of high f ract ion of inspired oxygen, opening at -elect at ic alveoli and improving
vent ilat ion-perf usion mismat ch. How ever, bag-mask vent ilat ion may rapidly
dist end t he st omach, increasing gast ric pressure up t o 30 cm H2 O . Cricoid
pressure may overcome t his increased gast ric pressure as not ed previously but
may be uncomf ort able in an aw ake pat ient .
Suggested Readings
Allman K. The eff ect of cricoid pressure applicat ion on airw ay pat ency. J Clin
Anest h 1995; 7: 195–199.
Asai T. Cricoid pressure applied af t er placement of t he laryngeal mask
prevent s gast ric insuff lat ion but inhibit s vent ilat ion. Br J Anaest h
1996; 76: 772–776.
Landsman I . Cricoid pressure: indicat ions and complicat ions. Pediat r Anest h
2004; 14: 43–47.
Vanner RG . Mechanism of regurgit at ion and it s prevent ion w it h cricoid

pressure. I nt J O bst et Anest h 1993; 2: 207–215.
Vanner RG , Hart silver EL. Airw ay obst ruct ion w it h cricoid pressure.
Anest hesia 2000; 55: 208–211.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 106 - D o not us e The P r es enc e of
E nd- Tidal C O 2 to R ule O ut E s ophageal Intubation
106
Do not use The Presence of End-Tidal CO2 to
Rule Out Esophageal Intubation
Scenario
You are called t o emergent ly int ubat e an obese w oman f or increasing respirat ory
dist ress w ho is f our hours post operat ive f rom an unevent f ul nephrect omy. She
w as successf ully ext ubat ed in t he operat ing room and init ially on admission t o
t he int ensive care unit (I CU) looked like she w ould be a st raight f orw ard one-night
I CU st ay. O n your arrival t o t he bedside, t he medical st udent is at t empt ing t o
mask-vent ilat e her. You ready your laryngoscope and t he endot racheal t ube and
proceed t o secure t he airw ay. You are conf ident t hat you int ubat ed t he t rachea
alt hough her body habit us made t he int ubat ion challenging. The respirat ory
t herapist hooks up t he port able end-t idal CO2 monit or, gives t he pat ient a
breat h, and t ells you “good job” because t he port able endt idal CO2 monit or
det ect s t he presence of CO2 . The nurse also t ells you he hears bilat eral breat h
sounds. Are you sat isf ied w it h t his assessment and ready t o have t he respirat ory
t herapist t ape t he endot racheal t ube?
Explanation
No! You should not be sat isf ied w it h t his assessment of t he placement of t he
endot racheal t ube. Det ect ion of CO2 in one breat h by t he CO2 monit or is not
suff icient t o ensure t hat t he endot racheal t ube is indeed in t he t rachea.
Port able CO2 monit ors cont ain lit mus paper t hat quickly changes color (i. e. , f rom
purple t o yellow ) in response t o exhaled CO2 . I t is import ant t o not e t hat some
air is alw ays f orced int o t he st omach w hen mask-vent ilat ing a pat ient bef ore
int ubat ion. During a diff icult mask vent ilat ion (as in t he previous scenario w it h an
inexperienced medical st udent and an obese pat ient ), a signif icant amount of air
can be int roduced int o t he st omach. I f t he esophagus is int ubat ed, t he CO2
det ect or w ill change color because it is det ect ing CO2 f rom t he st omach. I t is
only af t er at least 4 t o 5 breat hs w it h a cont inued color change on t he CO2
monit or t hat you can be conf ident t hat t he t rachea has been int ubat ed.
I n addit ion, auscult at ing f or bilat eral breat h sounds is not suff icient t o conf irm
endot racheal int ubat ion. Sounds f rom air moving
t hrough t he esophagus and st omach can be mist aken f or bilat eral breat h sounds.
Theref ore, it is imperat ive t o also include auscult at ion over t he gast ric region t o
ensure t he absence of breat h sounds w it h t he more obvious auscult at ion of t he
bilat eral lung f ields f or t he presence of breat h sounds.
O ne f inal not e concerns t he use of port able CO2 monit ors in cardiac arrest .
During cardiac arrest (in t he absence of adequat e cardiopulmonary resuscit at ion
[ CPR] ), t here is no blood f low t o t he lungs; t heref ore, CO2 is not being delivered
t o t he lungs. The CO2 det ect or w ill not change color even if t he endot racheal
t ube is correct ly posit ioned. During t he resuscit at ion, t he CO2 det ect or may
change color if eff ect ive chest compressions are being done and blood f low is
part ially rest ored t o t he lungs.
Suggested Readings
Hogg K, Teece S. Best evidence t opic report : colourmet ric CO2 det ect or
compared w it h capnography f or conf irming ET t ube placement . Emerg Med J
2003; 20: 265–266.
Miller R, ed. Miller's Anest hesia. 6t h Ed. Philadelphia: Churchill Livingst one;
2005: 1648.
Nellcor Purit an Bennet t , I nc. CO2 Det ect ion: int ubat ion and cardiopulmonary
resuscit at ion. Clinical Monographs. June 1, 2003. ht t p: / / w w w. nellcor. com
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 107 - Have a Low Thr es hold for
C ontac ting the Mos t E xper ienc ed Available Air w ay P r ofes s ional in P atients w ith D is eas es As s oc iated
w ith D iffic ult Air w ays
107
Have a Low Threshold for Contacting the Most
Experienced Available Airway Professional in
Patients with Diseases Associated with Difficult
Airways
Lauren C. Berkow MD
Pat ient s in t he int ensive care unit (I CU) may require airw ay management f or a
variet y of reasons, but t he most common is respirat ory dist ress or f ailure.
Pat ient s may also require int ubat ion f or an invasive procedure. Alt hough t he
majorit y of pat ient s can be saf ely int ubat ed by I CU personnel t rained in airw ay
management , cert ain pat ient s may require t he assist ance of an anest hesiologist .
Anest hesiologist s are t rained t o manage t he diff icult airw ay and can provide
addit ional resources and skills not rout inely f ound in t he I CU set t ing.
Watch Out For

A pat ient previously ident if ied as a diff icult int ubat ion in t he past may not easily
be int ubat ed by direct laryngoscopy, and enlist ing t he help of an anest hesiologist
if t he pat ient requires airw ay management may prevent complicat ions. I t is also
import ant t o keep in mind t hat a pat ient w ho requires reint ubat ion
post operat ively may have developed airw ay edema or laryngospasm t hat may
complicat e int ubat ion. Reint ubat ion may also be diff icult if t he pat ient has
undergone surgery involving or close t o t he airw ay (cervical f usion,
t hyroidect omy, carot id endart erect omy) or has received large volumes of
int ravenous f luids or blood product s t hat could lead t o airw ay edema.
I CU pat ient s in respirat ory dist ress or f ailure of t en have decreased respirat ory
reserve and may be hypoxic. These pat ient s desat urat e much more quickly
during airw ay management , so preoxygenat ion is crit ical, and t he amount of t ime
t o saf ely secure t he airw ay might be diminished. This is especially common in
pat ient s w it h obst ruct ive sleep apnea or pat ient s requiring cont inuous posit ive
airw ay pressure (CPAP) or bilevel posit ive airw ay pressure (BiPAP) t o maint ain
oxygenat ion. I n t hese pat ient s, if diff icult y w it h mask vent ilat ion or int ubat ion is
suspect ed, cont act ing your airw ay prof essional f or assist ance prior t o
int ervent ion may be prudent . I t may also be saf er t o secure t he airw ay early, and
in a cont rolled f ashion, prior t o t he development of acut e respirat ory dist ress or
f ailure.
Access t o t he pat ient f or airw ay management in t he I CU is of t en more diff icult
t han in t he operat ing room because of t he smaller size
of t he room. Many cables and plugs are connect ed t o an I CU bed and addit ional
monit ors or int ravenous pumps may limit access t o t he head of t he bed f or
airw ay manipulat ion; space f or addit ional airw ay equipment may be limit ed. I t
may also be diff icult f or addit ional personnel t o access t he pat ient , f or inst ance,
if a cricot hyroidot omy or t racheost omy needs t o be perf ormed. This can increase
t he challenge of diff icult airw ay management .
Pat ient s w it h specif ic diseases may have an increased likelihood of diff icult
int ubat ion (Tabl e 107. 1). Some of t hese pat ient s may be previously ident if ied as
a diff icult int ubat ion, but many may not . I n addit ion, t hese pat ient s may have
undergone procedures t hat may f urt her complicat e mask vent ilat ion or int ubat ion,
such as cervical f usion or uvulopalat oplast y. These pat ient s may also be more
likely t o require int ensive care management post operat ively af t er a major
procedure because of eit her t heir underlying disease st at e or t heir diff icult
airw ay st at us.
The I CU set t ing should cont ain equipment f or bot h t he rout ine and nonrout ine
int ubat ion in an easily accessible locat ion, and st aff members should be f amiliar
w it h t he locat ion and t he equipment . Specialized airw ay equipment such as a
f iberopt ic scope, int ubat ion st ylet s (Eschmann, gum elast ic bougie), laryngeal
mask airw ays, cricot hyrot omy set , and so on can be assembled int o a cart f or
quick access. Emergency cont act numbers f or anest hesiology and t rauma or ear,
nose, and t hroat (ENT) surgery should also be clearly post ed in t he I CU.
TABLE 107-1 DISEASE STATES ASSOCIATED WITH

DIFFICULT AIRWAY M ANAGEM ENT
CONGENITAL ACQUIRED
Pierre Robin syndrome Morbid obesity
Treacher Collins
Acromegaly
syndrome
Infections involving the
Goldenhar syndrome
airways (Ludwig angina)
Mucopolysaccharidoses Rheumatoid arthritis
Achondroplasia Obstructive sleep apnea
Micrognathia Ankylosing spondylitis
Tumors involving the airway

Down syndrome Trauma (airways, cervical
spine)
Adapted from Barash PG, Cullen BF, Stoelting RK, eds.

Clinical Anesthesia. Philadelphia: Lippincott W illiams &
W ilkins; 2001; and Benum of JL, ed. Airway
Management: Principles and Practice. St. Louis:
Mosby; 1996.
So, in a pat ient w it h a know n or suspect ed diff icult airw ay, w ho should perf orm
t he int ubat ion? I t is recommended t hat t he most skilled
prof essional should perf orm t he procedure. Depending on resources available

and t he urgency of t he int ubat ion, t his may be an I CU prof essional, an
anest hesiologist (pref erable), or a t rauma or ENT surgeon. Having a variet y of
airw ay equipment available at all t imes in t he I CU set t ing is crucial, especially if
addit ional resources are not quickly available. I n case of emergency, especially
if supraglot t ic pat hology is a possibilit y, t here should be a physician available
w ho is able t o perf orm a cricot hyroidot omy if needed.
Suggested Readings
Benum J. Airw ay Management : Principles and Pract ice. St . Louis:
Mosby; 1996: 129–130.
Pract ice guidelines f or management of t he diff icult airw ay. An updat ed report
by t he American Societ y of Anest hesiologist s Task Force on Management of
t he Diff icult Airw ay. Anest hesiology 2003; 98: 1269–1277.
St oet ling RK, Barash PG , Cullen BF. Clinical Anest hesia. Philadelphia:
Lippincot t Raven, 1997.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 108 - Alw ays Us e a Ver tic al Inc is ion
w hen P er for m ing a C r ic othyr oidotom y
108
Always Use a Vertical Incision when Performing
a Cricothyroidotomy
Brandon R. Bruns MD
Heidi L. Frankel MD
No sit uat ion in medicine is as st ressf ul as t he management of a pat ient w it h an
unst able airw ay af t er mult iple f ailed at t empt s at endot racheal int ubat ion.
Cricot hyroidot omy can save t he day if perf ormed correct ly. The indicat ion f or
cricot hyroidot omy is simple: t he inabilit y t o obt ain an airw ay by any ot her means.
The maneuver is t heoret ically simple t o perf orm. How ever, t he st ressf ul sit uat ion
surrounding a compromised airw ay requires f ull underst anding of t he t echnique
of cricot hyroidot omy and a st eady hand.
I f a prepackaged cricot hyroidot omy kit is available, t he operat or may request it .
O t herw ise, an ant isept ic, a scalpel, and an airw ay device are all t hat w ill be
required. Chlorhexidine is a bet t er ant isept ic t han is an iodine-based solut ion,
but t his is of secondary concern. The airw ay device may eit her be a small-
caliber t racheost omy t ube or an endot racheal t ube. The anat omical landmarks
used t o ident if y t he cricot hyroid membrane are t he t hyroid cart ilage (Adam's
apple) above and t he cricoid ring below. The t hyroid cart ilage is grasped in t he
operat or's nondominant hand w hile t he ot her hand is used t o make a generous
verti cal incision t hrough t he skin and subcut aneous t issue. A vert ical incision
avoids t he ant erior jugular veins t o minimize bleeding. Making a generous incision
w ill allow t he necessary exposure. O nce t he cricoid membrane is locat ed, it is
perf orat ed w it h t he t ip of t he scalpel, w hich can be moved caudally t o expose
t he airw ay opening and allow f or easy placement of t he t ube. O nce t he t ube has
been placed in t he airw ay, t he operat or or assist ant should maint ain cont rol of
t he airw ay by keeping at least one hand on it t o secure it in place. Tube
placement should be conf irmed by verif ying breat h sounds and checking end-t idal
carbon dioxide. I t should now be easy t o oxygenat e and vent ilat e t he pat ient . I t
is cont roversial w het her a cricot hyroidot omy must be convert ed t o a
t racheost omy if needed long t erm.
Experienced pract it ioners w ill report t hat t he most diff icult aspect of perf orming
a cricot hyroidot omy is making t he decision t o put one in. Alt hough avoiding t he
ant erior jugular veins is highly desirable, speed is of t he essence and should
t ake considerat ion over cosmesis and an inexact ent ry int o t he t rachea.
Suggested Readings
Boyd AD, Romit a MC, Conlan AA, et al. A clinical evaluat ion of
cricot hyroidot omy. Surg G ynecol O bst et 1979; 149: 365–368.
Francois B, Clavel M, Desachy A, et al. Complicat ions of t racheost omy

perf ormed in t he I CU: subt hyroid t racheost omy versus surgical
cricot hyroidot omy. Chest 2003; 123: 151–158.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 109 - Us e B r onc hos c opic Guidanc e for
B eds ide P er c utaneous D ilatational Tr ac heos tom y ( P D T)
109
Use Bronchoscopic Guidance for Bedside
Percutaneous Dilatational Tracheostomy (PDT)
Elliott R. Haut MD
Percut aneous dilat at ional t racheost omy (PDT) has quickly become t he st andard
met hod f or conversion f rom short -t erm endot racheal int ubat ion t o t racheosot omy
f or longer-t erm mechanical vent ilat ion. The procedure can be saf ely perf ormed in
t he I CU, avoiding t ransport of crit ically ill pat ient s t o t he operat ing room. The
perf ormance of a bedside PDT is cont ingent upon saf et y-minded advanced
planning and a coordinat ed t eam eff ort . G iven t he inherent risks t o t he
procedure, t hought f ul preparat ion is essent ial. I n t he best case, a bedside
t racheost omy proceeds smoot hly; in t he w orst case it can end in t ragedy.
Alt hough some providers perf orm bedside PDT w it hout direct visualizat ion, most
have now adopt ed t he prot ocol of using direct bronchoscopic visualizat ion.
Bef ore st art ing t his procedure, adequat e personnel must be assembled. O ne
person is designat ed t o manipulat e t he f lexible bronchoscope and st ands at t he
head of t he bed. A second person is chosen t o maint ain cont rol of t he airw ay
during manipulat ion of t he endot racheal t ube. Anot her person should be in charge
of sedat ion, anest hesia, and monit oring. O ne or t w o people are dedicat ed t o
perf orming t he t racheost omy it self .
There is no one best surgical t echnique f or perf orming bedside percut aneous
t racheost omy. A commonly used t echnique is t he sequent ial dilat ion t echnique
(i. e. , Blue Rhino by Cook). The f ollow ing st eps are used by many surgeons
experienced at t his t echnique:
Posit ion t he pat ient by placing a rolled sheet beneat h t he shoulders t o

ext end t he neck and expose it s ant erior st ruct ures.
Place a bedside t able above t he pat ient 's t orso and prep and drape t he
pat ient .
Test t he t racheost omy balloon and lay out each inst rument in t he order in
w hich t hey w ill be used.
Af t er inf ilt rat ing local anest het ic, use t he scalpel t o make a vert ical incision
t hrough t he skin int o t he subcut aneous t issue (about 2 cm superior t o t he
st ernal not ch).
Use a blunt clamp t o dissect t he midline connect ive t issues off t he ant erior
w all of t he t rachea (use of a blunt t echnique minimizes bleeding pot ent ial).
The bronchoscope operat or advances t he endoscope t o t ransilluminat e t he
sit e of t he ent ry int o t he t rachea (w hen posit ioned correct ly t he surgeons w ill
see a glow t hrough t he t rachea).
Advance a f inder needle and t hen a larger bore needle caref ully t hrough t he
t rachea bet w een t he f irst and second or second and t hird t racheal rings
under direct visualizat ion w it h t he broncohoscope.
Pass t he w ire t hrough t he needle and remove t he needle (Seldinger
t echnique).
Visualize t he w ire placement int o t he t rachea and direct ion of passage w it h
bronchoscopic visualizat ion.
Begin t he sequent ial dilat ions over t he w ire (Seldinger t echnique) and
event ually place t he t racheost omy t ube.
Conf irm t ube placement by visualizat ion of t he carina w hen t he bronchoscope
is placed dow n t he new t racheost omy t ube and ret urn of end-t idal CO2 w hen
vent ilat ed.
There are a number of t echniques t hat should keep t he operat or out of surgical
bleeding. First and f oremost , t he surgeon should know t he anat omy of t he neck
and st ay bet w een t he parallel ant erior jugular veins. These run on eit her side of
t he t rachea bet w een t he midline and st ernocleidomast oid, coursing f rom t he
region of t he hyoid inf eriorly t o drain t o t he ext ernal jugular vein or subclavian
vein. By st aying squarely in t he midline and using a vert ical incision, t he surgeon
can avoid t hese st ruct ures, w hich may bleed prof usely if ent ered and complet ely
obscure t he operat ive f ield. I f t hese veins are visualized and are likely t o
int erf ere w it h t he procedure, t hey may be sut ure ligat ed. Anot her pot ent ial
source of bleeding includes t he divided t hyroid ist hmus w hich can usually be
cont rolled w it h direct pressure. The most serious (but t hankf ully rare) source of
bleeding is t he innominat e art ery, w hich may necessit at e median st ernot omy f or
exposure and cont rol. The t eam should make every eff ort t o remain calm in t he
event of prof use bleeding. O f ut most import ance in t his sit uat ion is t o maint ain
t he airw ay and prot ect t he endot racheal t ube. Provided t his is done, t he t eam
can t hen decide on t he appropriat e next st ep, w hich may be a move t o t he
operat ing room w hile maint aining direct pressure on any sources of bleeding.
O f t en t his venous bleeding w ill st op f rom t amponade once t he t racheost omy is in
place.
O ne t ool t hat should never be used t o cont rol bleeding during t racheost omy is
elect rocaut ery. The combinat ion of t he elect ric current
w it h sparks and high-f low oxygen f rom t he endot racheal t ube comprise t he t w o
ingredient s f or combust ion, risking serious burns t o t he pat ient and st aff .
Alt hough surgeons have become accust omed t o using elect rocaut ery f or
hemost asis, t his is a case w here it is cont raindicat ed.
Suggested Readings
Bojar RM, Warner KG , eds. Manual of Perioperat ive Care in Cardiac Surgery.
3rd Ed. Blackw ell: Malden; 1999: 459–461.
Zollinger RM Jr, Zollinger RM Sr, eds. Zollinger's At las of Surgical

O perat ions. 8t h Ed. New York: McG raw -Hill; 2003: 374–375.
ht t p: / / w w w. cookgroup. com/ cook crit ical care/ f eat ures/ blue rhino. ht ml
Delaney A, Bagshaw SM, Nalos M. Percut aneous dilat at ional t racheost omy
versus surgical t racheost omy in crit ically ill pat ient s: a syst emat ic review and
met a-analysis. Crit Care. 2006; 10(2): R55. at
ht t p: / / ccf orum. com/ cont ent / 10/ 2/ R55
Freeman BD, I sabella K, Lin N, Buchman TG . A met a-analysis of prospect ive

t rials comparing percut aneous and surgical t racheost omy in crit ically ill
pat ient s. Chest . 2000; 118(5): 1412–1418.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 110 - C ons ider E ar ly Tr ac heos tom y in
S elec t P atients
110
Consider Early Tracheostomy in Select Patients
G eorge C. Velmahos MD, PHD
The t iming of t racheost omy present s an ongoing debat e. St udies are inconsist ent
on w hat is considered early or lat e. St udied populat ions are relat ively small and
met a-analysis is f law ed as t he crit eria of t racheost omy t iming vary f rom paper t o
paper. The risk of t he surgical procedure is not insignif icant . Direct procedural
complicat ions (such as bleeding), long-t erm complicat ions (such as t racheal
st enosis), and indirect complicat ions (such as risks of t ransport ing crit ically ill
pat ient s t o t he operat ing room f or open t racheost omy) should be balanced
against t he benef it s of bet t er oral and pulmonary t oilet , easier w eaning, and
reduct ion of airw ay resist ance. The risks of t ransport at ion f rom t he int ensive
care unit (I CU) t o t he operat ing room include t ube and line dislodgment ,
subopt imal monit oring during t ransport , and inabilit y t o manage crit ical event s.
Bedside t racheost omy by t he percut aneous dilat at ional t echnique has emerged
as a saf e alt ernat ive t o open t racheost omy. All w ould agree t hat if a
t racheost omy is event ually needed, it w ould bet t er be done earlier t han lat er.
The problem lies in our inadequacy t o predict accurat ely t he durat ion of
mechanical vent ilat ion. For pat ient s w ho w ould need prolonged mechanical
vent ilat ion (usually def ined as longer t han 2 w eeks), most surgeons agree t hat
an early t racheost omy is benef icial.
The exist ing prospect ive randomized t rials have f ailed t o show a survival
advant age of early t racheost omy. How ever, according t o f ive randomized cont rol
t rials, t he lengt h of mechanical vent ilat ion is signif icant ly reduced by a
t racheost omy perf ormed w it hin 6 days of hospit al admission compared w it h a
t racheost omy perf ormed af t er 14 days.
Suggested Readings
G riff it hs J, Barber VS, Morgan L, et al. Syst emat ic review and met a-analysis
of st udies of t he t iming of t racheost omy in adult pat ient s undergoing art if icial
vent ilat ion. BMJ 2005; 330: 1243–1247.
Moller MG , Slaikeu JD, Bonelli P, et al. Early t racheost omy versus lat e
t racheost omy in t he surgical int ensive care unit . Am J Surg 2005; 189: 293–
296.
Velmahos G C, Belzberg H, Chan L, et al. Fact ors predict ing prolonged

mechanical vent ilat ion in crit ically injured pat ient s: int roducing a simplif ied
quant it at ive risk score. Am Surg 1997; 63: 811–817.
Velmahos G C, G omez H, Boicey CM, et al. Bedside percut aneous

t racheost omy: prospect ive evaluat ion of a modif icat ion of t he current
t echnique in 100 pat ient s. World J Surg 2000; 24: 1109–1115.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 111 - P os ition the Tip of the
E ndotr ac heal Tube 4 C entim eter s Above the C ar ina
111
Position the Tip of the Endotracheal Tube 4
Centimeters Above the Carina
Leo Hsiao DO
Est ablishment of t he airw ay is an int egral part of any resuscit at ive algorit hm. I n
t he acut e set t ing it is of t en necessary t o int ubat e a pat ient in order t o secure a
means of providing vent ilat ion. To t his end, t he posit ion of t he endot racheal t ube
is paramount t o it s success. A malposit ioned endot racheal t ube can be
det riment al and pot ent ially lif e t hreat ening (Fi g. 111. 1).
To brief ly review, t he upper airw ay begins at t he nose and mout h and ends at t he
carina w here t he t rachea divides int o t he lef t and right main-st em bronchi. The
nose and mout h f orm a common passage post eriorly called t he pharynx, w hich is
divided int o t hree part s: nasopharynx, oropharynx, and laryngopharynx. The
pharynx leads t o t he t rachea ant eriorly and t he esophagus post eriorly. The
glot t is is t he inlet t o t he t rachea. I t is bordered by t he vocal cords lat erally and
epiglot t is superiorly. The t rachea, w hich corresponds t o t he C6 vert ebra
post eriorly, is a t ubular conduit support ed by semicircular cart ilaginous rings
leading dow n t o t he carina, t he branching point f or t he main-st em bronchi.
What to Do
I n det ermining t he proper placement of an endot racheal t ube, t he lengt h of t he
segment s t hrough w hich t he t ube must pass should be considered. The dist ance
f rom t he t eet h t o t he vocal cords is approximat ely 10 t o 15 cm. O nce t hrough t he
glot t ic apert ure, t he t rachea measures anot her 12 t o 15 cm t o it s end at t he
carina. Because t he t ip of t he endot racheal t ube is opt imally posit ioned at 4 cm
above t he carina, some aut hors advocat e empiric insert ion of t he endot racheal
t ube t o a dept h of 23 cm and 21 cm, measured at t he lips, f or men and w omen,
respect ively. Auscult at ion of t he bilat eral lung f ields and st omach should f ollow
int ubat ion. The presence of sust ained end-t idal CO2 is conf irmat ory of t racheal
placement . A chest radiograph t hen conf irms t he posit ion of t he t ube and t he
need f or subsequent reposit ioning. An improperly posit ioned endot racheal t ube
can be dangerous if lef t unchecked. Esophageal int ubat ion is diagnosed by
nonsust ained or absent end-t idal CO2 and epigast ric gurgling on vent ilat ion.
Vent ilat ion should be t erminat ed immediat ely t o avert insuff lat ion of t he st omach,
w hich can increase gast ric pressures and t he risk f or aspirat ion.
FIG URE 111. 1. Endot racheal t ube in right main bronchus. Port able
examinat ion, ant eropost erior view, show s bilat eral pulmonary edema and a
dense lef t low er lobe (l ong arrow). The endot racheal t ube is in t he bronchus
int ermedius (short arrow), obst ruct ing t he lef t main bronchus and causing lef t
low er lobe at elect asis. (Reused w it h permission f rom Umali CB. Chest
radiographic examinat ion. I n: I rw in RS, Rippe JM, eds. I rw in and Rippe's
I nt ensive Care Medicine. 5t h Ed. Philadelphia: Lippincot t Williams & Wilkins;
2003: 705. )
Even w hen t racheal int ubat ion is achieved, t he posit ion of t he t ube can be
hazardous depending on it s posit ion w it hin t he t rachea. I f t he endot racheal t ube
is t oo cephalad in posit ion t he danger of inadvert ent ext ubat ion and vocal-cord
injury becomes problemat ic. O n t he ot her hand, if t he endot racheal t ube is
advanced t oo deeply, t he risk of main-st em int ubat ion is of concern. Because t he
right main-st em bronchus f orms a more obt use angle w it h t he t rachea t han t he
lef t main-st em bronchus, it is more commonly int ubat ed on deep advancement .
Endobronchial int ubat ion, regardless of side, is diagnosed by unequal or
unilat eral breat h sounds, high peak airw ay pressures, and hypoxemia. I f lef t
unchecked t here is a pot ent ial f or bronchospasms, cont ralat eral at elect asis, and
ipsilat eral t ension pneumot horax.
O nce t he decision is made t o reposit ion t he endot racheal t ube, t he most
accurat e manner of det ermining proper posit ioning is w it h a f iberopt ic
bronchoscope. How ever, t his is of t en impract ical and t ypically t he t ube is
ret ract ed or advanced, w it h t he cuff dow n, by 1-cm increment s, t aking care t o
reassess by auscult at ion over t he bilat eral lung f ields. Alt ernat ively, if t here is a
chest radiograph available, t he degree of ret ract ion or advancement can be
est imat ed on f ilm.
Suggested Readings
Miller RD, ed. Miller's Anest hesia. 6t h Ed. Philadelphia: Churchill Livingst one;
2004: 1617–1618, 1633–1635, 1648–1649.
Morgan E, Mikhail M, Murray M, et al. , eds. Clinical Anest hesiology. 3rd Ed.
New York: McG raw -Hill; 2002: 72–74, 79–80.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 112 - R em em ber that E ven a Fully
Inflated C uff on the E ndotr ac heal Tube is not Adequate P r otec tion Agains t As pir ation
112
Remember that Even a Fully Inflated Cuff on the
Endotracheal Tube is not Adequate Protection
Against Aspiration
Ashita G oel MD
Case
A 45-year-old obese male present ed as an emergency t o t he operat ing room
(O R) f or gunshot w ound t o t he abdomen. The pat ient 's records indicat ed a
hist ory of severe sleep apnea. I n addit ion, t he pat ient had eat en a large meal
prior t o his injury. O n physical exam, t he pat ient had signs of a diff icult airw ay.
Necessary precaut ions w ere t aken t o decrease t he risk of aspirat ion by
employing an aw ake f iberopt ic t echnique f or int ubat ion. During t he case, t he
pat ient 's sat urat ion declined w it h copious part iculat e secret ions rising f rom t he
endot racheal t ube. A diagnosis of int raoperat ive aspirat ion w as made.
Discussion
Aspirat ion can be a f at al complicat ion of int ubat ed pat ient s. Alt hough a cuff ed
endot racheal t ube can decrease t he risk, it does not eliminat e t he risk of
aspirat ion. The exact incidence of clinically relevant aspirat ion is unknow n.
How ever, it is est imat ed t hat up t o 1 of 3, 000 operat ions is complicat ed by
int raoperat ive aspirat ion pneumonit is. Mult iple f act ors account f or t he
inexact ness, including subclinical microaspirat ion, misdiagnosis, and variat ions in
t he pat ient populat ion under st udy. The risk of aspirat ion increases in pat ient s
w it h alt ered airw ay ref lexes (e. g. , drugs, general anest hesia, st roke,
neuromuscular disease, encephalopat hy) and in pat ient s w it h alt ered anat omy
(e. g. , hiat al hernia, pregnancy, obesit y, poor low er esophageal sphinct er t one,
f ull st omach). Addit ional pat ient -relat ed risk f act ors include ext remes of age,
acuit y of surgery, and abdominal surgery.
The modern cuff ed endot racheal t ubes come in t w o major t ypes: high pressure,
low volume and low pressure, high volume. The low -pressure cuff is more
commonly used because it provides a great er cont act area bet w een t he cuff and
t he t racheal mucosa. As a result , t he low -pressure syst em decreases t he risk of
ischemic mucosal damage w hen compared w it h t he high-pressure syst em. The
low -pressure
cuff , how ever, increases t he risk of sore t hroat , aspirat ion, spont aneous
ext ubat ion, and diff icult insert ion. Classically, in t he low -pressure syst em, t he
endot racheal cuff is inf lat ed t o approximat ely 20 t o 25 mm Hg, w hich provides
reasonable prot ect ion f rom aspirat ion and limit s ischemic injury t o t he t rachea.
Tracheal perf usion pressure ranges f rom 25 t o 35 mm Hg.
I t is import ant t o not e if t here is evidence of aspirat ion, it does not necessarily
mean a pat ient w ill develop aspirat ion pneumonit is. The risk of pneumonit is is
based on gast ric volume (>25 mL or >0. 3 mL/ kg) and acidit y of gast ric f luid (pH
<2. 5). How ever, gast ric cont ent s can also damage lung t issue if it cont ains f ood
part iculat es or bact eria. Pat ient s on H2-blockers or prot on pump inhibit ors are
t hought t o be at increased risk f or gast ric colonizat ion by bact eria secondary t o
a more neut ral pH environment .
Signs and Symptoms

Aspirat ion pneumonit is result s in damage t o t he t racheobronchial t ree and t he
lung parenchyma. The init ial react ion is t he result of damage t o t issue f rom
caust ic gast ric secret ions. Secondary damage occurs f rom t he body's
inf lammat ory response, mediat ed by neut rophils and cyt okine release. Aspirat ion
can present as dyspnea, cyanosis, coughing, w heezing, pulmonary edema,
hypot ension, and hypoxia. Pat ient s can rapidly progress t o respirat ory dist ress
and even deat h. I t is est imat ed t hat 10% t o 30% of anest het ic deat hs are a
result of aspirat ion pneumonit is. How ever, t he majorit y of pat ient s w it h aspirat ion
are clinically asympt omat ic. Radiographically, pat ient s may show evidence of
at elect asis, alveolar edema, and int erst it ial edema.
I f aspirat ion is suspect ed, t he pat ient 's upper airw ay should be suct ioned. I n
int ubat ed pat ient s, t he low er airw ay should be suct ioned t hrough t he
endot racheal t ube. I nt ubat ion should be considered f or nonint ubat ed pat ient s
t hat have aspirat ed t o decrease t he risk of f urt her aspirat ion and t o assist w it h
vent ilat ory support . G enerally, prophylact ic ant ibiot ics are not recommended
because prophylact ic t reat ment may result in select ion of more resist ant
organisms. How ever, empiric t reat ment w it h ant ibiot ics is recommended f or
pat ient s w it h know n aspirat ion and gast ric dysmot ilit y or gast ric colonizat ion
(e. g. , small bow el obst ruct ion, gast ric paresis, prot on pump inhibit ors). Also,
empirical t herapy is recommended f or pat ient s w hose aspirat ion pneumonit is
sympt oms f ail t o resolve af t er 48 hours. Cort icost eroids have been used in t he
past in t he management of aspirat ion pneumonit is.
How ever, t here is limit ed dat a on t he role of t hese agent s as t herapy f or

aspirat ion pneumonit is.
Suggested Readings
Blunt MC, Young PJ, Pat il A, et al. G el lubricat ion of t he t racheal t ube cuff
reduces pulmonary aspirat ion. Anest hesiology 2001; 95: 377–381.
Marik P. Aspirat ion pneumonit is and aspirat ion pneumonia. N Engl J Med
2001; 344: 665–671.
Miller RD, ed. Miller's Anest hesia. 6t h Ed. Philadelphia: Elsevier/ Churchill
Livingst one; 2005: 1634–1636.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 113 - D o not O ver inflate the
E ndotr ac heal C uff
113
Do not Overinflate the Endotracheal Cuff
G regory Dalencourt MD
A large percent age of pat ient s w ho are cared f or in t he int ensive care unit (I CU)
are int ubat ed w it h an endot racheal t ube. Regardless of t he reason f or requiring
int ubat ion, key f act ors must be underst ood w hen caring f or such pat ient s. The
endot racheal t ube (ETT) is t ypically placed orally using a laryngoscope and is
advanced just past t he vocal cords unt il t he cuff disappears (t o t he view of t he
operat or). The cuff is t hen inf lat ed. The dist ension of t he pilot balloon, next t o
t he inject ion port , ref lect s inf lat ion of t he ETT cuff . The appropriat e posit ion of
t he ETT cuff is dist al t o t he cords w it h t he t ip 4 cm above t he carina.
Conf irmat ion of ETT placement in t he t rachea is best accomplished w it h t he use
of end-t idal carbon dioxide conf irmat ion. Auscult at ion of bilat eral breat h sounds
may conf irm placement in t he t rachea and is usef ul in evaluat ing w het her t he ETT
is t oo deep, result ing in endobronchial int ubat ion. This is clinically indicat ed by
loss of breat h sounds in one of t he lung f ields, t ypically t he lef t because of t he
st raight pat h t o t he right main-st em bronchus compared w it h t he sharp angle t o
t he lef t . The ETT is t hen secured w it h t ape or ot her ETT-securing devices.
Watch Out For

The purpose of t he ETT cuff is t o f acilit at e posit ive-pressure vent ilat ion and
prevent aspirat ion of secret ions by f orming a seal w it hin t he t rachea. The current
pract ice is use of a low -pressure, high-volume cuff . The risk of high-pressure
cuff s is t hat t he pressure t ransmit t ed t o t he underlying t racheal mucosa can
cause ischemia w hen t he cuff pressure exceeds t he capillary art eriolar pressure
(approximat ely 32 mm Hg). The low -pressure cuff s are pref erred t o limit t his risk
of ischemia and t he recommended maximum pressure of t he cuff is 25 mm HG .
Low -pressure, high-volume cuff s inf lat e more symmet rically t han high-pressure
cuff s, w hich may also decrease t he risk of t racheal dilat at ion, and t hey can
inf lat e t o t he shape of t he pat ient 's t rachea, opt imizing t he seal.
The f act t hat cuff s are “low pressure” does not mean t hat one cannot overinf lat e
t hem and increase t he pressure such t hat t here is risk of ischemia. Any t ype of
cuff can become a high-pressure cuff if enough
air is inject ed or if inadvert ent overinf lat ion occurs, as w it h t he diff usion of
nit rous oxide int o t he cuff over t ime. I n addit ion, t he pressure est imat ed by
palpat ing t he pilot balloon f requent ly underest imat es t he act ual cuff pressure, so
eit her cuff pressure should be measured and document ed on a regular basis or
t he cuff should be inf lat ed only t o t he point of eliminat ing t he leak during
posit ive-pressure vent ilat ion.
Commonly, a cuff leak is not ed w hen air leaks around t he cuff during posit ive-
pressure vent ilat ion. When t his occurs, t he f irst maneuver is t o add addit ional air
int o t he cuff . This procedure not only increases t he risk of cuff overinf lat ion and
mucosal ischemia but can also lead t o unident if ied herniat ion of t he cuff above
t he cords. When t here is a new leak around t he cuff , t he posit ion of t he ETT
should be det ermined. O n average, insert ion of an ETT t o 20 t o 22 cm beyond
t he lips in an average adult places t he ETT t ip in t he midt rachea. How ever, in
pat ient s w ho have had prolonged int ubat ion, t he ETT may sof t en and bulge int o
t he post erior pharynx, and t hus t he dept h marker may become less reliable.
When invest igat ing and correct ing a cuff leak, observe t he f ollow ing st eps:
Review t he most recent chest x-ray.

O bt ain a repeat x-ray if t ime permit s.
Suct ion t he post erior pharynx.
Def lat e t he balloon t o conf irm how much air has been inf lat ed.
Reinf lat e t he pilot balloon w it h t he minimum volume of air required t o prevent
a leak (t ypically, 3 t o 6 mL is suff icient w it h a maximum of 10 mL of air
volume required t o obt ain a seal).
I f >10 mL is aspirat ed, one must be suspicious t hat t he balloon has been
hyperinf lat ed and is eit her bridging t he cords or migrat ing supraglot t ic
(above t he cords) and put t ing t he pat ient at risk f or inadvert ent ext ubat ion.
Consider direct visualizat ion w it h laryngoscopy or a f iberopt ic scope and
advancement or reposit ioning of t he ETT.
Suggested Readings
Braz JR, Navarro LH, Takat a I H, et al. Endot racheal t ube cuff pressure: need
f or precise measurement . Sao Paulo Med J 1999; 117: 243–247.
Schmidt G A, Hall JB, Wood LDH. Principles of Crit ical Care. New York:
McG raw -Hill, Healt h Prof essions Division; 1992: 594–595.
St oelt ing RK, Miller RD. Basics of Anest hesia. 2nd ed. New York: Churchill
Livingst one; 1989: 163–166.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 114 - C hec k for a C uff Leak in
P atients w ho Might have Tr ac heal E D E MA B efor e E xtubation
114
Check for a Cuff Leak in Patients who Might
have Tracheal EDEMA Before Extubation
E. David Bravos MD
Crit ically ill pat ient s of t en require int ubat ion and mechanical vent ilat ion. Along
w it h t he morbidit ies of int ubat ion, problems f ollow ing ext ubat ion can occur.
Airw ay obst ruct ion f rom laryngeal or t racheal edema is a signif icant risk and can
result in respirat ory dist ress t hat may require emergent reint ubat ion. Risk
f act ors f or airw ay edema include t raumat ic or diff icult int ubat ion, durat ion of
int ubat ion, high balloon cuff pressure or large endot racheal t ube size,
inhalat ional injuries including burns or caust ic subst ances, and recent self -
ext ubat ion.
What to Do
The gold st andard f or det ect ion of airw ay edema is direct visualizat ion. How ever,
t he presence of an endot racheal t ube obscures t he view, making t his somet imes
diff icult . I ndirect met hods of det ermining airw ay edema such as t he cuff -leak t est
are commonly used. This t est measures w het her air is able t o pass around t he
out side of t he endot racheal t ube t hrough t he pat ient 's airw ay w hen t he cuff of
t he t ube is def lat ed and t he pat ient exhales. This can be assessed qualit at ively
and quant it at ively. Q ualit at ively, cooperat ive pat ient s should breat he w hen t he
cuff is def lat ed w it h assessment f or t he presence of st ridor. This is more
commonly done in pediat ric pat ient s w it h croup. More recent ly, quant it at ive
measurement s are done by comparing t he ret urned t idal volume of pat ient s on
volume-cont rol vent ilat ion w hen t he cuff is inf lat ed as w ell as def lat ed. The
diff erence in t he t w o values is t he cuff leak. I n pract ice, several t idal volumes
are recorded (usually f ive t o six) w it h t he cuff def lat ed and t he values averaged.
Addit ionally, t he size of t he cuff leak has been hypot hesized t o be inversely
proport ional t o t he amount of edema and pot ent ial obst ruct ion.
Sensit ivit y and specif icit y of t he cuff -leak t est have been variable in report ed
st udies but have been as high as 100% and 99%, respect ively. Addit ionally,
posit ive and negat ive predict ive values have been report ed t o be as high as 69%
and 98%, respect ively. Variabilit y in report ed values may ref lect variable pat ient
populat ions st udied. Theref ore, w hen deciding t o ext ubat e a pat ient , one should
use t he cuff -leak t est as a guide f or pat ient s w ho may be at risk f or
post ext ubat ion airw ay obst ruct ion rat her t han an absolut e crit eria f or ext ubat ion.
Addit ionally, it may also suggest w hich pat ient s may need closer monit oring af t er
ext ubat ion and might benef it f rom t reat ment f or st ridor w it h bronchodilat ors,
helium-oxygen mixt ures, or cort icost eroids.
Suggested Readings
De Backer D. The cuff -leak t est : w hat are w e measuring? Crit Care
2005; 9: 31–33.
Jaber S, Chanques G , Mat ecki S, et al. Post -ext ubat ion st ridor in int ensive
care unit pat ient s. Risk f act ors: evaluat ion and import ance of t he cuff -leak
t est . I nt ens Care Med 2003; 29: 69–74.
Miller RL, Cole RP. Associat ion bet w een reduced cuff leak volume and
post ext ubat ion st ridor. Chest 1996; 110: 1035–1040.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 115 - K now how to Meas ur e P lateau
P r es s ur e w hen Us ing P r es s ur e- R egulated Volum e C ontr ol Ventilation Mode and K now w hat to do w ith
the Value O nc e it is O btained
115
Know how to Measure Plateau Pressure when
Using Pressure-Regulated Volume Control
Ventilation Mode and Know what to do with the
Value Once it is Obtained
David N. Hager MD
Mechanical vent ilat ion has been show n t o cause vent ilat ion-induced lung injury by
at least t w o mechanisms. First , using a low lung volume allow s t he repeat ed
opening and closing of small airw ays and alveoli. This is associat ed w it h high
shear f orces on t he lung int erst it ium and alveolar surf act ant deplet ion, bot h of
w hich result in vent ilat ion-induced lung injury. To reduce t his low -volume injury,
t he use of posit ive end-expirat ory pressure (PEEP) is advocat ed, t hough t he
ideal amount of PEEP cont inues t o be t he t opic of int ense st udy. Second, t he use
of t oo much PEEP and/ or t he use of supraphysiologic t idal volumes have been
show n t o cause vent ilat ion-induced lung injury by overdist ent ion of alveoli.
Hist orically, alveolar overdist ent ion has been a concern w hen plat eau pressure
(Pplat ) is >30 cm H2 O . Alt hough much at t ent ion is of t en given t o high peak
airw ay pressure, it is Pplat t hat ref lect s alveolar pressure, value most
associat ed w it h vent ilat ion-induced lung injury. Tw o approaches t o limit ing
alveolar overdist ent ion are (a) t o reduce t idal volume unt il plat eau pressure ≤30
cm H2 O during assist volume cont rol (AC) vent ilat ion or (b) t o limit inspirat ory
airw ay pressure t o ≤30 cm H2 O during pressure-regulat ed volume cont rol
(PRVC) vent ilat ion. The ast ut e clinician w ill be able t o underst and (a) t he
diff erences bet w een AC and PRVC vent ilat ion modes, (b) t hat alveolar pressures
≤30cm H2 O might not be unif ormly saf e, and (c) t he appropriat e st rat egy t hat
limit s bot h low volume and overdist ent ion lung injury.
Controlled M odes of Ventilation

AC and PRVC are bot h “cont rolled” modes of mechanical vent ilat ion. For each,
t he clinician set s a minimum respirat ory rat e, w hich ensures a set number of
breat hs even if t he pat ient is apneic or paralyzed. A nonparalyzed/ nonapneic
pat ient w ho t riggers t he vent ilat or in excess of t his set rat e w ill receive f ull
vent ilat or support f or each addit ional breat h. How ever, t he modes diff er in how
t he magnit ude of t idal volumes are det ermined. During assist cont rol vent ilat ion,
t he
clinician set s a t idal volume, inspirat ory f low rat e, inspirat ory t ime, and
inspirat ory f low w avef orm. I rrespect ive of lung compliance, airw ay resist ance, or
simult aneous at t empt s at act ive exhalat ion by t he pat ient , t he set t idal volume
w ill be delivered. For t his reason, even in t he apneic pat ient , a minimum minut e
vent ilat ion (respirat ory rat e t imes t idal volume) is guarant eed, t hough t here is
t he pot ent ial f or high peak airw ay pressures. Plat eau pressure is measured at
end inspirat ion by occluding t he expirat ory valve of t he vent ilat or circuit . This
allow s airw ay pressure t o equilibrat e t hroughout t he lungs.
By cont rast , t he magnit udes of t idal volume during PRVC vent ilat ion are
dependent on t he level of t he inspirat ory airw ay pressure, lung compliance,
airw ay resist ance, and even pat ient eff ort . The vent ilat or aut omat ically adjust s
inspirat ory f low and f low w avef orm as it at t empt s t o maint ain a const ant
inspirat ory pressure. Though t he respirat ory rat e and inspirat ory pressures are
chosen t o achieve adequat e minut e vent ilat ion, t idal-volume size may vary f rom
breat h t o breat h. Despit e t his variabilit y, some argue t hat PRVC is at t ract ive
because airw ay pressures w ill never exceed t he prescribed level. I f , f or
example, t he inspirat ory pressure is set at 29 cm H2 O , t he “dangerous” t hreshold
of 30 cm H2 O is t heoret ically never reached. Plat eau pressures w ill alw ays be
l ess than or equal to t he prescribed inspirat ory pressure. O n some vent ilat ors,
obt aining t he plat eau pressure value cannot be done in t he PRVC mode and a
brief t ransit ion t o AC is required t o measure plat eau pressure.
Airw ay Pressures
As ment ioned earlier, clinicians have hist orically t ried t o avoid plat eau pressures
>30 cm H2 O . The f ocus on t his t hreshold pressure is likely based on t hree
observat ions. How ever, w hen each observat ion is review ed caref ully, t he concept
of a “saf e” plat eau pressure t hreshold appears t o be misguided. First , t hough
early animal st udies did show vent ilat ion-induced lung injury t o occur
predominant ly in animals vent ilat ed at airw ay pressures >30 cm H2 O , t hese
st udies w ere only a f ew hours long at most . When similar st udies w ere
conduct ed over a period of days, signif icant lung injury w as observed even in
animals vent ilat ed at t he report edly saf e inspirat ory pressure of 30 cm H2 O .
Second, t he normal t ranspulmonary pressure in humans at t ot al lung capacit y is
approximat ely 30 t o 35 cm H2 O . I t has been reasoned t hat if a person can
generat e t hese pressures spont aneously, such pressures are unlikely t o cause
lung injury. How ever, exposure t o pressures of t his magnit ude during a solit ary
maneuver t o achieve t ot al lung capacit y is
very diff erent t han being exposed t o such pressures several t housand t imes per
day, as may occur during mechanical vent ilat ion. Third, in t he cont ext of
randomized cont rolled t rials of low (6 mL/ kg) versus t radit ional (12 mL/ kg) t idal
volumes in pat ient s w it h acut e lung injury and t he acut e respirat ory dist ress
syndrome (ARDS), only t hose st udies in w hich mean plat eau pressures w ere ≥32
cm H2 O among t he t radit ional t idal-volume group w as a mort alit y benef it
associat ed w it h low t idal volumes. Some have t heref ore reasoned t hat if plat eau
pressure is <32 cm H2 O , t here is no benef it in t idal-volume reduct ion. How ever,
close inspect ion of t he dat a f rom t he largest and most rigorously conduct ed of
t hese st udies show s a st eady decline in mort alit y as plat eau pressure decreases
w ell below 30 cm H2 O , irrespect ive of randomizat ion group. I n ot her w ords,
pat ient s randomized t o t he low -t idal-volume group did bet t er even w hen
compared w it h pat ient s randomized t o t he t radit ional-t idal-volume group in w hom
plat eau pressure w as <30 cm H2 O . For t hese reasons, it seems inappropriat e t o
assume t hat an airw ay pressure of 30 cm H2 O is saf e, be it plat eau pressure
generat ed during AC vent ilat ion, or t he set inspirat ory pressure during PRVC.
Strategy
Because vent ilat ion at low lung volumes and overdist ent ion of alveoli are t he t w o
leading causes of vent ilat ion-induced lung injury, mechanical vent ilat ion
st rat egies should be t ailored t o limit t hese t w o t ypes of injury. I t is know n t hat
PEEP decreases low -volume injury, alt hough how one det ermines t he ideal PEEP
is a cont inued t opic of st udy. A good general guideline is t o st art all pat ient s w it h
PEEP = 5 cm H2 O and t it rat e upw ard according t o oxygenat ion requirement as
per t he Acut e Respirat ory Dist ress Syndrome (ARDS) Net w ork low -t idal-volume
st rat egy. Like PEEP, recommendat ions f or det ermining t he appropriat e t idal
volume are in evolut ion. I t is generally agreed t hat t idal volumes >10 mL/ kg are
excessive even in uninjured lungs. How ever, t he best evidence t o dat e show s
t hat among pat ient s w it h acut e lung injury or ARDS, t idal volume should be set t o
6 mL/ kg predict ed body w eight , even w hen plat eau pressures are not >30 cm
H2 O . This is easily done in t he AC mode by dialing in t he desired t idal volume.
How ever, during PRVC vent ilat ion, t idal volume is set by adjust ing t he inspirat ory
pressure. Though t his t oo is easily accomplished, it requires close f ollow -up and
f requent adjust ment s because a given inspirat ory pressure may ult imat ely deliver
t idal volumes t hat are t oo small in a pat ient w hose lung compliance is w orsening
or t oo big
in a. I t should not be assumed a pri ori t hat inspirat ory airw ay pressures ≤30 cm
H2 O are saf e.
Suggested Readings
Dreyf uss D, Saumon G . Vent ilat or-induced lung injury: lessons f rom
experiment al st udies. Am J Respir Crit Care Med 1998; 157: 294–323.
Hager DN, Krishnan JA, Hayden DL, et al. Tidal volume reduct ion in pat ient s
w it h acut e lung injury w hen plat eau pressures are not high. Am J Resp Crit
2005; 172: 1241–1245.
Vent ilat ion w it h low er t idal volumes as compared w it h t radit ional t idal volumes
f or acut e lung injury and t he acut e respirat ory dist ress syndrome. The Acut e
Respirat ory Dist ress Syndrome Net w ork. N Engl J Med 2000; 342: 1301–1308.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 116 - Us e P lateau or Mean P r es s ur e
as a m or e Ac c ur ate As s es s m ent of B ar otr aum a than P eak P r es s ur e
116
Use Plateau or Mean Pressure as a more
Accurate Assessment of Barotrauma than Peak
Pressure
Benjamin Kratzert MD
Monit oring lung mechanics during mechanical vent ilat ion has become an
import ant paramet er f or managing pulmonary disease and acut e changes in
respirat ory st at us in pat ient s. Addit ionally, t his monit oring is used t o adjust
vent ilat or set t ings f or prevent ion of alveolar overdist ent ion (or barot rauma) and
t o at t empt t o improve clinical out comes. The t w o most common airw ay-pressure
paramet ers used are peak inspirat ory pressure (PI P) and plat eau pressure
(Pplat ).
PI P is measured at t he end of inspirat ory inf lat ion (see Fi g. 116. 1) and is a
f unct ion of t he inf lat ion volume, t he f low resist ance of t he airw ays, and t he
compliance of t he lungs and chest w all. Since PI P is great ly inf luenced by
resist ance in t he upper airw ays and vent ilat or equipment , it represent s a poor
marker of alveolar pressures. The most accurat e measurement of alveolar
pressure is plat eau pressure. This value ref lect s airw ay pressure during a 1- t o
2-second inspirat ory
pause and t heref ore is less inf luenced by vent ilat or equipment . The Pplat
approximat es small airw ay and alveolar pressure more closely t han PI P. I t is
believed t hat cont rol of t he Pplat is import ant as excessive st ret ch of alveoli has
been implicat ed as t he cause of vent ilat or-induced lung injury. The recent Acut e
Respirat ory Dist ress Syndrome (ARDS) Net w ork mult icent er prospect ive
randomized t rial show ed t hat maint aining plat eau pressures at ≤30 cm H2 O , w it h
result ant permissive hypercapnia, result ed in a st at ist ically signif icant decrease
in t he number of vent ilat or days and improved mort alit y.
FIG URE 116. 1. Pressure, f low, and volume curves during measurement of
peak and plat eau pressures.
Diff erent vent ilat or st rat egies are usef ul in modif ying Pplat . These st rat egies
include low ering t idal volume, short ening inspirat ory t ime, decreasing inspirat ory
f low rat e, avoiding aut o posit ive endexpirat ory pressure (PEEP), permissive
hypercapnia, pressure cont rol vent ilat ion, pressure-regulat ed volume cont rol
vent ilat ion, and t reat ing pulmonary edema.
Suggested Readings
ARDS Net w ork. Vent ilat ion w it h low er t idal volumes as compared w it h
t radit ional t idal volumes f or acut e lung injury and t he acut e respirat ory
dist ress syndrome. N Engl J Med 2000; 342: 1301–1308.
Dreyf uss D, Saumon G . Vent ilat or-induced lung injury: lessons f rom
423.
Tobin MJ. Principles and Pract ice of Mechanical Vent ilat ion. New York:
McG raw -Hill; 1994.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 117 - C ons ider Air w ay P r es s ur e
R eleas e Ventilation for D eliver ing an O pen- Lung S tr ategy
117
Consider Airway Pressure Release Ventilation
for Delivering an Open-Lung Strategy
Airw ay pressure release vent ilat ion (APRV) is a new er mode of mechanical
vent ilat ion t hat alt ernat es bet w een t w o levels of cont inuous posit ive airw ay
pressure (CPAP). The set rat e f or t his mode of vent ilat ion is normally 20 or
f ew er breat hs per minut e t o prevent inverse rat io vent ilat ion. The levels of CPAP
are ref erred t o as t he high-pressure and low -pressure limit s and also need t o be
set t o allow adequat e oxygenat ion. I n addit ion, t he provider needs t o set t w o
t imes, w hich allow t he vent ilat or t o cycle bet w een t he high- and low -pressure
levels f or 0. 5 t o 1. 5 seconds. When t he machine cycles t o t he low er CPAP level,
exhalat ion occurs.
Since APRV is a new er mode of vent ilat ion, it has not been f ully evaluat ed in
clinical t rials, but early dat a are encouraging regarding it s eff ect iveness. The
value of APRV is derived f rom it s abilit y t o provide adequat e oxygenat ion at
low er pressure levels. I nt rapulmonary shunt w it h spont aneous breat hing appears
t o be reduced w it h APRV. The high-pressure limit provides f or an open-lung
st rat egy, and t he low -pressure limit prevent s shear st ress and cyclic alveolar
collapse w it h at elect asis, bot h of w hich cont ribut e t o vent ilat or-induced lung
injury. As t he pat ient 's condit ion improves, t he lung can be “def lat ed” by reducing
t he pressure limit s and t imes t o allow f or normal pressure-limit ed mechanical
vent ilat ion. When t he pat ient is able, t he vent ilat or rat e can be reduced t o allow
t he pat ient t o t ake on more of t he w ork of breat hing in preparat ion f or
ext ubat ion.
The major challenges occurring w it h APRV include t he need t o monit or t he level
of hypovent ilat ion, hypercarbia, and respirat ory acidosis. While permissive
hypercapnia is a valuable t echnique f or lung prot ect ion, pat ient s on APRV may
develop considerable hypercarbia w it h a result ing respirat ory acidosis. I n
addit ion, t he need t o maint ain t he pressure levels t o support an open-lung
st rat egy may lead t o hypot ension if t he pat ient is hemodynamically unst able or
has a reduced preload. Volume loading can assist w it h t his problem. How ever, on
some occasions, hemodynamic compromise may not allow t his mode of
vent ilat ion t o be used eff ect ively.
Suggested Readings
G oldman L, Ausiello D, eds. Cecil Text book of I nt ernal Medicine. 22nd Ed.
Philadelphia: Saunders; 2004: 410–424.
Wrigge H, Zinserling J, Neumann P, et al. Spont aneous breat hing w it h airw ay

pressure release vent ilat ion f avors spont aneous vent ilat ion in dependent lung
regions and count ers cyclic alveolar collapse in oleic-acid-induced lung injury:
a randomized cont rolled comput ed t omography t rial. Crit Care Med
2005; 9: 780–789.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 118 - R em em ber that S tatic
C om plianc e of the R es pir ator y S ys tem is not the s am e Thing as D ynam ic C om plianc e
118
Remember that Static Compliance of the
Respiratory System is not the same Thing as
Dynamic Compliance
Nirav G . Shah MD
Compliance is a measure of lung dist ensibilit y and is measured as a change in
lung volume per unit change in pressure. There are t w o diff erent t ypes of lung
compliance—st at ic and dynamic. St at ic lung compliance is represent ed by t he
slope of t he pressure-volume curve during def lat ion f rom t ot al lung capacit y. I n a
pat ient w ho is being mechanically vent ilat ed, t his can be measured by dividing
t he exhaled t idal volume by t he diff erence bet w een t he plat eau pressure (Pplat )
and t he posit ive end-expirat ory pressure (PEEP) af t er t emporary occlusion of
t he expirat ory limb of t he vent ilat or circuit . The peak pressure is t he most
accurat e ref lect ion of st at ic compliance.
Dynamic compliance is t he rat io of t he change in volume t o t he change in
pressure over a t idal breat h. I nsof ar as t here is no abnormalit y in t he exhaled
t idal volume, a decrease in dynamic lung compliance w ould suggest disease or
narrow ing of t he smaller airw ays. The plat eau pressure most accurat ely ref lect s
t his measurement .
When opt imizing t he set t ings f or mechanical vent ilat ion, it is import ant t o be able
t o cont rol t he t idal volumes as w ell as t he airw ay pressures. There have been
several large t rials suggest ing t hat a low -t idal-volume st rat egy improves
mort alit y. This is t hought t o result f rom t he pressure-volume curves in acut e
respirat ory dist ress syndrome (ARDS) pat ient s w ho show a f lat t ened curve
init ially at low lung volumes, indicat ing t hat a small increase in pressure result s
in large changes in volume. O ne goal is t o keep t he PEEP above t he point at
w hich compliance signif icant ly decreases. This is know n as t he low er inf lect ion
point .
There is great debat e about t he most opt imal mode of vent ilat ion chosen f or
pat ient s w it h poor lung compliance. Pressure-cont rol vent ilat ion cont rols airw ay
pressures but can result in large variat ions in t idal volume t o maint ain t he set
pressure. Volume-cont rol vent ilat ion cont rols t he t idal volume but can result in
increased airw ay pressures if f aced w it h air t rapping, decreased lung
compliance, or w orsening respirat ory mechanics. As a pat ient 's compliance
w orsens, it may be necessary t o alt ernat e modes of vent ilat ion t o achieve t he
needs of t he pat ient w hile minimizing t he eff ect s of acut e lung injury. Conversely,
as compliance improves, t he plat eau pressures and peak pressures may

decrease, indicat ing an improved readiness f or w eaning f rom mechanical
vent ilat ion, depending on t he vent ilat ion mode.
Suggested Readings
Byrne K, Cooper KR, Carey PD, et al. Pulmonary compliance: early
assessment of evolving lung injury af t er onset of sepsis. J Appl Physiol
1990; 69: 2290–2295.
Ware LB, Mat t hay MA. The acut e respirat ory dist ress syndrome. N Engl J
Med 2000; 342: 1334–1349.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 119 - R em em ber that R ec r uitm ent of
Alveoli Us ing an Inc r eas ed Level of P os itive E nd- E xpir ator y P r es s ur e c an Take 6 to 12 Hour s
119
Remember that Recruitment of Alveoli Using an
Increased Level of Positive End-Expiratory
Pressure can Take 6 to 12 Hours
David N. Hager MD
There are t w o f orms of posit ive end-expirat ory pressure (PEEP). These are
ext rinsic PEEP and int rinsic-PEEP (aka, aut o-PEEP). Ext rinsic PEEP is t he
pressure applied t o t he airw ays by t he vent ilat or at end expirat ion. I nt rinsic
PEEP is t he pressure remaining in t he airw ays in excess of ext rinsic PEEP upon
occluding t he expirat ory valve at end expirat ion. Ext rinsic PEEP is t ypically used
f or t hree reasons: (a) t o decrease w ork of breat hing in pat ient s w it h int rinsic
PEEP, (b) t o reduce or avoid at elect asis, and (c) t o improve art erial oxygenat ion
and avoid or limit low -volume lung injury. The purpose of t his chapt er is t o review
t he use of PEEP in each of t hese sit uat ions.
Work of Breathing and PEEP

The use of ext rinsic PEEP in t he cont ext of obst ruct ive lung disease may
decrease t he w ork of breat hing. At end expirat ion, many of t hese pat ient s w ill
have measurable int rinsic PEEP. They must t heref ore drop t heir pleural pressure
f ar enough t o (a) overcome t heir int rinsic PEEP and (b) t rigger t he vent ilat or. For
example, if a given pat ient has an int rinsic PEEP of 5 cm H2 O and t he vent ilat or
t rigger t hreshold is set at negat ive 2 cm H2 O , t he pat ient has t o decrease his or
her pleural pressure by 7 cm H2 O bef ore t he vent ilat or w ill deliver t he next
assist ed breat h. I n t his sit uat ion, t hat applicat ion of 5 cm H2 O of ext rinsic PEEP
w ill decrease t he pat ient 's w ork of breat hing. The pat ient w ill t hen need t o
decrease his or her pleural pressure by only 2 cm H2 O . The of t en cit ed concern
t hat t he use of l ow levels of ext rinsic PEEP (5 cm H2 O ) w ill subst ant ially
increase t he hazards of hyperinf lat ion have been overemphasized.
Atelectasis and PEEP

Vent ilat ion at low lung volumes w it h deep sedat ion predisposes pat ient s t o t he
development of at elect asis. Though t here are no convincing dat a t o suggest t hat
import ant out comes such as durat ion of mechanical vent ilat ion, morbidit y, or
mort alit y are reduced by t he
empiric use of low levels of PEEP (5 cm H2 O ) in healt hy pat ient s, many clinicians
have adopt ed t his pract ice based on physiologic measurement s t hat show t hat
such maneuvers reduce shunt .
Low -Volume Lung Injury, Oxygenation, and PEEP

I n pat ient s w it h acut e lung injury (ALI ) and t he acut e respirat ory dist ress
syndrome (ARDS), higher levels of PEEP w ill improve oxygenat ion by opening
(recruit ing) at elect at ic lung. I t has also been suggest ed t hat PEEP w ill limit low -
volume lung injury, w hich result s f rom t he repeat ed opening and closing of small
airw ays and alveoli during t idal vent ilat ion. This act ion is associat ed w it h high
shear f orces on t he lung int erst it ium and surf act ant deplet ion, bot h of w hich
exacerbat e vent ilat or-associat ed lung injury. Though t he use of PEEP is st rongly
advocat ed in t hese pat ient s based on observat ions in animal models, t he ext ent
t o w hich PEEP should be raised in adult s w it h ALI or ARDS is not clear despit e a
large body of lit erat ure t hat f ocuses on diff erent approaches t o t he ident if icat ion
of “opt imal” PEEP. Some invest igat ors have suggest ed adjust ing PEEP t o a
pressure slight ly above t he low er inf lect ion point on a pressure-volume curve
based on t he belief t hat t idal derecruit ment w ill not occur at t his end-expirat ory
pressure. How ever, it has been clearly show n t hat subst ant ial lung recruit ment
and derecruit ment cont inue t o occur even w hen PEEP exceeds t he low er
inf lect ion point of t he pressure-volume curve. Furt her, a quasist at ic pressure-
volume curve does not charact erize t he mechanical propert ies of t he lung alone.
Rat her, it charact erizes t he combined mechanical propert ies of t he lung and t he
chest w all. This is import ant t o underst and, as t he ext ent t o w hich t he chest w all
aff ect s t he pressure-volume curve is unpredict able.
The largest st udy t o dat e in ALI / ARDS pat ient s f ound no mort alit y benef it w it h
t he use of higher (average = ~15 cm H2 O ) versus low er (average = ~9 cm H2 O )
levels of PEEP. I n addit ion, t here is recent dat a t o suggest t hat some pat ient s
are PEEP responders and ot hers are PEEP nonresponders. Furt her invest igat ion
is clearly needed t o bet t er clarif y an eff ect ive w ay of est ablishing t he opt imal
level of PEEP f or individual pat ient s. For t he t ime being, a reasonable approach
t hat at t empt s t o balance t he benef it s and limit at ions of PEEP and f ract ion of
inspired oxygen (FI O2 ) is t he prot ocol used by t he ARDS Net w ork invest igat ors.
I n t his prot ocol, increases in FI O2 are alt ernat ed w it h increases in PEEP unt il t he
goal PaO2 of 55 t o 80 mm Hg is achieved.
Last ly, it should be not ed t hat t he st at ement in t his chapt er's t it le can be helpf ul
w hen t hinking about a pat ient 's response t o t he level of PEEP administ ered. I n
t he cont ext of t he previous discussion, it seems prudent t o apply PEEP
according t o t he ARDS Net w ork prot ocol t o achieve oxygenat ion goals. O ver
t ime, as PEEP is increased, one may observe t hat PaO2 dramat ically improves.
This eff ect may be t he result of lung unit s opening gradually over t he course of
several hours or resolving inf lammat ion.
Suggested Readings
Acut e Respirat ory Dist ress Syndrome Net w ork Vent ilat ion w it h low er t idal
volumes as compared w it h t radit ional t idal volumes f or acut e lung injury and
t he acut e respirat ory dist ress syndrome. N Engl J Med. 2000; 342(18): 1301–
1308.
Bindslev L, Hedenst ierna G , Sant esson J, et al. Vent ilat ion-perf usion
dist ribut ion during inhalat ional anaest hesia; eff ect s of spont aneous breat hing,
mechanical vent ilat ion, and posit ive end-expirat ory pressure. Act a
Anaest hesiol Scand 1981; 25: 360– 371.
Brow er RG , Lanken PN, MacI nt yre N, et al. Higher versus low er posit ive end-
expirat ory pressures in pat ient s w it h t he acut e respirat ory dist ress syndrome.
N Engl J Med 2004; 351: 327–336.
Dreyf uss D, Saumon G . Vent ilat or-induced lung injury; lesions f rom
G rasso S, Fanelli V, Caf arelli A, et al. Eff ect s of high versus low posit ive end-
expirat ory pressures in acut e respirat ory dist ress syndrome. Am J Respir Crit
Care Med 2005; 171: 1002–1008.
Levy MM. O pt imal PEEP in ARDS; changing concept s and current

cont roversies. Crit Care Clin 2002; 18: 15.
Smit h TC, Marini JJ. I mpact of PEEP on lung mechanics and w ork of
breat hing in severe airf low obst ruct ion. J Appl Physiol 1988; 65: 1488–1499.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 120 - Have a High S us pic ion of Auto
P os itive E nd- E xpir ator y P r es s ur e w hen Attem pting to W ean P atients w ith C hr onic O bs tr uc tive
P ulm onar y D is eas e
120
Have a High Suspicion of Auto Positive End-
Expiratory Pressure when Attempting to Wean
Patients with Chronic Obstructive Pulmonary
Disease
Jose Rodriguez-paz MD
Case
You are in t he int ensive care unit (I CU) and you are called t o see Mrs. Smit h, a
w oman w it h chronic obst ruct ive pulmonary disease (CO PD) w ho had been
admit t ed f or respirat ory f ailure secondary t o pneumonia. You quickly go t o her
bed t o evaluat e. You are t old t hat she w as doing w ell w it h t he planned w eaning
f rom mechanical vent ilat ion, but she became acut ely unst able. Upon your arrival
she is f ound t o be t achypneic, diaphoret ic, and hypot ensive. O f course, t oday is
your f irst day in t he I CU and you have not received sign-out yet . All of a sudden
t he I CU f ellow arrives and quickly disconnect s Mrs. Smit h f rom t he vent ilat or and
her blood pressure improves and she st art s looking bet t er. What happened?
Discussion
Pat ient s w it h CO PD have regions of t he lung w it h f low limit at ion, needing longer
t ime const ant s f or t heir alveoli t o def lat e (i. e. , t hey t ake longer t o complet e
expirat ion). I n t hese pat ient s, especially during mechanical vent ilat ion, t hese
areas t hat t ake longer t o def lat e may st ill be in expirat ion once t he lung is
making t he next inspirat ory eff ort , t heref ore t rapping air in t he alveoli. This
phenomenon is called aut o posit ive end-expirat ory pressure (PEEP) or int rinsic
PEEP (iPEEP). This basically means t hat t he pat ient 's expirat ory f low get s
int errupt ed bef ore t he alveoli are complet ely empt y, t hus creat ing a pressure
diff erence bet w een t he alveoli and t he proximal airw ay.
How do you measure if a pat ient has aut oPEEP? The easiest met hod in a
vent ilat ed pat ient is t o review t he expirat ory f low in t he vent ilat or graphics. I f t he
expirat ory f low does not reach zero and cont inues unt il t he onset of t he f ollow ing
inspirat ory cycle, your pat ient has aut oPEEP (Fi g. 120. 1). There are also more
sophist icat ed and more complicat ed w ays t o measure. Using an esophageal
balloon, t he int rapleural pressure can be measured and relat ed t o t he onset of
inspirat ory f low in pat ient s w ho t ake spont aneous breat hs. I f t he pat ient is not
t riggering t he vent ilat or, t he st at ic aut oPEEP can be measured by
occluding t he expirat ory port of t he vent ilat or, allow ing t he pressure of t he
alveoli t o equilibrat e w it h t he pressure of t hat port (not all vent ilat ors allow you
t o do t his f ancy measurement ). Anot her w ay t o measure aut oPEEP can be done
by applying an end-expirat ory pause f or 1 t o 2 seconds and measuring t he
pressure in excess of t he PEEP set on t he vent ilat or.
FIG URE 120. 1. The vent ilat or delivers t he breat h bef ore f low has ceased (it
does not reach t he baseline) (B); t heref ore Paw (peak airw ay pressure)
increases (A).
Watch Out For

There are several reasons aut oPEEP is import ant in CO PD pat ient s. First , t w o-
t hirds of mechanically vent ilat ed pat ient s w it h CO PD have aut oPEEP. This is
especially t rue in t hese pat ient s w ho are mechanically vent ilat ed w it h high minut e
vent ilat ion and reduced expirat ory t ime. Second, aut oPEEP is associat ed w it h
hemodynamic inst abilit y, by reducing venous ret urn w it h increased int rat horacic
pressure. I mport ant ly, aut oPEEP is a possible cause of pulseless elect rical
act ivit y in a mechanically vent ilat ed pat ient .
Third, aut oPEEP can cause diff icult y t riggering t he vent ilat or and increase t he
w ork of breat hing, w it h associat ed f ailure t o w ean f rom mechanical vent ilat ion.
Because of t his hyperinf lat ion t he pat ient needs t o generat e more negat ive
int rapleural pressure (increased
w ork of breat hing) t o overcome t he gradient of pressure generat ed by aut oPEEP

and achieve similar levels of t idal volume. Also, aut oPEEP can cause w orsening
of t he gas exchange w it h low er PaO2 . I n an at t empt t o decrease t he risk of t his
f rom occurring in a mechanically vent ilat ed pat ient , t he expirat ory t ime can be
increased t o account f or t he air t rapping in such pat ient s.
Several maneuvers can be implement ed t o opt imize t he abilit y t o successf ully
w ean pat ient s w it h CO PD. First , f low resist ance must be decreased. This can be
accomplished w it h bronchodilat ors and minimizing secret ions. Second, at t empt s
should be made t o overcome t he aut oPEEP by applying ext ernal PEEP. This
seems count erint uit ive but t he f unct ion of ext ernal PEEP is t o open t he airw ay t o
increase expirat ory f low and reduce t he w ork of breat hing. I t needs t o be
applied gent ly, w it h t he object ive of giving PEEP t hat is less t han or equal t o t he
aut oPEEP. This est ablishes equilibrium t hroughout t he airw ay and decreases t he
w ork of breat hing.
Anot her possibilit y f or CO PD pat ient s w ho are comf ort able and t olerat e minimal
set t ings is t o ext ubat e t o a mode of noninvasive vent ilat ion (NI V). O f course, t his
approach has t he risk of f ailure and need f or reint ubat ion, and t he risk-t o-benef it
rat io must be evaluat ed f or each individual pat ient and I CU set t ing depending on
local resources and experience.
Suggested Readings
I rw in RS, Rippe JM, eds. I rw in's and Rippe's I nt ensive Care Medicine. 5t h Ed.
Mughal MM. Aut o-posit ive end-expirat ory pressure: mechanisms and
t reat ment . Cleve Clin J Med 2005; 72: 801–809.
Plant PK, Elliot t MW. Chronic obst ruct ive pulmonary disease 9: management
of vent ilat ory CO PD. Thorax 2003; 58: 537–542.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 121 - B e C autious in Us ing P os itive
E nd- E xpir ator y P r es s ur e After S ingle- Lung Tr ans plants
121
Be Cautious in Using Positive End-Expiratory
Pressure After Single-Lung Transplants
Eric S. Weiss MD
Ashish S. Sanh MD
Case
A 63-year-old man w it h an ext ensive hist ory of cigaret t e smoking and severe
chronic obst ruct ive pulmonary disease (CO PD) (f irst expirat ory volume in 1
second [ FEV1 ] <25% predict ed and rest ing PaO2 <55 mm Hg) present s t o a
pulmonary specialist f or management . He st at es t hat he has been f ree of
cigaret t es f or almost 1 year but his poor oxygenat ion severely limit s his
f unct ional abilit y. He is t hus deemed a suit able candidat e f or lung
t ransplant at ion. Af t er a relat ively short period on t he t ransplant list , t he pat ient
receives a lef t single lung f rom an ot herw ise healt hy 24-year-old donor w ho died
f rom head injuries in a mot orcycle crash.
The operat ion progresses smoot hly and t he pat ient is t ransf erred t o t he cardiac
surgical int ensive care unit (CSI CU) post operat ively. I nit ial art erial blood gas
measurement show s t he art erial part ial pressure of oxygen t o be only 55 mm Hg
on a f ract ion of inspired oxygen (FI O2 ) of 100%. The resident physician orders
an increase in t he posit ive end-expirat ory pressure (PEEP) on t he vent ilat or f rom
5 t o 10. Subsequent blood gas show s a w orsening of oxygenat ion w it h a PaO 2
of 45 and PaCO2 of 55. I n an eff ort t o improve t he pat ient 's respirat ory st at us,
he is disconnect ed f rom t he vent ilat or and hand bagged. His oxygenat ion
cont inues t o decline and now his blood pressure drops precipit ously. He
becomes bradycardic and ult imat ely arrest s.
Discussion
First perf ormed in a human by Dr. James Hardy in 1963, lung t ransplant at ion has
become a lif esaving procedure f or many pat ient s w it h a variet y of pulmonary
disorders. Pat ient s w ho have severe disabling pulmonary disease but w ho are
ot herw ise healt hy are generally considered good candidat es f or lung
t ransplant at ion. This pat ient present ed w it h chronic obst ruct ive pulmonary
disease. The indicat ions f or lung t ransplant at ion in CO PD include a predict ed
FEV 1 of less t han 30%, rest ing part ial pressure of oxygen of less t han 55 mm
Hg,
development of pulmonary hypert ension, severe disabling sympt oms, or rapid

decline of f unct ion.
PEEP has an import ant role in t he early management of all lung-t ransplant
pat ient s. The new ly reperf used allograf t is st iff and sensit ive t o volume and
oxygen injury. As a result , PEEP helps t o maint ain alveolar recruit ment , limit
pulmonary overcirculat ion, and minimize t he inspired oxygen. All t his likely
reduces t he dreaded complicat ion of primary allograf t f ailure and may cont ribut e
t o bet t er long-t erm allograf t f unct ion. I n t he specif ic set t ing of single-lung
t ransplant at ion, how ever, vent ilat or management must t ake int o considerat ion t he
nont ransplant ed, nat ive lung. I n t his sit uat ion hyperinf lat ion, due t o excessive
PEEP, insuff icient expirat ory t ime, and air t rapping leads t o hypovent ilat ion and
mechanical compression of t he heart and lung allograf t . This let hal scenario is
f urt her w orsened by manual bagging. I n double-lung t ransplant pat ient s, unless
t here is an isolat ed airw ay problem, unilat eral hyperinf lat ion is unlikely. How ever,
even double-lung t ransplant pat ient s may air t rap if t hey have insuff icient
expirat ory t ime.
For most lung t ransplant s, experienced physicians of t en keep PEEP at 8 cm H2 O
on t he f irst night of surgery. This not only maint ains alveolar recruit ment (w hich
minimizes t he oxygen requirement ) but also limit s pulmonary blood f low and may
prevent reperf usion injury. I n pat ient s w it h pulmonary hypert ension, a higher
PEEP of 10 t o 12 cm H2 O is used t o f urt her limit blood f low and alveolar f luid
leak. Typically, PEEP is cont inued overnight unt il t he pat ient is ready f or
ext ubat ion, at w hich point convent ional set t ings are used as t olerat ed.
Suggested Readings
De Perrot M, Bonser RS, Dark J, et al. Report of t he I SHLT Working G roup
on Primary Lung G raf t Dysf unct ion part I I I : donor-relat ed risk f act ors and
markers. J Heart Lung Transpl 2005; 24: 1460–1467.
De Perrot M, I mai Y, Volgyesi G A, et al. Eff ect of vent ilat or-induced lung
injury on t he development of reperf usion injury in a rat lung t ransplant model.
J Thorac Cardiovasc Surg 2002; 74: 1658–1662.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 122 - S elec t an Initial P r es s ur e
S etting that is Jus t S lightly Higher than the P atient' s P eak P r es s ur e W hen Attem pting a P r es s ur e
S uppor t W ean
122
Select an Initial Pressure Setting that is Just
Slightly Higher than the Patient's Peak Pressure
When Attempting a Pressure Support Wean
Pressure-support vent ilat ionis a f low -cycled mode of vent ilat ion. During
inspirat ion, inspirat ory f low in t his mode decreases as lung volumes reach f ull
inhalat ion. This mode is also pat ient t riggered, meaning t hat t he pat ient is able
t o init iat e t he inspirat ory cycle. This leads t o increased pat ient comf ort by
enhancing pat ient -vent ilat or synchrony. Finally, pressure-support vent ilat ion is
considered t o be pressure limit ed, w hereby t he pressure-support vent ilat ion level
is det ermined by t he provider and programmed int o t he machine t o assist t he
pat ient . The pat ient cont rols t he rat e and t iming of respirat ions.
What to Do
As a mode t o assist w it h w eaning mechanical vent ilat ion, t he pressure-support
level can be adjust ed so t hat t he pat ient is achieving an adequat e t idal volume (6
t o 10 cm3 / kg) w it h a comf ort able respirat ory rat e (10 t o 14 breat hs per minut e).
As a st art ing point , t he pressure-support level can be placed 2 t o 5 cm H2 O
above t he peak pressure limit and adjust ed as necessary t o accomplish t he
previously recommended paramet ers. O f course, t he assessment of acid-base
balance and oxygenat ion by eit her art erial blood gas measurement s or end-t idal
CO 2 and pulse oximet ry w ill allow more def init ive adjust ment s.
Pressure-support vent ilat ion has some clear advant ages as a w eaning mode. I t
is more comf ort able t han ot her modes and of t en w ell t olerat ed by pat ient s.
How ever, it should be remembered t hat pressure-support vent ilat ion is a
spont aneous breat hing mode and t hat t here is no backup rat e if t he pat ient
should become apneic or more heavily sedat ed. O ne sensit ive met hod of
assessing a pat ient 's f ailure t o w ean on pressure-support vent ilat ion is t o
regularly assess t he respirat ory rat e. Minut e vent ilat ion is det ermined by t he
product of t he respirat ory rat e and t idal volume, as depict ed in t he f ollow ing
equat ion:
minut e vent ilat ion = respirat ory rat e × t idal volume
I f t he pat ient develops at elect asis or is insuff icient ly support ed by t he applied
pressure-support vent ilat ion, t idal volume w ill decrease and t he minut e
vent ilat ion w ill become compromised. This w ill manif est it self clinically as an
elevat ion in t he respirat ory rat e. I f pat ient s on pressure-support vent ilat ion
become t achypneic, considerat ion should be given t o increasing t he level of
pressure support or providing a backup rat e.
Suggested Readings
Brochard I , Rauss A, Benit o S, et al. Comparison of t hree met hods of gradual
w it hdraw al f rom vent ilat ory support during w eaning f rom mechanical
vent ilat ion. Crit Care Med 1994; 150: 896–903.
Fink MP, Abraham E, Vincent JL, et al. Text book of Crit ical Care. 5t h Ed.
Philadelphia: Elsevier Saunders; 2005: 120–123.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 123 - D o Not R ever s e Neur om us c ular
B loc kade Unles s the P atient is W ar m
123
Do Not Reverse Neuromuscular Blockade Unless
the Patient is Warm
James F. Weller MD
Neuromuscular blocking agent s (Tabl e 123. 1) are used rout inely during general
anest hesia t o f acilit at e endot racheal int ubat ion and t o provide opt imal surgical
condit ions. These drugs block t ransmission at t he neuromuscular junct ion t hrough
t heir eff ect s on t he post synapt ic acet ylcholine recept or. Since t hese drugs block
t ransmission of impulses f rom t he nerve t o t he muscle, t hey cannot prevent
muscle cont ract ion in response t o direct applicat ion of elect rical st imulat ion
(e. g. , f rom elect rocaut ery).
The agent s used t o provide muscle relaxat ion f all int o t w o classes: depolarizing
and nondepolarizing. The only depolarizing agent used clinically in t he Unit ed
St at es is succinylcholine. Binding of succinylcholine t o t he post synapt ic
acet ylcholine recept or causes t he channel t o open, leading t o an init ial w ave of
depolarizat ion t hat causes t he muscle t o f asciculat e. How ever, unlike
acet ylcholine, succinylcholine remains bound t o t he recept or, leading t o a period
of f laccid paralysis last ing several minut es. I t is degraded by plasma
pseudocholinest erase and cannot be reversed.
The nondepolarizing neuromuscular relaxant s w ork by compet it ive inhibit ion of
acet ylcholine at t he same recept or sit e. The choice of nondepolarizer is
generally made on t he basis of t ime of onset , durat ion of act ion, and side-eff ect
prof ile. Eliminat ion generally involves renal and/ or hepat ic met abolism, w hich
may also impact drug choice in some pat ient s. Except ions include mivacurium,
w hich like succinylcholine is degraded by plasma pseudocholinest erase, and
at racurium and cisat racurium, w hich are eliminat ed by Hoff man degradat ion (a
nonenzymat ic phenomenon t hat result s in breakdow n of t hese drugs at
physiologic t emperat ure and pH).
Because t heir physiologic eff ect depends upon compet it ive inhibit ion, t he
nondepolarizing agent s can be “reversed” by administ rat ion of
acet ylcholinest erase inhibit ors (e. g. , neost igmine, edrophonium), w hich lead t o
accumulat ion of acet ylcholine at t he neuromuscular junct ion and elsew here in t he
body. I n order t o prevent undesirable cholinergic side eff ect s out side t he
neuromuscular junct ion (e. g. , bradycardia, salivat ion, lacrimat ion), an
ant imuscarinic agent (e. g. , glycopyrrolat e, at ropine) is generally administ ered
concomit ant ly w it h t he cholinest erase inhibit or.
TABLE 123-1 COM PARISON OF COM M ON NEUROM U

AGENTS
T IME
CLINICAL
GENERIC T RADE OF MODE O
DURAT ION
NAME NAME ONSET ELIMINA
(MIN)
(MIN)
Plasma
Succinylcholine Anectine 1–1.5 5–8
choline
Renal
eliminat
d-tubocurarine 4–6 80–120
hepatic
clearan
Hoffman
Atracurium Tracrium 2–4 30–60
degrada
Hoffman
Cisatracurium Nimbex 2–4 30–60
degrada
Plasma
Mivacurium Mivacron 2–4 12–18
choline
Rocuronium Zemuron 1–2 30–60 Hepatic
metabo
Hepatic
metabo
Vecuronium Norcuron 2–4 60–90
renal
clearan
Renal
Pancuronium Pavulon 4–6 120–180
eliminat
Watch Out For

Among t he many delet erious eff ect s of perioperat ive hypot hermia is prolongat ion
of t he eff ect s of neuromuscular relaxant s. The eff ect s of hypot hermia on
paralyt ic agent s appear t o be pharmacokinet ic rat her t han pharmacodynamic.
The durat ion of act ion of vecuronium is doubled at 34°C; how ever, t here is no
change in t he concent rat ion eff ect (i. e. , pharmacodynamic) relat ionship at t he
same t emperat ure. Furt hermore, hypot hermia it self is associat ed w it h muscle
relaxat ion, even in t he absence of neuromuscular blocking agent s.
Because of prolongat ion of t he durat ion of act ion of neuromuscular blockade,
at t empt ed reversal in t he hypot hermic pat ient may be unsuccessf ul. Despit e
f ailing t o rest ore st rengt h adequat e f or ext ubat ion, how ever, early reversal may
unmask t he innat e physiologic shivering response t o hypot hermia. Alt hough
shivering does not increase met abolic demands by 400% as once t hought , it is
associat ed w it h an increase in met abolic oxygen demand. This may be
undesirable in pat ient s w it h limit ed cardiac reserve and is uncomf ort able f or
many pat ient s. Alt hough all opioids reduce shivering, t he most eff icacious choice
remains meperidine. O t her pharmacologic t reat ment s f or post operat ive shivering
include low -dose ket amine, clonidine, and physost igmine.
I n t he cont emporary pract ice of anest hesia, ext ensive eff ort s are made t o
monit or and maint ain core body t emperat ure in t he operat ing room. Nonet heless,
w hen a pat ient arrives t o t he int ensive care unit hypot hermic (e. g. , f ollow ing
massive resuscit at ion or prolonged exposure t o t he environment ), aggressive
measures should be inst it ut ed t o rew arm t he pat ient prior t o reversal of
neuromuscular blockade. These include applicat ion of ext ernal f orced-air
devices, w arming of int ravenous f luids, and humidif icat ion of inhaled gases.
While aw ait ing achievement of normot hermia, t he pat ient should be sedat ed
appropriat ely.
Suggested Readings
Hardman J, Limbird L, G oodman G ilman A, eds. G oodman and G ilman's t he
Pharmacologic Basis of Therapeut ics. 10t h Ed. New York: McG raw -Hill; 2001.
Heier T, Caldw ell JE, Sessler DI , et al. Mild int raoperat ive hypot hermia
increases durat ion of act ion and spont aneous recovery of vecuronium
blockade during nit rous oxide-isof lurane anest hesia in humans. Anest hesiology
1991; 74: 815–819.
Heier T, Caldw ell JE, Sharma ML, et al. Mild int raoperat ive hypot hermia does
not change t he pharmacodynamics (concent rat ion-eff ect relat ionship) of
vecuronium in humans. Anest h Analg 1994; 78: 973–977.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 124 - Us e Glyc opyr r olate B efor e Us ing
Neos tigm ine w hen R ever s ing Neur om us c ular B loc kade
124
Use Glycopyrrolate Before Using Neostigmine
when Reversing Neuromuscular Blockade
Leo Hsiao DO
During surgery it is of t en necessary t o achieve dense-muscle relaxat ion t o
f acilit at e surgical t echnique or t o ensure pat ient saf et y. To t his end, t he use of
neuromuscular blocking agent s have become an int egral part of a balanced
anest het ic t echnique. The t w o main classes of neuromuscular relaxing agent s are
t he depolarizing muscle relaxant s and t he nondepolarizing muscle relaxant s.
Depolarizing agent s mimic t he act ion of acet ylcholine at t he neuromuscular
junct ion (NMJ) and prevent t he normal recovery of t he muscle unit f rom it s
ref ract ory st at e, rendering it unexcit able. Depolarizing relaxat ion is t erminat ed
w hen t he agent diff uses f rom t he neuromuscular junct ion t o t he plasma, w here it
is met abolized by plasma pseudocholinest erases. I n cont rast , nondepolarizing
agent s w ork as compet it ive ant agonist s of acet ylcholine (ACH) at t he NMJ. Their
eff ect s dissipat e eit her w hen t he drug diff uses aw ay f rom it s sit e of act ion or
w hen t here is an eff ect ive surge of ACH at t he neuromuscular junct ion t hat
compet it ively displaces t he agent . The lat t er mechanism is t he principle behind
reversal of nondepolarizing neuromuscular blockade.
To brief ly review, during normal t ransmission of an act ion pot ent ial, inf lux of
calcium at t he nerve t erminal t riggers t he release of st orage vesicles cont aining
acet ylcholine. The membrane packages diff use across t he synapt ic clef t t o bind
w it h t he ACH recept ors at t he mot or endplat e. This binding result s in a
conf ormat ional change in t he ligand-act ivat ed recept or at t he mot or endplat e t hat
allow s t he inf lux of ions, producing a depolarizat ion of t he mot or endplat e. When
t he membrane depolarizat ion reaches a t hreshold, an act ion pot ent ial is
produced and t he muscle f iber f ires. St imulat ion is t erminat ed w hen ACH is
degraded by acet ylcholinest erase locat ed at t he neuromuscular junct ion.
Neost igmine is a pot ent inhibit or of cholinest erase enzyme act ivit y at t he NMJ.
The t ypical dose is w eight based at 0. 08 mg/ kg w it h a maximum dose of 5 mg.
By prevent ing acet ylcholine degradat ion, neost igmine indirect ly elevat es t he
concent rat ion of neurot ransmit t er at t he NMJ. Acet ylcholine in t urn compet it ively
displaces residual nondepolarizing drug and “reverses” t he w aning eff ect s of
blockade.
Unf ort unat ely, t his acet ylcholine surge ubiquit ously aff ect s bot h muscarinic and
nicot inic recept ors t hroughout t he body. This result s in side eff ect s such as
bradycardia or asyst ole, hyperperist alsis, nausea, and vomit ing, increased
bronchial secret ions, and urinary urge. These eff ect s are part icularly w orrisome
in a pat ient emerging f rom anest hesia and should be avert ed if possible.
What to Do
G lycopyrrolat e is an ant icholinergic-class medicat ion given in an int ravenous
f ormulat ion as an adjunct t o count er t he unt ow ard eff ect s of neost igmine during
neuromuscular blockade reversal. The t ypical dose of glycopyrrolat e f or t his
indicat ion is 0. 2 mg per 1 mg of neost igmine. Because of it s quat ernary
st ruct ure, glycopyrrolat e is essent ially devoid of cent ral eff ect s. I t does,
how ever, produce an impressive blockade of acet ylcholine at t he peripheral
muscarinic recept or sit es. The result is a const ellat ion of side eff ect s including
int ense drying of mucous membranes, signif icant rise in heart rat e, diminished
int est inal mot ilit y, and urinary ret ent ion.
Not surprisingly glycopyrrolat e's pharmacodynamic prof ile negat es many of t he
side eff ect s produced by neost igmine. This is a model example of an
advant ageous drug-drug int eract ion w hereby one drug count erbalances t he side
eff ect s of anot her drug w it hout negat ing it s principle t herapeut ic act ivit y. To
f acilit at e t his benef icial int eract ion, proponent s advocat e t he administ rat ion of
glycopyrrolat e minut es in advance of neost igmine. The rat ionale behind t his
sequence of administ rat ion is t o achieve a signif icant blockade of muscarinic
recept ors so as t o mit igat e t he impending surge of acet ylcholine. The result is a
smoot h reversal of muscle blockade. An alt ernat ive ant icholinergic is at ropine.
Suggested Readings
Miller RD, ed. Anest hesia. 6t h Ed. Philadelphia: Churchill Livingst one;
2005: 518–523.
Morgan E, Mikhail M, Murray M, et al. Clinical Anest hesiology. 3rd Ed. New
York: McG raw -Hill; 2002: 179–187, 199–204, 207–210.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 125 - D o Not Attem pt to R ever s e
Neur om us c ular B loc kade if ther e ar e no Tw itc hes
125
Do Not Attempt to Reverse Neuromuscular
Blockade if there are no Twitches
E. David Bravos MD
Paralysis t hrough t he use of neuromuscular blockers can provide opt imal
condit ions f or many surgical procedures and can help f acilit at e int ubat ion f or
pat ient s requiring mechanical vent ilat ion. These drugs can be broadly classif ied
as being depolarizing or nondepolarizing not only by t heir mechanism of
producing paralysis but by t heir response t o peripheral nerve st imulat ion and
reversal of blockade.
Neuromuscular blockers are st ruct urally similar t o acet ylcholine and produce
t heir eff ect s by int eract ing w it h acet ylcholine recept ors at t he neuromuscular
junct ion. Depolarizing neuromuscular blockers such as succinylcholine bind t o t he
acet ylcholine recept or and cause a conf ormat ional change result ing in act ivat ion
of t he recept or. This leads t o depolarizat ion of t hemuscle membrane and an
act ion pot ent ial if t he t hreshold pot ent ial is reached. This explains t hemuscle
f asciculat ion seen w it h use of succinylcholine. These drugs remain bound t o t he
recept or and prevent acet ylcholine f rom binding, t hus producing paralysis.
Nondepolarizing agent s bind t he acet ylcholine recept or as w ell; how ever, t hey do
not cause a conf ormat ional change and t heref ore do not cause depolarizat ion.
Similarly, acet ylcholine cannot bind t o any bound recept or w it h result ant
paralysis.
Watch Out For

O nce neuromuscular blockade has occurred, t he dept h of blockade can be
measured using peripheral nerve st imulat ion. Diff erent responses t o nerve
st imulat ion can be seen depending on t he t ype of neuromuscular blocker used,
as w ell as t he pat t ern of st imulat ion used. The most common pat t erns of
st imulat ion are t he t rain of f our (TO F) and t et any. Wit h depolarizing
neuromuscular blockade, TO F st imulat ion produces a decreased response t o
successive st imulat ion know n as f ade. This is t hought t o be due t o exhaust ion of
acet ylcholine w it h each successive st imulat ion in t he TO F (and t heref ore t here is
less acet ylcholine available t o overcome t he ant agonist ic block). Wit h t et anic
st imulat ion, f ade is seen as w ell. Depolarizing neuromuscular blockers do not
produce f ade w it h a single TO F. How ever, f ade may be seen w it h successive
TO Fs. Rat her, a const ant but diminished response
may be seen if more t han one t w it ch is present . The percent of recept or

blockade can be roughly est imat ed by t he number of t w it ches present . Three
t w it ches is roughly a 75% block; t w o t w it ches, an 80% block; and one t w it ch, a
90% block.
When t est ing f or t w it ches, t he adduct or pollicis and t he orbicularis oculi are
f requent ly used. Tw it ches in t hese muscles can be seen by st imulat ion of t he
ulnar and f acial nerves, respect ively. I n general, muscles have diff erent
sensit ivit ies t o neuromuscular blockers. More cent rally locat ed muscles including
t he laryngeal muscles and diaphragm are quicker t o respond t o neuromuscular
blockers t han more peripheral ones. Addit ionally, t he orbicularis oculi is relat ively
more resist ant compared w it h t he adduct or pollicis. The best area t o place t he
pads of t he nerve st imulat or is one t hat isolat es t he nerve and w ill not direct ly
st imulat e t he muscle. For t he orbicularis oculi, pads should be placed
superolat eral and inf erolat eral t o t he lat eral cant hus. For t he ulnar nerve, pads
should be placed just proximal and dist al t o t he ulnar nerve as it t ravels t hrough
t he groove bet w een t he medial epicondyle and t he medial aspect of t he
olecranon process of t he arm. When neuromuscular blockade is no longer
needed, reversal of t he blockade may be w arrant ed. Reversal is done w it h t he
use of nondepolarizing agent s, but succinylcholine should never be reversed and
t he use of ant icholinergics w hile succinylcholine is st ill present may act ually
prolong it s eff ect s. Bef ore reversal is at t empt ed, t he number of t w it ches should
be checked using t he TO F t o be sure at least one t w it ch is present . This
especially holds t rue w hen using longer-act ing agent s such as pancuronium.
When one t w it ch is present , spont aneous recovery is st art ing t o occur. How ever,
if no t w it ches are present t hen f ull blockade is st ill occurring and t he eff ect of
t he neuromuscular blocker may out last t he eff ect s of t he reversal agent . Pat ient s
may have generalized w eakness w it h possible respirat ory f ailure if ext ubat ed.
Suggested Readings
Miller RD, ed. Miller's Anest hesia. 6t h Ed. Churchill Livingst one; 2005: 518–
525.
Morgan G E. Clinical Anest hesiology. 3rd Ed. Lange/ McG raw -Hill; 2002: 120–
123.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 126 - R em ove C ontinuous P os itive
Air w ay P r es s ur e and B ilevel P os itive Air w ay P r es s ur e Mas ks P er iodic ally
126
Remove Continuous Positive Airway Pressure
and Bilevel Positive Airway Pressure Masks
Periodically
Cont inuous posit ive airw ay pressure (CPAP) and bilevel posit ive airw ay pressure
(BiPAP) are t w o noninvasive vent ilat ory modes t hat are commonly used in t he
int ensive care unit (I CU) as w ell as ot her set t ings. CPAP is commonly used on an
out pat ient basis t o alleviat e sleep apnea. Alt hough less common, BiPAP may also
be used f or t his purpose. While bot h of t hese modes may also be used f or
pat ient s in t he I CU w ho have t his diagnosis, t hese modes are also commonly
used t o assist I CU pat ient s w ho are having respirat ory diff icult y unrelat ed t o
sleep apnea. These sit uat ions include avoiding t he need f or invasive mechanical
vent ilat ion (e. g. , endot racheal int ubat ion) or t o help bridge pat ient s f rom invasive
vent ilat ion t o supplement al oxygen only. This may be done in a variet y of disease
set t ings, including cardiogenic pulmonary edema; chronic obst ruct ive pulmonary
disease exacerbat ions post operat ively; adult respirat ory dist ress syndrome; and
ot hers. Variable eff ect iveness has been described depending on t he disease
st at e.
Regardless of t he disease st at e being t reat ed or w het her one is at t empt ing t o
avoid int ubat ion or bridging f rom invasive vent ilat ion, close at t ent ion needs t o be
maint ained so t hat t he clinician is able t o recognize w hen t he pat ient is f ailing
t his noninvasive mode. While several st udies point t o t hese modes’ eff ect iveness
in prevent ion of or w eaning f rom invasive vent ilat ion, it is clear t hat early
recognit ion of t he pat ient 's f ailing t he noninvasive st rat egies is crucial t o
prevent ing morbidit y. I n general, a pat ient w ho is get t ing w orse af t er an hour or
so of noninvasive vent ilat ion, as det ermined by his or her blood gas values and
clinically assessed w ork of breat hing, should be considered f or int ubat ion. I f t he
pat ient is not improving but is not det eriorat ing, t he CPAP or BiPAP may be
cont inued, perhaps w it h adjust ment s t o t he pressure set t ings.
What to Do
CPAP uses a consist ent pressure set t ing t hroughout t he respirat ory cycle w hile
BiPAP uses a bilevel set t ing w it h t he inspirat ory phase being set higher t han t he
expirat ory phase. Values f or t he pressures set generally range f rom 5 t o 15 cm
H2 O . Delivery of t hese airw ay pressures requires t ight -f it t ing masks. These are
usually f ull-f ace masks t hat cover t he nose
and mout h, t hough t here are nasal masks t hat f it only over t he nose. For t hese
nasal masks t o f unct ion w ell t he pat ient must be cooperat ive enough t o keep his
or her mout h closed, or all of t he pressure w ill escape t hrough t he mout h and
vent ilat ion might be ineff ect ive.
The t ight -f it t ing nat ure of CPAP and BiPAP masks poses t he risk of pressure
necrosis of t he underlying t issues, part icularly over t he bridge of t he nose w here
t he skin is t hin and bone and cart ilage lie just near t he surf ace. This is because
t he t ight ness required t o overcome t he f low pressures and prevent leakage of
gas of t en w ill exceed t issue-perf usion pressure. I t is f or t his reason t hat every
f ew hours (4 hours is a commonly used int erval) t he mask should be removed t o
allow t he t issue t o receive adequat e blood f low. This t ime period is generally
less t han a half hour and t hen t he mask is reapplied unless t he pat ient
demonst rat es t he abilit y t o remain off support during t hat t ime. I f t he pat ient
does not t olerat e being off t he mask f or even a short t ime periodically, t he
clinician should consider moving t o int ubat ion.
Suggested Readings
Masip J, Roque M, Sanchez B, et al. Noninvasive vent ilat ion in acut e
cardiogenic pulmonary edema: syst emat ic review and met a-analysis. JAMA
2005; 294: 3124–3130.
Pet er JV, Moran JL, Phillips-Hughes J, et al. Eff ect of non-invasive posit ive
pressure vent ilat ion (NI PPV) on mort alit y in pat ient w it h acut e cardiogenic
pulmonary oedema: a met a-analysis. Lancet 2006; 367: 1155–1163.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 127 - E m pir ic ally C over the C om m on
Nos oc om ial Mic r obes in Ventilator - As s oc iated P neum onia until the C ultur es ar e R etur ned
127
Empirically Cover the Common Nosocomial
Microbes in Ventilator-Associated Pneumonia
until the Cultures are Returned
Nosocomial pneumonia is an import ant hospit al-acquired inf ect ion t hat occurs 48
hours or more af t er hospit al admission. I t is t he leading cause of deat h f rom
nosocomial inf ect ions. A subset of nosocomial pneumonia occurs in pat ient s
receiving mechanical vent ilat ion. These pat ient s have w hat is t ermed vent ilat or-
associat ed pneumonia (VAP). The use of a vent ilat or by t he pat ient increases t he
risk of developing pneumonia by a f act or of 10. As a result , VAP f requent ly
occurs in pat ient s requiring signif icant mechanical vent ilat ory support such as
pat ient s w it h acut e respirat ory dist ress syndrome (ARDS), shock syndromes,
and mult iorgan f ailure.
I t is import ant t o underst and t hat t he et iology of VAP is dependent upon t he
f undament al relat ionship bet w een t he host , t he microbial agent , and t he
environment . The host ref ers t o t he int ensive care unit (I CU) pat ient w ho may be
at part icular risk of acquiring a nosocomial inf ect ion. Alt erat ions in underlying
host def enses like t he endot racheal t ube, w hich suppresses t he normal abilit y of
t he body t o eliminat e respirat ory pat hogens by coughing; host diseases like
cancers; and underlying respirat ory diseases such as cyst ic f ibrosis aff ect bot h
t he colonizat ion of and predisposit ion t o VAP. The microbial agent s t hemselves
have specif ic propert ies t hat increase t heir pat hogenicit y. Pili, microbial t oxins,
and adherence f act ors all play a role in t he abilit y of part icular organisms t o
have a propensit y t o cause diseases. G ram-negat ive bact eria are more likely
t han gram-posit ive organisms t o cause VAP (60% vs. 40%). The gram-negat ive
microbial organisms t hat are responsible f or causing VAP include Pseudomonas
aerugi nosa, Escheri chi a col i , Kl ebsi el l a pneumoni ae, and Aci netobacter
species. The gram-posit ive spect rum primarily involves Staphyl ococcus species.
I n addit ion, t he environment of t he I CU, t he hospit al, and t he region w ill all aff ect
t he spect rum of microbials t hat colonize t he environment and predispose t he
pat ient t o inf ect ions.
Watch Out For
Host f act ors t hat increase t he likelihood of acquiring VAP include pat ient s w it h
ARDS, chronic obst ruct ive airw ays disease (CO PD),
malnut rit ion, burns, t rauma, a higher severit y of illness, and t he administ rat ion of
specif ic medicat ions and t reat ment s. The overuse of broad-spect rum
ant imicrobial agent s suppresses some organisms but increases t he risk of t he
pat ient t o ot her organisms t hat are prone t o inf ect t he respirat ory t ract . I n
addit ion, t he use of H2-blockers reduces t he gast ric pH and leads t o increased
colonizat ion of organisms t hat can cause nosocomial pneumonia.
There are st rat egies t hat can be used t o reduce t he risk of VAP. Aggressive
post pyloric, early f eeding has a benef icial eff ect on t he pat ient 's abilit y t o guard
against acquiring nosocomial pneumonia. Elevat ion of t he head of t he pat ient 's
bed has a benef icial eff ect on t he rat e of VAP. From an environment al
perspect ive, appropriat e hand w ashing is helpf ul in prevent ing all t ypes of
nosocomial inf ect ions. Healt h care w orkers are part icularly poor at using t his
met hod of prevent ion f or t heir pat ient s’ benef it . The respirat ory equipment it self
may lead t o increased rat es of pneumonia because of t he colonizat ion of
bact eria in t he t ubing and condensat e. Similarly, appropriat e cleaning of pat ient
rooms and mult ipat ient equipment needs t o be perf ormed. Finally, an
underst anding of t he common pat hogens in t he I CU and hospit al as w ell as t heir
ant ibiograms can be helpf ul in select ing appropriat e empiric t herapy.
What to Do
For pat ient s developing VAP, t he goals of t herapy include t he appropriat e
st abilizat ion, early diagnosis, and empiric t reat ment f ollow ed by specif ic t herapy
f or isolat ed organisms. The American Thoracic Societ y has proposed guidelines
f or t he diagnosis and t reat ment of VAP. When suspect ed, t he early cult ure of
blood and respirat ory secret ions may be benef icial in ident if ying an organism
prior t o t he init iat ion of ant imicrobial agent s. These cult ure t echniques may
include direct specimen examinat ion by use of bronchoscopy w it h prot ect ed
specimen brush specimens or bronchoalveolar lavage. I f t he cult ures are
posit ive, t hen t he init iat ion of ant ibiot ics can be helpf ul w it h a f ine-t uning of t he
ant imicrobial spect rum w hen t he sensit ivit y result s are know n. The lengt h of
ant ibiot ic coverage f or VAP should be 8 days (unless it is Pseudomonas,
requiring a 14-day course) and should include t he f ollow ing ant ibiot ic regimens:
Vancomycin plus piperacillin and t azobact am (Zosyn)

or
Vancomycin plus meropenem w it h or w it hout an aminoglycoside
or
Vancomycin plus cef epime
Suggested Readings
Fink MP, Abraham E, Vincent JL, et al. , eds. Textbook of Cri ti cal Care. 5t h
Ed. Philadelphia: Saunders; 2005: 542–544.
I brahim EH, Ward S, Sherman G , et al. Experience w it h a clinical guideline f or

t he t reat ment of vent ilat or associat ed pneumonia. Crit Care Med
2001; 29: 1109–1115.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 128 - K eep the Head of B ed E levated
at Leas t 30 degr ees for Intubated P atients if no C ontr aindic ations E xis t
128
Keep the Head of Bed Elevated at Least 30
degrees for Intubated Patients if no
Contraindications Exist
Deborah B. Hobson BSN
Sean M. Berenholtz MD, MHS
Mechanical vent ilat ion is a common int ervent ion in t he int ensive care unit (I CU).
I t places pat ient s at an increased risk f or deat h and complicat ions, such as
vent ilat or-associat ed pneumonia and gast roint est inal bleeding. Eff ort s t o
decrease t he morbidit y and mort alit y and improve t he qualit y and saf et y of care
f or pat ient s requiring mechanical vent ilat ion are paramount . O ne simple no-cost
int ervent ion t hat improves out comes in vent ilat ed pat ient s is t o keep t he head of
t he bed elevat ed ≥30 degrees. Elevat ing t he head of bed ≥30 degrees can
reduce t he f requency and risk f or nosocomial pneumonia compared w it h t he
supine posit ion, decrease durat ion of mechanical vent ilat ion, and decrease I CU
lengt h of st ay. The evidence support ing t his int ervent ion is f rom a w ell-done
randomized st udy of mechanically vent ilat ed pat ient s t hat f ound t he incidence of
pneumonia w as reduced f rom 38% in t he supine group t o 8% in t he ≥30-degrees
group. Days on t he vent ilat or and I CU lengt h of st ay w ere also reduced.
O t her st rat egies, in combinat ion w it h elevat ing t he head of t he bed ≥30 degrees,
have been show n in t he medical lit erat ure t o be very eff ect ive in decreasing t he
morbidit y, mort alit y, and cost of care f or pat ient s receiving mechanical
vent ilat ion. These t herapies are of t en t ermed t he “vent ilat or bundle” and include
head of bed elevat ion ≥30 degrees; appropriat e pept ic ulcer disease (PUD)
prophylaxis; appropriat e deep venous t hrombosis (DVT) prophylaxis; appropriat e
sedat ion; and daily assessment of abilit y t o ext ubat e. Each of t hese
int ervent ions is support ed by medical evidence and independent ly decreases
pat ient morbidit y and/ or mort alit y. A recent prospect ive mult icent er observat ional
t rial evaluat ing t he impact of improving compliance w it h t he vent ilat or bundle
f ound t hat f or t he 35 I CUs t hat consist ent ly collect ed dat a on vent ilat or bundle
compliance and VAP rat es, an average 45% reduct ion of VAP w as observed. I n
addit ion, t hree of t hese int ervent ions, elevat ing t he head of t he bed ≥30
degrees, PUD prophylaxis, and DVT prophylaxis, have been select ed by t he Joint
Commission on Accredit at ion of Healt hcare O rganizat ions (JCAHO ) f or inclusion
int o t he init ial core set of I CU qualit y measures.
Suggested Readings
Dodek P, Keenan S, Cook D, et al. Evidence-based clinical pract ice guideline
f or t he prevent ion of vent ilat or-associat ed pneumonia. Ann I nt ern Med
2004; 141: 305–313.
Drakulovic MB, Torres A, Bauer TT, et al. Supine body posit ion as a risk
f act or f or nosocomial pneumonia in mechanically vent ilat ed pat ient s: a
randomised t rial. Lancet 1999; 354: 1851–1858.
Kollef MH. The prevent ion of vent ilat or-associat ed pneumonia. N Engl J Med
1999; 340: 627–634.
Resar R, Pronovost P, Haraden C, et al. Using t he bundle approach t o

improve vent ilat or care processes and reduce vent ilat or-associat ed
pneumonia. Jt Comm J Q ual Pat ient Saf 2005; 31: 243–248.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 129 - Tr eat Ventilator - As s oc iated
P neum onia for 8 days
129
Treat Ventilator-Associated Pneumonia for 8
days
Vent ilat or-associat ed pneumonia (VAP) ref ers t o a pneumonia t hat arises af t er
48 hours of mechanical vent ilat ion. Unf ort unat ely, it is common, cost ly, and lif e
t hreat ening. I t is t he most f requent int ensive care unit (I CU)–acquired inf ect ion
among pat ient s receiving mechanical vent ilat ion, and it is t he second most
common nosocomial inf ect ion in t he Unit ed St at es. VAP has also been f ound t o
increase hospit al st ay by an average of 7 t o 9 days and result s in an increase in
healt h care–relat ed cost s of $12, 000 t o $40, 000 per case. I n addit ion,
approximat ely 50% of all ant ibiot ics prescribed in t he I CU are administ ered f or
respirat ory t ract inf ect ions. Crude I CU mort alit y associat ed w it h VAP ranges
f rom 30% t o 70% w it h an increased relat ive risk of mort alit y of approximat ely
2. 5.
Despit e t he undisput ed clinical import ance of VAP, lit t le consensus exist s
regarding it s prevent ion, diagnosis, or t reat ment . Some of t he risk f act ors f or
VAP are clearly know n: supine body posit ioning; t ransf usion of blood product s;
and durat ion of vent ilat ion. Durat ion of vent ilat ion has been closely associat ed
w it h an increased risk of pneumonia as a direct f unct ion of t ime. The daily risk of
VAP has been est imat ed t o st art at approximat ely 3% per day f or t he f irst f ive
days, f ollow ed by 2% per day f or days six t hrough t en, bef ore f inally decreasing
t o 1% per day f or every day of mechanical vent ilat ion t hereaf t er.
Signs and Symptoms

When it comes t o diagnosing vent ilat or-associat ed pneumonia, t here is lit t le
consensus in t he lit erat ure because of diff erent int erpret at ion of t he clinical signs
and sympt oms suggest ive of lung inf ect ion, variable diff erent iat ion bet w een
colonizat ion and inf ect ion of t he low er respirat ory t ract , and poorly specif ied
prior ant ibiot ic use in t he I CU. As clinical signs are nonspecif ic and subject ive,
no single clinical crit erion has been specif ically diagnost ic f or VAP. Fever,
t achycardia, and leukocyt osis are all relat ively nonspecif ic. Chest radiographs
are most helpf ul w hen normal, as t hey can rule out pneumonia. Air bronchograms
or alveolar opacit ies on a chest radiograph in pat ient s w it hout acut e respirat ory
dist ress syndrome (ARDS) are most
correlat ed w it h pneumonia and have a 68% diagnost ic accuracy. When chest

radiographs have been compared w it h subsequent aut opsy f indings, a localized
inf ilt rat e w as 87% sensit ive f or VAP but only 25% specif ic. Furt hermore,
approximat ely 10% of t he cases of pneumonia f ailed t o have a new or w orsening
inf ilt rat e at all.
The clinical pulmonary inf ect ion score (CPI S) w as developed in an at t empt t o
f ind more reliable crit eria f or t he diagnosis of VAP. The CPI S combines clinical,
radiographic, physiological, and micro-biologic dat a int o a single numerical
result . The score is derived by aw arding 0, 1, or 2 point s f or each of seven
variables: t emperat ure; w hit e blood count ; volume and qualit y of t racheal
suct ions; oxygenat ion; chest radiograph f indings; and semiquant it at ive cult ure of
t racheal aspirat e. When t he CPI S exceeds 6, good correlat ion w it h t he presence
of pneumonia, as def ined by quant it at ive cult ures of bronchoscopic and
nonbronchoscopic lavage specimens, has been f ound. When compared w it h
post mort em quant it at ive lung cult ures as t he ref erence st andard, t he CPI S has a
sensit ivit y of 77% and a specif icit y of 42% f or VAP.
I ncreased specif icit y at t he cost of sensit ivit y over t he CPI S may be obt ained by
using diagnost ic crit eria of a radiographic inf ilt rat e and at least t w o of t he t hree
f ollow ing clinical f eat ures of VAP: f ever great er t han 38C; leukocyt osis or
leukopenia; and purulent t racheal secret ions. Use of t hese crit eria result ed in a
69% sensit ivit y and a 75% specif icit y f or VAP, w hich represent s t he most
accurat e clinical crit eria f or st art ing empiric ant ibiot ic t herapy.
O nce t he diagnosis of VAP is being considered, a low er respirat ory t ract sample
f or cult ure t o guide t herapy should be collect ed bef ore ant ibiot ic administ rat ion.
I f cult ures are obt ained af t er init iat ion of ant ibiot ics, a 40% f alse negat ive rat e
ensues. I f t here is a high pret est probabilit y of pneumonia, or in t he 10% of
pat ient s w it h evidence of sepsis, prompt t herapy is required, regardless of
w het her bact eria are f ound on microscopic examinat ion of low er respirat ory t ract
samples.
The diagnosis and t reat ment of VAP is diff icult in part because of t he mult iple
et iologic agent s associat ed w it h t he disease. The causes of VAP are legion and
w ill vary by hospit al, pat ient populat ion, and t ype of I CU, emphasizing t he need
f or t imely, local surveillance dat a. O ne of t he main goals of surveillance and
guiding principles of t reat ment of VAP is t o det ermine w het her t he pat ient has
VAP associat ed w it h ant ibiot ic-resist ant organisms or not .
VAP associat ed w it h ant ibiot ic-resist ant microorganisms can usually be predict ed
by t he durat ion of vent ilat ion of 5 days or more; t he
lengt h of st ay bef ore t he onset of VAP; a hist ory of prior ant ibiot ic t reat ment ; or
know ledge of a high f requency of ant ibiot ic resist ance in t he communit y or
hospit al. Typically, ant ibiot ic-resist ant organisms associat ed w it h VAP are
Pseudomonas aerugi nosa, Aci netobacter, met hicillin-resist ant Staphyl ococcus
aureus (MRSA), and mult idrug-resist ant gram-negat ive bacilli. I n t hese pat ient s
suspect ed of having VAP due t o ant ibiot ic-resist ant organisms, init ial empiric
coverage should be broad.
Those w it hout risk f act ors f or resist ant organisms (less t han 5 days of
vent ilat ion, no prior ant ibiot ic exposure, no prior cont act w it h t he healt h care
syst em) f requent ly have VAP due t o ant ibiot ic-sensit ive bact eria such as
Haemophi l us i nf l uenzae, Streptococcus pneumoni ae, MSSA (met hicillin-sensit ive
S. aureus), or ant ibiot ic-sensit ive ent eric gram-negat ive bacilli. These pat ient s
may be empirically st art ed on ant ibiot ic monot herapy. Legi onel l a species,
anaerobes, f ungi, and even viruses should also be ment ioned as pot ent ial rare
causes of pneumonia in vent ilat ed pat ient s. This being said, it is not uncommon
t o isolat e Candi da species f rom endot racheal aspirat es of vent ilat ed pat ient s.
This dist inct ion arises because candidal colonizat ion of t he airw ays is common,
yet inf ect ion is rare, w it h a maximum of 20% of pat ient s w it h posit ive candidal
endot racheal aspirat es act ually having invasive candidal disease. Theref ore, a
posit ive endot racheal aspirat e rarely mandat es ant if ungal t herapy.
O nce t he diagnosis of VAP is made, select ion of an appropriat e empiric
ant ibiot ic regimen is key t o pat ient survival. I n one st udy, inappropriat e init ial
ant ibiot ic t herapy (pat hogen resist ant or not covered) f or VAP w as document ed
in as many as 33% of cases, and init ial inappropriat e t herapy independent ly
increased t he risk of mort alit y by nearly seven f old. This increased risk of
mort alit y persist ed even w hen t he t reat ment w as changed appropriat ely based
upon subsequent cult ure dat a.
The previous f act s drive t he approach t o ant ibiot ic t reat ment of VAP. To reduce
pat ient mort alit y, one must avoid inappropriat e t reat ment by init iat ing empiric,
broad-spect rum mult iagent ant ibiot ic t herapy. The goal is not f or redundant
coverage of a pat hogen, but t o increase t he chance t hat t he init ial ant ibiot ic
regimen is appropriat e f or t he et iologic agent once it becomes know n. There has
been no document ed clinical superiorit y of double coverage as opposed t o
appropriat e monot herapy. Theref ore, init ial ant ibiot ic t herapy should be t ailored
t o inst it ut ional pat t erns of ant ibiot ic resist ance and pat ient f act ors t hat have an
inf luence upon t he et iologic agent s and ant ibiot ic-resist ance pat t erns f or VAP in
t he individual.
What to Do
G iven t he above, ant imicrobial t herapy f or VAP t ypically f ollow s a t w o-st age
process. The f irst st age involves init ial broad-spect rum coverage t o avoid
inappropriat e t reat ment . O nce t he decision t o t reat is made, cult ures should be
obt ained and ant ibiot ic t herapy should be init iat ed. Delays in t he init iat ion of
appropriat e ant ibiot ic t herapy can increase t he mort alit y of VAP, and t hus
t herapy should not be post poned f or t he purpose of perf orming diagnost ic
st udies in pat ient s w ho are clinically unst able. I nit ial empiric t herapy is based
upon pat ient risk f or mult idrug-resist ant pat hogens. When it is det ermined t hat
t here is a low risk f or resist ant organisms, t hen monot herapy including
cef t riaxone, a f luoroquinolone, or ampicillin/ sulbact am should be init iat ed. For
pat ient s w it h risk f act ors f or mult idrug-resist ant pat hogens, empiric coverage
should be broad and init iat ed as f ollow s:
Cef epime/ cef t azidime or imipenem/ meropenem

and
Ciprof loxacin/ levof loxacin or amikacin/ gent amicin/ t obramycin
and
Vancomycin/ linezolid
Not e t hat t he goal of combinat ion t herapy is not t o provide redundant coverage
so much as t o ensure init ial appropriat e empiric t herapy.
The second st age of ant imicrobial t herapy f ocuses on maint aining appropriat e
coverage w it hout overusing ant ibiot ics. Therapy may be modif ied in pat ient s w it h
a low probabilit y of disease based upon clinical response on days t w o and t hree.
Ant ibiot ic coverage may be sw it ched t o monot herapy or ot herw ise narrow ed
based upon cult ure dat a, and t he durat ion of t herapy may be short ened t o 8
days. Ant ibiot ic spect ra should be narrow ed as cult ure dat a become know n.
O nce t he result s of respirat ory t ract and blood cult ures become available,
t herapy can of t en be narrow ed on t he basis of t he ident it y of specif ic pat hogens
and t heir suscept ibilit y t o specif ic ant ibiot ics.
O nce t reat ment f or VAP is init iat ed, a quest ion arises as t o w hen it is
appropriat e t o st op. Unf ort unat ely, t herapy cannot be guided by chest
radiographs as t hey are of limit ed value f or def ining clinical improvement in
severe pneumonia. I nit ial radiographic det eriorat ion is common, especially
among pat ient s w ho are bact eremic or w ho are inf ect ed w it h highly virulent
organisms. I n addit ion, radiographic improvement of t en lags behind clinical
paramet ers. Clinical improvement usually becomes apparent af t er t he f irst 48
hours of t herapy and, t heref ore, t he select ed ant imicrobial regimen should not be
changed
during t his t ime unless progressive det eriorat ion is not ed or init ial microbiologic
st udies so dict at e. Consequent ly, f ailure t o respond t o t herapy is not clinically
evident unt il t reat ment day t hree. Fort unat ely, t his is about t he t ime t hat
microbiologic dat a become available t o guide f urt her t herapy.
A recent st udy has show n t hat using t he CPI S syst em t o ident if ies pat ient s w it h
a low clinical suspicion of VAP (CPI S <6) w ho could be t reat ed w it h 3 days of
ant ibiot ics as opposed t o t he convent ional pract ice of 10 t o 21 days of ant ibiot ic
t herapy. Pat ient s receiving t he short er course of ant ibiot ics had bet t er clinical
out comes t han t hose receiving longer t herapy and had f ew er subsequent super
inf ect ions w it h ant ibiot ic-resist ant pat hogens. I n anot her recent t rial, pat ient s
w ho received appropriat e, init ial empiric t herapy f or VAP f or 8 days had
equivalent mort alit y and risk of pulmonary reinf ect ion as t hose of pat ient s w ho
received t herapy f or 15 days, w hile having signif icant ly more ant ibiot ic-f ree
days. A t rend t o great er rat es of relapse f or short -durat ion t herapy w as seen if
t he et iologic agent w as P. aerugi nosa or an Aci netobacter species. Hence t he
recommendat ion f or an 8-day course of ant ibiot ics f or VAP except in t he case of
a pseudomonal or Aci netobacter pneumonia, w here t reat ment should be t aken
out t o 14 days.
Suggested Readings
American Thoracic Societ y; I nf ect ious Diseases Societ y of America.
G uidelines f or t he management of adult s w it h hospit al-acquired, vent ilat or-
associat ed, and healt hcare-associat ed pneumonia. Am J Respir Crit Care Med
2005; 171: 388–416.
Chast re J. Conf erence summary: vent ilat or-associat ed pneumonia. Respir

Care 2005; 50: 975–983.
Fink MP, Abraham E, Vincent JL, Kachanek PM, eds. Text book of Crit ical
Care. 5t h Ed. Philadelphia: Elsevier Saunders; 2004: 663–677.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 130 - D o not R outinely E xtubate on
C linic al P ic tur e Alone
130
Do not Routinely Extubate on Clinical Picture
Alone
The benef it s of separat ing pat ient s f rom mechanical vent ilat ion include
decreasing t he risk of vent ilat or-associat ed pneumonia and vent ilat or-induced
lung injury, increasing pat ient comf ort , and decreasing cost . These benef it s,
how ever, must be w eighed against t he complicat ions of removing mechanical
vent ilat ion t oo soon, w hich include t he pot ent ial f or diff icult y in re-est ablishing
t he airw ay and in complicat ions relat ed t o impaired gas exchange.
Decisions t o discont inue mechanical vent ilat ion cannot be based on t he clinical
pict ure alone. “Foot of t he bed” assessment is not oriously unreliable in
predict ing pat ient suit abilit y f or w eaning and ext ubat ion. St udies have
document ed sensit ivit ies as low as 35%, implying t hat of t en pat ient s w ho may
successf ully w ean are missed. O verall, clinical assessment is bet t er at
predict ing t hose w ho w ill f ail t o w ean but st ill has only a specif icit y of less t han
80% at best . To t his end, guidelines f or vent ilat or w eaning have been adopt ed by
several prof essional organizat ions including t he American College of Chest
Physicians, t he Societ y of Crit ical Care Medicine, and t he American Associat ion
f or Respirat ory Care.
The guidelines are based on t he best current ly available evidence and
recommend crit eria t o det ermine w hom should be considered f or discont inuat ion
of vent ilat ion, principles t o guide t hat assessment , how t o t reat pat ient s w ho f ail
t heir w eaning assessment , as w ell as considerat ions f or long-t erm vent ilat or
dependence and t he design and use of w eaning prot ocols by nonphysicians.
Crit eria suggest ed prior t o a f ormal assessment of discont inuat ion include
evidence of some reversal of t he underlying reason f or t he pat ient 's respirat ory
f ailure; adequat e oxygenat ion (PaO2 / FI O 2 ≥200 mm Hg, PEEP ≤5–8 cm H2 O ,
FI O 2 ≤0. 4–0. 5, and pH ≥7. 25; FI O2 , f ract ion of inspired oxygen; PEEP, posit ive
end-expirat ory pressure); hemodynamic st abilit y; and pat ient abilit y t o generat e
an inspirat ory eff ort . Spont aneous breat hing t rials (SBTs) may be at t empt ed
w hen t hese crit eria are met . Trials may be carried out on eit her T-piece or w it h
so-called “minimal” vent ilat or set t ings of 5 cm H2 O cont inuous posit ive airw ay
pressure (CPAP) and 5 cm H2 O pressure support . The decision t o use one
met hod or t he ot her does not seem t o aff ect out come. The f ormer has t he
advant age of removing all support ive posit ive pressure f rom t he airw ay but has
t he disadvant age of removing t he pat ient f rom any apnea alarms on t he

vent ilat or. At t hese set t ings, concept ually CPAP w ill maint ain recruit ment of
alveoli and pressure support w ill overcome t he resist ance t o t he vent ilat or
circuit . The pat ient should be observed f or respirat ory pat t ern, qualit y of gas
exchange, hemodynamic changes, and subject ive comf ort . Pat ient s w ho pass an
SBT of 30 t o 120 minut es durat ion should be considered f or separat ion f rom t he
vent ilat or. Pat ient s successf ully complet ing a 2-hour t rial have a 90% chance of
st aying off mechanical vent ilat ion at 48 hours. Pat ient s w ho f ail should be placed
on a mode of vent ilat ion t hat w ill provide rest and be reassessed f or w eaning in
24 hours. O t her considerat ions should include t he amount and qualit y of
secret ions combined w it h t he pat ient 's abilit y t o prot ect t he airw ay and t he
pot ent ial f or airw ay edema.
What to Do
Many paramet ers have been proposed and used t o assess pat ient s f or
discont inuat ion of mechanical vent ilat ion. Commonly used maneuvers t o t est
vent ilat ory muscle st rengt h include measurement of t he peak inspirat ory
pressure or negat ive inspirat ory f orce (NI F) and t he vit al capacit y (VC). Peak
inspirat ory pressure is perf ormed by measuring t he pressure generat ed by t he
pat ient w it h maximal inspirat ory eff ort s against a closed shut t er over 20
seconds. Values of negat ive 20–30 cm H2 O are t hought t o be predict ive of
success. This t echnique predominant ly t est s isomet ric muscle st rengt h as
opposed t o endurance, and t he predict ive value is poor. Vit al-capacit y
maneuvers may give some indicat ion of reserve capacit y. Vit al-capacit y values of
great er t han 1 lit er or great er t han t w o t imes t he t idal volume suggest suff icient
reserve f unct ion t o at t empt an SBT. The poor predict ive value of individual t est s
like t he NI F and VC has led t o t he development of int egrat ive indices. The rapid
shallow breat hing index (RSBI ; aka Tobin index) is one of t he most w idely used
because of it s ease of calculat ion and improved predict ive value. I t is usef ul as a
screening t ool t o assess pat ient s f or t he abilit y t o proceed t o an SBT. The RSBI
is calculat ed by count ing t he number of breat hs in 1 minut e divided by t he
average t idal volume in lit ers. Values great er t han 105 are predict ive of inabilit y
t o w ean w hile values less t han 80 are associat ed w it h maint ained separat ion
f rom t he vent ilat or at 24 hours. Pat ient s w it h values bet w een 80 and 105 are at
increased risk f or reint ubat ion.
I n summary, no single t est is predict ive of success in discont inuat ion of
mechanical vent ilat ion. Pat ient s should be assessed daily f or abilit y t o w ean and
proceed t o an SBT if indicat ed. Proper implement at ion of prot ocols can aid in
ident if ying pat ient s w ho are ready t o
w ean, t hus decreasing t he complicat ions associat ed w it h bot h unnecessarily

prolonged mechanical vent ilat ion and premat ure separat ion f rom t he vent ilat or.
Suggested Readings
Est eban A, Frut os F, Tobin MJ, et al. A comparison of f our met hods of
w eaning pat ient s f rom mechanical vent ilat ion. Spanish Lung Failure
Collaborat ive G roup. N Engl J Med 1995; 332: 345–350.
MacI nt yre NR. Respirat ory mechanics in t he pat ient w ho is w eaning f rom t he
vent ilat or. Respir Care 2005; 50: 275–286; discussion 284–286.
> Table of C ontents > Ventilator s /Air w ay/Intubation/E xtubation > 131 - C ons ider Us ing Heliox in the
Mec hanic ally Vented As thm atic P atient
131
Consider Using Heliox in the Mechanically
Vented Asthmatic Patient
Anthony D. Slonim MD DrPH
Therapeut ic st rat egies in pulmonary medicine f requent ly employ bronchodilat ors
or st eroids t o decrease airw ay resist ance and improve airf low. As an alt ernat ive
st rat egy, t he viscosit y and densit y of inspired air may be manipulat ed t o achieve
t his same desired eff ect . Air consist s of 21% oxygen, w it h most of t he remainder
being nit rogen. O xygen and nit rogen have nearly equivalent densit ies, and t hus
alt ering t he percent age of FI O2 (f ract ion of inspired oxygen) does not
appreciably change t he physical propert ies of t he gas mixt ure. How ever,
replacing nit rogen w it h t he less-dense helium can dramat ically decrease t he
overall air densit y. This in t urn reduces airw ay resist ance, even w it hout
anat omical changes in airw ay caliber.
There are no evidence-based guidelines direct ing t he use of heliox. How ever,
given t hat heliox does not have any signif icant side eff ect s, t herapeut ic t rials
may be used in pat ient s w it h obst ruct ive upper-airw ay condit ions and respirat ory
compromise. Low -densit y gas mixt ures have t he great est benef icial impact in
large airw ays w it h t urbulent densit y-dependent f low, result ing in improved
alveolar empt ying and reduced plat eau pressures. Considering t his, pat ient s w it h
severe ast hma (requiring mechanical vent ilat ion) may be ideal candidat es f or
heliox t herapy. I ndeed, in t rials of int ubat ed pat ient s w it h st at us ast hmat icus,
t hose t reat ed w it h heliox demonst rat ed decreased airw ay pressures, decreased
PaCO 2 (part ial pressure of carbon dioxide, art erial), and improved acidosis.
Pat ient s w it h ot her severe upper-airw ay diseases such as post ext ubat ion
obst ruct ion or chronic obst ruct ive pulmonary disease may benef it as w ell.
What to Do
I t is recommended t hat heliox t herapy be init iat ed early t o reduce respirat ory
muscle f at igue. Commonly used mixt ures cont ain 80: 20, 70: 30, or 60: 40 helium-
t o-oxygen rat ios. Clinically, t he best st rat egy is t o use t he highest helium cont ent
t hat maint ains an oxygen sat urat ion ≥90%. Using increasing helium
concent rat ions linearly maximizes t he benef it of low -densit y gas inhalat ion.
Conversely, adequat e oxygen f ract ions are necessary t o maint ain alveolar
oxygenat ion. The f avorable eff ect s of heliox t herapy are of t en seen w it hin
minut es. I mport ant ly, vent ilat ors
and inst rument at ion must be calibrat ed f or heliox use. There are published
algorit hms f or t hese calibrat ions.
Suggested Readings
G luck EH, O norat o DJ, Cast riot t a R. Helium-oxygen mixt ures in int ubat ed
pat ient s w it h st at us ast hmat icus and respirat ory acidosis. Chest 1990; 98: 693.
Kass JE, Cast riot t a RJ. Heliox t herapy in acut e severe ast hma. Chest
1995; 107: 757.
Rodrigo G , Rodrigo C, Pollack CV. Use of helium-oxygen mixt ures in t he

t reat ment of acut e ast hma: a syst emat ic review. Chest 2003; 123: 891–896.
> Table of C ontents > Infec tious D is eas e > 132 - C ons ider P ar vovir us B 19 Infec tion in P atients w ith
Anem ia or P anc ytopenia
132
Consider Parvovirus B19 Infection in Patients
with Anemia or Pancytopenia
Jayme E. Locke MD
Parvovirus B19 is a nonenveloped virus w it h single-st randed DNA t hat has a
t ropism f or eryt hroid progenit or cells. The virus replicat es w it hin t hese cells and
is cyt ot oxic, result ing in a pure red cell aplasia. I t is a common viral disorder
t hat is t ransmit t ed via droplet s, t ransplacent ally, and t hrough blood t ransf usions.
I t is acquired early in lif e and is so common t hat great er t han 90% of t he elderly
are seroposit ive. I n general, a normal host clears t he inf ect ion and at most
develops a t ransient clinically insignif icant anemia. This inf ect ion, how ever,
becomes hemat ologically signif icant in pat ient s w it h chronic hemolyt ic anemias,
such as sickle cell, t halassemia, and congenit al spherocyt osis; in pat ient s w it h
acut e malaria; and in pat ient s w it h human immunodef iciency virus/ acquired
immunodef iciency syndrome (HI V/ AI DS).
Parvovirus B19 can pose considerable risk t o a crit ically ill pat ient . Any pat ient
w it h pancyt openia or red cell aplasia of unknow n origin should be t est ed f or
parvovirus B19 inf ect ion. The virus can be det ect ed in a pat ient 's blood using a
dot -plot analysis and conf irmed w it h DNA hybridizat ion. The virus can also be
det ect ed in bone-marrow aspirat es, w here giant proeryt hroblast s w it h
eosinophilicn uclear inclusion bodies are seen. O nce diagnosed, parvovirus B19
is easily t reat ed w it h administ rat ion of immunoglobulin. Treat ment result s in
prompt ret iculocyt osis and resolut ion of t he anemia.
Suggested Readings
St rickland G T, ed. Hunt er's Tropical Medicine and Emerging I nf ect ious
Diseases. 8t h ed. Philadelphia: WB Saunders; 2000: 37.
Young NS, Abkow it z JL, Luzzat t o L. New insight s int o pat hophysiology of
acquired cyt openias. Hemat ology (Am Soc Hemat ol Educ Program).
2000: 18â 3 1.
> Table of C ontents > Infec tious D is eas e > 133 - C ons ider P r ophylac tic Antibiotic s w hen Leec hes ar e
Applied to Fr ee Flaps
133
Consider Prophylactic Antibiotics when Leeches
are Applied to Free Flaps
Benjamin A. Mandel MD
The use of leeches f or medicinal purposes dat es back at least 2, 000 years. This
pract ice peaked in t he 18t h and 19t h cent uries w hen pract it ioners employed t he
parasit es t o t reat pat hologies including gout , headaches, ment al illness, obesit y,
and w hooping cough among ot hers. Leech t herapy rapidly declined by t he
beginning of t he 20t h cent ury w hen t he int ervent ion w as f ound t o have lit t le
recognizable benef it .
More recent ly, how ever, modern medicine has seen a resurgence in t he
applicat ion of leeches. Surgeons now commonly use t he parasit es t o t reat acut e
venous congest ion af t er f ree-t issue t ransf er or replant at ion. This microvascular
complicat ion st ems f rom inadequat e venous drainage f rom t he t issue in quest ion.
Q uit e lit erally, blood backs up and compromises art erial inf low, w hich leads t o
hypoxia and ult imat ely t o surgical f lap demise. Signs of acut e venous congest ion
include dusky or bluish-purple discolorat ion, edema, brisk capillary ref ill, and
relat ive t issue w armt h.
To combat t his venous congest ion, medicinal leeches (Hi rudo medi ci nal i s) are
applied t o t he compromised t issue, allow ing t he collect ed blood an alt ernat ive
rout e of egress. This decongest ion allow s t he ret urn of art erial f low and
prevent s t issue ischemia and necrosis. Several leeches w ill consume f rom 5 t o
15 mL of blood over approximat ely 25 minut es. Af t er parasit e det achment , up t o
anot her 50 mL of blood may exit f rom t he bit e w ound; it is t his secondary blood
loss t hat provides t he major benef it f or t issue salvage. I n large-volume
bloodlet t ing, t ransf usion may be necessary.
The leech salivary gland secret es several subst ances t hat f acilit at e t he
benef icial eff ect s of t his t herapy. Leech saliva cont ains ant icoagulant s including
hirudin, a select ive t hrombin inhibit or, w hich prevent t hrombus f ormat ion. I n
addit ion, an anest het ic component renders t he bit e painless, and a hist aminelike
vasodilat or increases blood f low t o t he parasit e. I t is generally accept ed t hat t he
hirudin is largely responsible f or t he t herapeut ic post det achment blood loss.
The leech gut cont ains Aeromonas hydrophi l a, a gram-negat ive anaerobic
bacillus t hat aids in t he digest ion of consumed blood. This penicillin-resist ant
bact erium can be t ransmit t ed t o t he host via regurgit at ion of gut cont ent s,
leading t o cellulit is, abscess f ormat ion, f lap
f ailure, and sepsis. For t his reason, many experienced t issue-t ransf er surgeons
advocat e a 10- t o 14-day ant ibiot ic prophylaxis f or leech t herapy w it h
t rimet hoprim-sulf amet hoxazole, f luoroquinolones, or a t hird-generat ion
cephalosporin, alt hough t here are f ew high-qualit y clinical t rials support ing t his
pract ice.
Suggested Readings
Clark NM, Femino JE, Chenow et h CE. Aeromonas inf ect ion af t er medicinal
leech t herapy: case report s and review of t he lit erat ure. I nf ect Dis Clin Pract .
2001; 10(4): 221–218.
Weinf eld AB, Yuksel E, Bout ros S, et al. Clinical and scient if ic considerat ions
in leech t herapy f or t he management of acut e venous congest ion: an updat ed
review. Ann Plast Surg. 2000; 45(2): 207–212.
> Table of C ontents > Infec tious D is eas e > 134 - Tr eat Methic illin- R es is tant S taphyloc oc c us Aur eus
w ith a Minim um of 14 days of Antibiotic s
134
Treat Methicillin-Resistant Staphylococcus
Aureus with a Minimum of 14 days of Antibiotics
There are approximat ely 2 million nosocomial inf ect ions in t he Unit ed St at es
each year. About 30% of t hese inf ect ions are by Staphyl ococcus aureus, of
w hich 40% are met hicillin-resist ant S. aureus (MRSA). I n int ensive care unit s
(I CUs), t he percent age of ant ibiot ic-resist ant St aph inf ect ions approaches 60%.
The major reservoir of MRSA in inst it ut ions is colonized and inf ect ed inpat ient s.
Nasal carriage of MRSA in I CU pat ient s w as associat ed w it h an MRSA
bact eraemia rat e of 38%, f our t imes higher t han met hicillin-sensit ive S. aureus.
O ne-t hird of colonized pat ient s become inf ect ed and one-half of t hese have
pneumonia or bloodst ream inf ect ion. Mort alit y rat es f or nosocomial-acquired
MRSA inf ect ions may reach 50% f or bloodst ream inf ect ions and 33% f or
pneumonia.
What to Do
Vancomycin (a glycopept ide) is t he st andard t reat ment f or MRSA. Vancomycin
15 mg/ kg q 12h, or usually 1g qd, is administ ered int ravenously f or 14 t o 21
days. Recent randomized cont rol st udies have show n linezolid t o be as
eff icacious as vancomycin and w it h a predict able bact eriost at ic act ivit y f or
suspect ed or proven MRSA inf ect ions. Linezolid, an oxazolidinone ant ibiot ic, is
available in an int ravenous and an oral f orm, w hich is 100% bioavailable.
Linezolid's dose is 600 mg int ravenously or orally (I V/ PO ) q 12h. How ever,
vancomycin is st ill pref erred by some aut horit ies in order t o increase longevit y of
linezolid's act ivit y. I n Europe, t eicoplanin (also a glycopept ide) is commonly
used.
I n cases of vancomycin-t reat ment f ailure or allergy, ot her drugs used f or t reat ing
MRSA bact eremia are quinuprist in-dalf oprist in, 7. 5 mg/ kg I V q12h and
dapt omycin, 6 mg/ kg I V qd.
MRSA bact eremia should be t reat ed w it h at least a 14-day course of ant ibiot ics
if t here is prompt (w it hin 72 hours) clinical response. I f t here is persist ent
bact eremia (more t han 3 days), some advocat e 4 w eeks of t reat ment even w it h
a negat ive echocardiogram. I n t he case of endocardit is, t he t reat ment course
should increase t o at least
6 w eeks. Recurrence or relapse rat es are high if pat ient s are t reat ed f or less
t han t he suggest ed course.
Suggested Readings
Cepeda JA, Whit ehouse T, Cooper B, et al. Linezolid versus t eicoplanin in t he
t reat ment of G ram-posit ive inf ect ions in t he crit ically ill: a randomized,
double-blind, mult icent re st udy. J Ant imicrob Chemot her. 2004; 53(2): 345–355.
Chaix C, Durand-Zaleski I , Albert i C, et al. Cont rol of endemic met hicillin-

resist ant Staphyl ococcus aureus: a cost -benef it analysis in an int ensive care
unit . JAMA. 1999; 282: 1745–1751.
Haddadin AS, Fappiano SA, Lipset t PA. Met hicillin resist ant Staphyl ococcus
aureus (MRSA) in t he int ensive care unit . Post grad Med J. 2002; 78(921): 385–
392.
Li JZ, Willke RJ, Rit t enhouse BE, Rybak MJ. Eff ect of linezolid versus
vancomycin on lengt h of hospit al st ay in pat ient s w it h complicat ed skin and
sof t t issue inf ect ions caused by know n or suspect ed met hicillin-resist ant
st aphylococci: result s f rom a randomized clinical t rial. Surg I nf ect (Larchmt ).
2003; 4(1): 57–70.
Low y FD. Staphyl ococcus aureus inf ect ions. N Engl J Med. 1998; 339: 520–
552.
Merrer J, Sant oli F, Appere de Vecchi C, et al. “Colonizat ion pressure” and
risk of acquisit ion of met hicillin-resist ant Staphyl ococcus aureus in a medical
int ensive care unit . I nf ect Cont rol Hosp Epidemiol. 2000; 21: 718–723.
Pujol M, Pena C, Pallares R, et al. Nosocomial Staphyl ococcus aureus

bact eremia among nasal carriers of met hicillin-resist ant and met hicillin-
suscept ible st rains. Am J Med. 1996; 100: 509–516.
Romero-Vivas J, Rubio M, Fernandez C, et al. Mort alit y associat ed w it h

nosocomial bact eremia due t o met hicillin-resist ant Staphyl ococcus aureus.
Clin I nf ect Dis. 1995; 21(6): 1417–1423.
Sharpe JN, Shively EH, Polk HC Jr. Clinical and economic out comes of oral
linezolid versus int ravenous vancomycin in t he t reat ment of MRSA-
complicat ed, low er-ext remit y skin and sof t -t issue inf ect ions caused by
met hicillin-resist ant Staphyl ococcus aureus. Am J Surg. 2005; 189(4): 425–
428.
St evens DL, Herr D, Lampiris H, et al. Linezolid versus vancomycin f or t he

t reat ment of met hicillin-resist ant Staphyl ococcus aureus inf ect ions. Clin I nf ect
Dis. 2002; 34(11): 1481–1490.
Vincent JE, Bojaro DJ, Sut er PM, et al. The prevalence of nosocomial
inf ect ion in int ensive care unit s in Europe. JAMA. 1995; 274: 639–644.
Weigelt J, I t ani K, St evens D, et al. Linezolid versus vancomycin in t reat ment

of complicat ed skin and sof t t issue inf ect ions. Ant imicrob Agent s Chemot her.
2005; 49(6): 2260–2266.
Weigelt J, Kaaf arani HM, I t ani KM, et al. Linezolid eradicat es MRSA bet t er
t han vancomycin f rom surgical-sit e inf ect ions. Am J Surg. 2004; 188(6): 760–
766.
> Table of C ontents > Infec tious D is eas e > 135 - B e Aler t for Thr om boc ytopenia and Neutr openia w ith
Linez olid
135
Be Alert for Thrombocytopenia and Neutropenia
with Linezolid
Shaytone Nichols MD
Linezolid is an ant ibiot ic in t he class oxazolidinones and it is eff ect ive against all
G ram-posit ive inf ect ions. I t is usually reserved f or resist ant G ram-posit ive
organisms, especially vancomycin-resist ant Enterococcus (VRE) and met hicillin-
resist ant Staphyl ococcus aureus (MRSA). I t is also somet imes considered if
out pat ient t reat ment w it h oral medicines is required.
Linezolid exhibit s a mechanism of act ion t hat is diff erent t han any ot her
ant imicrobial class, t hus making cross-resist ance w it h ot her ant ibiot ics
uncommon. The drug w orks by inhibit ing t he bact erial t ranslat ion process.
Linezolid binds t o t he 23S pept idylt ransf erase of t he 50S subunit , t hus
prevent ing t he f ormat ion of a f unct ional 70S init iat ion complex, an essent ial st ep
in t he bact erial t ranslat ional process. This prevent s bact eria f rom mult iplying.
Watch Out For

The most signif icant side eff ect of linezolid is myelosuppression, causing
t hrombocyt openia, neut ropenia, and anemia. The t hrombocyt openia is reversible
upon discont inuat ion of t he drug. I n phase I I I t rials, t he rat e of t hrombocyt openia
w as f ound t o be 2. 4%. How ever, recent prospect ive and ret rospect ive st udies
have f ound rat es as high as 47% in pat ient s t reat ed f or longer t han 10 days w it h
t his ant ibiot ic. There have also been recent report s of t oxic opt ic neurit is in
pat ient s receiving linezolid f or 10 mont hs. These sympt oms also resolved w it h
discont inuat ion of t he ant ibiot ic. I n addit ion, since linezolid possesses
monoamine oxidase inhibit or (MAO I ) act ivit y, it should not be administ ered w it h
adrenergic and serot onergic agent s such as t he ant idepressant select ive
serot onin reupt ake inhibit ors (SSRI s).
I f linezolid must be st opped, t here are several alt ernat ive ant imicrobials current ly
available t o t reat resist ant G ram-posit ive organisms. Q uinuprist in-dalf oprist in
w as approved in 1999 and is a st rept ogramin ant ibiot ic w it h act ivit y against
G ram-posit ive organisms. I t is eff ect ive against bot h met hicillin- and vancomycin-
resist ant S. aureus and it can be used t o t reat vancomycin-resist ant
Enterococcus f aeci um but not vancomycin-resist ant E. f aecal i s. Resist ance
develops rapidly w it h it s use and as a result is a less reliable ant imicrobial t han
linezolid.
Dapt omycin (t he only ant ibiot ic in t his class) w as recent ly approved in 2003 and
has show n good ut ilit y. I t is a cyclic lipopept ide and is eff ect ive against all
G ram-posit ive microbes including VRE, MRSA, and VRSA. I n 2005 t igecycline
w as approved and has ext ended spect rum coverage including all G ram posit ives
and resist ant G ram posit ives. I t also has act ivit y against anaerobes and G ram
negat ives, including ext ended-spect rum bet a-lact amases.
Suggested Readings
Rao N, Ziran B, Wagener MM, et al. Similar hemat ologic eff ect s of long t erm
linezolid and vancomycin t herapy in a prospect ive observat ional st udy of
pat ient s w it h ort hopedic inf ect ions. Clin I nf ect Dis. 2004; 38: 1058–1064.
Senneville E, Legout L. Risk f act ors f or anemia in pat ient s on prolonged

linezolid t herapy f or chronic ost eomyelit is: a case-cont rol st udy. J Ant imicrob
Chemot her. 2004; 54: 798–802.
> Table of C ontents > Infec tious D is eas e > 136 - Have a High Thr es hold for Us ing C as pofungin and
Vor ic onaz ole in P atients w ith Liver D is eas e
136
Have a High Threshold for Using Caspofungin
and Voriconazole in Patients with Liver Disease
Leo Hsiao DO
Caspof ungin is a member of a class of ant if ungals know n as echinocandins. This
f amily of ant if ungal exert s it s eff ect by dest abilizing t he f ungal cell w all t hrough
noncompet it ive inhibit ion of β -(1, 3)-D-glucan synt hase. I nhibit ion of t his enzyme
prevent s t he synt hesis of glucan, w hich is a key component of cell-w all st abilit y.
Not surprisingly, t he relat ive amount of β -(1, 3)-D-glucan making up t he cell w all
of various f ungal species correlat es t o it s suscept ibilit y t o caspof ungin. O f not e,
caspof ungin's ant if ungal mechanism limit s it s act ivit y t o branch point s w here
t here is new cell-w all synt hesis.
Caspof ungin is eff ect ive against a w ide range of f ungi including Aspergi l l us
species (f umi gatus, f l avus, terreus), Pneumocysti s cari ni i, and Candi da
species. I t s eff icacy has been evaluat ed in t he t reat ment of invasive candidiasis
w here caspof ungin demonst rat ed a 74% success rat e versus a 62% success rat e
of convent ional t herapy w it h amphot ericin B. Because of limit at ions in
f ormulat ion, caspof ungin's role in t he t reat ment of oral or esophageal candidiasis
should be limit ed eit her t o pat ient s w ho cannot t olerat e azole t herapy or t o
cases of document ed f ailure t o azole t herapy. I n t he set t ing of empiric t reat ment
of f ebrile neut ropenia, caspof ungin has been show n t o be as eff ect ive as
liposomal amphot ericin B and w it h f ew er side eff ect s. Last ly, caspof ungin w as
init ially approved f or salvage t herapy f or t he management of invasive
aspergillosis. Alt hough t here are some dat a t o support it s eff icacy as primary
t herapy f or invasive aspergillosis, t here is concern t hat use of a f ungist at ic agent
in immunocompromised pat ient s might lead t o relapse of disease.
Watch Out For

Common adverse react ions t o caspof ungin include f ever, nausea, vomit ing, and
venous irrit at ion. Caspof ungin has also been show n t o cause rash,
bronchoconst rict ion, w heezing, and f acial edema and f lushing, denot ing a
probable correlat ion t o hist amine release. Alt erat ions in liver f unct ion t est s are
most ly t ransient ; how ever, in a pat ient w it h moderat e pre-exist ing liver disease,
def ined as a Child-Pugh score of 7 t o 9, dosage adjust ment s should be
inst it ut ed. Since t here are no good
dat a available t o support t he saf et y of caspof ungin in pat ient s w it h Child-Pugh

scores >9, it may prudent t o limit t he use of caspof ungin in t his pat ient
populat ion. Also, because caspof ungin is met abolized by N-acet ylat ion in t he
liver, it s use in cirrhot ic pat ient s pot ent ially lacking t his essent ial met abolic
pat hw ay should be judicious.
Voriconazole is a second-generat ion t riazole derivat ive of f luconazole. I t exert s
it s ant if ungal act ivit y by inhibit ing t he cyt ochrome P-450 enzyme 14-α
demet hylase. This inhibit ion prevent s t he conversion of lanost erol t o ergost erol.
Ergost erol is an import ant const it uent of t he f ungal cell membrane and t heref ore
it s absence, coupled w it h t he accumulat ion of t oxic st erol precursors, ult imat ely
leads t o f ungal demise.
The spect rum of act ivit y and clinical eff icacy of voriconazole has been w ell
document ed in bot h in vit ro t est ing and clinical t rials. For example, voriconazole
is highly eff ect ive against most st rains of Candi da despit e indicat ions of pot ent ial
cross-resist ance w it h f luconazole among cert ain isolat es. I n t he t reat ment of oral
candidiasis, voriconazole has proven t o be as eff icacious as f luconazole.
Moreover, t here is evidence t hat voriconazole is eff ect ive against invasive
candidiasis bot h as primary and salvage t herapy and t hat it s saf et y prof ile is
superior t o t hat of amphot ericin B. Likew ise, voriconazole can be considered an
alt ernat ive t o amphot ericin B f or t he t reat ment of Cryptococcus neof ormans.
Voriconazole demonst rat es impressive in vit ro act ivit y against Aspergi l l us
species, even against isolat es resist ant t o it raconazole and amphot ericin B.
These dat a have been borne out in clinical t rials in w hich voriconazole has show n
t o out perf orm amphot ericin B against invasive aspergillosis. Last ly, voriconazole
has show n promise against less common f ungal inf ect ions including t he
Scedospori um, Pseudal l escheri a, and Fusari um species.
Watch Out For

Adverse react ions t o voriconazole w ere evaluat ed in recent clinical t rials. The
most common react ions w ere visual dist urbances (blurry vision, alt ered color
percept ion, and phot ophobia), rash, and alt erat ions in liver-f unct ion t est s. O t her
react ions include hypoglycemia, pneumonit is, w orsening psoriasis, and visual or
audit ory hallucinat ions. O f not e, up t o 20% of st udy part icipant s developed
elevat ion in liver-f unct ion t est s (def ined as t hree t imes t he upper limit of normal).
Alt hough most of t hese elevat ions normalized af t er discont inuat ion of drug, t here
have been report s of hepat ic f ailure and deat h associat ed w it h
t he use of voriconazole. I t is t heref ore prudent t o consider pre-exist ing liver

disease as a relat ive cont raindicat ion t o voriconazole t herapy.
I n addit ion t o caspof ungin ot her members of t he echinocandin f amily of drugs
include micaf ungin and anidulaf ungin. These drugs may be considered
alt ernat ives t o caspof ungin alt hough f urt her st udies need t o be done t o bet t er
def ine t heir clinical role. The same can be said of voriconazole w it h respect t o
it s sist er second-generat ion azole derivat ives posaconazole and ravuconazole.
Last ly, combinat ion t herapy ut ilizing bot h voriconazole and caspof ungin/ liposomal
amphot ericin B has show n promise in t he set t ing of invasive aspergillosis, giving
yet anot her alt ernat ive t o single-agent t herapy.
Suggested Readings
Kof la G , Runhke M. Voriconazole: review of a broad spect rum t riazole
ant if ungal agent . Exp O pin Pharmacot her. 2005; 6(7): 1215–1229.
Pacet t i S, G elone S. Caspof ungin acet at e f or t reat ment of invasive f ungal

inf ect ions. Ann Pharmacot her. 2003; 37(1): 90–98.
Zaas A, Alexander B. Echinocandins: role in ant if ungal t herapy, 2005. Exp

O pin Pharmacot her. 2005; 6(10): 1657–1668.
> Table of C ontents > Infec tious D is eas e > 137 - D o Not Us e C as pofungin or Vor ic onaz ole to Tr eat
Yeas t in the Ur ine B ec aus e a Ver y S m all Am ount of Thes e D r ugs ar e E xc r eted in the Ur ine
137
Do Not Use Caspofungin or Voriconazole to
Treat Yeast in the Urine Because a Very Small
Amount of These Drugs are Excreted in the
Urine
Harjot K. Singh MD
Lesia K. Dropulic MD
Candiduria, or yeast in t he urine, is commonly encount ered in t he int ensive care
unit (I CU) set t ing. Risk f act ors f or candiduria include increased age, f emale sex,
urinary drainage cat het ers, ant ibiot ic use, prior surgical procedures, and
diabet es mellit us. Det ect ion of t he Candi da species of yeast in t he urine can
represent colonizat ion or inf ect ion and dist inguishing bet w een t he t w o can be
diff icult because of t he lack of appropriat e diagnost ic t est s. Clinical f indings
used t o diagnose bact erial inf ect ions of t he low er urinary t ract are not applicable
t o candiduria because pat ient s are of t en asympt omat ic and yeast colony count s
and pyuria do not correlat e w it h inf ect ion. Most of t he t ime, candiduria is a
benign process. How ever, caut ion is w arrant ed because candiduria may
represent disseminat ed or upper urinary t ract inf ect ion in cert ain high-risk I CU
pat ient s.
The I nf ect ious Diseases Societ y of America's guidelines f or t reat ment of
candidiasis recommend ant if ungal t reat ment of candiduria in t he f ollow ing
circumst ances: sympt omat ic pat ient s; pat ient s w ho w ill undergo urinary t ract
manipulat ion; pat ient s w it h neut ropenia; pat ient s w it h renal allograf t s; and
inf ant s w it h low birt h w eight . Treat ment of t hese select pat ient s might reduce t he
risk of ascending inf ect ion and/ or disseminat ed inf ect ion or might t reat occult
disseminat ed candidiasis. I n addit ion, ant if ungal t reat ment should be considered
f or pat ient s w it h candiduria w ho have urinary t ract obst ruct ion or an abnormal
urinary t ract , pat ient s w ho are t o have surgery t o implant prost het ic devices
[ e. g. , valves, joint s, vent ricular assist devices (VADs)] , and crit ically ill pat ient s
w ho already have implant ed VADs.
Fluconazole, dose adjust ed f or renal insuff iciency, is t he ant if ungal agent of
choice f or t reat ment of f luconazole-suscept ible Candi da species isolat ed f rom
urine. Seven t o 14 days of t herapy is recommended and durat ion depends on t he
ext ent of sympt oms and on t he severit y of t he underlying disease. C. al bi cans is
t he most common isolat e and is suscept ible t o f luconazole. C. gl abrata isolat es
might exhibit dose-dependent suscept ibilit y or may be resist ant t o f luconazole.
C. krusei is int rinsically resist ant t o f luconazole. For such
f luconazole-resist ant isolat es, amphot ericin B deoxycholat e is recommended at a

range of doses and f or a range of days (0. 3 t o 1. 0 mg/ kg per day f or 1 t o 7
days).
Caspof ungin and voriconazole are not recommended f or t he t reat ment of
candiduria because a very small amount of t hese drugs is excret ed in t he urine.
I n addit ion, bladder irrigat ion w it h amphot ericin only t ransient ly clears candiduria
and is not recommended. Relapse of candiduria is f requent af t er ant if ungal
t reat ment and commonly occurs w it h cont inued use of a urinary cat het er or
st ent . Discont inuat ion of cat het er use alone may result in eradicat ion of
candiduria in up t o 40% of pat ient s. I f possible, cat het ers and st ent s should be
removed concomit ant ly w it h inst it ut ion of ant if ungal t herapy.
Persist ent candiduria in immunocompromised pat ient s requires evaluat ion of t he
kidneys w it h ult rasound or comput ed t omography scan. Since candiduria can be
a manif est at ion of disseminat ed inf ect ion, blood cult ures should be obt ained in
pat ient s w it h syst emic signs and sympt oms.
Suggested Readings
Kauff man CA. Candiduria. Clin I nf ect Dis. 2005; 41 (suppl 6): S371–S376.
Pappas PG , Rex JH, Sobel JD, et al. G uidelines f or t reat ment of candidiasis.
Clin I nf ect Dis. 2004; 38: 161–189.
> Table of C ontents > Infec tious D is eas e > 138 - Adm inis ter a D os e of Antibiotic B efor e the B ile
S ys tem is Ins tr um ented or Manipulated
138
Administer a Dose of Antibiotic Before the Bile
System is Instrumented or Manipulated
Kelly O lino MD
Bile duct s are normally st erile. How ever, in cases w here t he duct s become
obst ruct ed or inst rument ed, t he presence of bact erobilia w it h some series
report ing 90% w it h posit ive bile cult ures. The most common organisms f ound are
Escheri chi a col i , Kl ebsi el l a, Enterobacter, Enterococcus, Pseudomonas,
Candi da species, gramposit ive cocci, bact eroides, and Cl ostri di um. I n addit ion,
in immunocompromised populat ions, one should keep in mind unusual organisms
such as crypt osporidium and in pat ient s f rom t ropical locat ions, parasit es such
as Ascari s.
What to Do
When a pat ient is undergoing a biliary t ract procedure such as percut aneous
t ranshepat ic cholangiography (PTC), endoscopic ret rograde cholangiopan-
creat ography (ERCP), or open or laparoscopic surgical operat ions,
preprocedural ant ibiot ic t reat ment is indicat ed. When select ing appropriat e
ant ibiot ic coverage it is crucial t o look at local inst it ut ional microbiological
prof iles as w ell as ident if ying risk f act ors f or bact erobilia. For low -risk pat ient s,
namely t hose w ho are not severely ill, have not had previous biliary procedures,
and have communit y-acquired illness, a one-t ime preprocedural t reat ment w it h 1
g of cef ot et an int ravenously (I V) or a f luoroquinolone I V w it h addit ional
anaerobic coverage w it h met ronidazole 500 mg I V is suff icient , w it h no proven
benef it f or post procedural ant ibiot ics. I n pat ient s w ho are severely ill, w it h
previous inst rument at ion or hospit al-acquired illness, broader-spect rum
t reat ment w it h piperacillin/ t azobact am 3. 375 mg I V or a f luoroquinolone w it h
addit ional anaerobic coverage is indicat ed. A one-t ime dose is suff icient , except
in cases w here adequat e drainage is not achieved or in cases of sepsis or
cholangit is, w here anyw here f rom 5 t o 14 days of t reat ment may be required.
To brief ly review, cholangit is may present secondary t o obst ruct ion or af t er
biliary procedures, w it h up t o a 1% t o 3% incidence af t er ERCP. Common
et iologies leading t o obst ruct ion include gallst ones, neoplasms, benign
st rict ures, post operat ive st rict ure, sclerosing cholangit is, papillary st enosis, and
malf unct ion of indw elling
st ent s. There is a broad spect rum of present ing signs and sympt oms of
cholangit is, ranging f rom around t w o-t hirds of pat ient s w it h at least t w o
sympt oms of Charcot t riad (right upper quadrant pain, f ever, jaundice) t o 5% of
pat ient s w it h Reynold's pent ad (Charcot 's t riad plus shock and change in ment al
st at us). Most pat ient s w ill have an elevat ed w hit e blood cell, jaundice, mild right
upper quadrant pain, and abnormalit ies in liver-f unct ion t est s. The t reat ment of
cholangit is is cent ered around appropriat ely t imed drainage of t he inf ect ed bile,
I V f luid hydrat ion, and ant ibiot ic t reat ment , and t he mort alit y rat e has been
described t o be 5%.
Suggested Readings
Cameron JL, ed. Current Surgical Therapy. 8t h ed. New York: Mosby; 2004.
Lipset t PA, Pit t H. Acut e cholangit is. Front Biosci. 2003; (Sept . 1): 1229–
1239.
The American Societ y f or G ast roint est inal Endoscopy. Complicat ions of ERCP.
G ast roint est Endosc. 2003; 57(6): 633–638.
> Table of C ontents > Infec tious D is eas e > 139 - Adm inis ter an Antibiotic B efor e Ur inar y Tr ac t
O bs tr uc tion is R elieved
139
Administer an Antibiotic Before Urinary Tract
Obstruction is Relieved
Bact eriuria can become bact eremia w it h associat ed syst emic complicat ions in
t he set t ing of urinary t ract obst ruct ion. Most bact eria ent er t he urinary t ract
syst em t hrough t he uret hra f rom t he f ecal reservoir. I n addit ion, t he kidneys can
be secondarily inf ect ed by Staphyl ococcus aureus or Candi da by hemat ogenous
spread. Escheri chi a col i (most common), Proteus, Kl ebsi el l a, and Enterococcus
f aecal i s are f requent ly f ound t o cause inf ect ion in t he urinary t ract .
When t here is an increased pressure in t he upper urinary t ract , pyelovenous
backf low may allow inf ect ed urine t o ent er t he bloodst ream. I n addit ion, t here is
a change in renal blood f low and decreased neut rophil delivery. This is t he
proverbial “pus under pressure. ” Clinical signs and sympt oms of urinary t ract
obst ruct ion are dependent on t he t ime course of t he obst ruct ion, w het her t he
obst ruct ion is complet e or unilat eral, and t he cause of t he obst ruct ion. Frequent
sympt oms include f lank pain, nausea, and vomit ing. Some clinical signs include
rising blood urea nit rogen (BUN) or creat inine, elect rolyt e abnormalit ies, and
hypert ension.
What to Do
Urinary t ract inf ect ion in a set t ing of urinary obst ruct ion is an emergency. The
pat ient should be hydrat ed and st art ed on int ravenous (I V) ant ibiot ics as soon as
possible. I f t he pat ient is sept ic, broad-spect rum ant ibiot ics such as ampicillin (2
gm I V q6) and gent amicin (1. 5 mg/ kg q8) or a f luoroquinolone should be given as
soon as blood cult ures and urine cult ure are sent . I n any pat ient w it h a
suspect ed urinary t ract obst ruct ion, a Foley cat het er should be placed t o provide
maximal urinary drainage. O nce st abilized, t he pat ient should undergo eit her
uret eral st ent placement or percut aneous nephrost omy t ube placement . O nce t he
obst ruct ion is relieved, t he pat ient should cont inue t o be observed closely f or
signs of sepsis. An obst ruct ed urinary syst em should not be inst rument ed w it hout
t he prot ect ive ant ibiot ics on board, as t he bact erial load released int o t he
bloodst ream can cause acut e and prof ound physiological derangement s.
The pat ient should also bemonit ored f or post obst ruct ive diuresis, w hich might be
manif est ed by polyuria w it h salt w ast ing, hypot ension, or rising BUN. I t is
t ypically def ined as marked polyuria af t er t he
relief of uret eral obst ruct ion. This may be physiologic in w hich case, renal
f unct ion quickly ret urns t o normal and t he diuresis ceases. I n some cases it may
be pat hologic in nat ure w it h cont inued impairment of sodium reabsorpt ion and
impaired renal concent rat ing abilit y. The pat ient s most likely t o exhibit
post obst ruct ive diuresis are t hose w it h a hist ory of congest ive heart f ailure,
hypert ension, edema, and chronic obst ruct ion. Af t er t he release of t he
obst ruct ion, t he pat ient should be monit ored f or urine out put >200 cc per hour
f or more t han 2 hours. I f brisk diuresis cont inues, t he pat ient should have
replacement of half of t he hourly urinary out put w it h D5 ½ normal saline w it h
serial elect rolyt e checks. Typically, BUN and creat inine w ill ret urn close t o
baseline in 24 t o 48 hours.
Suggested Readings
Rubenst ein JN, Schaeff er AJ. Managing complicat ed urinary t ract inf ect ions.
The urologic view. I nf ect Dis Clin N Am. 2003; 17: 333–351.
Walsh PC, ed. Campbell's Urology. 8t h ed. Philadelphia: WB Saunders;

2002: 452–454, 515–516, 3340.
> Table of C ontents > Infec tious D is eas e > 140 - R em em ber that E nter oc oc c us is a R ar e Invas ive
P ulm onar y Tr ac t Infec tion
140
Remember that Enterococcus is a Rare Invasive
Pulmonary Tract Infection
Ent erococci rarely cause low er respirat ory t ract inf ect ions, t hough t hey are
capable of colonizing t he oropharynx. Respirat ory t ract inf ect ions due t o
ent erococci are exceedingly unusual, but case report s of ent erococcal
pneumonia and lung abscess exist in pat ient s w it h severe and debilit at ing
diseases. These rare cases of ent erococcal pneumonia have been described in
t he set t ing of broad-spect rum ant imicrobial t herapy, especially w it h
cephalosporins, along w it h ent eric f eeding in severely debilit at ed pat ient s.
How ever, most experienced clinicians w ill not at t ribut e a pneumonia t o
Enterococcus even if a respirat ory cult ure ret urns result s posit ive f or t his
organism.
I t is know n t hat ent erococci are not as int rinsically virulent as organisms such as
Staphyl ococcus aureus and Streptococcus pyogenes. They do not secret e
exot oxins or produce superant igens. How ever, st udies of pat ient s w it h
ent erococcal bact eremia have show n an at t ribut ed mort alit y of 31% t o 37% and
some st udies suggest vancomycin resist ance t o be an independent predict or of
mort alit y.
Alt hough ent erococci do not generally cause t rouble in t he respirat ory t ract ,
inf ect ion at ot her sit es can be clinically signif icant . Ent erococci can cause
cat het er-relat ed bloodst ream inf ect ions. These organisms adhere t o heart valves
and renal epit helial cells and t hus can cause endocardit is and urinary t ract
inf ect ions. Ent erococci are commonly f ound in cult ures of int ra-abdominal and
pelvic inf ect ions. I t is t hought t hat ent erococci act synergist ically w it h ot her
bact eria in int ra-abdominal sepsis t o enhance morbidit y and mort alit y.
Ent erococci by t hemselves rarely, if ever, cause cellulit is or ot her sof t -t issue
inf ect ions. They are f requent ly isolat ed f rom mixed cult ures in surgical w ound
inf ect ions, decubit us ulcers, and diabet ic f oot inf ect ions, but t heir signif icance is
diff icult t o assess in t hese cases.
I n cont rast t o t he Enterococcus, Streptococcus pneumococcus remains t he
leading cause (16% t o 60%) of acut e communit y-acquired pneumonia (CAP) in
most series. Classically, communit y-acquired pneumonia present s w it h a sudden
onset of a chill f ollow ed by f ever, pleurit ic chest pain, and cough w it h purulent
sput um. Three percent t o 38% of cases of CAP are caused by Haemophi l us
i nf l uenzae, and S. aureus account s f or 2% t o 5% of cases. Aerobic G ram-
negat ive bact eria, mixed
aerobic and anaerobic inf ect ions, and at ypical organisms (Legi onel l a,
Mycopl asma, and Chl amydi a) cause most of t he remaining cases of acut e
communit y-acquired pneumonia.
The clinical f eat ures of nosocomial pneumonia are nonspecif ic and may include
f ever, changing chest radiographs, purulent sput um, and an elevat ed w hit e blood
cell count in a crit ically ill pat ient . Approximat ely 60% of cases of nosocomial
pneumonia are caused by aerobic G ram-negat ive bacilli, w it h members of t he
f amily Ent erobact eriaceae (Kl ebsi el l a pneumoni ae, Escheri chi a col i , Serrati a
marcescens, and Enterobacter species) and Pseudomonas species account ing
f or t he majorit y of t hese. S. aureus causes 13% t o 40% of cases and is more
common in burn unit s, in pat ient s w it h w ound inf ect ions, and in pat ient s on a
vent ilat or af t er neurosurgery or head t rauma. S. pneumoni ae causes only 3% t o
20% of nosocomial pneumonias, mainly early in t he hospit al course. Anaerobic
bact eria play a role w hen aspirat ion is likely t o have occurred. Pneumonia
caused by Legi onel l a species may occur sporadically or as part of out breaks,
especially in immunosuppressed pat ient s. Tw ent y-f ive percent t o 46% of
vent ilat or-associat ed pneumonia may be polymicrobial. Nosocomial viral
inf ect ions are most commonly caused by respirat ory syncyt ial virus, inf luenza,
and parainf luenza, w hich can cause pneumonit is.
Suggested Readings
Mandell G L, Bennet t JE, Dolin R, eds. Mandell, Douglas, and Bennet t 's
Principles and Pract ice of I nf ect ious Diseases. 6t h ed. New York: Churchill
Livingst one; 2005: 831–836, 2411–2417.
> Table of C ontents > Infec tious D is eas e > 141 - K now How to C alc ulate the C linic al P ulm onar y
Infec tion S c or e
141
Know How to Calculate the Clinical Pulmonary
Infection Score
Vent ilat or-associat ed pneumonia (VAP) ref ers t o pneumonia t hat occurs more
t han 48 hours af t er pat ient s are int ubat ed and t reat ed w it h mechanical
vent ilat ion, and it is t he second most common nosocomial inf ect ion in t he Unit ed
St at es. Despit e t his clinical f requency, VAP has been surprisingly diff icult t o
diagnose as t here is no single clinical crit erion t hat is specif ic t o VAP. I n
addit ion, t here is variat ion in t he lit erat ure as t o w hat const it ut es t he clinical
diagnosis of VAP. Depending upon t he st udy review ed, t here have been varying
int erpret at ions of clinical signs and sympt oms suggest ive of lung inf ect ion,
diff erent def init ions of colonizat ion versus inf ect ion, and diff erent use of
ant ibiot ics in t he int ensive care unit (I CU).
Consist ent and t imely diagnosis is no small mat t er as early and appropriat e
t reat ment of VAP has been direct ly linked t o f avorable pat ient out comes.
Conversely, lat e or inappropriat e t herapy has been f ound t o increase risk of in-
hospit al mort alit y by as much as a f act or of seven. Typically, pneumonia is
suspect ed if t he pat ient has a radiographic inf ilt rat e t hat is new or progressive
concurrent w it h clinical f indings suggest ing inf ect ion: new onset of f ever, purulent
sput um, leukocyt osis, or a decline in oxygenat ion. Unf ort unat ely, t hese clinical
signs are nonspecif ic and subject ive. At best , no single radiographic sign has
great er t han 68% diagnost ic accuracy.
I n an eff ort t o enhance diagnost ic sensit ivit y and specif icit y f or VAP, and t o
provide some consist ency w it hin t he lit erat ure, Pugin et al. developed t he clinical
pulmonary inf ect ion score (CPI S). This clinical scoring syst em est ablishes t he
likelihood of VAP based upon several clinical paramet ers t hat can be scored 0,
1, or 2 point s (as adapt ed by Singh et al. ) (Tabl e 141. 1).
When t he t ot al CPI S exceeds 6, t here is a 93% sensit ivit y and a 96% specif icit y
f or t he presence of pneumonia, as def ined by quant it at ive cult ures of
bronchoscopic and nonbronchoscopic bronchoalveolar lavage (BAL) specimens.
The sensit ivit y and specif icit y of t he CPI S does vary t o some degree upon t he
ref erence st andard used f or diagnosis of pneumonia. For inst ance, w hen
compared w it h post mort em quant it at ive lung cult ures as t he ref erence st andard,
t he CPI S has a low er sensit ivit y of 72% t o 77% and specif icit y of 42% t o 85%.
Varying
sensit ivit y and specif icit y aside, t he overall int erpret at ion based upon t hese
st udies is t hat t he CPI S is reasonably accurat e f or t he clinical diagnosis of VAP.
Furt hermore, if t he CPI S is >6 t here should be suspicion of VAP and empiric
ant ibiot ics should be st art ed.
TABLE 141-1 CLINICAL PULM ONARY INFECTION

SCORE (CPIS)
PARAMET ER POINT S
Temperature (C)
36.5–38.4 0
38.5–38.9 1
<36.4 or >39.0 2
W BC Count
4,000–11,000 0
<4,000 or >11,000 1
with >50% bands +1
Tracheal Secretions
None 0
Nonpurulent 1
Purulent 2
Oxygenation: PaO2 /FIO 2 (mm Hg)
>240 or ARDS 0
<240 without ARDS 2
CXR
No infiltrate 0
Diffuse/patchy infiltrate 1
Localized infiltrate 2
Infiltrate Progression
No progression 0
Progression (in absence of CHF or ARDS) 2
ET Suction Culture
No/light growth 0
Heavy growth 1
Same growth on Gram stain +1

TOTAL (Range) 0–14
ARDS, acute respiratory distress syndrome; CHF,

congestive heart failure; CXR, chest x-ray; ET,
endotracheal; FIO2 , fraction of inspired oxygen; W BC,
white blood cell.
I n addit ion t o being used f or t he diagnosis of VAP, t he CPI S has also been used
t o guide t he durat ion of ant ibiot ic t herapy. Singh
et al. explored w het her t he CPI S could serve as a t ool t o limit ant ibiot ic overuse
in t he I CU. Subject s w ho w ere less likely t o have an inf ect ion (CPI S ≤6) w ere
randomized t o eit her “st andard, ” 10- t o 21- day ant ibiot ic t herapy or a 3-day
course of empiric ciprof loxacin, f ollow ed by re-evaluat ion, w it h discont inuat ion of
t reat ment if t he CPI S remained ≤6. O ut come, measured as mort alit y and lengt h
of I CU st ay, w as t he same f or bot h t he short -course group and t he longer-course
group receiving st andard t herapy. I t is post ulat ed t hat t he short er-course
pat ient s w ill have a decrease in superinf ect ions and a decrease in ant ibiot ic
resist ance. Thus, in pat ient s w it h an init ial CPI S ≤6 and an improving clinical
course, longer ant ibiot ic regimens of 10 t o 21 days may not be necessary.
Suggested Readings
American Thoracic Societ y; I nf ect ious Diseases Societ y of America.
G uidelines f or t he management of adult s w it h hospit al-acquired, vent ilat or-
associat ed, and healt hcare-associat ed pneumonia Am J Respir Crit Care Med.
2005; 171(4): 388–416.
Pugin J, Auckent haler R, Mili N, et al. Diagnosis of vent ilat or-associat ed

pneumonia by bact eriologic analysis of bronchoscopic and non-bronchoscopic
“blind” bronchoalveolar lavage f luid. Am Rev Respir Dis. 1991; 143: 1121–1129.
Singh N, Rogers P, At w ood CW, et al. Short -course empiric ant ibiot ic t herapy
f or pat ient s w it h pulmonary inf ilt rat es in t he int ensive care unit : a proposed
solut ion f or indiscriminat e ant ibiot ic prescript ion. Am J Respir Crit Care Med.
2000; 162: 505– 511.
> Table of C ontents > Infec tious D is eas e > 142 - R em em ber that Lac k of P os itive B lood C ultur es does
not R ule O ut B ac ter ial E ndoc ar ditis
142
Remember that Lack of Positive Blood Cultures
does not Rule Out Bacterial Endocarditis
Harjot K. Singh MD
I nf ect ive endocardit is (I E) is generally def ined by veget at ive cardiac lesions in
t he set t ing of posit ive blood cult ures. How ever, t he lack of a single diagnost ic
t est has led t o several diagnost ic algorit hms, of w hich t he Duke crit eria are t he
most w idely accept ed. The Duke crit eria have been show n t o have a specif icit y
of 99% and a negat ive predict ive value of 92%. Despit e t he diagnost ic diff icult y,
w hen inf ect ive endocardit is is suspect ed, empiric t herapy is w arrant ed because
of it s high morbidit y and mort alit y.
Signs and Symptoms

The signs and sympt oms of inf ect ive endocardit is can vary great ly and have a
low sensit ivit y and specif icit y f or inf ect ive endocardit is. Fevers, chills, and
sw eat s are common. Fat igue, syncope, congest ive heart f ailure, and embolic
event s can occur. O n exam, pat ient s can have f ever, new or louder pre-exist ent
murmurs, splint er hemorrhages, Janew ay lesions, O sler nodes, or Rot h spot s.
Risk f act ors endocardit is include presence of prost het ic valves, st ruct ural heart
disease, int ravenous drug use, and indw elling cat het ers.
Def init ive diagnosis using t he Duke crit eria requires f ulf illment of t w o major
crit eria, one major and t hree minor, or f ive minor crit eria.
M ajor Criteria
Tw o separat e posit ive blood cult ures f or a t ypical organism (communit y-
acquired Staphyl ococcus aureus or ent erococci w it hout anot her inf ect ious
f ocus, Streptococcus vi ri dans, Streptococcus bovi s, or HACEK1 group), or
persist ent ly posit ive blood cult ures w it h any organism (e. g. , t w o specimens
draw n 12 hours apart , all of t hree separat e cult ures at least 1 hour apart , or
most of ≥ f our cult ures)
Veget at ion/ abscess/ new prost het ic valve dehiscence on echocardiogram, or
new valvular regurgit at ion (increase or change in preexist ing murmur is not
suff icient )
M inor Criteria
Predisposing condit ion (previous endocardit is, inject ion drug use, prost het ic
valve, st ruct ural heart disease)
Fever >38C
Embolic event
I mmunologic phenomenon (e. g. , O sler nodes, glomerulonephrit is, posit ive
rheumat oid f act or)
Even w it h caref ul diagnost ic t echniques, blood cult ures remain negat ive in
approximat ely 5%–10% of pat ient s w it h inf ect ive endocardit is. Negat ive cult ures
are especially common w hen HACEK group organisms, Coxi el l a burneti i, or
ot her at ypical organisms (Chl amydi a, Mycopl asma, Brucel l a) are responsible
because t hey do not grow in st andard medium or require longer grow t h of 7 t o
21 days. Most laborat ories can t est f or t hese if asked. The t ypical organisms
t hat cause inf ect ive endocardit is include S. aureus, S. bovi s, S. vi ri dans, and
Enterococcus species. G ram-negat ive bacilli and f ungi are less common causes
of endocardit is.
What to Do
Theorganism, valvular locat ion, and presence of prost het ic devices det ermine
t reat ment of inf ect ive endocardit is. A long durat ion of parent eral ant ibiot ics is
necessary because of t he high densit y of micro-organisms w it hin t he veget at ion.
Uncomplicat ed right -sided inf ect ive endocardit is w it h penicillin-sensit ive
st rept ococci can be t reat ed eit her w it h penicillin (12 t o 18 million unit s/ day) and
gent amicin (1 mg/ kg int ravenously q8h) f or 2 w eeks (most ot her regimens
require longer t herapy) or w it h penicillin alone f or 4 w eeks. S. vi ri dans and S.
bovi s are t reat ed w it h penicillin (18 mu/ day) and gent amicin, w hile ent erococci
are t reat ed w it h ampicillin (12 g/ day) and gent amicin f or 4 t o 6 w eeks.
Penicillinallergic pat ient s eit her undergo desensit izat ion or vancomycin is used.
Met hicillin-sensit ive S. aureus is best t reat ed w it h naf cillin or oxacillin 2 g I V q4h
f or 4 t o 6 w eeks, w it h t he opt ional addit ion of gent amicin 1 mg/ kg I V q8h f or 3 t o
5 days. For MRSA or penicillin allergy, vancomycin is used at 15 mg/ kg I V q12h.
I f prost het ic mat erial is present , at least 6 w eeks of t herapy is required w it h
naf cillin, oxacillin, or vancomycin; rif ampin is added f or synergy f or 6 w eeks and
gent amicin is used f or 2 w eeks. Surgical int ervent ion is usually required in
addit ion f or prost het ic valve I E. HACEK organisms are t reat ed w it h cef t riaxone
(2 g/ day) f or 4 w eeks.
Suggested Readings
Bart let t JG , ed. Pocket Book of I nf ect ious Disease Therapy, 2005–6.
Principles and Pract ice of I nf ect ious Diseases. 6t h ed. New York: Churchill
Livingst one; 2005: 975–1002.
> Table of C ontents > Infec tious D is eas e > 143 - Tr eat B lac k Lips or a B lac k S pot on the O r al Muc os a
as a S ur gic al E m er genc y
143
Treat Black Lips or a Black Spot on the Oral
Mucosa as a Surgical Emergency
Mucormycosis is t he common name given t o several diff erent diseases caused
by f ungi of t he order Mucorales. I t is also called zygomycosis af t er t he class
Zygomycet es, w hich are molds t hat grow in t he environment and in t issue as
hyphal f orms. Rhi zopus species are t he most commonly isolat ed agent s of
mucormycosis, f ollow ed by Rhi zomucor and Cunni nghamel l a.
Mucormycosis is most ly limit ed t o people w it h severe immunocompromise,
diabet es mellit us, or t rauma. Solid-organ and hemat opoiet ic st em-cell t ransplant
recipient s represent a grow ing populat ion at risk. More t han half of pat ient s have
rhinocerebral disease; approximat ely 10% have pulmonary, cut aneous, or
disseminat ed disease; and 2% have kidney or gast roint est inal involvement .
Three-quart ers of t ransplant pat ient s w it h mucormycosis also had diabet es or
had received ant ireject ion t herapy. I t is rarely seen in immunocompet ent host s.
Most commonly, t he f ungus gains ent ry t o t he body t hrough t he respirat ory t ract
and is inhaled f rom t he nasal t urbinat es. I n t he case of primary cut aneous
mucormycosis, spores are int roduced direct ly int o abraded skin. They t hen
prolif erat e and can invade more w idely. O nce t he f ungus begins t o grow, t he
hyphae invade t issue and have a special aff init y f or blood vessels. Direct
penet rat ion and grow t h t hrough t he blood-vessel w all explain t he propensit y f or
t hrombosis and t issue necrosis, t w o major hallmarks of t he hist opat hology of
mucormycosis.
Rhinocerebralmucormycosis is most of t en f ound in pat ient s w it h diabet es
mellit us, part icularly in t he presence of acidosis, and in pat ient s w it h leukemia
w ho have been neut ropenic f or long periods and w ho have been receiving broad-
spect rum ant ibact erial drugs. Pat ient s complain of f acial pain and/ or headache,
and f ever and orbit al cellulit is occur. I nvasion of t en involves t he palat e and
perioral t issues, f irst w it h eryt hema and t hen t he development of ulcerat ion
(w here it can mimic a herpet ic lesion), w hich may t urn black, ref lect ing t he
charact erist ic t issue necrosis. I t is import ant t o consider t he possibilit y of
mucormycosis bef ore t he development of t he necrot ic ulcer. Wit h invasion of t he
orbit loss of ext raocular muscle f unct ion, propt osis, and
marked sw elling of t he conjunct iva become evident . Loss of visionmay result

f rom t hrombosis of t he ret inal art ery and t he development of cranial nerve
dysf unct ion is manif est ed by pt osis and pupillary dilat at ion, w hich represent s a
serious prognost ic event . Cerebral abscess, cavernous sinus, and int ernal
carot id art ery t hrombosis are addit ional complicat ions.
Pulmonarymucormycosis is marked by f ever, dyspnea, and cough. Wit h cont inued
t issue necrosis, hemopt ysis may develop; should a major blood vessel be
eroded, f at al pulmonary hemorrhage can result .
The hallmarks of mucormycosis are vascular invasion and t issue necrosis; black
eschars and discharges should be aggressively sought . The presence of a black
nasal discharge should not be dismissed as merely dried blood as it may ref lect
t issue necrosis. Similarly, black necrot ic lesions of t he nasal mucosa or hard
palat e may ref lect invasive mucormycosis. Diagnosis depends on demonst rat ion
of t he organism in t he t issue of a biopsy specimen. Sw abs of discharge or
abnormal t issue are not appropriat e and of t en result in erroneous inf ormat ion.
Fungal hyphae can be seen on pot assium hydroxide preparat ions of t ouch slides
prepared f rom t he biopsy specimen. Fixed t issue can be st ained w it h rout ine or
special f ungal st ains t o demonst rat e f ungal hyphae.
What to Do
I n mucormycosis aggressive correct ion of hyperglycemia and acidemia should be
pursued. I f possible, doses of immunosuppressive drugs, including st eroids,
should be decreased or st opped. The prognosis of t en depends on t he abilit y t o
reverse t he predisposing condit ions. The st andard t herapy f or invasive
mucormycosis is t reat ment w it h amphot ericin B. Because t he f ungus is relat ively
ref ract ory t o medical t reat ment , t he maximum t olerat ed dose of amphot ericin B
deoxycholat e is usually recommended, t ypically 1. 0 t o 1. 5 mg/ kg/ day. High doses
are not usually t olerat ed bef ore renal f unct ion det eriorat es; t heref ore, lipid
preparat ions such as amphot ericin B lipid complex (Abelcet ) and liposomal
amphot ericin B (AmBisome) are recommended at a dose of at least 5 mg/ kg
daily.
None of t he current ly available azoles (ket oconazole, it raconazole, f luconazole,
or voriconazole) or echinocandins has a role in t he t reat ment of mucormycosis. A
new, broad-spect rum t riazole, posaconazole (not yet clinically available), has
been show n t o be act ive in a murine model of mucormycosis. Successf ul use of
posaconazole has been report ed af t er init ial t herapy w it h amphot ericin B and
of t en surgery.
Alt hough report s have appeared in t he lit erat ure of recovery of pat ient s w it h
mucormycosis w it h ant if ungal t herapy alone, t hese are clearly t he except ion and
aggressive surgical debridement of necrot ic t issue is advisable. Repeat ed
operat ions may be required f or sat isf act ory removal of cont inuously appearing
necrot ic t issue. The overall hist orical mort alit y rat e has been about 50%,
alt hough higher survival rat es (up t o 85%) have been report ed more recent ly.
Suggested Readings
Principles and Pract ice of I nf ect ious Diseases. 6t h ed. London: Churchill
Livingst one; 2005: 2973â 2 981.
> Table of C ontents > Infec tious D is eas e > 144 - C hec k for C r yptos por idium in Im m unos uppr es s ed
P atients w ith C hr onic , S ever e, or R efr ac tor y D iar r hea
144
Check for Cryptosporidium in
Immunosuppressed Patients with Chronic,
Severe, or Refractory Diarrhea
Ala’ S. Haddadin MD
Diarrhea, t he principal manif est at ion of int est inal inf ect ion among t he crit ically ill,
aff ect s approximat ely one-t hird of all pat ient s (report ed incidence of 2% t o 63%)
admit t ed t o t he int ensive care unit (I CU). Alt hough many def init ions exist in t he
lit erat ure, diarrhea is best def ined as bow el movement s t hat , because of
increased f requency, abnormal consist ency (t he normal w at er cont ent of st ool is
60% t o 85% of t ot al w eight ), or increased volume (st ool volume out put <250 mL
or 250 g/ day), are delet erious t o t he w ell-being of t he pat ient . Pot ent ially
delet erious consequences in t hese pat ient s include perianal and sacral skin
ulcers w it h secondary superinf ect ions and decreased absorpt ion of ent erally
administ ered medicat ions. For t he crit ically ill pat ient , t he dehydrat ion t hat
accompanies severe diarrhea st rains a circulat ory syst em already limit ed by
impaired cardiac f unct ion and sept ic hemodynamics, w hich might ult imat ely
culminat e in mult isyst em organ f ailure. Diarrhea can also precipit at e met abolic
derangement s including elect rolyt e imbalances and anion gap acidosis w it h all
t heir debilit at ing consequences.
Three main t ypes of diarrhea include secret ory, osmot ic, and inf lammat ory
diarrhea.
Secretory Diarrhea
I n t his disorder, t here is bot h act ive int est inal secret ion of f luids (commonly
described as “w at ery” diarrhea) and elect rolyt es as w ell as decreased
absorpt ion. Common causes of secret ory diarrhea include ent erot oxins (e. g. ,
cholera, rot avirus, Escheri chi a col i); hormones (e. g. , vasoact ive int est inal
pept ide in t he Verner-Morrison syndrome); bile salt s (in t he colon) f ollow ing ileal
resect ion; f at t y acids (in t he colon) f ollow ing ileal resect ion; and laxat ives (e. g. ,
a docusat e sodium). Secret ory diarrhea occurs even w hen t he pat ient is f ast ing
because t he secret ory process is independent of ent eral int ake or t he absorpt ive
process.
Osmotic Diarrhea
This t ype of diarrhea result s f rom consuming nonabsorbable solut es by mout h,
nasogast ric t ube, or nasoent eral t ube. This t ype of diarrhea resolves once t he
osmot ic load is eliminat ed (i. e. , on t he out set of f ast ing). Some medicat ions
causing t his t ype of diarrhea include sorbit ol; a solut ion of polyet hylene glycol
and elect rolyt es (G oLY TELY ); and magnesium-cont aining medicat ions. I t can
also be secondary t o malabsorpt ion and incomplet e digest ion of prot ein
(azot orrhea), f at s (st eat orrhea), or carbohydrat es.
Inflammatory Diarrhea (M ucosal Destruction)

I nf lammat ory diarrhea occurs secondary t o damage t o t he int est inal mucosal cell
so t hat t here is loss of f luid and/ or blood. I n addit ion, t here is def ect ive
absorpt ion of f luids and elect rolyt es. Common causes are inf ect ive condit ions
(e. g. , clost ridial inf ect ions, dysent ery due t o Shi gel l a, part icularly S.
dysenteri ae, and Cryptospori di um) and inf lammat ory condit ions (e. g. ,
inf lammat ory bow el disease). This t ype of diarrhea w orsens af t er f eeding and
persist s af t er f ast ing. Fecal leukocyt es are present .
Crypt osporidiosis is caused by Cryptospori di um parvum, a microscopic one-
celled parasit e t hat can live in t he int est ines of humans, f arm animals, w ild
animals, and pet s. The parasit e is prot ect ed by an out er shell called an oocyst .
This prot ect ive shell allow s it t o survive out side t he body f or an ext ended period.
C. parvum is highly inf ect ious, requiring only 101 t o 103 oocyst s t o cause human
disease (50% inf ect ious dose, 102 ). The oocyst s are inf ect ious immediat ely, and
t he lif e cycle of t he parasit e produces f orms t hat invade t he int est ine. Locat ion
of t he parasit e in t he int est ine is int racellular but ext racyt oplasmic, w hich might
cont ribut e t o t he marked resist ance of C. parvum t o t reat ment . Large numbers
of oocyst s are excret ed and are resist ant t o harsh condit ions, including chlorine
in levels usually applied in w at er t reat ment . Crypt osporidiosis is a f requent
cause of inf ect ious diarrhea, especially in immunocompromised pat ient s w hose
CD4 + lymphocyt e count f alls below 200 cells/ µL. Crypt osporidiosis parasit es are
passed in t he st ool of inf ect ed persons and animals. People get
crypt osporidiosis w hen t hey sw allow t he parasit es. Sources of disease include
parasit e-cont aining st ool, f ood, and w at er.
The mechanism by w hich C. parvum causes diarrhea is not know n f or cert ain, but
it appears t o be a combinat ion of secret ory and malabsorpt ive processes. I n
pat ient s w ho are not immunocompromised, t he inf ect ion is limit ed t o t he jejunum.
I n pat ient s w it h acquired
immunodef iciency syndrome (AI DS), t he ent ire gast roint est inal t ract and
respirat ory t ract may be involved. Af t er an incubat ion period of 7 t o 10 days
(range 5 t o 28 days), t he pat ient develops diarrhea and may have abdominal
cramps and a low -grade f ever. Pat ient s w it h AI DS can have voluminous diarrhea
(up t o 17 L/ d). Biliary inf ect ion in pat ient s w it h AI DS is associat ed w it h right
upper quadrant pain, nausea, and vomit ing. Physical f indings are nonspecif ic.
Temperat ure higher t han 39°C is not charact erist ic of crypt osporidiosis and
w arrant s invest igat ion f or ot her inf ect ions.
There are many diagnost ic t est s ordered f or pat ient s w it h t he possibilit y of
Cryptospori di um. O n st ool microscopy modif ied acid-f ast st aining of st ool show s
red-st ained round oocyst s against a bluegreen background. Whit e and red blood
cells should not be seen in t he st ool. St ool cult ures are usually perf ormed t o rule
out t he presence of ot her bact erial pat hogens: Elevat ed alkaline phosphat ase
and glut amyl t ranspept idase w it hout hyperbilirubinemia are t ypical signs of
biliary inf ect ion. CD4+ lymphocyt e count s predict t he durat ion of disease in
pat ient s inf ect ed w it h human immunodef iciency virus (HI V). When t he count s are
great er t han 200, t he diarrhea is likely t o resolve spont aneously. Abdominal
ult rasound w ill show dilat ed or irregular int rahepat ic and ext rahepat ic bile duct s,
along w it h a t hickened gallbladder in biliary involvement . I dent if icat ion of
Cryptospori di um in bile on endoscopic ret rograde cholangiopancreat ography
conf irms t he diagnosis of crypt osporidiosis. Papillary st enosis may be present .
Result s of ant ibiot ic t reat ment f or crypt osporidiosis in pat ient s w it h AI DS have
been disappoint ing. Most experienced clinicians use a prot ocol involving
at ovaquone, azit hromycin, and/ or paromomycin. Fort unat ely, t he incidence of
crypt osporidiosis is reduced great ly w here HAART (highly act ive ant iret roviral
t herapy) is used w idely. Sympt omat ic t herapy includes t reat ment w it h
ant idiarrheal agent s such as loperamide or diphenoxylat e-at ropine. Treat ment
usually is not required f or crypt osporidiosis in pat ient s w ho are
immunocompet ent and no clinical t rials have convincingly demonst rat ed t he
eff icacy of ant imicrobials. Ent eral nut rit ion f requent ly exacerbat es t he diarrhea.
Suggested Readings
Dionsio D. Crypt osporidiosis in HI V-inf ect ed pat ient s. J Post grad Med.
2002; 48(3): 215–216.
Hunt er PR, Nichols G . Epidemiology and clinical f eat ures of crypt osporidium
in immunocompromised pat ient s. Clin Microbiol Rev. 2002; 15(1): 145–154.
Xiao L, Ryan UM. Crypt osporidiosis: an updat e in molecular epidemiology.

Curr O pin I nf ect Dis. 2004; 17(5): 483–490.
> Table of C ontents > Infec tious D is eas e > 145 - P ay Attention to the Mor phology R epor ted B y the
Mic r obiology Lab for Fungal C ultur es
145
Pay Attention to the Morphology Reported By
the Microbiology Lab for Fungal Cultures
Shelley S. Magill MD
William G . Merz PhD
Fungi cause a w ide variet y of diseases in t he crit ically ill pat ient , f rom
mucocut aneous inf ect ions t o bloodst ream inf ect ion and deep-seat ed, t issue-
invasive disease. Thehighmort alit y of invasive f ungal inf ect ions and t he
availabilit y of new ant if ungal agent s w it h diff erent spect ra of act ivit y make
prompt , accurat e, specif ic diagnosis crit ical. When an invasive f ungal inf ect ion is
suspect ed, appropriat e specimens, such as blood, body f luids, and/ or t issue,
should be sent t o t he laborat ory specif ically f or f ungal microscopic examinat ion
and f ungal cult ure. The microscopic examinat ion involves preparing a slide of t he
specimen w it h pot assium hydroxide, a Calcof luor Whit e st ain, or a G ram st ain.
There are t w o common morphologic f orms of f ungi t hat may be seen on
microscopic examinat ion and in t he cult ure plat e or t ube: yeast s and molds (or
f ilament ous f ungi). Yeast s are f ungi t hat reproduce by budding or, in rare cases,
by binary f ission. I f t he laborat ory report s yeast like f orms on a microscopic
examinat ion, t he most common organisms t o consider w ould be Candi da species
and Cryptococcus species, alt hough endemic f ungi such as Hi stopl asma (agent
of hist oplasmosis), Bl astomyces (agent of blast omycosis), and Paracocci di oi des
(agent of paracoccidioidomycosis) w ould also be possible, but less likely.
Candi da species are common colonizers of sit es such as t he oropharynx,
gast roint est inal t ract , and f emale genit ourinary t ract , but t hey may also cause
devast at ing invasive inf ect ion. Cryptococcus neof ormans should alw ays be
considered a pat hogen. I f t he laborat ory report s t he grow t h of yeast in cult ure,
t he same organisms should be considered, alt hough Hi stopl asma, Bl astomyces,
and Paracocci doi des w ould be much less likely since t hey are dimorphic f ungi
and more of t en grow as molds.
Watch Out For

I f t he laborat ory report s t he det ect ion of f ungal hyphae by microscopic
examinat ion, t his means t hat t he f ungus in quest ion is a mold. I t is import ant t o
dist inguish hyphae t hat are morphologically consist ent w it h Aspergi l l us species
and relat ed molds f rom t hose t hat are consist ent
w it h Zygomycet es (e. g. , Rhi zopus, Cunni nghamel l a, Mucor), since t he

t reat ment s diff er. Zygomycot ic hyphae are w ide (up t o 10 t o 12 microns in
diamet er), have w avy/ ribbony cell w alls, inf requent branching at right angles, and
relat ively inf requent sept a. The agent s of hyalohyphomycosis (Aspergi l l us,
Pseudal l escheri a, Fusari um) are narrow er (up t o 8 microns in diamet er), have
parallel cell w alls, f requent branching at acut e angles, and usually f requent
sept a.
When molds are recovered in cult ure, t hey f orm colonies made up of masses of
hyphae f rom w hich spores are produced. The colony morphology and
microscopic morphology of cult ured molds allow t hem t o be ident if ied in t he
laborat ory. Because molds t end t o grow slow ly, ident if icat ion t o t he genus level
may t ake several days or even longer. Many molds may be deadly pat hogens in
t he proper clinical cont ext , but t hey may also be colonizers or laborat ory
cont aminant s.
Some f ungi are dimorphic, meaning t hat t hey exist in t w o diff erent morphologic
f orms, most commonly yeast like and mold f orms. These f ungi t ypically exist in
t heir yeast like st at e in human t issues and grow in nat ure (and in cult ure in t he
laborat ory) as molds. Examples of w ell-know n dimorphic f ungi t hat f it t his
def init ion are Hi stopl asma capsul atum, Bl astomyces dermati di ti s, Sporothri x
schencki i , and Paracocci di oi des brasi l i ensi s. Cocci di oi des i mmi ti s is also a
dimorphic f ungus but f orms large saclike st ruct ures called spherules (not budding
yeast cells) in t issue. These dimorphic f ungi are alw ays pat hogens.
I f t here is uncert aint y regarding t he clinical signif icance of a yeast or mold
appearing on a microscopic examinat ion or grow ing f rom a cult ure of a pat ient
specimen, prompt inf ect ious disease consult at ion should be obt ained.
Suggested Readings
Doct or Fungus. ht t p: / / w w w. doct orf ungus. org
Merz WG , Hay RJ, ed. Topley and Wilson's Microbiology and Microbial
I nf ect ions: Medical Mycology. 10t h ed. Washingt on, DC: ASM Press; 2005.
Murray PR, Baron EJ, Pf aller MA, et al. , eds. Manual of Clinical Microbiology.
8t h ed. Vol. 2. Washingt on, DC: ASM Press; 2003.
> Table of C ontents > Infec tious D is eas e > 146 - C ons ider pos s ible fungal infec tion in patients w ith
hypother m ia and br adyc ar dia
146
Consider possible fungal infection in patients
with hypothermia and bradycardia
Suneel Khetarpal MD
Andrew J. Kerwin MD
I nf ect ions in t he int ensive care unit (I CU) remain a major source of morbidit y and
mort alit y. Typically inf ect ions are bact erial. How ever, w it h an increasing level of
pat ient acuit y, more advanced procedures, and more aggressive t herapy, f ungal
inf ect ions are increasingly common. Cert ain academic surgical I CUs now f ind
f ungal isolat es as t he f ourt h most common bloodst ream inf ect ion. Furt hermore,
t he at t ribut able mort alit y f or such inf ect ions remains bet w een 20% and 60%.
Cont ribut ing t o t his mort alit y is t he diff icult y in diagnosing t his ent it y. Despit e a
bet t er underst anding of f ungal inf ect ions, t here remains a delay in diagnosis as
t here is no rapidly available microbiological markers and cult ures suggest ive of
syst emic f ungal inf ect ions are posit ive in only 30% t o 50% of cases. O f t en, t hey
are not posit ive unt il lat e in t he process.
Signs and Symptoms

How ever, despit e t he diff icult y in diagnosis, t he key t o t reat ment of f ungal
inf ect ions is early recognit ion. Risk f act ors f or f ungal inf ect ion include vent ilat ed
pat ient s, prolonged I CU st ays, previous ant ibiot ics, t ot al parent eral nut rit ion
(TPN), high APACHE (Acut e Physiological and Chronic Healt h Evaluat ion) scores,
and previous abdominal surgery. Because of t he delay in diagnosis, any
individual w it h an at ypical present at ion of sepsis should be considered t o have
t he possibilit y of a syst emic f ungal inf ect ion. Such present at ions include:
hypot hermia; bradycardia; ongoing elevat ion in t emperat ure or w hit e blood cell
count ; or ongoing sepsis despit e being on t he “appropriat e” empiric
ant imicrobiological t herapy. I n order t o increase t he accuracy of diagnosis, t he
role of colonizat ion has been ext ensively evaluat ed. Emerging evidence has
show n t hat t he risk of candidemia may be relat ed t o t he densit y and ext ension of
f ungal colonizat ion over t ime. The risk of deat h has act ually been f ound t o be
similar in pat ient s w it h mult iple-sit e f ungal colonizat ion and t hose w it h conf irmed
invasive candidiasis. As a result t he concept of early empiric t herapy has
emerged. This t herapy recognizes t he high-risk pat ient w ho begins t o show
evidence of Candi da colonizat ion by posit ive cult ures f rom
mult iple sit es. I t has been suggest ed t hat such high-risk pat ient s be st art ed on
ant if ungal t herapy, not as prophylaxis, but as t reat ment . O n t he ot her hand,
some experienced clinicians believe prophylact ic st rat egies against f ungal
inf ect ions have proven t o be ineff ect ive at signif icant ly reducing syst emic f ungal
inf ect ions in an I CU. Select groups of pat ient s such as pancreat ic t ransplant
pat ient s may, how ever, benef it f rom such st rat egies.
The most common f ungal inf ect ion seen in I CUs are t hat of Candi da species,
w it h 50% being C. al bi cans. An increasing number of isolat es are nonalbicans
species such as C. gl abrata, C. parapsi l osi s, C. tropi cal i s, and C. krusei . First -
line t herapy remains f luconazole (400 t o 800 mg/ d). How ever, w it h t he increasing
nonalbicans species t he limit at ion of f luconazole t o t reat C. gl abrata and C.
krusei must be recognized. Current ly t here is a second-generat ion azole
(voriconazole) w it h improved eff iciency against nonalbicans species. Anot her new
class of ant if ungals are t he echinocandins (caspof ungin, anidulaf ungin), w hich
can be used f or t herapy against candidiasis. I t should be not ed t hat most
experienced clinicians no longer consider t he use of amphot ericin B t o be an
appropriat e f irst - or second-line agent .
Alt hough rare, mold inf ect ions such as Aspergi l l us species remain a problem in
t he I CU. High-risk pat ient s are usually signif icant ly immunosuppressed, such as
solid-organ t ransplant or bone-marrow t ransplant pat ient s; how ever, any pat ient
w hose normal immune syst em has been alt ered can be considered at risk. O nce
again early recognit ion plays a key role in successf ul t reat ment . Treat ment of
syst emic molds has been t radit ionally by amphot ericin; how ever, recent lit erat ure
suggest s t hat second-generat ion azole (voriconazole) or t he echinocandins may
have improved eff iciency. Despit e t hese recent advances, t he mort alit y rat e f or
Aspergillus inf ect ions remains high.
Suggested Readings
I banez-Nolla J, Nolla-Salas M, Leon MA, et al. Early diagnosis of candidiasis
in non neut ropenic crit ically ill pat ient s. J I nf ect . 2004; 48(2): 181–192.
Kullberg BJ, Sobel JD, Ruhnke M, et al. Voriconazole versus a regimen of

amphot ericin B f ollow ed by f luconazole f or candidaemia in non-neut ropenic
pat ient s: a randomised non-inf eriorit y t rial. Lancet . 2005; 366(9495): 1435–
1442.
Vardakas KZ, Samonis G , Michalopoulos A, et al. Ant if ungal prophylaxis w it h

azoles in high-risk, surgical int ensive care unit s: a met a-analysis of
randomized, placebocont rolled t rials. Crit Care Med. 2006; 34(4): 1216.
> Table of C ontents > Infec tious D is eas e > 147 - Give S pec ial C ons ider ation to the E xtended- S pec tr um
B eta- Lac tam as e- P r oduc ing O r ganis m s B efor e Adm inis ter ing Antibiotic s
147
Give Special Consideration to the Extended-
Spectrum Beta-Lactamase-Producing Organisms
Before Administering Antibiotics
John J. Lewin III PharmD
Ext ended-spect rum bet a-lact amase (ESBL)-producing organisms represent one
of t he largest inf ect ious disease–relat ed challenges in t he int ensive care unit
(I CU) environment w orldw ide. The β -lact amase enzymes are capable of
hydrolyzing t he bet a-lact am ring of penicillins and relat ed compounds, rendering
t hem inact ive. To dat e, t here have been hundreds of diff erent t ypes of β -
lact amase enzymes ident if ied. There have been many classif icat ion schemes
proposed f or β -lact amases. The classif icat ion init ially int roduced by Ambler in
1980 is based on t he amino-acid sequence and recognizes f our molecular
classes designat ed A t o D. This scheme is based on amino-acid similarit y, w it h
classes A, C, and D being serine-β -lact amases and class B being met allo-β -
lact amases. The scheme proposed by Bush is a f unct ional classif icat ion t hat
def ines f our groups according t o t heir subst rat e and inhibit or prof iles.
The ESBL-producing organisms are not ew ort hy as t hey represent signif icant
challenges in prevent ion, ident if icat ion, and t reat ment . The ESBL enzyme is
capable of hydrolyzing essent ially all β -lact am ant ibiot ics except carbapenems
and cephamycins (cef ot et an and cef oxit in). The f irst report s of ESBLs came in
t he early 1980s, af t er t he development and w idespread use of t hird-generat ion
cephalosporins. O f not e, ESBL enzymes are plasmid-mediat ed, rendering t he
DNA encoding t he ESBL enzyme (as w ell as ot her t ypes of resist ance) easily
t ransf erable. Theref ore, ESBL-producing organisms are f requent ly resist ant t o
ot her classes of ant ibiot ics suchas f luoroquinolones, aminoglycosides, and
t rimet hoprim/ sulf amet hoxazole.
Watch Out For

I dent if icat ion of ESBLs can be problemat ic f or many clinical microbiology labs.
The Nat ional Commit t ee f or Clinical Laborat ory St andards has developed
st andards t o improve upon t he early and reliable recognit ion of ESBLs. How ever,
some st udies indicat e an inabilit y f or many labs t o reliably ident if y ESBL isolat es
and as such, clinicians should be f amiliar w it h t heir microbiology laborat ory's
met hods and abilit ies t o
det ect and report ESBLs. I t is not uncommon t o see an init ial sensit ivit y prof ile
indicat ing suscept ibilit y t o t hird-generat ion cephalosporins (especially cef t riaxone
and cef ot axime) in vit ro. How ever, using one of t hese agent s t o act ively t reat an
ESBL inf ect ion is likely t o lead t o t reat ment f ailure in vivo. As such, if an ESBL-
producing organism is suspect ed, cephalosporins should not be ut ilized.
Because of t he ESBLs’ abilit y t o hydrolyze most β -lact ams, and t heir plasmid-
mediat ed t ransmission of resist ance t o ot her classes of ant ibiot ics, t reat ment
choices are severely limit ed. Carbapenems are considered t he drugs of choice
f or severe inf ect ions due t o ESBL producing organisms. Theses agent s exert t he
most consist ent act ivit y against t hese organisms and are highly resist ant t o t he
hydrolyt ic eff ect s of ESBLs.
Most ESBLs are eff ect ively inhibit ed by β -lact amase inhibit ors (clavulanic acid,
sulbact am, and t azobact am) t o varying degrees. How ever, t he β -lact am/ β -
lact amase-inhibit or combinat ions are not recommended in t he management of
inf ect ions due t o ESBL-producing organisms. Many ESBL-producing organisms
may also generat e non-ESBL β -lact amase resist ant t o t he β -lact amase inhibit ors
or generat e mult iple t ypes of ESBL, rendering t he β -lact amase inhibit ors
inact ive, and t he ant imicrobial agent suscept ible t o hydrolysis. I n pat ient s w it h a
severe bet a-lact am allergy in w hom an alt ernat ive t o carbapenems is desired,
f luoroquinolones and aminoglycosides can be considered if t he organisms are
show n t o be suscept ible.
Suggested Readings
Ambler RP, Coulson AF, Frere JM, et al. A st andard numbering scheme f or t he
class bet a-lact amases. Biochem J. 1991; 276 (Pt 1): 269–270.
Bush K, Jacoby G A, Medeiros AA. A f unct ional classif icat ion scheme f or bet a
lact amases and it s correlat ion w it h molecular st ruct ure. Ant imicrob Agent s
Chemot her. 1995; 39(6): 1211–1233.
Pat erson DL, Bonomo RA. Ext ended-spect rum β -Lact amases: a clinical
updat e. Clin Microbiol Rev. 2005; 18(4): 657–686.
St ew ard CD, Wallace D, Hubert SK, et al. Abilit y of laborat ories t o det ect
emerging ant imicrobial resist ance in nosocomial pat hogens: a survey of
project I CARE laborat ories. Diagn Microbiol I nf ect Dis. 2000; 38(1): 59–67.
> Table of C ontents > Infec tious D is eas e > 148 - Have a High Thr es hold for Thor ac entes is W hen
Looking for a S our c e of Infec tion
148
Have a High Threshold for Thoracentesis When
Looking for a Source of Infection
Deba Sarma MD
The onset of a new f ever or presence of a new or increasing leukocyt osis is a
problem t hat is encount ered f requent ly in any int ensive care unit (I CU) and can
lead t o an ext ensive w orkup, w hich subject s pat ient s t o many t est s and of t en
produces inconclusive result s. Fevers may be a generalized response t o a
noninf ect ious inf lammat ory st at e such as t hose relat ed t o post operat ive changes,
alcohol or drug w it hdraw al, t ransf usion of blood product s, f ever, pancreat it is,
adrenal insuff iency, deep venous t hrombosis, and various ot her et iologies.
I nf ect ions, how ever, are a much more common cause of f ever w it h t he
prevalence of nosocomial inf ect ions in I CUs ranging f rom 3% t o 31%. The most
commonly report ed inf ect ions in t he I CU are vent ilat or-associat ed pneumonia,
bloodst ream inf ect ion, and cat het er-relat ed inf ect ions. The onsent of new f ever
w ill usually result in a w orkup obt aining t est s t hat are direct ed t o t he most
common and likely et iologies. This w ill of t en include cult ures obt ained f rom
blood, urine, and sput um and a chest radiograph t o assess f or t he presence of
at elect asis or in-f ilt rat es. I f t he et iology is not obvious based on preliminary
st udies, f urt her w orkup such comput ed t omography (CT) scans t o rule out f luid
collect ions w ill be done based on t he next most likely sources. By t his point
noninf ect ious et iologies are considered, as are less common inf ect ious causes
such as cent ral nervous syst em (CNS) inf ect ions and inf ect ed pleural eff usion
(parapneumonic eff usions or empyemas).
During rout ine daily t est ing of I CU pat ient s, t he presence of a pleural eff usion is
of t en not ed on chest radiograph. I n a recent st udy of medical I CU pat ient s, it
w as not ed t hat most of t he I CU pat ient s w ere admit t ed f or condit ions ot her t han
pleural disease, but t he presence of pleural eff usions w as common secondary t o
pleural eff ect s of pulmonary parnenchymal disorders and dysf unct ion of ot her
organ syst ems. I n t his st udy, pleural eff usions result ed f rom noninf ect ious
causes in 82% of pat ient s f ound t o have an eff usion. O f t hese pat ient s t he most
common cause of pleural eff usions w as heart f ailure diagnosed as eit her a
primary condit ion or developing af t er aggressive f luid resuscit at ion. O nly 11% of
pat ient s in t his st udy w ere det ermined t o have an inf ect ed pleural eff usion. The
suspicion f or inf ect ion w as
raised w hen a pat ient remained t oxic despit e ant iobiot ic coverage. O f all t he
st udy pat ient s w it h an eff usion, 21% of t he pat ient s underw ent a t horacent esis t o
rule out a malignant eff usion or an inf ect ion. O nly t hree of eight w it h a suspect ed
inf ect ion had a successf ul t horacent esis, w it h ot hers being unsuccessf ul
secondary t o small size of eff usion or t erminat ion of procedure secondary t o
inst abilit y. The pat ient s w ho did not undergo a t horacent esis had resolut ion of
t heir pleural eff usion af t er init iat ion or changes in ant ibiot ics. Anot her st udy
looking specif ically at f ebrile medical I CU pat ient s and assessing t he ut ilit y of
ult rasound f or diagnosis of empyemas f ound t hat 62% of pat ient s w it h f evers and
a pleural eff usion did have an inf ect ious exudat e. The prevalence of empyemas
(a pot ent ially lif e-t hreat ening condit ion), how ever, w as only 16%. I n t his
part icular st udy, t he specif ic f indings such as complex, sept at ed pat t erns on
ult rasound increased t he likelihood t hat an eff usion w as t ruly an empyema t hat
w ould require drainage. The remainder of pleural eff usions w ere f ound t o be
parapneumonic eff usions or t ransudat ive (noninf ect ious/ noninf lammat ory
collect ions), w hich have been f ound t o resolve w it h appropriat e ant ibiot ic
coverage and observat ion, respect ively.
Anot her recent st udy assessing t he f requency of posit ive cult ure result s f rom
pleural eff usions show ed t hat t he posit ive yield of microbiologic cult ures is low.
I n t his st udy t he st rongest predict or of a t rue posit ive result w as t he presence of
loculat ed pleural eff usions. O f t he remaining eff usions t hat w ere f ree-f low ing,
only 1. 1% w ere posit ive f or a t rue pat hogen. Even w hen analyzing f luid f rom
pat ient s expect ed t o have t he highest likelihood of having an inf ect ious cause f or
t heir eff usion, t he f requency of t rue posit ive result s w as only 18%, suggest ing
t hat t he sensit ivit y of pleural f luid cult ures is low, t hereby arguing against
f requent analysis of pleural f luid.
Watch Out For

O f t he f ew st udies t hat have looked at t he presence of inf ect ed pleural eff usions
in t he I CU, none has looked at t he t he yield of t horacent esis during a f ever
w orkup in surgical I CU pat ient s. I t is know n, how ever, t hat in surgical pat ient s,
especially af t er t horacic or cardiac procedures and even int ra-abdominal
surgeries, t he presence of pleural eff usions as seen in chest radiographs is f airly
common and inconsequent ial. Furt hermore, t he post operat ive resuscit at ion is
of t en aggressive and result s in f air amount of pulmonary congest ion leading t o
pulmonary eff usions. As is t he case in t he st udies seen in medical I CUs, t he
f ormat ion of pleural eff usions is very common in pat ient s in posit ive f luid
balance, usually resolves w it hin days, and, as such, is not a likely source of
inf ect ion in I CU pat ient s w it h a f ever. Furt hermore, a t horacent esis is not a
benign procedure, and t hough it can be perf ormed in t he I CU even on
mechanically vent ilat ed pat ient s, it does run t he risk of several complicat ions.
Major complicat ions include pneumot horax, hemopneumot horax, hypot ension due
t o vasovagal response, hemorrhage, and re-expansion pulmonary edema.
What to Do
Thus, t he recommendat ion in various st udies is t o have a high clinical suspicion
and possibly f urt her w orkup such as ult rasound or CT bef ore doing a
t horacent esis, as t he incidence of inf ect ion and t he yield of t horacent esis are
f airly low. Some st udies generally recommend t horacent esis if ot her sources of
f ever are ruled out and if t he pat ient has a large loculat ed eff usion or source
such as t hose relat ed t o possible malignancy, pancreat it s, possibilit y of a
chylot horax, or esophageal leak/ rupt ure. These are t he exudat ive eff usions
result ing f rom local inf lammat ion or movement of f luid f rom t he perit oneal space
and drainage is usually required t o aid in diagnosis and relieve t he inf lammat ory
process.
Suggested Readings
Barnes TW, O lson EJ, Morgent haler TI , et al. Low yield of microbiologic
st udies on pleural f luid specimens. Chest . 2005; 127: 916–921.
Marik PE. Fever in t he I CU. Chest . 2000; 117: 855–869.
Mat t ison LE, Coppage L, Alderman DF, et al. Pleural eff usions in t he medical
I CU: prevalence, causes, and clinical implicat ions. Chest . 1997; 111: 1018–
1023.
Tu C, Hsu W, Hsia T, et al. Pleural eff usions in f ebrile medical I CU pat ient s:
chest ult rasound st udy. Chest . 2004; 125: 1274–1280.
> Table of C ontents > Infec tious D is eas e > 149 - Aim for a P eak of Ten Tim es the Minim um Inhibitor y
C onc entr ation ( MIC ) to K ill P s eudom onas W hen Us ing an Am inoglyc os ide
149
Aim for a Peak of Ten Times the Minimum
Inhibitory Concentration (MIC) to Kill
Pseudomonas When Using an Aminoglycoside
Aminoglycosides (gent amicin, t obramycin, and amikacin) are bact ericidal
ant ibiot ics t hat are act ive against aerobic G ram-negat ive and G ram-posit ive
organisms. Clinically t hese agent s are ut ilized f or serious G ram-negat ive
inf ect ions, as w ell as in combinat ion w it h bet alact ams f or G ram-posit ive synergy.
This chapt er discusses how t o dose aminoglycosides f or serious G ram-negat ive
inf ect ions.
When designing dosing regimens, one must consider t he pharmacodynamics of
t he agent in quest ion. For aminoglycosides, t his means underst anding how t hese
agent s exert t heir bact ericidal eff ect s. Aminoglycosides display concent rat ion-
dependent killing (Tabl e 149. 1), w hich simply means t hat t he higher t he peak
concent rat ion increases above t he minimum inhibit ory concent rat ion (MI C), t he
bet t er t he killing eff ect iveness. Time-dependent agent s are t he opposit e. Their
killing eff ect iveness is solely dependent on t he amount of t ime t he drug
concent rat ion remains above t he MI C and not t he degree t o w hich t hey are over
t he MI C. To t ake advant age of t his, t ime-dependent drug doses are administ ered
mult iple t imes during t he day t o keep t he serum concent rat ions above t he MI C as
much as possible.
Mult iple st udies have show n a dose-response relat ionship w it h aminoglycosides,
bot h clinically and experiment ally. A series of papers w ere published in t he mid
1980s t hat show init ial “t herapeut ic” peaks w ere import ant predict ors of
successf ul t reat ment out comes w it h aminoglycosides. The last paper in t his
series, published by Moore et al. (1987), discussed t he import ance of t he
peak: MI C rat io. Maximum peak: MI C rat io, def ined as >10, w as one of t w o
st at ist ically signif icant variables show n t o be a predict or of a f avorable out come
(t he ot her w as a f avorable underlying prognosis). These st udies w ere done w it h
init ial doses of gent amicin 2 mg/ kg, and amikacin 8 mg/ kg.
Subsequent st udies have ut ilized higher doses of aminoglycosides f or t he
t herapy of t hese serious inf ect ions. A st udy of a dose 3 mg/ kg, based on eit her
an ideal or adjust ed body w eight , of gent amicin or t obramycin w as evaluat ed in
crit ically ill surgical pat ient s. This
st udy show ed t hat t he increased dose result ed in a higher init ial peak (8. 1
µg/ mL), but t his w as achieved in only 50% of pat ient s. A dose of nearly 4 mg/ kg
w as ext rapolat ed f rom st udy dat a t o achieve a peak 10 µg/ mL and can be
reliably used. Post inf usion levels can be obt ained, and even experienced
providers w ill consult an I CU PharmD f or discussion of t he next dose.
TABLE 149-1 DIFFERENTIATION OF ANTIM ICROBIAL

AGENTS BASED ON PHARM ACODYNAM IC PROFILE
T IME-DEPENDENT CONCENT RAT ION-

AGENT S DEPENDENT AGENT S
Penicillins Aminoglycosides
Cephalosporins Quinolones
Carbapenems Metronidazole
Monobactams Daptomycin
Vancomycin
Linezolid
Clindamycin
An example of how t o dose one of t he aminoglycosides f ollow s, but keep in mind

t hat doses may need t o be adjust ed in order t o maint ain t he peak: MI C rat io >10
if MI C sat any part icular inst it ut ion are higher t han 1 µg/ mL. The st epw ise
prot ocol is as f ollow s:
1. Af t er t he gent amicin dose is administ ered, obt ain 1-hour peak concent rat ion
and 8-hour post inf usion concent rat ion.
2. Use t he f ollow ing equat ions in t his sequence t o calculat e w hen t he next dose
can be saf ely given and how much t o give:
a.
I f you give a 400-mg dose of gent amicin and t he result ing peak is 8
mg/ dL (8 µg/ mL), t hen t he Vd = 50 L.
b.
The levels af t er your 400-mg dose are a peak of 8 mg/ dL and an 8-hour
level of 4 mg/ dL. The result ing Ke w ould be (ln [ 8/ 4] )/ 7 hours = 0. 099
hours - 1 .
c.
The half -lif e f or t his example w ould be 0. 693/ 0. 099=~7 hours. I t is
generally saf e t o redose in t hree t o f our half -lives, so a q24h regimen
w ould be appropriat e if renal f unct ion remains st able.
d. Next dose
You know t hat you can give t he next dose 24 hours af t er your f irst dose.
At t hat t ime, t he serum concent rat ionis approximat ely 1 µg/ mL. I f you are
t arget ing a peak of 10 µg/ mL, t he next dose is calculat ed by rearranging
t he Vd equat ion: Dose (mg) = Vd × goal peak. I n det ermining t he f ollow -
up doses, realize t hat t he serum levels are not 0 µg/ mL, but usually
bet w een 1 and 2 µg/ mL. The t rough level must be t aken int o
considerat ion w hen giving t hese f ollow -up doses. For t his example:
Dose (mg) = 50 L × 9µg/ mL
Dose (mg) = 450 mg (not 500 mg)
Suggested Readings
Leggert JE, Ebert S, Fant in B, et al. Comparat ive dose eff ect relat ionships at
several dosing int ervals f or bet a-lact am, aminoglycoside, and quinolone
ant ibiot ics against G ram-negat ive bacilli in murine t high inf ect ion and
pneumonit is models. Scand J I nf ect Dis Suppl 1990; 74: 179–184.
Moore RD, Smit h CR, Leit man PS. Associat ion of aminoglycoside plasma
levels w it h t herapeut ic out come in G ram-negat ive pneumonia. Am J Med
1984; 77: 657–662.
Moore RD, Smit h CR, Leit man PS. The associat ion of aminoglycoside plasma
levels w it h mort alit y in G ram-negat ive bact eremia. J I nf ect Dis 1984; 149: 443–
448.
Moore RD, Leit man PS, Smit h CR. Clinical response t o aminoglycoside
t herapy: import ance of t he rat io of peak concent rat ion t o minimal inhibit ory
concent rat ion. J I nf ect Dis 1987; 155: 93–99.
> Table of C ontents > Infec tious D is eas e > 150 - K now the D efinition of a C atheter - R elated
B loods tr eam Infec tion
150
Know the Definition of a Catheter-Related
Bloodstream Infection
Cat het er-relat ed bloodst ream inf ect ions (CRBSI s) are a major cause of
morbidit y and mort alit y in crit ical care unit s w it h an est imat ed 28, 000 pat ient s
dying f rom t his complicat ion every year in t he Unit ed St at es alone. How ever, t his
complicat ion is largely prevent able. Recent evidence show s t hat a process t hat
incorporat es evidence based guidelines f rom t he Cent ers f or Disease Cont rol
and Prevent ion (CDC), t he Societ y of Crit ical Care Medicine, and ot hers int o a
comprehensive saf et y-based program can virt ually eliminat e CRBSI s.
A f ormal def init ion of a CRBSI can be f ound t hrough t he CDC. The pat ient must
have a cent ral venous cat het er in place at t he t ime of t he suspect ed inf ect ion
and t here must be no ot her ident if iable source of inf ect ion present . Bot h t he
cat het er and peripheral blood must grow t he same organism w hile t he cat het er
w as in place. Proper cult ure of t he cat het er requires t hat t he int radermal port ion
of t he cat het er be cult ured, not t he t ip as is commonly and incorrect ly done. I f at
all possible, t he peripheral blood draw n should be at least 10 mL, draw n f rom
t w o separat e peripheral sit es. Blood draw n t hrough t he suspect cat het er is
invalid unless it is done t hrough a special prot ocol and t he lab perf orms
quant it at ive cult ures. This is usually done only f or cert ain permanent cat het ers
such as Hohn and Hickman cat het ers and occasionally w it h special lines
dedicat ed f or hyperaliment at ion. Colony count s f or regular cent ral cat het ers t hat
report >15 colony-f orming unit s f or t he cat het er are considered posit ive.
Prevent ion of CRBSI s requires a mult iple-element program w hose most import ant
component is educat ion and reinf orcement . The f irst st ep is developing
aw areness of t he magnit ude and implicat ions of t he problem among t he st aff .
Creat ing a cent ral locat ion w here all necessary supplies are immediat ely
available is essent ial. I deally t his w ould be mobile such as a cart . This helps
reduce breaches of st erilit y by prevent ing t he need t o obt ain mat erials once t he
procedure of placing a cat het er has begun.
What to Do
The supplies required include f ull gow n, st erile gloves, mask, and hair covering
just as f or surgery. Chlorhexidine f or skin prep should be used, as it has been
show n t o be eff ect ive in reducing t he incidence of inf ect ion. Fullbody draping of
t he pat ient and bed, not just t he immediat e procedure sit e, is a must . I t is
t hought t hat a major source of cont aminat ion is secondary t o t he w ire used f or
t he Seldinger t echnique cont act ing nonst erile equipment near t he f ield.
Widespread coverage w it h a f ull drape helps t o eliminat e t his. All part icipant s in
t he st erile port ion of t he procedure must w ash t heir hands prior t o gow ning and
gloving. Chlorhexidine soap or alcohol-based hand w ashes are appropriat e.
St rict st erilit y must be maint ained at all t imes. Any t ime t here is a breach in t his
sequence of st eps, t he procedure should be st opped and t he sit uat ion rect if ied
even if it means st art ing t he w hole procedure over. Nurses should be empow ered
t hrough educat ion and a checklist t o ensure t hat st rict adherence t o pract ice is
maint ained. Wit h t he except ion of dialysis or possible dialysis pat ient s, t he
subclavian sit e select ion is pref erred as it is associat ed w it h t he low est risk of
CRBSI s. Limit ing t he number of lumens t o t he low est needed may also help.
Af inal component t o best pract ice f or reducing CRBSI s is t o evaluat e daily
w het her a part icular cat het er is needed and, if not , remove it as soon as
possible. The risk of cat het er inf ect ion and CRBSI increases w it h t he passage of
t ime and cat het ers t hat have been in place f or great er t han or equal t o 7 days
are especially at risk and ones present f or more t han 21 days are quit e prone t o
get t ing inf ect ed. Cat het ers t hat are placed in a nonst erile f ashion such as
emergent lines or t rauma lines should be removed w it hin 24 hours and replaced,
if necessary, w it h cent ral lines in accordance w it h t he guidelines described here.
Suggested Readings
Berenholt z, SM Pronovost P, Lipset t PA, et al. Eliminat ing cat het er relat ed
bloodst ream inf ect ions in an int ensive care unit . Crit Care Med 2004; 32: 2014–
2020.
Mermal LA. Prevent ion of int ravascular cat het er relat ed inf ect ions. Ann I nt ern
Med 2000; 132: 391–402.
> Table of C ontents > Infec tious D is eas e > 151 - S tr ongly c ons ider S topping P r ophylac tic Antibiotic s
after 24 hour s in P enetr ating Abdom inal Tr aum a
151
Strongly consider Stopping Prophylactic
Antibiotics after 24 hours in Penetrating
Abdominal Trauma
G eorge C. Velmahos MD, PHD
Despit e abundant evidence t o t he cont rary, mult iple and prolonged ant ibiot ics are
st ill used f ollow ing penet rat ing abdominal t rauma. I n 515 randomized pat ient s,
Fabian et al. show ed t hat 1 day of a second-generat ion cephalosporin
prophylaxis w as equally eff ect ive as 5 days of t he same ant ibiot ic. Cornw ell et
al. , in a prospect ive randomized st udy of 63 high-risk pat ient s w it h penet rat ing
colon injury, f ound no diff erence in inf ect ious morbidit y bet w een 1 and 5 days of
a second-generat ion cephalosporin. Dellinger et al. randomized 116 pat ient s w it h
small or large bow el injuries t o 12 hours versus 5 days of ant ibiot ics and f ailed
t o ident if y any signif icant diff erences in morbidit y or mort alit y. Velmahos et al. ,
in a prospect ive st udy of 250 crit ically injured pat ient s, f ound t hat ant ibiot ic
prophylaxis f or penet rat ing abdominal t rauma longer t han 1 day did not decrease
sept ic morbidit y and w as an independent risk f act or f or delayed resist ant
inf ect ion.
What to Do
Mult iple st udies have been perf ormed comparing diff erent t ypes of ant ibiot ics in
penet rat ing abdominal t rauma. Moxalact am, gent amicin w it h clindamycin,
cef oxit in, cef ot et an, cef ot axime, azt reonam, ampicillin/ sulbact am, and
piperacillin/ t azobact am are some of t he many prophylact ic agent s t hat have been
t est ed. The general conclusion is t hat a single broad-spect rum ant ibiot ic is as
eff ect ive as mult iple ant ibiot ics. Second-generat ion cephalosporins have show n a
decreasing eff icacy in some st udies, probably because of bact erial resist ance
f rom prolonged use or because of inabilit y of ent erococcal coverage.
Ampicillin/ sulbact am seems t o be cost -eff ect ive and provides adequat e
coverage.
St andard dosing of prophylact ic ant ibiot ics in t he majorit y of t rauma pat ient s
w it h abdominal t rauma is suff icient . How ever, t he pharmacodynamics of t hese
agent s in crit ical illness is poorly underst ood. I ncreased volume of dist ribut ion,
met abolic changes aff ect ing drug excret ion, and mult iple drug int eract ions alt er
t he required doses. More of t en t han not , crit ically ill pat ient s are underdosed.
Alt hough higher t arget s and monit oring algorit hms have been est ablished f or t he
aminoglycosides and vancomycin, such research is lacking f or most ot her

ant ibiot ics used f or prophylaxis. Alt hough most experienced pract it ioners
consider a single day of a single broad-spect rum ant ibiot ic t o be adequat e f or
prophylaxis f ollow ing penet rat ing abdominal t rauma, a low t hreshold should exist
f or increasing t he dose and/ or short ening t he dosing int erval (according t o t he
class of ant ibiot ic) f or pat ient s w ho are crit ically injured.
O ne f inal not e is t hat considerat ion should be given t o redosing ant ibiot ics in t he
f ace of cont inuing blood loss (> 120 t o 1, 500 cc) in t he int ensive care unit or at
repeat laparot omy.
Suggested Readings
Cornw ell EE I I I , Doughert y WR, Berne TV, et al. Durat ion of ant ibiot ic
prophylaxis in high-risk pat ient s w it h penet rat ing abdominal t rauma: a
prospect ive randomized st udy. J G ast roint Surg. 1999; 3: 648–653.
Dellinger EP, Wert z MJ, Lennard ES, et al. Eff icacy of a short -course
ant ibiot ic prophylaxis af t er penet rat ing int est inal injury: a prospect ive
randomized t rial. Arch Surg. 1986; 121: 23–30.
East ern Associat ion f or t he Surgery of Trauma (EAST) evidence-based

guidelines: prophylact ic ant ibiot ics in penet rat ing abdominal t rauma.
ht t p: / / w w w. east . org
Fabian TC, Croce MA, Payne LW, et al. Durat ion of ant ibiot ic t herapy f or
penet rat ing abdominal t rauma: a prospect ive t rial. Surgery. 1992; 112: 785–
792.
Hooker KD, DiPiro JT, Wynn JJ. Aminoglycoside combinat ions versus bet a-
lact ams alone f or penet rat ing abdominal t rauma: a met a-analysis. J Trauma.
1991; 31: 1155– 1160.
Velmahos G C, Tout ouzas KG , Sarkisyan G , et al. Severe t rauma is not an

excuse f or prolonged ant ibiot ic prophylaxis. Arch Surg. 2002; 137: 537–541.
Weigelt JA, Easley SM, Thal SR, et al. Abdominal surgical w ound inf ect ion
w it h improved perioperat ive ent erococcus and bact eroides t herapy. J Trauma.
1993; 34: 579– 584.
> Table of C ontents > Infec tious D is eas e > 152 - Us e C lindam yc in in Nec r otiz ing Fas c iitis to C over
Gr oup a S tr eptoc oc c us
152
Use Clindamycin in Necrotizing Fasciitis to
Cover Group a Streptococcus
Carrie A. Sims MD, MS
Patrick K. Kim MD
Necrot izing f asciit is is a lif e-t hreat ening surgical emergency. I mmediat e w ide
surgical incision, drainage, and debridement is t he cornerst one of t herapy.
Broad-spect rum int ravenous ant ibiot ics should be inst it ut ed as soon as t he
diagnosis is suspect ed. At operat ion, t issue cult ures should be obt ained and
ant ibiot ic coverage should be t ailored t o cult ure grow t h. The most common
causat ive agent s of necrot izing f asciit is are Staphyl ococcus aureus, Cl ostri di um
species, group A Streptococcus, ent erococci, and Bacteroi des species. Many
necrot izing inf ect ions are mixed aerobic/ anaerobic in origin. Cont roversy st ill
remains about t he use of hyperbaric t herapy in t hese cases. How ever, if it is
ut ilized, hyperbaric t reat ment should never delay aggressive surgical t reat ment .
I nf ect ions caused by group A Streptococcus (e. g. , S. pyogenes, aka “f lesh-
eat ing bact eria”) deserve special ment ion. G roup A Streptococcus is responsible
f or a range of skin and sof t -t issue inf ect ions, including impet igo, erysipelas,
cellulit is, and necrot izing f asciit is. G roup A Streptococcus necrot izing f asciit is is
commonly associat ed w it h sept ic shock and mult iorgan f ailure—t he st rept ococcal
t oxic shock syndrome. This is secondary t o t he presence of M prot ein, a
virulence f act or present on t he bact erial surf ace t hat is highly ant igenic and
inhibit s phagocyt osis. I n addit ion, st rept ococcal pyrogenic exot oxins cause f ever
and cont ribut e t o organ f ailure and shock by st imulat ing host synt hesis of t umor
necrosis f act or α , int erleukin-1 β , and int erleukin-6.
What to Do
Penicillins are generally eff ect ive f or t he st aphylococcal and st rept ococcal
inf ect ions of erysipelas, impet igo, and cellulit is. How ever, penicillin monot herapy
is much less eff ect ive in deep necrot izing inf ect ions. Experiment ally, penicillins
lose eff ect iveness in t he presence of large numbers of organisms and w hen
bact eria are in t he st at ionary grow t h phase. I n deep inf ect ions, combinat ion
penicillin and clindamycin t herapy improves survival compared w it h penicillin
monot herapy. This clinical eff ect is support ed by several experiment al f indings:
clindamycin suppresses synt hesis of group A Streptococcus exot oxins; inhibit s M-
prot ein
synt hesis; suppresses synt hesis of prot eins involved in bact erial cell-w all
synt hesis; and might suppress host t umor necrosis f act or synt hesis. I t s
mechanism is independent of bact erial inoculum and grow t h st age. Finally, group
A Streptococcus resist ance t o clindamycin is ext remely rare.
Suggested Readings
Bisno AL, St evens DL. St rept ococcal inf ect ions of skin and sof t t issues. New
Engl J Med. 1996; 334: 240–245.
St evens DL. The f lesh-eat ing bact erium: w hat 's next ? J I nf ect Dis. 1999; 179
(Suppl 2): S366–S374.
St evens DL, Bisno AL, Chambers HF, et al. Pract ice guidelines f or t he
diagnosis and management of skin and sof t -t issue inf ect ions. Clin I nf ect Dis.
2005; 41(10): 1373– 1406.
> Table of C ontents > Infec tious D is eas e > 153 - B e C autious in Us ing Antibiotic s for Uninfec ted
P anc r eatitis
153
Be Cautious in Using Antibiotics for Uninfected
Pancreatitis
Benjamin Braslow MD
The best w ay t o t hink about acut e pancreat it is is t o equat e it t o an int ernal
chemical burn. The acut e resuscit at ion, complicat ed f luid and elect rolyt e
abnormalit ies, and inf ect ious complicat ions associat ed w it h acut e pancreat it is
must be addressed in a similar f ashion t o t he management of severe ext ernal
burns. The pat hophysiology of acut e pancreat it is involves a cascade of event s
init iat ed by acinar cell injury and pancreat ic duct obst ruct ion. These processes
allow t he inappropriat e ext racellular leakage of act ivat ed digest ive enzymes and
t he consequent aut odigest ion of pancreat ic and ext rapancreat ic t issues.
A w ide range of et iologies of acut e pancreat it is have been ident if ied. I n t he
Unit ed St at es, more t han 75% of cases are at t ribut able t o eit her gallst ones or
alcohol. O t her less common causes include iat rogenic causes such as
endoscopic ret rograde pancreat ography (ERCP), cardiopulmonary bypass, and
abdominal operat ions. Pat ient -based causes include blunt or penet rat ing
abdominal t rauma, periampullary neoplasm, pancreas divisum, sphinct er of O ddi
spasm, hyperlipidemia, hypercalcemia, and ischemia. More t han 85 medicat ions
have also been implicat ed in causing acut e pancreat it is. The highest incidence is
not ed w it h immunosuppressive agent s (azat hioprine and 6-mercapt opurine) and
t he ant iviral didanosine. O t her drugs incriminat ed include est rogen, nonst eroidal
ant i-inf lammat ory drugs (NSAI Ds) (sulf asalazine, sulindac, salicylat es), some
diuret ics (f urosemide, t hiazide diuret ics, et hacrynic acid), numerous ot her
ant ibiot ics (pent amidine, met ronidazole, t et racycline, t rimet hoprim-
sulf amet hoxazole, nit rof urant oin), valproic acid, procainamide, and several
angiot ensin-convert ing enzyme (ACE) inhibit ors.
Most episodes of acut e pancreat it is (80%) do not require any signif icant
int ervent ion, since t hey are mild and self -limit ing. How ever, approximat ely 20%
of pat ient s go on t o develop a severe f orm of acut e pancreat it is associat ed w it h
mult isyst em organ f ailure and/ or local complicat ions like necrosis, abscess
f ormat ion, or hemorrhage. These pat ient s have prolonged int ensive care unit
(I CU) st ays and hospit alizat ions w it h an in-house mort alit y exceeding 30%
despit e improvement s in diagnost ic and t reat ment modalit ies.
Signs and Symptoms

The clinical diagnosis of acut e pancreat it is is considered af t er t he t ypical
present at ion of severe epigast ric pain radiat ing t hrough t he back. Associat ed
nausea and vomit ing are f requent ly seen. Low -grade f evers are common; high-
grade f evers are unusual in t he absence of localized or syst emic inf ect ion.
Depending on t he causat ive et iology (i. e. , gallst ones), jaundice may be present .
Biochemical evidence of pancreat ic injury helps t o conf irm t he diagnosis of acut e
pancreat it is. I n t he unexaminable, obt unded pat ient , laborat ory abnormalit ies
may be t he f irst clue of pancreat ic injury. I n acut e pancreat it is a variet y of
digest ive enzymes escape f rom acinar cells and ent er t he syst emic circulat ion.
Amylase and lipase are t he most w idely assayed t o conf irm t he diagnosis.
Amylase levels rise w it hin several hours af t er t he onset of sympt oms and
t ypically remain elevat ed f or 3 t o 5 days. How ever, because of t he short serum
half -lif e of amylase (2. 5 t o 3. 0 hours), levels may normalize w it hin 24 hours of
disease onset . Lipase has a longer serum half -lif e and may be usef ul f or
diagnosing acut e pancreat it is lat er in t he course of an episode. I t is import ant t o
remember t hat t he magnit ude of increases in amylase or lipase concent rat ions
does not correlat e w ell w it h t he severit y of pancreat it is. High levels do not
predict w orse disease. O t her serum biomarkers such as C-react ive prot ein,
neut rophil elast ase, int erleukin-6, procalcit onin, and urinary concent rat ions of
t rypsinogen-act ivat ing pept ide (TAP) t end t o correlat e bet t er w it h disease
severit y. How ever, because assays of t hese markers are not w idely available
t hey are of limit ed clinical ut ilit y t o predict out comes or t riage pat ient s f or I CU
admission.
Radiographic imaging is of t en essent ial t o rule out ot her pot ent ial et iologies,
conf irm t he diagnosis, and st age t he severit y of acut e pancreat it is. Cont rast -
enhanced comput ed t omography (CT) is t he pref erred st udy f ollow ing adequat e
f luid resuscit at ion (Fig. 153. 1). I t has 90% sensit ivit y f or det ect ion of pancreat ic
necrosis, w hich is a f inding t hat is predict ive of severe disease. O n cont rast -
enhanced CT scan, pancreat ic necrosis appears as f ocal or diff use zones of
nonenhanced parenchyma. This f inding may not be evident unt il 48 t o 72 hours
af t er present at ion and delaying t he scan several days t o ident if y local
complicat ions may be benef icial. The import ance of int ravenous (I V) cont rast dye
cannot be overemphasized. How ever, f or pat ient s unable t o receive it , t he
diagnosis of acut e pancreat it is can be inf erred f rom homogeneous glandular
enlargement and t he presence of peripancreat ic f luid collect ions. Magnet ic
resonance imaging (MRI ) depict s necrosis
vividly w hen present and is an excellent alt ernat ive t o CT f or st able pat ient s
unable t o receive I V cont rast dye.
FIG URE 153. 1. Cont rast -enhanced comput ed t omographic scan at t he t ime
of t ransf er show s inf ect ed pancreat ic necrosis w it h a w at er-densit y mass
cont aining gas bubbles, complet ely replacing t he normal pancreas. (Reused
w it h permission f rom Solomkin JS, Moult on JS, Luchet t e FA. Diagnosis and
management of int raabdominal sepsis. I n: I rw in RS, Rippe JM, eds. I rw in
and Rippe's I nt ensive Care Medicine. 5t h ed. Philadelphia: Lippincot t
Williams & Wilkins; 2003: 1670.)
What to Do
The most import ant component of init ial management of acut e pancreat it is is
f luid resuscit at ion and elect rolyt e replacement . I n severe cases, up t o one-t hird
of t he plasma volume can be sequest ered due t o t hird space losses. Urine out put
must be closely monit ored via placement of an indw elling urinary cat het er and
cent ral venous pressure monit oring may by helpf ul in guiding volume
replacement . Mult isyst em organ f ailure is common and early support ive
management (e. g. , mechanical vent ilat ion, renal replacement t herapy,
vasopressors, nut rit ion) is imperat ive.
Pat ient s w it h severe acut e pancreat it is are severely hypercat abolic and
nut rit ional support must st art early t o avoid a malnourished st at e. Accumulat ing
evidence st rongly support s ent eral f eeding over t he parent eral rout e (t ot al
parent eral nut rit ion, or TPN). I t is st rongly recommended t hat ent eral nut rit ion be
init iat ed early af t er init ial
resuscit at ion via t he nasojejunal rout e. I f t his is not possible, nasogast ric f eeds
are also accept able. Parent eral nut rit ion should be implement ed only w hen
at t empt s at ent eral nut rit ion have f ailed af t er a 5- t o 7-day t rial secondary t o
proximal obst ruct ion or ileus. Ent eral nut rient s delivered t o t he jejunum appear t o
have minimal st imulat ory eff ect s on pancreat ic exocrine secret ion and convey t he
t heoret ic advant age of maint aining t he int egrit y of t he int est inal mucosal barrier.
This mechanism pot ent ially limit s bact erial t ranslocat ion, w hich is of t en credit ed
w it h inf ect ing st erile pancreat ic necrosis. Regardless of t he rout e of nut rit ional
support , supplement al glut amine and t ight glycemic cont rol are associat ed w it h
decreased complicat ion rat es. Clinical t rials of agent s t hat inhibit act ivat ed
pancreat ic enzymes (gabexat e mesilat e, aprot inin), inhibit pancreat ic secret ions
(somat ost at in, oct reot ide), or blunt inf lammat ion (Lexipaf ant , ant i–t umor necrosis
f act or α ) have f ailed t o demonst rat e improved out comes in pat ient s w it h acut e
pancreat it is.
I nf ect ion of t he necrot ic pancreas develops in 30% t o 50% of pat ient s w it h CT-
scan-document ed necrosis. Alt hough somet imes evident w it hin a w eek of
present at ion, it s incidence t ends t o peak in t he t hird w eek of t he disease and t he
associat ed mult isyst em organ f ailure and mort alit y escalat e. Alt hough t he exact
mechanism by w hich t he necrot ic pancreat ic t issue becomes inf ect ed is unclear,
gut bact erial t ranslocat ion appears t o be involved. Recent evidence suggest s
t hat pat ient s w it h proven pancreat ic necrosis should receive ant ibiot ic
prophylaxis w it h eit her imipenem or meropenem f or a 14-day course t o reduce
t he risk of inf ect ed necrosis. This prophylact ic t herapy has not been show n t o
increase t he incidence of subsequent f ungal inf ect ions as once f eared.
I nf ect ion of necrot ic pancreat ic t issue is usually suspect ed in pat ient s w ho
develop clinical signs of sepsis. Subsequent CT scans may ident if y air bubbles
w it hin previous areas of st erile necrosis. Ult rasound- or CT-guided f ine-needle
aspirat ion (FNA) of pancreat ic or peripancreat ic t issue w it h G ram st ain and
cult ure should be perf ormed once inf ect ion is suspect ed. I f t he FNA is posit ive
f or t he presence of bact eria, surgical debridement (necrosect omy) and drainage
is indicat ed. O f t en mult iple subsequent debridement s are necessary. Pat ient s
w it h st erile necrosis should not be surgically explored unless t hey display signs
of overw helming clinical det eriorat ion. I f possible, delaying operat ive
necrosect omy and/ or drainage at least 2 t o 3 w eeks conf ers low er mort alit y
rat es and f ew er debridement s.
Suggested Readings
Heinrich S, Schaf er M, Rousson V, et al. Evidence-based t reat ment of acut e
pancreat it is: a look at est ablished paradigms. Ann Surg. 2006; 243(2): 154–
168.
Nat hens AB, Curt is JR, Beale RJ, et al. Management of t he crit ically ill pat ient
w it h severe acut e pancreat it is. Crit Care Med. 2004; 32(12): 2524–2536.
Uhl W, Warshaw A, I mrie C, et al. I nt ernat ional Associat ion of Pancreat ology.
I AP guidelines f or t he surgical management of acut e pancreat it is.
Pancreat ology. 2002; 2(6): 565–573.
> Table of C ontents > Infec tious D is eas e > 154 - W as h Your Hands
154
Wash Your Hands
Sandra Swoboda MSN
“Wash your hands! ” Most healt h care providers have heard t hese simple w ords
over and over again t heir ent ire lives—w het her f rom t heir mot her in childhood or
at w ork f rom inf ect ion-cont rol colleagues. Despit e t he f act t hat t his simple act ion
decreases nosocomial inf ect ion and const it ut es only a f ract ion of a provider's
day, compliance w it h hand hygiene is dismal. Alarmingly t he int ensive care unit
(I CU) is associat ed w it h t he low est overall hand-hygiene compliance rat es (about
50%) in t he hospit al, despit e t he f act t hat t he need f or f requent and eff ect ive
hand hygiene is paramount . The consequences of poor hand hygiene are
unif ormly injurious and cont ribut e t o nosocomial inf ect ions t hat are an import ant
source of morbidit y and mort alit y in pat ient s. O n average, inf ect ions can
complicat e 8% t o 10% of all hospit al admissions; inf ect ion rat es in t he I CU occur
at an even increased incidence. St udies have show n t hat as hand-hygiene rat es
improve, nosocomial inf ect ion rat es go dow n. Reasons associat ed w it h poor
compliance f or proper hand hygiene include higher prof essional st at us;
reluct ance t o experience skin irrit at ion; poor access t o supplies; “being t oo
busy; ” inat t ent ion t o exist ing prot ocols; and t he w earing of gloves (w hich is
t hought t o “negat e” t he need f or hand hygiene).
To brief ly review, hand hygiene encompasses bot h hand w ashing w it h soap and
w at er and hand rubbing w it h w at erless alcohol-based solut ions. Many healt h
care w orkers f alsely believe t hat t radit ional hand w ashing is bot h more
eff icacious and gent le on t he skin. How ever, alcohol-based solut ions are
microbiologically superior, gent ler on skin w it h less disrupt ion t o t he skin's lipid
layers, and more likely t o be used correct ly w hen compared w it h soap-and-w at er
w ashing.
Some common def init ions pert aining t o hand hygiene are as f ollow s:
Hand hygiene: hand w ashing, ant isept ic hand w ash, ant isept ic hand rub,
surgical hand ant isepsis
Hand w ashing: w ashing hands w it h plain (nonant imicrobial) soap and w at er
Hand ant isepsis: ant isept ic hand w ash or ant isept ic hand rub
Ant isept ic hand w ash: w ashing hands w it h soap and w at er or det ergent s
cont aining an ant isept ic agent
Ant isept ic hand rub: applicat ion of w at erless ant isept ic agent t o hands. Does
not require t he use of w at er (alcohol-based product )
The Cent ers f or Disease Cont rol and Prevent ion (CDC) guidelines f or hand
hygiene in t he healt h care set t ing suggest t he f ollow ing:
I f hands are visibly dirt y or cont aminat ed or soiled w it h blood or ot her body
f luids, w ash hands w it h nonant imicrobial soap and w at er or an ant imicrobial
soap and w at er.
I f hands are not visibly soiled, use a w at erless ant isept ic agent (an alcohol-
based hand rub) f or rout ine decont aminat ion.
Perf orm hand hygiene af t er t ouching blood, body f luids, secret ions,
excret ions, and cont aminat ed it ems, w het her or not gloves are w orn.
Perf orm hygiene immediat ely af t er gloves are removed, bet w een pat ient
cont act s, and w hen ot herw ise indicat ed t o avoid t ransf er of micro-organisms
t o ot her pat ient s or environment s. I t may be necessary t o perf orm hand
hygiene bet w een t asks and procedures on t he same pat ient t o prevent cross-
cont aminat ion of diff erent body sit es.
List ed below are some of t he “nast y bugs” t hat surround pat ient s in t he I CU and
procedures f or perf orming hand hygiene w hen providers are exposed t o t hem.
M ethicillin-Resistant Staphylococcus Aureus (M RSA)

Staphyl ococcus aureus is a common bact erium f ound on t he skin, and st rains
t hat are resist ant t o met hicillin are becoming increasingly common. MRSA can be
present in t he nose, on t he skin, or in t he blood or urine and is spread t hrough
physical cont act . MRSA is t ransf erred t o pat ient s by t he hands of healt h care
w orkers w ho have been cont aminat ed by cont act w it h pat ient s, surf aces in t he
w orkplace, or medical devices. This organism can survive f or an hour or more on
environment al surf aces such as f loors, sinks, and blood-pressure cuff s. Hand
hygiene w it h w at erless ant isept ic agent s is recommended.
Vancomycin-Resistant Enterococci (VRE)

There are t w o common clinical isolat es of ent erococci t hat are resist ant t o
vancomycin (Enterococcus f aecal i s and Enterococcus f aeci um). These st rains
can survive f or 60 minut es on gloved and ungloved f ingers. E. f aecal i s has been
recovered on count ert ops f or up t o 5 days and E. f aeci um f or 7 days. Bot h can
survive on bedrails f or up t o 24 hours;
f or 60 minut es on a t elephone receiver; and f or 30 minut es on t he diaphragm of
a st et hoscope. VRE can be spread person t o person by t he hands of healt h care
w orkers or in direct ly on cont aminat ed environment al surf aces and pat ient care
equipment . Hand hygiene w it h w at erless ant isept ic agent s is recommended.
Clostridium Difficile
Cl ostri di um di f f i ci l e ( C. di f f i ci l e) is a spore-f orming, G ram-posit ive anaerobic
bacillus t hat produces t w o endot oxins: endot oxin A and endot oxin B. I t account s
f or 15% t o 25% of all episodes of ant ibiot ic-associat ed diarrhea. I t is shed in
f eces and any surf ace, device, or mat erial (bedpan, t oilet , t hermomet er, bedside
commode, equipment , bedrails, et c. ) t hat is cont aminat ed w it h f eces serves as a
reservoir. Spores are t ransf erred t o pat ient s by t he hands of healt h care
w orkers w ho have t ouched t he cont aminat ed skin of a pat ient , a cont aminat ed
surf ace, or a cont aminat ed piece of equipment .
Alcohol-based product s are not ef f ecti ve against t his spore-f orming bact erium;
t hus t he use of nonant imicrobial soap and w at er or ant imicrobial soap and w at er,
w hich helps t o physically remove spores f rom t he surf ace of cont aminat ed
hands, is required. This is indicat ed especially in an out break sit uat ion.
Bacillus Anthracis (Anthrax)

Baci l l us anthraci s is a spore-f orming bact erium. I t is most commonly f ound in t he
w ild in cat t le, sheep, goat s, camels, ant elopes, and ot her herbivores. How ever,
it can also occur in humans w hen t hey are exposed t o inf ect ed animals or t issue
f rom inf ect ed animals.
Alcohol-based product s do not have act ivit y against t his spore so hands must be
w ashed w it h eit her ant imicrobial or nonant imicrobial soap and w at er.
Acinetobacter
Aci netobacter is a group of G ram-negat ive bact eria t hat is commonly f ound in
soil and w at er. I t can also be f ound on t he skin of healt hy people, especially
healt h care w orkers. While t here are many t ypes or species of Aci netobacter
and all can cause human disease, inf ect ions w it h Acinet obact er are f ort unat ely
not common. Aci netobacter baumanni i is a w at erborne organism t hat is sensit ive
t o f ew ant ibiot ics. This mult iresist ant pat hogen is of t en cult ured f rom pat ient
sput um, respirat ory secret ions, w ounds, and urine. I t has also been f ound in
irrigat ing
solut ions. O ut breaks t ypically occur in I CUs and t his organism can be spread by
person-t o-person cont act , cont act w it h cont aminat ed surf aces, or environment al
exposure. I t can live on t he skin and survive on environment al surf aces f or
several days. Use of alcohol-based hand rub bet w een and bef ore pat ient and
equipment cont act has been show n t o decrease t ransmission during out breaks.
Suggested Readings
Cent ers f or Disease Cont rol and Prevent ion. Hand hygiene in healt hcare
set t ings. ht t p: / / w w w. cdc. gov/ handhygiene/
Noskin G A, St osor V, Cooper I , et al. Recovery of vancomycin-resist ant

ent erococci on f ingert ips and environment al surf aces. I nf ect Cont rol Hosp
Epidemiol. 1995; 16(10): 577–581.
Pit t et D. Hand hygiene: improved st andards and pract ice f or hospit al care.
Curr O pin I nf ect Dis. 2003; 16: 327–335.
Shaf ie S, Alishaq M, G arcia M. I nvest igat ion of an out break of mult i-drug-
resist ant Aci netobacter baumanni i in t rauma int ensive care unit . J Hosp
I nf ect . 2004; 56: 101– 105.
Sw oboda S, Earsing K, St rauss K, et al. Elect ronic monit oring and voice
prompt s improve hand hygiene and decrease nosocomial inf ect ions in an
int ermediat e care unit . Crit Care Med. 2004; 32(2): 358–363.
Sw oboda S, Lipset t P. Handw ashing compliance depends on prof essional

st at us. Surg I nf ect (Larchmt ). 2001; 2(3): 241–245.
> Table of C ontents > Infec tious D is eas e > 155 - C ons ider Ventr ic uloper itoneal S hunt Infec tion in
P atients w ith S eps is
155
Consider Ventriculoperitoneal Shunt Infection in
Patients with Sepsis
Jose I. Suarez MD
Hydrocephalus is a common problem in pat ient s w it h int racranial diseases. Under
normal circumst ances, cerebrospinal f luid (CSF) product ion and absorpt ion are in
dynamic equilibrium w it h approximat e balance bet w een t hem. Many int racranial
processes w ill lead t o eit her increased product ion or decreased absorpt ion of
CSF, overw helming t he balance and t hus leading t o hydrocephalus.
O perat ionally, hydrocephalus can be classif ied as obst ruct ive and
communicat ing. O bst ruct ive hydrocephalus is usually due t o a mass lesion
prot ruding int o t he vent ricular syst em or arising f rom t he ependymal lining.
Communicat ing hydrocephalus can be seen in pat ient s w it h processes in t he
subarachnoid space such as hemorrhage, inf ect ion, or inf lammat ion.
Most CSF (about 50% t o 80%) is produced in t he choroid plexuses t hat line t he
w alls of t he lat eral vent ricles and roof of t he t hird and f ourt h vent ricles. The
brain's ependymal lining and t he brain parenchyma it self are also sources of CSF
product ion. O nce f ormed, CSF circulat es t hroughout t he vent ricular syst em, exit s
t he f oramina of Magendie and Luschka in t he f ourt h vent ricle, circulat es t hrough
t he subarachnoid space of t he brain and spinal cord, and is event ually absorbed
by t he arachnoid villi int o t he venous syst em.
I n adult s sympt oms of hydrocephalus include headaches, diplopia, and ment al
st at us changes. Clinical signs may include papilledema and impaired upw ard
gaze. Sudden deat h may occur w it h severe increases in int racranial pressure. I n
pat ient s w it h chronic hydrocephalus (i. e. , normal-pressure hydrocephalus among
ot hers), sympt oms may include gait dist urbance, urinary incont inence, and
memory loss w it h or w it hout any ot her signs or sympt oms of elevat ed int racranial
pressure as described previously.
Acute CSF Diversion

Acut e hydrocephalus is best t reat ed w it h insert ion of an int ravent ricular cat het er
(I VC). I VCs are placed via burr holes or similar t echniques int o t he vent ricular
syst em, usually t he lat eral vent ricles (Fi g. 155. 1). I VCs can be used t o
t emporarily drain CSF in t he hope t hat
normal CSF circulat ion is rest ored (e. g. , in t he case of int ravent ricular
hemorrhage). The insert ion sit e of I VCs is usually t he right f ront al or post erior
pariet al region. The right side is usually chosen, as it is rarely t he dominant
hemisphere. I VC placement has an overall inf ect ion rat e of 4% t o 10%. The
opt imal durat ion of an I VC in order t o decrease inf ect ion has not been clearly
est ablished. How ever, many pract it ioners recommend insert ing a new I VC by day
5 if t here is st ill a need. The common organisms include coagulase-negat ive
Staphyl ococcus, met hicillin-resist ant Staphyl ococcus, and G ram-negat ive
bact eria such as Escheri chi a Col i , Kl ebsi el l a, and Pseudomonas aerugi nosa. I f
a pat ient w it h an I VC becomes f ebrile and/ or sept ic, CSF sampling should be
obt ained and appropriat e ant ibiot ic coverage should be empirically init iat ed.
Recommended empiric ant ibiot ics include t he combinat ion of a t hird-generat ion
cephalosporin and vancomycin, or meropenem and vancomycin. St erile t echnique
should be used f or CSF sampling and it should be obt ained f rom t he sideport of
t he ext ernal t ubing syst em.
FIG URE 155. 1.
Permanent Shunt Placement

Most permanent shunt s in adult s include vent riculoperit oneal shunt s (VPSs).
O t her t echniques include vent riculoat rial or vent riculopleural shunt s. A complet e
descript ion of t he surgical t echnique is beyond t he scope of t his chapt er.
How ever, a summary of VPS insert ion is present ed in (Fi gures 155. 2 and 155. 3. )
O ne import ant point t o realize is t hat VPSs have a valve and a reservoir t hat are
at t ached t o t he t ubing. The valve is used t o grade t he pressure at w hich CSF is
drained and current devices have programmable valves t hat can be easily
adjust ed. The I VC reservoir is usually f elt near t he ent ry burr hole. The side arm
of t he reservoir is at t ached t o t he proximal end of t he valve and subsequent ly t o
t he perit oneal t ubing. There are several t ypes of VPSs and each has it s ow n
peculiarit ies. How ever, most VPSs have a pumping mechanism, w hich is usually
locat ed at t he valve, or a reservoir t hat should be used t o t est proper
f unct ioning. The most common complicat ions af t er VPS insert ion are obst ruct ion,
disconnect ion, and inf ect ion.
VPS Infection
I nf ect ion remains an import ant complicat ion of VPSs. Mort alit y has been
report ed as being high (up t o 40% in one series). Thus, it is very import ant t o
have a high index of suspicion t o diagnose t hese inf ect ions. The overall inf ect ion
rat e af t er VPS insert ion has been est imat ed t o be around 7% w it h an acut e rat e
of inf ect ion of about 6%. Approximat ely 70% of VPS inf ect ions w ill manif est
t hemselves w it hin t he f irst 2 mont hs af t er insert ion. This underscores t he f act
t hat most VPS inf ect ions probably result f rom direct cont aminat ion at t he t ime of
surgery w it h f ew st emming f rom ot her sources such as hemat ogenous seeding,
ext ension f rom cont iguous t issue, or direct exposure of t he t ubing. Thus, it is not
surprising t hat most organisms f ound are skin commensals such as
Staphyl ococcus epi dermi di s. How ever, met hicillin-resist ant S. aureus and G ram-
negat ive bact eria such as E. Col i , Kl ebsi el l a pneumoni ae, and P. aerugi nosa are
also import ant organisms t o consider, since t hey have been report ed in up t o
20% of cases.
The clinical present at ion of VPS inf ect ions varies. Pat ient s may experience
sw elling and redness over t he VPS t ract or have signs of perit onit is. Sept icemia
is also possible but is more common in pat ient s w it h vent riculoat rial shunt s.
Pat ient s can present acut ely w it h headache, nausea, vomit ing, f ever, and
decreased level of consciousness. Frank meningeal signs can also be seen.
Because t he clinical signs of VPS inf ect ions can be nonspecif ic, pract it ioners
should have a high index of suspicion and secure CSF
sampling f or cult ure as soon as f easible. This is even more import ant in pat ient s
w it h unexplained sepsis, unexplained f ever, or shunt malf unct ion. A single
negat ive CSF cult ure is not suff icient t o rule out VPS inf ect ion; many expert s
recommend serial CSF sampling (alt hough t he number of cult ures needed has not
been det ermined) t o increase t he yield f or det ect ing t he culprit organisms. CSF
sampling should be done under st erile t echniques and by direct punct uring of t he
reservoir. CSF should be sent f or aerobic and anaerobic bact eria and f ungal
organisms. I t is equally import ant t o send cult ures f rom t he various incision sit es
along t he VPS t ract .
FIG URE 155. 2.
Treat ment of VPS inf ect ions usually involves medical and surgical management .
The medical management usually consist s of prompt init iat ion of ant ibiot ics and
support ive general care. Ant ibiot ic t herapy should include a t hird-generat ion
cephalosporin (because of good CSF penet rat ion) and vancomycin (usually at 1. 5
t imes t he usual
dosage). O t her alt ernat ives include meropenem and vancomycin. I t is common
pract ice t o add int rat hecal ant ibiot ics (usually gent amicin or vancomycin) daily.
Ant ibiot ics should be cont inued f or at least 2 t o 3 w eeks af t er negat ive CSF
cult ures are obt ained. I n addit ion t o ant ibiot ic administ rat ion, VPS hardw are
should be removed and a new I VC insert ed. O ld VPS t ips should be sent f or
cult ure. O nce t he cult ures are negat ive f rom daily CSF sampling f rom t he I VC,
t hen a new VPS can be reinsert ed. Many neurosurgeons w ait f or CSF cult ures t o
remain negat ive f or 72 t o 120 hours prior t o considering VPS reinsert ion.
FIG URE 155. 3.
Suggested Readings
Borgbjerg BM, G jerris F, Albeck MJ, et al. Risk of inf ect ion af t er
cerebrospinal f luid shunt : an analysis of 884 f irst -t ime shunt s. Act a Neurochir
(Wien). 1995; 136: 1–7.
Edw ards MS, Baker CJ, But ler KM, et al. Penet rat ion of cef uroxime int o
vent ricular f luid in cerebrospinal f luid inf ect ions. Ant imicrob Agent s
Chemot her. 1989; 33: 1108–1110.
> Table of C ontents > Infec tious D is eas e > 156 - R ec ogniz e that Vanc om yc in has ver y P oor C entr al
Ner vous S ys tem P enetr ation
156
Recognize that Vancomycin has very Poor
Central Nervous System Penetration
Eliahu S. Feen MD
Jose I. Suarez MD
Watch Out For

Unt il t he lat e 1960s, penicillin w as considered opt imal ant ibiot ic t herapy f or
meningit is secondary t o Streptococcus pneumoni ae and some st aphylococcal
species. Beginning w it h t he earliest report s in t he lat e 1960s, t here has been
increasing recognit ion of penicillin-resist ant pneumococcal and st aphylococcal
meningit is. For many years cef ot axime and cef t riaxone, broad-spect rum
cephalosporins, w ere adequat e t o t reat most st rains of t hese penicillin-resist ant
bact eria. How ever, w it h t he advent of st rains of pneumococcus t hat had even
higher levels of penicillin resist ance, alt ernat ive t herapies w ere sought .
Vancomycin is one of t he most eff ect ive alt ernat ive ant ibiot ics f or penicillin- and
cephalosporin-resist ant pneumococcal meningit is. I t has also been a mainst ay of
t herapy f or met hicillin-resist ant Staphyl ococcus aureus. The problem w it h using
vancomycin t o t reat meningit is lies in it s errat ic penet rat ion of t he blood–brain
barrier (BBB). Evidence f rom bot h animal models and human pat ient s
consist ent ly demonst rat es t hat w hen vancomycin is administ ered int ravenously,
only about 10% of t he serum level is f ound in t he cerebrospinal f luid (CSF).
I nf lammat ion w it hin t he meninges seems t o play a role in inhibit ing crossing of
t he BBB. The use of st eroids in t he acut e t reat ment of meningit is has become
more prevalent in light of recent evidence in bot h children and adult s t hat st eroid
use acut ely lessens t he burden of neurologic sequelae. How ever, t he use of
st eroids also seems t o limit t he amount of vancomycin t hat passes t hrough t he
BBB. As an example, in animal models, t he addit ion of dexamet hasone has been
f ound t o low er t he concent rat ion of vancomycin in t he CSF. Eit her t hrough
st abilizat ion of t he BBB by decreasing inf lammat ion or t hrough some ot her
mechanism, penet rat ion of t he BBB by vancomycin is impaired.
What to Do
I n order t o count er t his impaired penet rat ion of t he BBB, one st rat egy has been
t o use higher dosages of vancomycin. When t he vancomycin dose f or t reat ing
st andard syst emic inf ect ions is used, CSF levels are usually inadequat e
t o achieve appropriat e bact ericidal act ivit y. By increasing t he dose by one-t hird,
some researchers have show n clinical success in t reat ing penicillin-resist ant
pneumococcal meningit is. How ever, w hile using higher dosages of vancomycin
can increase t he CSF levels, it can be diff icult t o maint ain consist ent ly high
levels, and in addit ion t here is increased risk of vancomycin-induced hypoacusia
and ot her side eff ect s. Vancomycin alone as t herapy f or pneumococcal and
st aphylococcal disease, t heref ore, is not recommended (some experienced
int ensive care unit [ I CU] pharmacist s f eel t hat vancomycin should never be used
alone or in conjunct ion w it h dexamet hasone). By adding vancomycin t o
cef ot axime or cef t riaxone (or anot her broad-spect rum cephalosporin), a
synergist ic eff ect is achieved. This is one st rat egy t o employ vancomycin in t he
t reat ment of bet a-lact am-resist ant organisms.
Some clinicians use rif ampin as an alt ernat ive t o vancomycin f or resist ant
pneumococcal meningit is. Rif ampin alone is inadequat e, bot h because of t he
report s of rif ampin resist ance now ext ant and also because of t he rapid
development of bact erial resist ance t hat can occur in cert ain set t ings. Animal
and in vit ro evidence support s t he addit ion of rif ampin t o broad-spect rum
cephalosporins. Some research has show n t hat t he addit ion of rif ampin t o t he
cephalosporins has as much eff icacy as adding vancomycin t o t he
cephalosporins. I n summary, w hen a pat ient has pneumococcal or st aphylococcal
meningit is, if t here is any evidence t hat t he causat ive organism carries bet a-
lact am resist ance, vancomycin or rif ampin should be added t o a regimen of
ant ibiot ics including broad-spect rum cephalosporins such as cef ot axime or
cef t riaxone.
Suggested Readings
Ahmed A, Jaf ri H, Lut sar I , et al. Pharmacodynamics of vancomycin f or t he
t reat ment of experiment al penicillin- and cephalosporin-resist ant
pneumococcal meningit is. Ant imicrob Agent s Chemot her. 1999; 43: 876–881.
American Academy of Pediat rics, Commit t ee on I nf ect ion Diseases. Therapy

f or children w it h invasive pneumococcal inf ect ions. Pediat rics. 1997; 99: 289–
299.
G ump DW. Vancomycin f or t reat ment of bact erial meningit is. Rev I nf ect Dis.
1981; 3: S289–S292.
Paris MM, Ramilo O , McCracken G H Jr. Management of meningit is caused by

penicillin-resist ant St rept ococcus pneumoniae. Ant imicrob Agent s Chemot her.
1995; 39: 2171–2175.
Viladrich PF, G udiol F, Linares J, et al. Evaluat ion of vancomycin f or t herapy

of adult pneumococcal meningit is. Ant imicrob Agent s Chemot her.
1991; 35: 2467–2472.
> Table of C ontents > Infec tious D is eas e > 157 - B e Aler t for S eiz ur es w ith Im ipenem Us e
157
Be Alert for Seizures with Imipenem Use
Anthony D. Slonim MD, DRPH
I mipenem, meropenem, and ert apenem are members of t he class of ant ibact erial
agent s named carbapenems. Carbapenems are one of f our t ypes of bet a-lact am
ant ibiot ics. The ot her t hree are t he penicillins, cephalosporins, and
monobact ams. I mipenem shares many of t he ot her charact erist ics of t hese bet a-
lact am ant ibact erials: a bact ericidal mechanism of act ion; renal excret ion; and a
broad spect rum of act ivit y t hat covers many G ram-posit ive and G ram-negat ive
aerobic organisms. I n t he case of imipenem and many of t he penicillins, act ivit y
also includes anaerobes. I mipenem, in it s commercial f ormulat ion, is complexed
w it h cilast in t o reduce it s renal met abolism.
Watch Out For

Carbapenems have a number of adverse eff ect s t hat can be classif ied by t he
aff ect ed syst em. Carbapenems can cause hypersensit ivit y react ions. These
react ions w ill of t en present as skin rashes, urt icaria, or even St evens-Johnson
syndrome in severe cases. G ast roint est inal side eff ect s include nausea,
vomit ing, and abnormal liver-f unct ion t est s and t ransaminases. From a bone-
marrow perspect ive, marrow suppression and a posit ive Coombs t est can result .
Renal dysf unct ion can result f rom imipenem use and all bet a-lact am ant ibact erial
agent s (including imipenem) require dosage adjust ment in t he set t ing of renal
insuff iciency and f ailure because of t heir renal excret ion. Neurologically,
imipenem can cause myoclonus and seizures. Seizures are a f airly common
complicat ion and result f rom a low ering of t he seizure t hreshold in suscept ible
pat ient s. Fort unat ely, t his complicat ion is not nearly as common w it h t he use of
meropenem. Seizures also occur in t he set t ing of new -onset renal dysf unct ion if
t he carbapenem dose is not adjust ed.
I n pat ient s w here t he risk of seizures is prohibit ed, ot her agent s t hat can be
used in place of imipenem include t he broad-spect rum cephalosporins or t he
penicillins. I t must be not ed t hat cephalosporins w ill not provide t he coverage of
ent erococcus or anaerobes t hat imipenem provides and penicillins, part icularly
t he ext ended-spect rum penicillins, w hile providing adequat e ant imicrobial
coverage might also cause seizures and t w it ching. Q uinolones may be anot her
alt ernat ive ant ibact erial, but t hey are also know n t o cause seizures.
Suggested Readings
G oldman L, Ausiello D, eds. Cecil Text book of I nt ernal Medicine. 22nd ed.
Philadelphia: Saunders; 2004: 1753–1764.
Helinger WC, Brew er NS. Carbapenems and monobact ams: imipenem,

meropenem, and azt reonam. Mayo Clin Proc. 1999; 74: 420–434.
> Table of C ontents > Infec tious D is eas e > 158 - R em em ber if ther e is a Nor m al P latelet C ount it
C annot be Hantavir us P ulm onar y S yndr om e
158
Remember if there is a Normal Platelet Count it
Cannot be Hantavirus Pulmonary Syndrome
Several species of Hant avirus (f amily Bunyaviridae) can cause disease in
humans, but in t he Unit ed St at es, Sin Nombre virus is t he most not orious and
severe. Sin Nombre virus w as f irst report ed in 1993 in t he Four Corners region
of t he sout hw est ern Unit ed St at es (Arizona, Colorado, New Mexico, Ut ah).
Physicians w orking f or t he I ndian Healt h Service in t hat region not ed a clust er of
young healt hy persons w ho developed severe respirat ory illnesses charact erized
by rapidly progressive, noncardiogenic pulmonary edema, and a high case
f at alit y rat e. Though it w as f irst clinically det ect ed in t he Unit ed St at es in 1993,
ret rospect ive serologic dat a as early as 1959 indicat es t hat Sin Nombre virus
w as present decades earlier. From 1993 t o 2005, 396 cases of hant avirus
pulmonary syndrome (HPS) w ere report ed in t he Unit ed St at es, w it h 20 t o 50
cases occurring annually.
The epidemiology of HPS in t he Unit ed St at es usually aff ect s healt hy adult s
(mean age 38 years) residing in rural areas. Approximat ely 77% of aff ect ed
pat ient s are Caucasian and 19% American I ndian. Hant aviruses are t ransmit t ed
t o humans by rodent s via aerosolizat ion of inf ect ed excret a. While cont aminat ed
urine is most commonly implicat ed, rodent droppings, saliva, and nest ing
mat erials can also aerosolize t o cause illness. Rarely, hant aviruses may also be
t ransmit t ed t hrough direct mucous-membrane cont act or inoculat ion. These
viruses do not cause apparent illness in t heir rat reservoirs. Several murine
species can t ransmit hant aviruses, but in t he Unit ed St at es, deer mice
(Peromyscus mani cul atus) are t he primary reservoir f or Sin Nombre virus. The
cot t on rat , rice rat , and w hit e-f oot ed mouse t ransmit ot her hant avirus species in
t he Unit ed St at es. Theref ore, given t hat t here is neit her an eff ect ive vaccine nor
t herapy, t he primary st rat egy f or combat ing HPS is rat cont rol. No human-t o-
human cases have been report ed, nor have pet s or f arm animals been implicat ed
in disease t ransmission.
The diff erent ial diagnosis includes ot her inf ect ious diseases t hat at t ack
immunocompet ent pat ient s w ho have t he environment al exposure. This includes
pulmonary ant hrax, plague, t ularemia, hist oplasmosis, coccidioidomycosis,
lept ospirosis, Lyme disease, ehrlichiosis, chlamydia, mycoplasma, Q f ever, and
Legionnaires disease.
The incubat ion period of HPS is approximat ely 3 w eeks af t er t he pat ient inhales
t he virus, during w hich t ime t he pat ient is generally asympt omat ic. The prodrome
last s f or 3 t o 5 days w it h nonspecif ic sympt oms of f ever, chills, myalgias,
headache, and nausea, vomit ing, and diarrhea. There may also be cough,
dizziness, art hralgias, dyspnea, t achypnea, and t achycardia. G enerally,
rhinorrhea and sore t hroat are absent as are conjunct ival and cut aneous
sympt oms.
Watch Out For

During t he prodrome phase t he plat elet count is almost alw ays normal but a
dramat ic decrease in t he plat elet s marks t he beginning of t he pulmonary edema
phase. I n addit ion t o t hrombocyt openia, t here is a leukocyt osis w it h a marked
lef t shif t and t he presence of at ypical lymphocyt es. There is also of t en
hypoalbuminemia, elevat ed t ransaminases, and elevat ed serum lact at e.
Clinically, t he drop in plat elet s corresponds w it h a severe and acut e
decompensat ion w it h a rapid progression t o sept ic shock w it h a low cardiac
out put and an increased syst emic vascular resist ance. Pat ient s may progress t o
sinus bradycardia and malignant vent ricular arryt hmias. Pat ient s invariably
require int ubat ion and show a bilat eral diff use int erst it ial edema and pleural
eff usion.
What to Do
The t reat ment of HPS is largely support ive. While t he ant iviral compounds have
been show n t o inhibit hant aviruses in vit ro, t heir clinical eff icacy has not been
proven. Early aggressive int ensive care unit (I CU)-level care w it h caref ul
monit oring of oxygenat ion and hemodynamic paramet ers is imperat ive, as
decompensat ion and deat h can occur w it hin hours. The pat ient should receive
f luid resuscit at ion, inot ropic support , and mechanical vent ilat ion. Blood and
sput um samples should be sent t o a ref erence laborat ory (e. g. , Cent ers f or
Disease Cont rol and Prevent ion) f or det ect ion of t he Hant avirus RNA sequence
by polymerase chain react ion (PCR) or det ect ion of Hant avirus-specif ic I gM or a
rising I gG t it er. Even w it h excellent management , mort alit y st ill ranges f rom 30%
t o 40%. Somew hat surprisingly, HPS pat ient s w ho survive t he inf ect ion have a
rapid recovery and are of t en ext ubat ed and discharged w it hin days of t he acut e
decompensat ion.
Suggested Readings
Dolin R. Principles and Pract ice of I nf ect ious Diseases. Philadelphia: Elsevier;
2005.
O ut break of acut e illnessâ s out hw est ern Unit ed St at es, 1993. Morb Mort al
Wkly Rep. 1993; 42: 421.
> Table of C ontents > Infec tious D is eas e > 159 - C ons ider E m pir ic Helic obac ter P ylor i Tr eatm ent w hen
Gas tr ic or D uodenal Ulc er s ar e Found
159
Consider Empiric Helicobacter Pylori Treatment
when Gastric or Duodenal Ulcers are Found
Pept ic ulcer disease is a common condit ion t hat manif est s it s changes on eit her
t he gast ric or duodenal mucosa and account s f or nearly half of all upper
gast roint est inal bleeds admit t ed t o t he int ensive care unit (I CU). The t w o major
predisposing f act ors f or pept ic ulcer disease are inf ect ions w it h Hel i cobacter
pyl ori and nonst eroidal ant i-inf lammat ory drug (NSAI D) use. The lesions caused
by t hese t w o agent s depend not only on t he agent s but on t he charact erist ics of
t he gast roint est inal mucosa in t he area t hat t hey aff ect . H. pyl ori is a G ram-
negat ive organism t hat grow s w ell in t he acidic environment of t he duodenal
bulb. The damage caused by t he bact eria in t his locat ion leads t o f urt her injury
and colonizat ion by addit ional bact eria, w hich perpet uat es t he inf lammat ion. I n
cont rast , NSAI Ds inhibit cyclo-oxygenase, t he enzyme t hat assist s in
prost aglandin product ion. Prost aglandins are an import ant def ense mechanism
f or t he gast ric mucosa and inhibit ion result s in a propensit y f or ulcer f ormat ion.
O f import ance f or pat ient s w it h NSAI D-induced ulcer disease, t reat ment of H.
pyl ori inf ect ion, even empirically, might help in achieving a cure and alleviat ing
recurrences. Smoking and st ress cont ribut e t o t he occurrence of bot h t ypes of
pept ic ulcer disease and pat ient s should be encouraged t o reduce bot h t ypes of
insult .
H. pyl ori inf ect ion is a chronic condit ion. The organism can be ident if ied by using
several diff erent diagnost ic t est s. Noninvasive diagnost ic t echniques include an
enzyme-linked immunosorbent assay (ELI SA) t hat is available f or t he det ect ion
of ant ibodies t o H. pyl ori, a st ool ant igen t est , and a urea breat h t est t hat can
be used t o det ect t he presence of organisms in gast rit is or pept ic ulcer disease.
These t est s have sensit ivit ies and specif icit ies t hat exceed 85%, but clinicians
need t o realize t hat f alse posit ive and f alse negat ive t est s can st ill occur,
part icularly w hen pat ient s have been pret reat ed w it h ant ibiot ics. I n addit ion,
endoscopic specimens can be t est ed f or t he det ect ion of H. pyl ori on hist ologic
examinat ion or cult ure, but t hese t est s require t hat t he pat ient undergo
endoscopy.
What to Do
The t reat ment of pept ic ulcer disease needs t o include t herapy f or t he
hyperacidit y and t he underlying H. pyl ori inf ect ion. How ever, during acut e
bleeding episodes in t he I CU, t he f ocus should be on resuscit at ion, t he
rest orat ion of hemodynamic st abilit y, and t he t reat ment of hyperacidit y. The
t reat ment of H. pyl ori w ill occur af t er t he acut e gast roint est inal bleeding
subsides. This t reat ment should include hyperacidit y t reat ment and t reat ment
w it h t w o ant ibiot ics. A number of eff ect ive regimens exist (e. g. t he “prevpack”)
and are able t o achieve cure rat es of up t o 90% f or t hose pat ient s w ho complet e
t he 2-w eek course of t herapy.
Accept able regimens f or t he t reat ment of H pylori inf ect ion are (all f or one
w eek):
PPI bid
amoxicillin 1 gm bid
clarit hromycin 500 mg bid
or
PPI bid
met ronidazone 500 mg bid
clarit hromycin 250 mg bid
or
PPI bid
Pept o Bismol 2 t abs qid
met ronidazole 250 mg qid
t et racycline 500 mg qid
Suggested Readings
Conrad SA. Acut e upper gast roint est inal bleeding in crit ically ill pat ient s:
causes and t reat ment modalit ies. Crit Care Med. 2002; 30: S365–S368.
Huang JQ , Sridhar S, Hunt RH. Role of Hel i cobacter pyl ori inf ect ion and non
st eroidal ant i-inf lammat ory drugs in pept ic ulcer disease: a met a analysis.
Lancet . 2002; 35: 14– 22.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 160 - Adm inis ter E m pir ic B r oad- S pec tr um Antibiotic s
w hen a P atient m ay be in S eptic S hoc k
160
Administer Empiric Broad-Spectrum Antibiotics
when a Patient may be in Septic Shock
William R. Burns III MD
Sepsis is a st at e of physiologic derangement at t ribut ed t o eit her know n or
suspect ed inf ect ion (Tabl e 160. 1). When accompanied by organ dysf unct ion
and/ or hypoperf usion, t he condit ion is t ermed severe sepsis. Sept ic shock is
diagnosed if severe sepsis and hypot ension persist in spit e of f luid resuscit at ion.
Despit enumerous at t empt s t o minimize t he morbidit y and mort alit y associat ed
w it h t hese condit ions, t he proper administ rat ion of empiric, broad-spect rum
ant imicrobials is one of t he f ew int ervent ions t o have demonst rat ed ut ilit y.
What to Do
O nce severe sepsis or sept ic shock is recognized, resuscit at ive measures should
be init iat ed immediat ely, if not already under w ay. Eff icient delivery of broad-
spect rum ant imicrobial t herapy is a t reat ment of crit ical import ance. Pot ent ial
sources of inf ect ion, common pat hogens, pat ient risk f act ors, and local
resist ance pat t erns should be considered t o det ermine t he ideal drugs f or f irst -
line administ rat ion. These should almost alw ays include broad bact erial coverage
(G ram-posit ives, G ram-negat ives, and anaerobes), may of t en be direct ed at
resist ant G ram-posit ives such as met hicillin-resist ant Staphyl ococcus aureus
(MRSA) and/ or vancomycin-resist ant ent erococci (VRE), and may occasionally
t arget virulent G ram-negat ive bacilli and/ or f ungi. I n pat ient s w it h severe sepsis,
t he use of an ant ipseudomonal penicillin (such as piperacillin/ t azobact am), a
f ourt h-generat ion cephalosporin (such as cef epime), or a carbapenem (such as
meropenem) is recommended; alt ernat ively, an ant ipseudomonal f luoroquinolone
(such as ciprof loxacin) in addit ion t o met ronidazole is suit able f or t hose w it h a
bet a-lact am allergy. The use of vancomycin (f or MRSA coverage) or linezolid (f or
VRE coverage), aminoglycosides (f or ext ended G ramnegat ive coverage), and
f luconazole (f or ant if ungal coverage) is of t en indicat ed f or pat ient s experiencing
sept ic shock or specif ic cases of severe sepsis.
While medicat ions are being prepared (all orders f or ant ibiot ics should be w rit t en
on a “st at ” basis), appropriat e cult ures should be obt ained. Typically blood f rom
t w o peripheral sit es, urine, and sput um
are sampled. Furt hermore, samples f rom vascular access and percut aneous
drainage cat het ers may be analyzed as w ell. Microbial t it ers are highest at t he
t ime of diagnosis of severe sepsis or sept ic shock and t he abilit y t o isolat e
pat hogens becomes increasingly more diff icult w it h ant ibiot ic t reat ment . As such,
acquisit ion of cult ures should occur expedit iously. This process, how ever, should
not delay t he administ rat ion of ant imicrobial t herapy, w hich should ideally be
delivered w it hin t he f irst hour of diagnosis.
TABLE 160-1 PHYSIOLOGICAL STATES
Presumed or
documented T>38° C or T<36° C;
infection, HR>90; RR>20 or
Sepsis resulting in 2 or PCO 2 <32; W BC>12,000 or
more of the W BC<4,000 or bands>10%
following of neutrophils
criteria
Criteria for Organ dysfunction,

Severe
sepsis, plus the hypoperfusion, or
sepsis
following hypotension
Hypotension despite
Criteria for resuscitation, perfusion
Septic severe sepsis, abnormalities (lactic
shock plus the acidosis, oliguria, acute
following alteration in mental status),
vasopressor use
T, temperature; HR, heart rate; RR, respitory rate;

W BC, white blood cells.
Beyond ant ibiot ic t herapy, sept ic pat ient s require support ive t herapy and
int ensive monit oring. Addit ional cult ures must be obt ained in t he case of
w orsening physiologic derangement s, and considered every 24 hours. Cult ure
dat a and sensit ivit y inf ormat ion should be review ed regularly t o consider
narrow ing ant imicrobial t herapy and/ or t he inst it ut ion of coverage f or unt reat ed
organisms. I f t here is no evidence of MRSA and/ or VRE w it hin 48 t o 72 hours,
ant ibiot ics specif ically t arget ing t hese pat hogens should be discont inued.
O t herw ise, it is of t en prudent t o cont inue a short course of empiric t herapy w hile
init ial and f ollow -up cult ures are pending, keeping in mind t hat f ungi of t en t ake 7
days t o grow in cult ure (if t hey grow at all).
Selected Readings
I brahim EH, Sherman G , Ward S, et al. The inf luence of inadequat e
ant imicrobial t reat ment in bloodst ream inf ect ions on pat ient out comes in t he
I CU set t ing. Chest . 2000; 118: 146–155.
Leibovici L, Shraga I , Drucker M, et al. The benef it of appropriat e empirical

ant ibiot ic t reat ment in pat ient s w it h bloodst ream inf ect ion. J I nt ern Med.
1998; 244: 379–386.
Souba WW, Fink MP, Jurkovich G J, et al. , eds. ACS Surgery: Principles and
Pract ice. New York: WebMD Prof essional Publishing; 2006: 1260–1270.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 161 - D o not Adm inis ter the C os yntr opin tes t w ithin
24 Hour s of Us ing E tom idate
161
Do not Administer the Cosyntropin test within 24
Hours of Using Etomidate
Meghan C. Tadel MD
Hypoadrenalism is a lif e-t hreat ening syndrome in crit ically ill pat ient s. I t can be
f ound in pat ient s w it h chronic adrenal insuff iciency as w ell as t hose w it h adrenal
suppression w ho are simply unable t o mount t he expect ed st ress response t o
common condit ions in t he int ensive care unit (I CU), including inf ect ion, sepsis,
t rauma, and surgery. Adrenal insuff iciency can manif est as a broad number of
signs and sympt oms including w eakness, hyponat remia, hypoglycemia, disrupt ion
of acidbase st at us, and hypot ension (specif ically hypot ension ref ract ory t o
t herapy w it h pressors). Pat ient s suff ering f rom hypoadrenalism have signif icant ly
reduced mort alit y w hen t his condit ion is diagnosed and t herapy init iat ed.
The w idely accept ed t est f or primary adrenal insuff iciency is t he cosynt ropin
st imulat ion t est . Cosynt ropin is a synt het ic analogue of adrenocort icot ropic
hormone (ACTH), consist ing of t he f irst 24 amino acids of t he nat urally occurring
ACTH. This t est consist s of a baseline blood sample sent f or cort isol level,
f ollow ed by a single int ravenous (I V) dose of 250 µg of cosynt ropin, w it h f urt her
blood samples collect ed at 30 and 60 minut es af t er administ rat ion, alt hough t he
30-minut e measure may be superf luous as t he cort isol levels peak at 60 minut es.
There exist s some cont roversy over w hat const it ut es a posit ive and negat ive
st imulat ion t est , but even t he more conservat ive proponent s seem t o agree t hat
adrenal insuff iciency is not present if t he baseline cort isol level is >20 µg/ dL and
a normal response t o cosynt ropin st imulat ion is present (i. e. t he cort isol level
increases by great er t han 9 µg/ dL in response t o t he above t est ). I f a pat ient
meet s t hese crit eria, he or she does not have adrenal insuff iciency and no
supplement at ion of adrenal hormones is necessary.
O ne cause of adrenal insuff iciency t hat is of t en seen in t he I CU is t he prior
administ rat ion of et omidat e. Et omidat e is an imidazole, nonbarbit urat e, sedat ive-
hypnot ic. I t w as f ormerly used as a long-t erm sedat ive inf usion f or I CU pat ient s
but w as f ound t o have high rat es of associat ed hypoadrenalism w it h poor
out comes and such use has been most ly abandoned. How ever, et omidat e is st ill
f requent ly used, part icularly in t he I CU set t ing as an agent f or induct ion of
anest hesia, part icularly f or int ubat ing unst able pat ient s. This is because
et omidat e
is accept ed as having a more hemodynamically st able prof ile. More recent

invest igat ions of usage of et omidat e demonst rat e t hat even t his single int ubat ing
dose of et omidat e (st andard dose 0. 3 mg/ kg) can cause adrenal suppression f or
anyw here f rom 5 t o 24 hours. This is because et omidat e inhibit s bot h 11 β -
hydroxylase and cholest erol side chain cleavage enzyme, bot h of w hich are
act ive in adrenal st eroid synt hesis. Thus, t he use of et omidat e w it hin t he
previous 24 hours w ill obscure t he result s of a cosynt ropin st imulat ion t est .
Suggested Readings
Annane D. I CU physicians should abandon t he use of et omidat e! I nt ens Care
Med. 2005; 31: 325–326.
Miller R, ed. Miller's Anest hesia. 6t h ed. New York: Elsevier/ Churchill
Livingst one; 2005: 350–355, 1041.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 162 - S w itc h fr om Hydr oc or tis one to D exam ethas one
if the C os yntr opin S tim ulation Tes t is to be Adm inis ter ed
162
Switch from Hydrocortisone to Dexamethasone
if the Cosyntropin Stimulation Test is to be
Administered
Meghan C. Tadel MD
Pat ient s diagnosed w it h adrenal insuff iciency are generally t reat ed w it h
hydrocort isone in t he range of 60 t o 80 mg int ravenously (I V), 3 t o 4 t imes daily
f or a t ot al dose of 200 t o 300 mg per day. How ever, in a pat ient w ho has not had
a cosynt ropin st imulat ion t est t o conf irm t he diagnosis of adrenal insuff iciency,
hydrocort isone should be sw it ched t o dexamet hasone 2 t o 4 mg I V every 6 t o 8
hours f or t hree doses unt il t hat t est can be complet ed. Since t herapy w it h
supraphysiologic doses of any cort icost eroid f or even as f ew as f ive days can
result in adrenal suppression f or up t o 1 year f ollow ing cessat ion of t herapy, use
of st eroids should not be undert aken light ly and as such a st imulat ion t est should
be considered. The reason f or sw it ching f rom hydrocort isone t o dexamet hasone
is t hat hydrocort isone direct ly int erf eres in t he assay used by labs t o det ermine
cort isol levels in blood samples w hile dexamet hasone does not .
Addit ionally, t here may be sit uat ions w here a pat ient 's clinical pict ure
(hypot ension ref ract ory t o pressor inf usion) w arrant s immediat e t herapy w it h
glucocort icoids even bef ore t he st imulat ion t est can be init iat ed. I n t hese
inst ances, t herapy can begin immediat ely w it h dexamet hasone 4 t o 8 mg I V
bolus w it h considerat ion given t o adding a f ludrocort isone (a mineralocort icoid)
50 µg (PO ) per day. O nce t he cosynt ropin t est has been complet ed, t herapy can
be convert ed t o hydrocort isone. Mineralocort icoids are equally aff ect ed by
adrenal insuff iciency and play an import ant role in volume st at us and elect rolyt e
balance so t hey should not be overlooked in t he t herapy of hypoadrenalism.
Dexamet hasone is purely a glucocort icoid w it h no mineralocort icoid eff ect s. O n
t he ot her hand, at doses great er t han 50 t o 100 mg daily, hydrocort isone exhibit s
bot h glucocort icoid and mineralocort icoid eff ect s. Doses in t his range are
recommended in t he acut e t herapy of adrenal crisis or w hen a pat ient present s
in sepsis. Af t er t hat t ime, w hen hydrocort isone doses may be t apered t o a
st anding 20 t o 30 mg daily, f ludrocort isone 100 µg PO should be added f or
mineralocort icoid t herapy.
Suggested Readings
Brunt on L, Lazo J, Parker K, eds. G oodman and G ilman's t he Pharmacological
Basis of Therapeut ics. 11t h ed. New York: McG raw -Hill; 2006: 1599–1606.
G oldman L, Ausiello D, eds. Cecil Text book of Medicine. 22nd ed.

Philadelphia: WB Saunders; 2004: 1416–1418.
LarsenP, ed. Williams Text book of Endocrinology. 10t h ed. Philadelphia: WB

Saunders; 2003: 528–532.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 163 - Us e Vas opr es s or s Ins tead of Lar ge- Volum e
R es us c itation in the Tr eatm ent of S hoc k fr om Mas s ive P ulm onar y E m bolis m
163
Use Vasopressors Instead of Large-Volume
Resuscitation in the Treatment of Shock from
Massive Pulmonary Embolism
Massive pulmonary embolism (PE) occurs w hen more t han 50% of t he pulmonary
circulat ion is acut ely obst ruct ed by a venous embolus, usually originat ing in t he
low er ext remit ies. I n massive PE, severe hemodynamic derangement s result
f rom pulmonary art ery obst ruct ion; prof ound pulmonary vasoconst rict ion due t o
local mediat or release; and hypoxemia-induced vasoconst rict ion. The result ing
increased right vent ricular (RV) af t erload result s in acut e right vent ricular
dilat at ion as t he normally t hin-w alled RV is t asked w it h overcoming high
pulmonary vascular resist ance. O t her consequences of acut ely increased RV
pressures include t ricuspid regurgit at ion and somet imes a signif icant ly shif t ed
int ravent ricular sept um. This can lead t o impaired lef t vent ricular (LV) f illing and
ult imat ely decreased cardiac out put and hypot ension. I n addit ion, decreased
aort ic diast olic pressures and increased RV pressure may result in RV ischemia
due t o poor coronary perf usion, f urt her reducing RV w ork and exacerbat ing t he
shock st at e. This scenario may occur so rapidly t hat t he pat ient immediat ely
experiences syncope or deat h manif est ed by pulseless elect rical act ivit y.
The diagnosis of t his physiological process is made by recognizing t he clinical
scenario of an acut ely ill pat ient , w it h agit at ion, dyspnea, shock, and hypoxemia
in an appropriat e host w it h risk f act ors f or deep venous t hrombosis including
immobilit y, major surgeries, bed rest , malignancy, f amily hist ory, oral
cont racept ives use, and/ or t rauma. Usually, in massive PE, t here is scant t ime
f or conf irmat ory t est s; of t en t he pat ient is t oo unst able f or t ransport . How ever, a
variet y of modalit ies exist if t ime permit s, t he pat ient is st able, and t he clinical
sit uat ion is uncert ain. Probably t he most import ant noninvasive exam w ould be an
emergent comput ed t omography (CT) scan t o bot h suggest and/ or rule out ot her
mimics of PE. O t her t est s t hat can be used t o suggest t he presence of a PE are
an echocardiogram, art erial blood gases (may reveal hypoxemia and hypocarbia
due t o increased minut e vent ilat ion), and a d-dimer (may be of some value in
ruling out a PE if it is negat ive, alt hough a posit ive does not conf irm t he
diagnosis in all cases).
What to Do
The usual init ial t reat ment f or suspect ed or conf irmed massive pulmonary
embolism is int ravenous (I V) f luid administ rat ion. An init ial bolus of 500 mL t o
1, 000 mL of isot onic cryst alloid may be especially helpf ul if / w hen pat ient s are
int ubat ed w it h result ing decrease in right vent ricular preload due t o increased
int rat horacic pressures f rom posit ive pressure vent ilat ion. How ever, int ravenous
f luids, especially if given t oo quickly or in t oo large a volume, may w orsen RV
dilat at ion, increase RV w all st ress, decrease RV perf usion, and shif t t he
vent ricular sept um lef t w ard, result ing in decreased LV out put and w orsening t he
cycle of RV f ailure.
An alt ernat ive t reat ment is t o judiciously init iat e a vasoconst rict or w it h some
inot ropic eff ect , such as norepinephrine at 0. 05 t o 0. 2 µg/ kg/ min or perhaps
dopamine at 5 t o 15 µg/ kg/ min. Several animal st udies have suggest ed t hat
t hese agent s improve syst emic blood pressure, cardiac out put , pulmonary
vascular resist ance, and right vent ricular pressure compared w it h volume
resuscit at ion alone in models of massive PE. Ult imat ely, t hrombolyt ic agent s,
surgical or int ervent ional radiologic clot removal, and/ or ant icoagulat ion are t he
t reat ment s of choice in t he set t ing of massive pulmonary embolism w it h
hypot ension and/ or severe hypoxemia.
Suggested Readings
Benot t i JR, Dalen JE. The nat ural hist ory of pulmonary embolism. Clin Chest
Med. 1984: 5(3): 403–410.
G hignone M, G irling L, Prew it t RM. Volume expansion versus norepinephrine

in t reat ment of low cardiac out put complicat ing an acut e increase in right
vent ricular af t erload in dogs. Anest hesiology. 1984; 60(2): 132–135.
Riedel M. Venous t hromboembolic disease. Acut e pulmonary embolism 1:

pat hophysiology, clinical present at ion, and diagnosis. Heart . 2001; 85: 229–
240.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 164 - C ons ider the D is eas es that Mim ic S eptic
S hoc k in the D iffer ential of this C ondition
164
Consider the Diseases that Mimic Septic Shock
in the Differential of this Condition
Sept ic shock is an example of a part icular t ype of shock know n as dist ribut ive or
hyperdynamic shock. I n t his condit ion, t he pat ient demonst rat es a high cardiac
out put and vasodilat at ion result ing in t he maldist ribut ion of blood f low. Despit e a
circulat ory syst em t hat is w orking vigorously, t he cells and t issues are unable t o
receive t he appropriat e nut rient s and subst rat e necessary f or appropriat e
f unct ioning. The cells begin t o convert f rom aerobic t o anaerobic met abolism and
a lact ic acidosis develops.
Hemodynamically, t hese pat ient s may have a decreased or nearnormal f illing
pressure (cent ral venous pressure or pulmonary capillary w edge pressure)
init ially. When t he vasodilat at ion (manif est ed by a reduct ion in t he syst emic
vascular resist ance) and maldist ribut ion become severe enough, a relat ive
hypovolemia occurs and t he mean art erial pressure decreases. The cardiac
out put remains charact erist ically increased in t his t ype of shock and is
accompanied by increased venous oxygen sat urat ion.
Watch Out For

A number of condit ions are of t en associat ed w it h t his hemodynamic prof ile. When
pat ient s present w it h a shocklike st at e t hat is charact erist ic of hyperdynamic or
dist ribut ive shock, t he diff erent ial diagnosis and t reat ment st rat egies need t o be
broad. Adrenal insuff iciency, anaphylaxis, t hyrot oxicosis, neurogenic shock,
cirrhot ic liver disease, drug int oxicat ions, and several inf ect ious diseases can all
resemble t his hemodynamic prof ile. Addit ional clues t o t he presence of adrenal
insuff iciency include t he sodium and pot assium levels. The administ rat ion of
cort icost eroids can be lif esaving in t his circumst ance. Thyrot oxicosis can
manif est it self in pat ient s w it h pre-exist ing t hyroid disease or on t hyroid
medicat ion. Neurogenic shock can accompany major t rauma w it h neck or back
injuries and spinal-cord cont usion or t ransect ion. Cirrhot ic liver disease pat ient s
w it h impending liver f ailure can also present w it h t his hemodynamic prof ile.
Finally, drug int oxicat ions, part icularly t he t oxidromes associat ed w it h
ant icholinergics, can precipit at e t his spect rum of f indings.
I n addit ion t o t he numerous medical condit ions list ed previously, t here are a
number of inf ect ious agent s t hat can present w it h a similar spect rum of disease.
Malaria can be accompanied by hypot ension and shock w it h an elevat ed cardiac
index. This usually occurs w it h a high parasit emia. Ricket t sial diseases t hat are
due t o int racellular bact eria like Rocky Mount ain spot t ed f ever, babesiosis, and
ehrlichiosis can also present w it h a vasodilat ory shock of t en accompanied by a
skin rash. The last t w o condit ions should be suspect ed in pat ient s in shock w it h
a recent or suspect ed Lyme disease diagnosis and it must be not ed t hat Lyme
disease is now f ound in all part s of t he cont inent al Unit ed St at es and is endemic
in much of Canada and West ern and East ern Europe.
Suggested Readings
Fink MP, Abraham E, Vincent JL, et al. Text book of Crit ical Care. 5t h ed.
Philadelphia: Elsevier Saunders; 2005: 1367–1382, 897–904.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 165 - B e Aler t for a Lar ge S ys tem ic Inflam m ator y
R es pons e after B ac k S ur ger y
165
Be Alert for a Large Systemic Inflammatory
Response after Back Surgery
Ashita G oel MD
SI RS (syst emic inf lammat ory response syndrome) is a complex immune response
t hat is charact erized by hyperdynamic cardiac f unct ion and low af t erload. This
syndrome is manif est ed by signs of inf lammat ion including hypert hermia or
hypot hermia, t achycardia, t achypnea, and elevat ed or depressed w hit e blood cell
count . This st at e lies in a spect rum w it h sept ic shock w it h t he not able except ion
t hat sept ic shock is associat ed w it h t he presence of an inf ect ion. The mult iple
t riggers of SI RS include t rauma, surgery, medicat ions, and t ransf usions. The
proinf lammat ory react ion of SI RS correlat es t o t he severit y of t issue damage.
Depending on t he injury, t he response can progress t o a shock st at e including
hypot ension, oliguria, ment al st at us changes, and coagulopat hies.
O ne of t he st rongest surgical t riggers of SI RS is spine surgery. During spine
surgery, dest ruct ion of bone and sof t t issue act ivat es macrophages t o release
proinf lammat ory cyt okines such as t umor necrosis f act or α (TNF-α ), and
int erleukin 1 (I L-1). These f act ors t hen induce t he product ion of I L-6 and I L-8
f rom endot helial and epit helial cells, f ibroblast s, and monocyt es, propagat ing
SI RS. The onset of t hese f act ors occurs immediat ely f ollow ing t he procedure
and is manif est ed clinically as elevat ions of body t emperat ure and eryt hrocyt e
sediment at ion rat e (ESR), leukocyt osis, and an increase in C-react ive prot ein
(CRP). I n severe react ions, pat ient s may experience hypot ension,
coagulopat hies, and shif t ing of int ravascular f luid t o t he int erst it ial space. Much
of t he proinf lammat ory phase peaks on post operat ive day one but can cont inue
t o linger up t o post operat ive day f our. ESR w ill cont inue t o be elevat ed f or
w eeks f ollow ing surgery. The ext ent of SI RS is aff ect ed by t he durat ion and
ext ent of spinal surgery, t he volume of blood loss, t he level of post operat ive
pain, and t he degree of inst rument at ion during t he procedure. I nst rument at ion
result s in a dramat ic increase in t he inf lammat ory response possibly f rom
increased t issue dest ruct ion w it h implant placement or a direct physiological
react ion t o t he implant mat erial.
Treat ment during t he proinf lammat ory phase usually involves adequat e pain
cont rol and f luid resuscit at ion. How ever, in severe cases,
pressor agent s may be needed t o cont rol t he loss of syst emic vascular
resist ance. Also during t his period, pat ient s may experience hemorrhage
secondary t o coagulopat hy. Management may require t ransf usion of blood
f act ors or plat elet s. I n an ext ensive bleed or crit ically locat ed hemat omas,
pat ient s may need surgical t reat ment f or t he source of bleeding.
Follow ing t he init ial proinf lammat ory phase, a subacut e phase occurs, ref erred
t o as a compensat ory ant i-inf lammat ory react ion syndrome (CARS). This
react ion dow n regulat es t he SI RS response and is carried out by I L-10, I L-13,
I L-14, and t ransf orming grow t h f act or β (TG F-β ). This phase result s in t he
reversal of SI RS manif est at ions including t he hypot ension, leukocyt osis,
coagulopat hies, and f evers. How ever, it can lead t o a depression of t he immune
syst em and place pat ient s at risk f or inf ect ions.
Watch Out For

I n pat ient s recovering f rom spinal surgery, an inf lammat ory response beyond
post operat ive day f our may suggest t he presence of occult or overt inf ect ion.
G ram-posit ive bact eria such as Staphyl ococcus aureus are t he common causes
of inf ect ion; how ever, G ram-negat ive bact eria can result in severe inf ect ions.
Many G ram-negat ive bact eria possess endot oxins t hat can result in t he
development of severe sepsis. CRP levels are more accurat e markers of
inf ect ion t han ESR. ESR remains elevat ed f or w eeks f ollow ing uncomplicat ed
surgery; how ever, CRP should normalize by day f our in uncomplicat ed spinal
cases.
The post operat ive response of spinal procedures t hat involve neoplast ic
resect ions have variable courses depending on t he nat ure of t he t umor and t he
procedure. A SI RS response is generally predict ed but can be complicat ed in
pat ient s on chemot herapy. Also, coagulopat hies can be more pronounced in t his
group.
Suggested Readings
Lanken P, ed. The I nt ensive Care Unit Manual. Philadelphia: WB Saunders:
2001: 93– 101.
Takahashi J, Ebara S, Kamimura M, et al. Early-phase enhanced inf lammat ory

react ion af t er spinal inst rument at ion surgery. Spine. 2001; 26(15): 1698–1704.
Takahashi J, Ebara S, Kamimura M, et al. Pro-inf lammat ory and ant i-

inf lammat ory cyt okine increases af t er spinal inst rument at ion surgery. J Spinal
Disord Tech. 2002; 15(4): 294–300.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 166 - B e Aler t for the D evelopm ent of Abdom inal
C om par tm ent S yndr om e
166
Be Alert for the Development of Abdominal
Compartment Syndrome
Awori J. Hayanga MD
The t erms i ntra-abdomi nal hypertensi on and abdomi nal compartment syndrome
have somet imes been used int erchangeably. How ever, it is import ant t o
recognize t here is a dist inct ion bet w een t hese ent it ies. I nt ra-abdominal
hypert ension exist s w hen int ra-abdominal pressure exceeds a measured numeric
paramet er. This paramet er has generally been set at bet w een 20 and 25 mm Hg.
Abdominal compart ment syndrome exist s w hen int ra-abdominal hypert ension is
accompanied by manif est at ions of organ dysf unct ion, w it h reversal of t hese
pat hophysiologic changes upon abdominal decompression. These include t he
pulmonary, cardiovascular, renal, splanchnic, musculoskelet al/ int egument ary
(abdominal w all), and cent ral nervous syst em.
The exact incidence of abdominal compart ment syndrome is yet t o be
est ablished, but it is clearly increased in cert ain populat ion groups. These
include pat ient s w it h t he f ollow ing:
severe blunt and penet rat ing abdominal t rauma

rupt ured abdominal aort ic aneurysms
ret roperit oneal hemorrhage
pneumoperit oneum
neoplasm
pancreat it is
massive ascit es
liver t ransplant at ion
massive f luid resuscit at ion
accumulat ion of blood and clot
bow el edema
f orced closure of a noncompliant abdominal w all
circumf erent ial abdominal burn eschars
I n one prospect ive series of 145 pat ient s w ho w ere ident if ied as being at risk f or
development of t he abdominal compart ment syndrome, t he report ed incidence
w as 14%. The incidence f ollow ing primary closure af t er repair of rupt ured
abdominal aort ic aneurysm w as report ed in one series as 4%.
I n t he t rauma populat ion, t he group t hat is especially at risk includes t hose
pat ient s undergoing abbreviat ed or “damage cont rol”
laparot omy, especially w it h int ra-abdominal packing. I t must be not ed t hat having
an open abdomen does not necessarily preclude t he diagnosis of int ra-abdominal
hypert ension or abdominal compart ment syndrome, part icularly w here int ra-
abdominal packing has been used.
Signs And Symptoms

Clinical manif est at ions of organ dysf unct ion in abdominal compart ment syndrome
include respirat ory f ailure t hat is charact erized by impaired pulmonary
compliance t hat result s in elevat ed airw ay pressures w it h progressive hypoxia
and hypercapnia. I n t his sit uat ion, ext remely high driving pressures may be
required t o maint ain minimally suff icient t idal volumes, of t en w it h loss of
delivered t idal volume by dist ension of vent ilat ory t ubing. The high airw ay
pressures are needed t o overcome high ext rat horacic pressure exert ed t hrough
t he diaphragmand not t o overcome an int rinsic lung problem. Some aut hors
report pulmonary dysf unct ion and elevat ed peak airw ay pressures as t he earliest
manif est at ion of abdominal compart ment syndrome. Chest radiography may show
elevat ed hemidiaphragms w it h loss of lung volume.
Hemodynamic indicat ors consist ent w it h abdominal compart ment syndrome
include elevat ed heart rat e, hypot ension, elevat ed pulmonary art ery w edge
pressure and cent ral venous pressure, reduced cardiac out put , and elevat ed
syst emic and pulmonary vascular resist ance. I n t his sit uat ion measurement of
right vent ricular end-diast olic volume may be a more accurat e predict or of a
pat ient 's posit ion on t he St arling curve. The pat hophysiology is relat ed t o
decreased venous ret urn.
I n abdominal compart ment syndrome impairment in renal f unct ion is manif est ed
by oliguria progressing t o anuria w it h result ant azot emia. Renal insuff iciency as a
result of int ra-abdominal hypert ension is only part ly reversible by f luid
resuscit at ion. Renal f ailure in t he absence of pulmonary dysf unct ion is not likely
t o be t he result of int ra-abdominal hypert ension. Elevat ed int racranial pressure
is an addit ional clinical manif est at ion of abdominal compart ment syndrome.
Clinical conf irmat ion of int ra-abdominal hypert ension requires bedside
measurement s indicat ive of int ra-abdominal pressure. Experiment al and clinical
dat a indicat e t hat int ra-abdominal hypert ension is present w hen int ra-abdominal
pressure exceeds 20 mm Hg. Cont emporary measurement of int ra-abdominal
pressure out side of t he laborat ory is accomplished by a variet y of means. These
include direct measurement of int ra-abdominal pressure by means of an
int raperit oneal cat het er, as is done during laparoscopy. Bedside measurement
of int ra-abdominal pressure has been accomplished by t ransduct ion of pressures

f rom indw elling f emoral vein, rect al, gast ric, and urinary bladder cat het ers. O f
t hese met hods, measurement of urinary bladder and gast ric pressures are t he
most common clinical applicat ions. I n 1984, Kron et al. report ed a met hod t o
measure int ra-abdominal pressure at t he bedside w it h t he use of an indw elling
Foley cat het er as f ollow s:
St erile saline (50 t o 100 cc) is inject ed int o t he empt y bladder t hrough t he
indw elling Foley cat het er.
The st erile t ubing of t he urinary drainage bag is clamped just dist al t o t he
cult ure aspirat ion port .
The end of t he drainage bag t ubing is connect ed t o t he Foley cat het er.
The clamp is released just enough t o allow t he t ubing proximal t o t he clamp
t o drain f luid f rom t he bladder, t hen reapplied.
A 16-gauge needle is t hen used t o Y-connect a manomet er or pressure
t ransducer t hrough t he cult ure aspirat ion port of t he t ubing of t he drainage
bag.
Finally, t he t op of t he symphysis pubic bone is used as t he zero point w it h
t he pat ient supine.
An alt ernat ive bedside t echnique has been described in w hich int ragast ric
pressure measurement s are t aken f rom an indw elling nasogast ric t ube. This
met hod has been validat ed and f ound t o vary w it hin 2. 5 cm H2 O of urinary
bladder pressures. O f t hese t echniques, measurement of urinary bladder
pressure appears t o have gained w idest clinical accept ance and applicat ion.
O ne f inal not e is t hat t he clinician must know if t he pressure readings have been
report ed in cm H2 O or mmHg. To convert cm H2 O t o mm Hg, mult iply t he
report ed value by 0. 736.
Suggested Readings
Bailey J, Shapiro MJ. Abdominal compart ment syndrome. Crit Care.
2000; 4(1): 23–29.
G racias VH, Braslow B, Johnson J, et al. Abdominal compart ment syndrome in
t he open abdomen. Arch Surg. 2002; 137(11): 1298–1300.
Kron I L, Hart man PK, Nolan SP. The measurement of int ra-abdominal pressure
as a crit erion f or abdominal re-explorat ion. Ann Surg. 1984; 199: 28–30.
Saggi BH, Sugerman HJ, I vat ury RR, et al. Abdominal compart ment syndrome.
J Trauma. 1998; 45: 597–609.
Sugrue M. Abdominal compart ment syndrome. Curr O pin Crit Care.

2005; 11(4): 333–338.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 167 - Tr eat Abdom inal P ain O ut of P r opor tion To
P hys ic al E xam as Mes enter ic Is c hem ia Until P r oven O ther w is e
167
Treat Abdominal Pain Out of Proportion To
Physical Exam as Mesenteric Ischemia Until
Proven Otherwise
Rajan G upta MD
Acut e mesent eric ischemia is an abdominal emergency w it h high mort alit y and
devast at ing morbidit y. Perioperat ive mort alit y is 20% t o 80%, depending upon
t he et iology of t he acut e process, and t he overall mort alit y rat e remains above
60%. I n acut e mesent eric ischemia, init ial t issue injury occurs as a result of
compromised end-organ perf usion. Secondary injury at t he cellular level likely
occurs t hrough amplif icat ion of an inf lammat ory response w it h f ree-radical
generat ion during reperf usion. Acut e mesent eric ischemia is usually divided int o
f our cat egories, based upon et iology. Art erial embolism is t he most common t ype
and represent s 40% t o 50% of all acut e mesent eric ischemia cases. Emboli most
commonly originat e f rom a cardiac source and occlude t he superior mesent eric
art ery (SMA). Acut e mesent eric art erial t hrombosis account s f or 25% t o 30% of
acut e mesent eric ischemia. The proximal SMA is t he most common sit e of
t hrombosis. Nonocclusive mesent eric ischemia represent s approximat ely 20% t o
30% of t he acut e mesent eric ischemia cases. This condit ion is usually a result of
relat ive hypovolemia, low cardiac out put , and splanchnic vaso-const rict ion.
Mesent eric vein t hrombosis is t he least common cause of acut e mesent eric
ischemia, account ing f or approximat ely 5% t o 10% of t he cases. These cases
are of t en associat ed w it h a concomit ant int ra-abdominal process such as
malignancy or inf lammat ion, or w it h primary clot t ing disorders such as f act or V
Leiden mut at ion.
Signs and Symptoms

Diagnosis of acut e mesent eric ischemia may be diff icult because t he signs and
sympt oms of t en mimic ot her int ra-abdominal processes. This of t en result s in
signif icant delays in diagnosis, especially in t he int ensive care unit (I CU).
Because t hese delays can adversely aff ect out come, height ened clinical
suspicion is paramount . The f our t ypes of acut e mesent eric ischemia present
w it h diff erent pat t erns of signs and sympt oms. An embolic event usually causes
a sudden onset of sympt oms, w hile acut e mesent eric ischemia due t o t hrombosis
may present suddenly or w it h a progressive hist ory
of post prandial pain, nausea, and w eight loss. Nonocclusive mesent eric ischemia
present s less acut ely and of t en occurs in elderly pat ient s in t he I CU suff ering
f rom global hypoperf usion. Pat ient s w it h mesent eric vein t hrombosis classically
present lat e w it h previous sympt oms of nonspecif ic abdominal pain, anorexia,
and diarrhea.
Abdominal pain out of proport ion t o physical exam is t he hallmark of acut e
mesent eric ischemia. Physical exam is of t en not helpf ul in t he diagnosis of acut e
mesent eric ischemia and can be unreliable in t he I CU. I n f act , early in t he course
of acut e mesent eric ischemia, physical exam can be normal or remarkable only
f or dist ent ion. Perit oneal signs are lat e f indings t hat develop only af t er inf arct ion
has occurred. O ccult blood in t he st ool can also indicat e acut e mesent eric
ischemia; how ever, bloody diarrhea is more common w it h colonic ischemia.
Laborat ory t est s in pat ient s w it h acut e mesent eric ischemia w ill commonly
demonst rat e an increased w hit e blood cell count and acidosis. How ever, t hese
f indings are usually nonspecif ic and manif est t oo lat e in t he disease process t o
be t ruly helpf ul t o make a t imely diagnosis. I n t he I CU, abdominal pain
accompanied by met abolic acidosis should prompt early considerat ion of acut e
mesent eric ischemia. Findings on plain abdominal radiograph are of t en
nonspecif ic (e. g. , ileus) or occur lat e in t he course (e. g. , pneumoperit oneum,
pneumat osis, port al venous air). Colonoscopy or sigmoidoscopymay be usef ul in
suspect ed cases of colonic ischemia. Duplex ult rasonography is of t en limit ed
because of bow el gas pat t erns. Comput ed t omography (CT) has become t he
gold st andard t o evaluat e abdominal pain and has t he advant age of being able t o
evaluat e t he ent ire abdomen and ident if y processes t hat may clinically mimic
acut e mesent eric ischemia. Wit h t he advent of mult islice CT, t he sensit ivit y of
t his modalit y using CT angiography has improved f or primarily diagnosing acut e
mesent eric ischemia. I nt ravenous (I V) cont rast is necessary t o obt ain
appropriat e result s. Magnet ic resonance imaging can be employed, especially if
CT is cont raindicat ed; how ever, it does not appear t o have any addit ional value
w hen compared w it h CT angiography. Select ive mesent eric angiography remains
t he gold st andard in t he diagnosis of acut e mesent eric ischemia. Convent ional
angiography has t he added benef it of off ering pot ent ially t herapeut ic
int ervent ions such as angioplast y, st ent placement , embolect omy, and t he
inf usion of t hrombolyt ics and vasodilat ors.
The t reat ment f or acut e mesent eric ischemia begins w it h volume resuscit at ion t o
opt imize mesent eric (and global) perf usion. Broad-spect rum ant ibiot ics should be
administ ered and syst emic ant icoagulat ion should be considered. I n pat ient s
w it hout perit onit is, select ive
mesent eric angiography should be perf ormed and appropriat e t herapeut ic
measures should be init iat ed based on t he f indings. Thrombolyt ic t herapy is
indicat ed f or acut e mesent eric occlusion secondary t o embolus or t hrombosis.
Vasodilat ors (e. g. , papaverine) can be used in pat ient s w it h bot h occlusive and
nonocclusive mesent eric ischemia. They are a f irst -line t herapy in pat ient s w it h
nonocclusive mesent eric ischemia and can decrease occlusive vasospasm in
ot her causes of acut e mesent eric ischemia as w ell. For pat ient s w ho f ail
int ervent ional angiographic t herapy f or t hromboembolic disease, surgical
int ervent ion is required t o re-est ablish adequat e mesent eric circulat ion. Surgical
embolect omy is usually perf ormed f or embolic disease and bypass graf t ing is
commonly necessary f or t hrombot ic disease. O t her t han resect ion of nonviable
int est ine, t here is no surgical opt ion f or nonocclusive mesent eric ischemia.
Surgical int ervent ion f or mesent eric vein t hrombosis (e. g. , venous embolect omy
f or superior mesent eric vein or port al vein t hrombosis) is associat ed w it h poor
out comes.
Pat ient s w ho present w it h or develop perit onit is during t he diagnost ic and
resuscit at ion phase need emergent laparot omy w it h resect ion of nonviable
int est ine. A planned “second look” operat ion 24 t o 48 hours lat er is of t en
required t o det ermine t he t rue ext ent of bow el necrosis. Long-t erm
ant icoagulat ion is indicat ed in pat ient s w it h embolic or t hrombot ic causes of
acut e mesent eric ischemia. The use of ant iplat elet agent s post operat ively in
cases of acut e mesent eric ischemia is less w ell est ablished.
Suggested Readings
Agaoglu N, Turkyilmaz S, O vali E, et al. Prevalence of prot hrombot ic
abnormalit ies in pat ient s w it h acut e mesent eric ischemia. World J Surg.
2005; 29(9): 1135–1138.
O ldenburg WA, Lau LL, Rodenberg TJ, et al. Acut e mesent eric ischemia: a
clinical review. Arch I nt ern Med. 2004; 164(10): 1054–1062.
Schoot s I G , Levi MM, Reekers JA, et al. Thrombolyt ic t herapy f or acut e

superior mesent eric art ery occlusion. J Vasc I nt erv Radiol. 2005; 16(3): 317–
329.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 168 - D o not B olus Fluids that C ontain D extr os e
168
Do not Bolus Fluids that Contain Dextrose
Awori J. Hayanga MD
I nt ravenous f luids t hat cont ain dext rose have quest ionable ut ilit y in t he int ensive
care unit (I CU) and should not be administ ered at a rat e great er t han
maint enance f luids. Bolusing w it h dext rose-cont aining f luids causes a rapid
change in ext racellular osmolarit y. This can occur quit e rapidly and may cause
convulsions and coma. At t he very least , bolusing w it h dext rose w ill cause an
unaccept ably high serum glucose, w hich is det riment al in int ensive care pat ient s.
Accept able f luids t o bolus w it h include normal saline, hypert onic saline, and
Lact at ed Ringer's. Lact at ed Ringer's is an excellent isot onic salt solut ion f or
replacing isot onic gast roint est inal losses and pre-exist ing volume def icit s w hen a
pat ient has a normal or near normal elect rolyt e level. I sot onic sodium (0. 9%
saline) is ideal f or t he init ial correct ion of deplet ed ext racellular f luid volume
associat ed w it h hyponat remia, hypochloremia, and met abolic alkalosis. O ne
concern w it h normal saline is t hat t he high chloride concent rat ion may exceed t he
capacit y of t he kidney t o excret e it and has t he pot ent ial t o cause dilut ional
acidosis.
Suggested Readings
Boldt J. Fluid choice f or resuscit at ion of t he t rauma pat ient : a review of t he
physiological, pharmacological, and clinical evidence. Can J Anaest h.
2004; 51(5): 500–513.
Miller RD, ed. Miller's Anest hesia. 6t h ed. New York: Elsevier/ Churchill
Livingst one; 2005: 1783–1795.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 169 - R em em ber that D iur es is May not be the B es t
Tr eatm ent for Hyper kalem ia in the E ar ly P os t C ar diopulm onar y B ypas s P er iod
169
Remember that Diuresis May not be the Best
Treatment for Hyperkalemia in the Early Post
Cardiopulmonary Bypass Period
Frank Rosemeier MD MRCP (UK)
Hyperkalemia is a common elect rolyt e abnormalit y f ollow ing cardiac surgery
w hich can result in serious complicat ions. Pot ent ial causes of hyperkalemia are
list ed in Table 169. 1. While evaluat ing pat ient s f or clinical sequelae f rom t his
dist urbance; t hese underlying causes should be caref ully considered.
Decreased Cl earance of Potassi um. Preexist ing renal insuff iciency or evolving
acut e t ubular necrosis is t he most concerning cause of hyperkalemia in t he
post operat ive period. Cardiopulmonary bypass (CPB) is charact erized by alt ered
renal blood f low because of non-pulsat ile pump f low and int raoperat ive
hypot ensive episodes. Vasodilat at ion induced by CPB and syst emic inf lammat ory
response syndrome (SI RS) may result in diminished renal f low, pooling of blood
in venous capacit ance vessels, and capillary leakage w it h result ing hypot ension.
Shear f orces of CPB may cause red cell dest ruct ion result ing in int ravascular
hemolysis, w hich in t urn can lead t o hemoglobinurea, hyperkalemia and renal
f ailure. Management st rat egies include t he rest orat ion and maint enance of
normovolemia, renal blood f low, and perf usion pressure t o limit f urt her insult .
ACE inhibit ors are know n t o result in pot assium ret ent ion secondary t o a
low ering of plasma aldost erone levels and t hus a decrease in pot assium
clearance. This phenomenon is not e t o occur in individuals w it h ot her risk f act ors
f or hyperkalemia, including disrupt ion of homeost asis or congest ive heart f ailure.
Furt hermore, ACE inhibit ors may also cont ribut e t o renal insult in sit uat ions
w here MAP f alls below 50-65 mmHg, hypovolemia, or inadequat e cardiac out put .
Theref ore prior t o t reat ment of hyperkalemia w it h pot assium w ast ing diuret ics,
you must make cert ain t hat t he pat ient is adequat ely resuscit at ed so as not t o
cont ribut e t o ongoing renal insult .
Increased Intake of Potassi um. Pot assium is commonly replet ed in post -bypass
pat ient s t o a level of 4. 0 t o 4. 5 mEq/ L t o decrease t he risk of perioperat ive
arrhyt hmias. Cont inuous I V insulin inf usions f or t he management of perioperat ive
hyperglycemiapromot e cellular upt ake
of pot assium increasing t he scale of pot assium replet ion. As glycemic cont rol
improves, and t he insulin requirement s decrease, subsequent hyperkalemia may
occur because of increased t ot al body st ores.
TABLE 169-1 POTENTIAL CAUSES OF

PERIOPERATIVE HYPERKALEM IA IN CARDIAC
SURGERY PATIENTS
1. ATN
2. Chronic renal failure
3. Preop ACE inhibitors
4. Exogenous potassium administration
5. NSAIDs via the aldosterone pathway
6. Preop spironolactone or potassium sparing diuretics
7. Intraoperative use of high-potassium cardioplegia
8. Heparin via an aldosterone effect
9. Epsilon-aminocaproic acid
O ther. O t her causes of hyperkalemia include preoperat ive administ rat ion of
spironolact one or pot assium sparing diuret ics, high-pot assium-cont aining
cardioplegia and large volume red cell t ransf usion. Rare causes of hyperkalemia
are t he administ rat ion of t he lysine analogue epsilon-aminocaproic acid and
heparin via an aldost erone like eff ect . Finally, bot h respirat ory and met abolic
acidemia (t he lat t er f or example caused by hypoperf usion) may cause pot assium
shif t f rom int racellular t o ext racelluar w it h an increase of 0. 5 mEq K+ f or a pH
decrease by 0. 1.
Treatment. Wit hin t he set t ing of acut e post operat ive hyperkalemia it is
paramount t o opt imize normovolemia and adequat e renal perf usion pressure t o
prevent f urt her renal insult . Alt hough f urosemide or ot her pot assium w ast ing
diuret ics may support t he clearance of pot assium, it may, in f act , increase
relat ive hypovolemia t hus cont ribut ing t o acut e t ubular necrosis. How ever, if
pot assium exceeds 6meq/ L or t here are clinical sequelae not ed f rom
hyperkalemia w it h peaked T-w ave being t he f irst change seen on ECG , t reat ment
must be init iat ed. Temporizing measures include t he int ravenous administ rat ion of
calcium t o st abilize myocardial membranes, and sodium bicarbonat e, insulin
and/ or bet a agonist , all of w hich promot e an int racellular shif t of pot assium.
Furt her measures are t hose direct ed at decreasing t he t ot al body pot assium
st ores. These include pot assium w ast ing diuret ics (f urosemide) and hemodialysis
or hemof ilt rat ion. Caut ion is advised t o t he use of oral or rect ally administ ered
Kayexalat e as it can cause ischemic bow el
t hrough t he osmot ic load of it s sorbit ol component especially in t he set t ing of

hemodynamic inst abilit y and hypoperf usion.
Suggested Readings
Day JR, Chaudhry AN, Hunt I , Taylor KM. Heparin-induced hyperkalemia af t er
cardiac surgery. Ann Thorac Surg 2002; 74(5): 1698–700.
Hessell EA, Hild PG . Pat hophysiology of Cardiopulmonary Bypass. I n: Hensley

FA, Mart in DE, G ravlee G P, eds. A Pract ical Approach t o Cardiac Anest hesia.
3rd ed. Philadelphia: Lippincot t Williams & Wilkins; 2003: 537–54.
Weber DO , Yarnoz MD. Hyperkalemia complicat ing cardiopulmonary bypass:

analysis of risk f act ors. Ann Thorac Surg 1982; 34(4): 439–45.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 170 - R em em ber that Mos t P atients R ec eive
Mannitol W hen Going O n the P um p s o P os toper ative Ur ine O utput is not a Mar ker for Volum e S tatus or
P er fus ion After C ar diac S ur ger y
170
Remember that Most Patients Receive Mannitol
When Going On the Pump so Postoperative
Urine Output is not a Marker for Volume Status
or Perfusion After Cardiac Surgery
Current t echniques in cardiac surgery somet imes require t he use of
cardiopulmonary bypass. The essent ial goal of cardiopulmonary bypass is t o
divert blood around a nonbeat ing heart t hrough an ext racorporeal circuit t hat
f unct ionally replaces t he heart and lungs. There are several st eps t o be
perf ormed w hen placing a pat ient on bypass. Af t er administ rat ion of heparin and
at t aining adequat e act ivat ed clot t ing t ime (ACT), t he art erial cannula (most
commonly ascending aort a) is insert ed f irst , t o allow inf usion of volume in case
of hemorrhage associat ed w it h venous cannulat ion. When t he art erial cannula is
insert ed, t he syst olic blood pressure should be decreased t o 90 t o 100 mm Hg t o
reduce t he risk of aort ic dissect ion and t o f acilit at e cannulat ion. Venous blood
ret urning t o t he right heart is drained by gravit y t o t he venous reservoir t hrough
eit her t w o venous cannulas separat ely insert ed int o t he inf erior vena cava (I VC)
and superior vena cava (SVC) or, more commonly, a single cannula insert ed
direct ly int o t he right at rium.
The composit ion of pump-priming solut ion generally consist s of elect rolyt es,
colloids, mannit ol, and heparin. Depending upon t he hemat ocrit of t he pat ient ,
blood or packed red blood cells may be added t o t he priming solut ion. The
volume of priming solut ion ranges f rom 1, 800 mL t o 2, 200 mL. Af t er t he init iat ion
of t ot al cardiopulmonary bypass, mild t o moderat e hypot hermia is induced (30° C
t o 32° C). The aort a is cross-clamped and cardioplegic solut ion is inf used
ant egrade t hrough t he aort ic root and ret rograde via t he coronary sinus t o arrest
t he heart .
When cryst alloid solut ions alone are used t o prime t he cardiopulmonary bypass
circuit , t he result ant low ering of plasma oncot ic pressure result s in ext ravascular
f luid shif t s and t he requirement of addit ional f luid during t he procedure. Mannit ol
is added t o t he prime t o help reduce t his eff ect . Mannit ol is an osmot ic diuret ic
and a f ree-radical scavenger. I n t he Unit ed St at es, t he dose of mannit ol in t he
priming solut ion ranges f rom 12. 5 g t o 25 g.
Wit hin 10 minut es of administ rat ion, t he majorit y of t he administ ered mannit ol
diff uses int o t he int erst it ial f luid. Very lit t le, if any, penet rat es t he cells and so
t he raised ext racellular t onicit y result s in t he w it hdraw al of f luid f rom t he cells.
I n addit ion, it act s t o prevent f urt her cellular edema, w hich w ould ot herw ise have
t o be t reat ed by t he addit ion of f urt her f luid during bypass. The plasma-volume
expansion due t o mannit ol has been report ed as being bet w een 1 mL and 3 mL
per kg of body w eight per 5 g of mannit ol inf used. During cardiopulmonary
bypass, t hese act ions result in a great er and more st able volume of blood st ored
in t he oxygenat or.
At t he same t ime, mannit ol is ent ering t he kidney w here it is f reely f ilt ered at t he
glomerulus int o t he renal t ubules. Because it cannot ent er t he cells of t he renal
t ubules, it is only poorly reabsorbed and so st ays w it hin t he lumen of t he
t ubules. The increased osmot ic pressure caused by t he mannit ol remaining in t he
proximal t ubules reduces w at er reabsorpt ion, w hich result s in an increased
diuresis.
The addit ion of mannit ol t o t he pump prime may increase diuresis f or about 12
hours post operat ively. Because of t his, t he urine out put is not reliable as an
indicat or f or volume st at us in pat ient s af t er cardiac surgery. The cent ral venous
pressure and pulmonary capillary w edge pressures, as det ermined w it h a
pulmonary art ery cat het er, are t he pref erred measurement s t o be used at t he
bedside. I n addit ion, venous oxyhemoglobin sat urat ion can be measured
int ermit t ent ly by t rans-cat het er aspirat ion or cont inuously by f iberopt ic
pulmonary art ery cat het ers. Finally, cent ral venous (right at rial) oxygen
sat urat ions are t ypically measured, and rest ing values are usually 75%.
Assuming a 100% art erial sat urat ion, t his represent s unloading of one-f ourt h of
t he available oxygen. Venous desat urat ion usually indicat es inadequat e t issue
perf usion w het her or not t he cardiac index is depressed.
Suggested Readings
Est af anous FG , ed. Cardiac Anest hesia Principles and Clinical Pract ice. 2nd
ed. Philadelphia: Lippincot t Williams & Wilkins; 2001: 415–446.
St oelt ing RK, ed. Pharmacology and Physiology in Anest het ic Pract ice. 3rd
ed. Philadelphia: Lippincot t Williams & Wilkins; 1999: 440–441.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 171 - D o not us e Ur ine O utput as a Meas ur e of
Volum e S tatus in P atients w ho ar e C old
171
Do not use Urine Output as a Measure of Volume
Status in Patients who are Cold
Juan N. Pulido MD
Daniel R. Brown MD, PhD
Hypot hermia is a clinical ent it y def ined as a core body t emperat ure less t han 35°
C (95° F) and is classif ied in f our st ages depending on t he t emperat ure,
sympt omat ology, and impact on specif ic organ physiology (Table 171. 1).
All organs are ult imat ely aff ect ed by hypot hermia, including t he kidneys. The
renal response t o cold is rapid and varies w it h t he diff erent st ages of
hypot hermia. I nit ially, peripheral vasoconst rict ion result s in relat ive cent ral
hypervolemia, producing an increase in urine out put . This response, t ermed col d
di uresi s, has been described even in pat ient s w it h mild t o moderat e
hypot hermia. The et iology of t his phenomenon is mult if act orial and includes an
init ial increase in cardiac out put and renal blood f low due t o hypot hermia-induced
changes in vascular capacit ance. O t her import ant cont ribut ors are nonosmot ic
suppression of ant idiuret ic-hormone (ADH) release by t he hypot halamus and
subsequent decreased renal t ubular reabsorpt ion. These responses usually begin
as soon as t he core body t emperat ure reaches 35° C and become more
pronounced unt il moderat e hypot hermia, w hen decreased renal blood f low and
glomerular f ilt rat ion rat e (reduced 50% at 27° t o 30° C) may lead t o renal
f ailure.
Watch Out For

Even in t he set t ing of a large diuresis (t he urine is usually dilut e w it h osmolarit y
<300 mO sm/ L and specif ic gravit y <1. 003), t he kidneys are unable t o handle
nit rogenous w ast e due t o t ubular dysf unct ion. Alt hough uncommon, elect rolyt e
dist urbances including hypernat remia, hyperchloremia, and hyperkalemia can
occur and are more f requent as hypot hermia progresses in durat ion and/ or
severit y. Cold diuresis is exacerbat ed by et hanol ingest ion and w at er
submersion, w hich may coexist w it h hypot hermia and can pot ent iat e
inappropriat e diuresis by inhibit ing ADH secret ion.
I t is import ant t o underst and t he pat hophysiology of t his phenomenon w hen
making clinical decisions regarding f luid management in hypot hermic pat ient s.
The cold diuresis can be massive and generally creat es a hypovolemic st at e t hat
w orsens w it h rew arming because
of t he reverse changes in vascular t one as core body t emperat ure is raised. I f

overlooked or underappreciat ed, t his phenomenon can exacerbat e elect rolyt e
dist urbances, cont ribut e t o hypot ension, and result in prerenal st ress.
TABLE 171-1 SEQUELAE OF HYPOTHERM IA
CORE
STAGE T EMPERAT URE Characteristics
°C (°F)
Increased metabolic rate,

hypertension,
32 (89.6)–35 tachycardia, shivering,
Mild
(95) cold diuresis, CNS
hyperexcitability,
coagulopathy
Decreased cardiac
output, hypoventilation,
28 (82.4)–32 CNS depression, atrial
Moderate
(89.6) arrhythmias, ↑O2
consumption (~25% to
50%)
Progressive hypotension
and bradycardia,
ventricular arrhythmias,
22 (71.6)–28 VF, decreased CBF,
Severe areflexia, loss of bulbar
(82.4)
reflexes, decreased O2
consumption (<50% of
baseline)
Asystole, EEG burst

Profound <22 (71.6)
suppression
CNS, central nervous system; VF, ventricular

fibrillation; CBF, cerebral blood flow; EEG,
electroencephalogram.
I nt ravascular volume st at us should be closely monit ored t o avoid complicat ions

of t his “physiologically inappropriat e” renal response. I nit ially, it should be
assumed t hat t he pat ient is signif icant ly dehydrat ed. Frequent measurement s of
elect rolyt es and hemat ocrit w ill help guide f luid t herapy and elect rolyt e
replacement and help monit or f or dehydrat ion. Cent ral venous access should be
considered t o allow f or saf er elect rolyt e replacement and rapid volume
administ rat ion, t hough elect rolyt e abnormalit ies may increase cardiac irrit abilit y
and arrhyt hmia risk during cat het er placement . I nvasive art erial blood pressure
monit oring should also be considered t o f acilit at e laborat ory det erminat ions and
t o evaluat e f luid responsiveness.
Suggested Readings
Auerbach PS. Wilderness Medicine. 4t h ed. Philadelphia: Mosby; 2001: 135–
155.
I rw in RS, Rippe JM, eds. I rw in and Rippe's I nt ensive Care Medicine. 5t h ed.
Morgan ML. Mechanism of cold diuresis in t he rat . Am J Physiol.

1983; 244(2): F210–F216.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 172 - B e C ar eful to Not O ver hydr ate P os toper ative
Liver - Tr ans plant P atients
172
Be Careful to Not Overhydrate Postoperative
Liver-Transplant Patients
Dorry L. Segev MD
Warren R. Maley MD
The resuscit at ion of pat ient s immediat ely f ollow ing a liver t ransplant is similar in
concept t o t hat of cirrhot ic individuals w ho have or are experiencing a bout of
variceal hemorrhage. Like t he cirrhot ic pat ient w ho is suff ering f rom
gast roint est inal bleeding, t he t ransplant ed pat ient should be replet ed t o
euvolemia t o maximize end-organ perf usion and f unct ion. Excessive hydrat ion of
t he variceal bleeder w ill exacerbat e blood loss by f urt her elevat ing cent ral
venous pressure (CVP) and port al hypert ension. Likew ise, overaggressive
volume resuscit at ion of t he liver recipient produces elevat ion of t he CVP,
result ing in dist ent ion of t he t ransplant ed organ and diminished liver perf usion
pressure.
Allocat ion of livers f or t ransplant at ion is current ly t riaged according t o t he Model
f or End-st age Liver Disease (MELD) syst em. Under MELD, each pat ient is
assigned a score calculat ed f rom his or her most recent blood values of
creat inine, bilirubin, and prot hrombin t ime. MELD scores range f rom 6 t o a
capped maximum of 40 and correlat e direct ly w it h t he risk of 3-mont h mort alit y
over a w ide spect rum of et iologies of cirrhosis. Those w it h t he highest scores
w it hin t heir respect ive blood group are t he next individuals t o be t ransplant ed.
I nclusion of renal f unct ion (as measured by serum creat inine) in t he MELD
equat ion recognizes t he signif icant cont ribut ion of hepat orenal syndrome t o
mort alit y, as a lat e manif est at ion of severe cirrhot ic liver disease. Using t his
allocat ion algorit hm result s in a signif icant proport ion of recipient s undergoing
t ransplant at ion w it h renal insuff iciency or even acut e renal f ailure secondary t o
changes in renal f unct ion precipit at ed by t heir advanced liver f ailure. Thus,
urinary out put , of t en t he best measure of t issue perf usion in t he t rauma or
post surgical pat ient , may f ail t o accurat ely represent an adequat e volume of
resuscit at ion in t he liver-t ransplant recipient . I n addit ion, int raoperat ive blood
loss and hypot ension, as w ell as t he eliminat ion of massive amount s of ascit es
and result ant ref ormat ion, may f urt her w orsen t he pat ient 's already t enuous
renal st at us. I n liver-t ransplant recipient s, t rends in CVP, pulmonary art ery
diast olic or w edge pressure, and cardiac out put may bet t er represent t he
hydrat ion st at us of t he pat ient .
Watch Out For

I n liver-t ransplant recipient s, int erpret at ion of t he Sw an-G anz paramet ers must
be predicat ed upon know ledge of t he pat hophysiology of t he pre-exist ing
cirrhot ic st at e. Prior t o t ransplant at ion, pat ient s w it h cirrhosis exhibit an
increased plasma volume, low syst emic vascular resist ance, and a hyperdynamic
cardiac out put . This physiologic condit ion persist s f or a signif icant period of t ime
f ollow ing liver t ransplant at ion and can easily be misint erpret ed as being
consist ent w it h sepsis. O rdinary measures of adequat e st roke volume may
underest imat e t he need f or f urt her resuscit at ion given t he vasodilat ed
hyperdynamic st at e of t he chronic liver-f ailure pat ient . Blood product s provide
t he best volume replet ion and are indicat ed in t he set t ing of coagulopat hy,
anemia, or t hrombocyt openia. O t herw ise, many liver-t ransplant surgeons pref er
albumin solut ions t o cryst alloid, especially in pat ient s w it h a hist ory of
preoperat ive ascit es and hypoalbuminemia. The rat ionale f or t his is t hat a liver
t ransplant in a pat ient w it h preoperat ive ascit es is like a large-volume
parascent esis, w hich is one of t he f ew sit uat ions in w hich albumin use has show n
clinical benef it . During volume resuscit at ion, a t ransf usion requirement beyond 4
unit s of blood merit s st rong considerat ion of operat ive explorat ion.
O nce euvolemia is ascert ained by f requent serial measures of CVP and
pulmonary pressures, vasopressors may be indicat ed in t he pat ient w ho remains
hypot ensive. I f blood product s are f urt her required t o correct anemia,
coagulopat hy, or t hrombocyt openia in a recipient w ho is already adequat ely
resuscit at ed, diuresis may be necessary t o prevent an excessive rise in CVP and
t he at t endant w orsening of hepat ic congest ion. Markedly elevat ed CVPs are
associat ed w it h diminished hepat ic perf usion pressure, pot ent ially leading t o
graf t ischemia, hepat ic art ery t hrombosis, or even parenchymal rupt ure. I n
pat ient s w it h renal insuff iciency unresponsive t o diuret ics, considerat ion should
be given t o early cont inuous venovenous hemodialysis (CVVHD).
Suggested Readings
Henriksen JH, Fuglsang S, Bendt sen F, et al. Art erial hypert ension in
cirrhosis: art erial compliance, volume dist ribut ion, and cent ral
haemodynamics. G ut . 2006; 55(3): 380–387.
Kamat h PS, Wiesner RH, Malinchoc M, et al. A model t o predict survival in

pat ient s w it h end-st age liver disease. Hepat ology. 2001; 33(2): 46–470.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 173 - R em em ber that Ur em ia Alone R ar ely C aus es
An Anion Gap to be Gr eater Than
173
Remember that Uremia Alone Rarely Causes An
Anion Gap to be Greater Than
The serum anion gap (AG ) is an import ant t ool in assessing t he acid-base st at us
of a pat ient and narrow ing t he diff erent ial diagnosis of et iologies responsible f or
acid-base dist urbance. The anion gap is a measure of serum elect rical neut ralit y.
Tot al serum cat ions alw ays equal serum anions. Thus, sodium + pot assium +
calcium + magnesium = bicarbonat e + chloride + phosphat e + sulf at e + prot ein +
organic acid anions. To simplif y t he equat ion, pot assium, calcium, and
magnesium are considered unmeasured cat ions (UC), and phosphat e, sulf at e,
prot ein, and organic acids are considered unmeasured anions (UA). Thus, UA -
UC = sodium - chloride - bicarbonat e = serum anion gap. The normal range of
anion gap is bet w een 3 t o 11 mEq/ L, w it h modern t echniques of measuring
elect rolyt es. A reduct ion of 2. 5 t o 3 mEq in AG should be made f or every 1-gram
reduct ion in t he serum albumin below 4.
A decrease in t he anion gap is seen w it h a reduct ion in unmeasured cat ions or
w it h increased anions. Albumin is an anion, and t hus a low albumin w ill decrease
t he AG . Lit hium, hypercalcemia, and cat ionic immunoglobulins cont ribut e
unmeasured cat ions, t hus also low ering t he AG .
Elevat ions in t he anion gap can be seen w it h a decrease in unmeasured cat ions,
w hich w ould include magnesium, calcium, and pot assium. I ncreases in
unmeasured anions can occur f rom organic acids (phosphat e, sulf at e), inorganic
acids (ket o acids, lact at e, uremic anions), and exogenous t oxins (salicylat e).
Typical causes of AG elevat ion are as f ollow s:
Ket oacidosis (diabet ic and alcoholic)

Renal f ailure
Lact ic acidosis
Rhabdomyolysis
Toxins
Met hanol
Et hylene glycol
Paraldehyde
Salicylat es
Suggested Readings
Haber R. A pract ical approach t o acid-base disorders. West J Med.
1992; 155: 146–151.
Salem M, Mujais S. G aps in t he anion gap. Arch I nt ern Med. 1992; 152: 1625–
1629.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 174 - D o not R eplete C alc ium in the S etting of High
P hos phor us or P hos phor us in the S etting of High C alc ium
174
Do not Replete Calcium in the Setting of High
Phosphorus or Phosphorus in the Setting of
High Calcium
Praveen Kalra MD
Mehmet S. O zcan MD
Calciphylaxis is an unusual condit ion t hat involves an abnormalit y in calcium and
phosphat e met abolism. I t is charact erized by occlusion of small subcut aneous
art erioles w it h calcium deposit s leading t o ischemic skin lesions as severe as dry
gangrene. Secondary hyperparat hyroidism caused by chronic renal f ailure (CRF)
is t he most common set t ing f or t he disorder alt hough it is report ed in pat ient s
w it h mult iple myeloma, primary hyperparat hyroidism, vit amin Dint oxicat ion, and
hypercalcemia of malignancy. About 1% of all pat ient s w it h CRF and 4%
undergoing hemodialysis are report ed t o develop calciphylaxis. I t carries a very
high mort alit y rat e (60% t o 78%), making prevent ion of paramount import ance.
Watch Out For

Concurrent elevat ion of calcium and phosphat e is crucial f or t he development of
calciphylaxis. Mult iplicat ion of serum calcium and phosphat e levels (in mg/ dL)
w it h t he t ot al equal t o or great er t han 56 has been associat ed w it h an increased
risk. I n chronic renal f ailure, serum phosphat e is generally elevat ed secondary t o
impaired excret ion. Calcium levels init ially f all because of decreased int est inal
absorpt ion but subsequent ly increase as a result of secondary
hyperparat hyroidism. This leads t o medial calcif icat ion and int imal hyperplasia of
subcut aneous art erioles. Since not all pat ient s at risk develop calciphylaxis,
presence of ot her f act ors such as local t rauma and hypercoagulable st at es (e. g. ,
prot ein C and S def iciencies) have been suggest ed. O t her condit ions t hat are
report ed t o be associat ed w it h calciphylaxis include diabet es mellit us, obesit y,
cort icost eroid use, and ant icoagulat ion w it h w arf arin.
Diagnosis of calciphylaxis requires clinical suspicion in t he high-risk pat ient .
Painf ul skin lesions w it h purple discolorat ion are t he usual init ial present at ion,
w hich progress t o cut aneous ulcers and subsequent eschar f ormat ion. I nf ect ion
of t he ulcers and dry gangrene are t he main causes of morbidit y and mort alit y.
Lesions have been described in various body part s w it h a proximal locat ion
associat ed w it h a w orse prognosis t han a dist al locat ion. I n many pat ient s,
laborat ory reveals a calcium-phosphat e product great er t han 56 and increased
parat hyroid hormone levels. Biopsy conf irms t he clinical diagnosis, w hich
t ypically show s a medial deposit ion of calcium in art eries w it h a diamet er of 40
t o 100 µm.
Management of calciphylaxis includes medical and surgical opt ions as w ell as
hyperbaric oxygen t herapy. As a prevent ive measure, maint aining a calcium-
phosphat e product less t han 56 mg2 / dL 2 is recommended in CRF and ot her high-
risk pat ient s. A low -phosphat e diet w it h or w it hout phosphat e binders is an
import ant t ool t o achieve t his goal. Surgical opt ions include w ound management
w it h debridement as w ell as parat hyroidect omy. Alt hough not a universally
agreed upon t echnique, parat hyroidect omy seems t o off er a bet t er prognosis
t han conservat ive management . I n addit ion, hyperbaric oxygen may aid in
t reat ment by inducing angiogenesis and f ibroblast prolif erat ion. Anot her benef it
of hyperbaric t herapy may be improvement of host def enses against inf ect ions.
Suggested Readings
Beus KS, St ack BC Jr. Calciphylaxis. O t olaryngol Clin Nort h Am.
2004; 37: 941–948.
Block G A. Prevalence and clinical consequences of elevat ed Ca x P product in

hemodialysis pat ient s. Clin Nephrol. 2000; 54: 318–324.
Duff y A, Schurr M, Warner T, et al. Long-t erm out comes in pat ient s w it h
calciphylaxis f rom hyperparat hyroidism. Ann Surg O ncol. 2006; 13: 96–102.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 175 - C hec k P os toper ative and S er ial S er um Levels
of P hos phor us and Aggr es s ively R eplete
175
Check Postoperative and Serial Serum Levels of
Phosphorus and Aggressively Replete
Phosphat e (PO4 ) is a t rivalent anion t hat is most not able f or t he energy-rich
bonds it f orms in adenosine t riphosphat e (ATP). Phosphat e is also int imat ely in
involved red blood cell f unct ion (2, 3-diphosphoglycerat e), plat elet f unct ion,
immune f unct ion, and t he cent ral nervous syst em (CNS). Phosphat e homeost asis
is closely linked w it h t hat of calcium, w hich t oget her are cont rolled by
parat hyroid hormone, calcit onin, and vit amin D.
Pat ient s in t he int ensive care unit (I CU) can have eit her hyperphosphat emia or
hypophosphat emia. Post operat ive hypophosphat emia is classically linked w it h
liver resect ion and living-donor hepat ect omies f or liver t ransplant at ion. Classic
t eaching on t his subject highlight s t he ut ilizat ion of t ot al body phosphat e st ores
f or t he regenerat ion of t he liver; how ever, a recent st udy concluded t he presence
of hyperphosphat uria point s t ow ard a renal origin of phosphat e loss. Low serum
phosphat e has also been report ed af t er ot her surgeries, such as colorect al
surgery, renal t ransplant at ion, abdominal aort ic aneurysm repair, and open heart
surgery. I f not prevent ed or t reat ed, hypophosphat emia w ill f requent ly develop
af t er major surgery on post operat ive day one and reach a nadir on post operat ive
day t w o.
Nonsurgical et iologies of hypophosphat emia can be broadly grouped int o t hree
main causes: decreased int est inal absorpt ion; t he redist ribut ion of phosphat e
int racellularly; and increased renal excret ion. Examples of int est inal
malabsorpt ion include vomit ing, diarrhea, vit amin D def iciency, and st arvat ion.
Cellular redist ribut ion can be observed w it h respirat ory alkalosis (such as t hat
seen w it h anxiet y or sepsis), t reat ed diabet ic ket oacidosis, and w it h t he eff ect s
of various hormones such as insulin, cat echolamines, and glucocort icoids.
I ncreased renal excret ion is seen w it h hyperparat hyroidism, hyperaldost eronism,
and ot her renal t ubular abnormalit ies. I at rogenic causes are usually relat ed t o
t ot al parent eral nut rit ion w it h inadequat e amount s of phosphat e ordered.
Mild hypophosphat emia (PO4 less t han 2. 5 mg/ dL) is usually asympt omat ic but
has been associat ed w it h signif icant ly higher post operat ive complicat ions
(pulmonary compromise, pancreat it is, inf ect ion,
ileus, et c. ) af t er liver resect ion. Severe hypophosphat emia (PO4 less t han 1. 0
mg/ dL) is lif e t hreat ening and can result in serious morbidit y including muscle
w eakness and respirat ory insuff iciency, cardiomyopat hy, eryt hrocyt e dysf unct ion,
CNS manif est at ions (conf usion, seizures, parest hesia), rhabdomyolysis, and
skelet al demineralizat ion. Hypophosphat emia has been linked w it h prolonged
mechanical vent ilat ion, prolonged I CU st ay, and prolonged hospit alizat ion.
The t reat ment of hypophosphat emia includes t reat ing t he underlying cause w hile
replacing elect rolyt es. Phosphat e salt s are available f or oral and parent eral use.
Sodium phosphat e and pot assium phosphat e are most commonly used, w it h t he
choice based on need f or replet ion of t he ot her elect rolyt es. As much as 45 t o
60 mmol per day of phosphat e preparat ions can be administ ered per 24-hour
period; overaggressive inf usions can lead t o hypocalcemia, hypot ension, and
renal f ailure. Replet ion t o a phosphat e level of at least 2. 5 mg/ dL is not
support ed by high-qualit y evidence but is probably rat ional given t he
complicat ions report ed f rom even mild hypophosphat emia.
Suggested Readings
Buell JF, Berger AC, Plot kin JS, et al. The clinical implicat ions of
hypophosphat emia f ollow ing major hepat ic resect ion or cryosurgery. Arch
Surg. 1998; 133(7): 757–761.
Salem RR, Tray K. Hepat ic resect ion-relat ed hypophosphat emia is of renal

origin as manif est ed by isolat ed hyperphosphat uria. Ann Surg.
2005; 241(2): 343–348.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 176 - C ons ider E lec tr olyte D is tur banc es W hen
Ther e is a C hange of Mental S tatus
176
Consider Electrolyte Disturbances When There
is a Change of Mental Status
Nirav G . Shah MD
Toxic-met abolic encephalopat hy result s in a change of ment al st at us and is a
common diagnosis in t he int ensive care unit (I CU). Elect rolyt e dist urbances need
t o be high on t he list of diff erent ial diagnoses w hen evaluat ing t he I CU pat ient
w it h alt ered ment al st at us. Some commonly encount ered dist urbances include
hyponat remia, hypernat remia, hypoglycemia, hyperglycemia, hypermagnesia,
acidosis, and alkalosis.
Watch Out For

Clinical manif est at ions of hyponat remia include dysf unct ion of t he cent ral nervous
syst em and are dependent on t he severit y and rat e of development . For
inst ance, t he acut e development of hyponat remia (<24 hours) as w ell as
ext remely low sodium concent rat ions (<120 mEq/ L) may manif est w it h severe
sympt oms including conf usion, agit at ion, delirium, let hargy, and seizures. The
most common et iologies include syndrome of inappropriat e ant idiuret ic hormone
(SI ADH), t reat ment w it h t hiazide diuret ics, polydipsia, and inappropriat e
administ rat ion of hypot onic int ravenous f luids. Treat ment consist s of f ree-w at er
rest rict ion and, in severe cases, administ rat ion of hypert onic saline t o correct
t he hyponat remia. To prevent cent ral pont ine myelinolysis, care must be t aken t o
prevent t oo rapid a rise in serum sodium. This disorder result s in quadriplegia
and pseudobulbar palsy and is prevent able w it h t he judicious correct ion of serum
sodium. The goal in pat ient s w it h chronic hyponat remia and in asympt omat ic
pat ient s should be a gradual correct ion of <10 mEq/ L per 24 hours.
Similarly, hypernat remia can result in alt ered ment al st at us. The most common
sympt oms include generalized muscle w eakness, let hargy, conf usion, and coma.
As w it h hyponat remia, sympt oms are dependent on t he degree and rat e of rise
of serum sodium. Et iologies cont ribut ing t o t his disorder include diabet es
insipidus, loop diuret ics, gast roint est inal losses, and hypert onic sodium.
Treat ment consist s of t reat ing t he underlying illness and correct ing t he
hypert onicit y. The lat t er is achieved by administ rat ion of hypot onic saline t o
low er t he serum sodium concent rat ion by 1 mEq/ L per hour in acut e
hypernat remia and 0. 5 mEq/ L per hour in chronic hypernat remia (goal
reduct ion of 10 mEq/ L per day in chronic cases). A relat ively slow correct ion
helps t o prevent cerebral edema and seizures.
Hypo- and hyper-glycemia can also cause a st at e of alt ered ment al st at us. I n
cases of hypoglycemia, sympt oms may include conf usion, t remulousness, coma,
and seizures. The cause can be inappropriat e insulin or oral hypoglycemic agent
administ rat ion, insulinomas, liver f ailure, and inf ect ion. I t is t reat ed by
administ rat ion of a dext rose load (usually one ampule of D50), w hich should
result in resolut ion of clinical sympt oms. Hyperglycemia can result in visual
changes, let hargy, coma, and seizures and most commonly occurs as part of t he
clinical spect rum of diabet ic ket oacidosis or nonket ot ic hyperglycemia.
Treat ment requires adequat e f luid resuscit at ion, int ravenous insulin t herapy,
correct ion of elect rolyt e derangement s (part icularly pot assium), and t reat ment of
t he underlying cause.
The most common side eff ect s of hypermagnesemia are t he neurologic
sympt oms of obt undat ion, loss of deep t endon ref lexes, and muscle paralysis.
They are most f requent ly seen in pat ient s w it h renal f ailure and w omen on high-
dose magnesium inf usions f or eclampsia. I n pat ient s w it h compromised renal
f unct ion, eit her hemodialysis or perit oneal dialysis must be init iat ed. I n addit ion,
w hen rapid reversal is required, calcium may be given int ravenously as a
magnesium ant agonist . For w omen on magnesium inf usions and w it h normal renal
f unct ion, cessat ion usually result s in a f airly rapid ret urn t o normal serum levels.
Finally, w hen assessing alt ered ment at ion t he diff erent ial should alw ays include
hypoxia, acidosis and alkalosis. An art erial blood gas w ill indicat e t he degree of
acidosis or alkalosis and can assist w it h det ermining w het her it s origin is
respirat ory, met abolic, or a mixed et iology. I n t he I CU set t ing, t he most common
et iologies are sepsis, uremia, hepat ic f ailure, and elect rolyt e abnormalit ies.
Treat ment consist s of correct ing t he underlying problem w hile providing
support ive care f or t he acidosis or alkalosis.
Suggested Readings
Adrogué HJ, Madias NE. Hyponat remia. N Engl J Med. 2000; 342: 1581–1589.
Adrogué HJ, Madias NE. Primary care: hypernat remia. N Engl J Med.
2000; 342: 1493–1499.
Bolt on CF, Young G B, eds. Baillere's Clinical Neurology. London: Balliere

Tindall; 1996: 577.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 177 - K eep the S er um P otas s ium at High or Nor m al
Levels W hen Attem pting to C or r ec t a Metabolic Alkalos is
177
Keep the Serum Potassium at High or Normal
Levels When Attempting to Correct a Metabolic
Alkalosis
Eric S. Weiss MD
Case
A 45-year-old at t orney t ravels t o t he Caribbean islands f or vacat ion. While aw ay
she develops prof use vomit ing and diarrhea due t o an unknow n gast roint est inal
(G I ) pat hogen. She is able t o f ly home but becomes prof oundly dehydrat ed,
unable t o keep up w it h her G I f luid losses. O n her second day back f rom t he
t rip, her husband ret urns home f rom w ork t o f ind her somnolent and minimally
responsive w it h decreased respirat ory eff ort . He immediat ely calls 911 and an
ambulance arrives t o t ransport her t o her local emergency depart ment (ED).
I n t he ED she is clearly dehydrat ed w it h dry mucous membranes and loss of her
normal skin t urgor. Her heart rat e is 124 and her blood pressure is measured at
100/ 70 mm Hg. I n addit ion, she appears somnolent and has a severely
diminished respirat ory drive, requiring endot racheal int ubat ion f or her severe
hypovent ilat ion. Her init ial art erial blood gas show s a pH of 7. 55 (normal 7. 40 t o
7. 44), w it h a part ial pressure of carbon dioxide (PCO2 ) of 66 mm Hg (normal
40). Her art erial part ial pressure of oxygen is normal at 95 mm Hg.
Wit h concern f or dehydrat ion, t he covering ED physician places a cent ral venous
cat het er and begins int ravenously administ ering f luids t o replace volume loss.
Af t er a half hour of volume administ rat ion, t he pat ient ret urns t o st able vit al
signs and again begins t o produce urine. How ever, t he nurse is concerned
because despit e t his improvement , t he pat ient remains w eak and let hargic,
unable t o f ollow commands. I n addit ion, t he cardiac monit or show s w hat t he
nurse believes t o be an abnormal rhyt hm. The resident orders an
elect rocardiogram (ECG ), w hich show s prominent U w aves. A serumelect rolyt e
panel is ordered, w hich reveals a pot assium level of 1. 7 mg/ dL (nl 3. 5 t o 5. 0
mg/ dL). Af t er aggressive pot assium replet ion, t he pat ient once again begins
f ollow ing commands and regains her st rengt h, able t o t olerat e event ual
ext ubat ion. She is discharged home on hospit al day number t w o.
Discussion
Met abolic alkalosis is t he most common acid-base abnormalit y seen in hospit al
inpat ient s. Severe alkalosis is an ext remely serious problem as
mort alit y rat es can exceed 50%. Alkalosis can lead t o diff use art erial
const rict ion w it h perf usion reduct ion. Thus, common signs of alkalosis include
decreased ment al st at us and seizures. I n addit ion, severe alkalosis can cause
hypovent ilat ion, leading t o hypoxemia. Finally, alkalosis can lead t o lif e-
t hreat ening hypokalemia due t o cellular shif t s of pot assium f or hydrogen ions in
an at t empt t o correct serum pH.
Met abolic alkalosis is def ined as a pat hologic increase in t he serum bicarbonat e
(HCO 3- ) concent rat ion. This occurs eit her as a gain in bicarbonat e or a loss of
acid. Compensat ion occurs t hrough act ion of t he respirat ory syst em, w it h
hypovent ilat ion t o lead t o ret ained CO2 and increased acid. Typical blood gas
values show a pH value around 7. 50 w it h PCO2 values around 60 mm Hg.
There are many causes of met abolic alkalosis. First is loss of hydrogen ions,
leading t o bicarbonat e excess. Hydrogen losses occur commonly in t he
gast roint est inal t ract . G I losses include vomit ing and nasogast ric suct ion. A
second source of alkalosis occurs as a response t o hypokalemia. I n t his set t ing,
pot assium w ill shif t out of cells as compensat ion. Hydrogen ions shif t
int racellularly, leading t o alkalosis. Finally, a t hird et iology of met abolic alkalosis
is due t o loss of bicarbonat e-poor, chloride-rich ext racellular f luid as can occur
w it h dehydrat ion in t he set t ing of diuret ic t herapy. The low er f luid levels cont ain
a relat ive excess in serumbicarbonat e, leading t o a “cont ract ion alkalosis. ”
This pat ient developed a met abolic alkalosis due t o a combinat ion of G I losses
f rom vomit ing and diarrhea and also had cont ract ion due t o ext racellular f luid
losses. Her present at ion w as t ypical in t he sense t hat she w as somnolent and
hypovent ilat ed t o compensat e f or her alkalosis. I n t he case described previously,
t here w as f ailure t o replet e a low pot assium level in t he set t ing of met abolic
alkalosis. Pat ient s w it h met abolic alkalosis have int racellular shif t ing of
pot assium, w hich allow s hydrogen ions t o shif t ext racellularly t o help correct pH.
For t his reason, pat ient s w it h met abolic alkalosis can manif est severe
hypokalemia. No pot assium w as checked unt il t he pat ient developed ECG
changes. The principal clinical manif est at ions of hypokalemia include let hargy
and muscle w eakness. I n addit ion, U w aves are of t en seen on ECG t racings.
Hypokalemia by it self is unlikely t o cause lif e-t hreat ening arrhyt hmias. How ever,
one must alw ays be cognizant of ot her elect rolyt e abnormalit ies associat ed w it h
hypokalemia, such as hypomagnesemia, w hich is know n t o cause serious
arrhyt hmias. I t is, how ever, import ant t o be caut ious w it h pot assium replet ion, as
overaggressive replet ion in t he set t ing of alkalosis can lead t o lif e-t hreat ening
hyperkalemia. This occurs w hen t he alkalosis is correct ed and int racellularly
st ored pot assium ret urns t o t he ext racellular compart ment . This can be
especially problemat ic in t he set t ing of acut e renal f ailure, w hich can occur in
dehydrat ed pat ient s. Thus, vigilance w it h caut ious pot assium replet ion of
pot assium t o normal t o high normal levels is of paramount import ance f or
pat ient s w it h met abolic alkalosis.
Suggested Readings
Whit t ier WL, Rut ecki G W. Primer on clinical acid-base problem solving. Dis
Mon. 2004; 50(3): 122–162.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 178 - D o Not R eplete C alc ium in R habdom yolys is
Unles s a P atient is in Tetany
178
Do Not Replete Calcium in Rhabdomyolysis
Unless a Patient is in Tetany
Rhabdomyolysis is a condit ion t hat is associat ed w it h t he dest ruct ion of skelet al
muscle. There are a number of causes of rhabdomyolysis. These are classif ied
as heredit ary or nonheredit ary. The heredit ary causes of rhabdomyolysis usually
involve t he def iciency of a part icular muscle enzyme (e. g. , carnit ine), w hich
prevent s t he abilit y t o generat e adenosine t riphosphat e (ATP) by eit her f at t y-
acid oxidat ion or anaerobic glycolysis during exercise. These pat ient s of t en have
a f amily hist ory of rhabdomyolysis or experience repeat ed bout s of
rhabdomyolysis. The nonheredit ary causes of rhabdomyolysis include t rauma
w it h large amount s of t issue damage (burns, crush injuries), sepsis, medicat ions
(st at ins, cocaine, amphet amines, ant iconvulsant s, serot onin, and prot ease
inhibit ors), elect rolyt e abnormalit ies (hypokalemia, hypophosphat emia), and
sepsis.
The diagnosis of rhabdomyolysis usually begins w it h an appropriat e clinical
hist ory t hat of t en includes excessive exercise associat ed w it h muscle pain and
cramping. Associat ed sympt oms are nonspecif ic and include nausea, w eakness,
and dehydrat ion. The laborat ory analysis is import ant in making t he diagnosis.
Usually, t here is a met abolic acidosis, an elevat ed serum creat ine kinase
concent rat ion >10, 000 I U/ L, hyperkalemia, hyperphosphat emia, and/ or
hypocalcemia. These abnormalit ies usually result f rom t he breakdow n product s
of t he skelet al muscle. A clue t o t he diagnosis includes a urinalysis t hat w ill be
posit ive f or blood but w ill not demonst rat e eryt hrocyt es on t he microscopic
examinat ion.
The t reat ment of rhabdomyolysis is f ocused on removing t he off ending agent and
providing support ive care. O ne of t he major consequences of rhabdomyolysis is
acut e renal f ailure and t he t reat ment of t he syndrome is direct ed at prevent ing
t his complicat ion. Aggressive int ravenous hydrat ion is indicat ed t o improve t he
solubilit y of t he by-product s, improve t he dehydrat ion, and improve urine f low.
Alkalinizat ion has been used t o improve t he solubilit y of myoglobin in t he renal
t ubules and prevent t he onset of acut e renal f ailure. Sodium bicarbonat e has
been used int ravenously t o alkalinize t he urine, but it cont ribut es t o t he
precipit at ion of calcium in ect opic locat ions and has
not been show n t o improve out comes. This ect opic precipit at ion of calcium is a
problem in rhabdomyolysis and increases w it h increasing muscle degradat ion.
Hypocalcemia, alt hough common in t his condit ion, is generally not t reat ed even
w hen sympt omat ic (parest hesias, cramping, mood changes) unless severe (e. g. ,
ionized hypocalcemia, evidence of t et any, convulsions, or hyperkalemia). This is
because t he serum albumin is also low and art if icially reduces t he serum calcium
and t he danger of ect opic calcium deposit ion is a common side eff ect , w hich may
lead t o hypercalcemia w hen t he rhabdomyolysis is subsiding.
Suggested Readings
Vanholder R, Sever MS, Erek E, et al. Rhabdomyolysis J Am Soc Nephrol.
2000; 11(8): 1553â 1 561.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 179 - C ons ider E xc es s C hlor ide as a C aus e of an
Unexplained Non- Anion- Gap Metabolic Ac idos is
179
Consider Excess Chloride as a Cause of an
Unexplained Non-Anion-Gap Metabolic Acidosis
Met abolic acidosis is t radit ionally classif ied int o an elevat ed anion-gap met abolic
acidosis or a non-anion-gap (hyperchloremic) met abolic acidosis. The import ance
of t his classif icat ion is t hat it provides a diff erent ial diagnosis and guidance f or
priorit ies of t reat ment . The anion gap is t he diff erence in charge bet w een t he
sodium and t he sum of t he bicarbonat e and chloride (Na-[ Cl+HCO 3 ] ). The normal
anion gap ranges f rom 10 t o 14. When t he anion gap is elevat ed in a met abolic
acidosis, t he diff erent ial diagnosis is charact erized by t he mnemonic MUDPI LES,
w hich st ands f or met hanol, uremia, diabet ic ket oacidosis, paraldehyde,
iron/ isoniazid, lact ic acidosis, et hanol/ et hylene glycol, and salicylat es.
A non-anion-gap met abolic acidosis is usually classif ied in t hree diff erent w ays.
First , a renal t ubular acidosis (RTA) must be considered and ruled out w hen a
non-anion-gap met abolic acidosis present s it self . These disorders may be
associat ed w it h eit her hypokalemia (proximal t ype I I RTA) or hyperkalemia (t ype
I V RTA). Second, a hyperchloremic met abolic acidosis may occur f rom t he loss
of f luids t hat are low in chloride. The most common f luids t o consider are
excessive st ool out put , drainage f rom pancreat ic or ileost omy drains, and urinary
diversions. Finally, a hyperchloremic met abolic acidosis may simply occur
because of excessive administ rat ion of a chloride cont aining salt , such as sodium
chloride, pot assium chloride, or ammonium chloride. When hyperchloremia
develops, t he kidney exchanges bicarbonat e t o maint ain elect rical neut ralit y and
t he met abolic acidosis develops.
What to Do
I n t he administ rat ion of t ot al parent eral nut rit ion (TPN), it is common t o provide
diff erent element al salt s in t he f orm of chloride salt s. For example, sodium,
pot assium, and calcium are usually added t o TPN complexed t o a chloride ion.
Af t er 3 t o 4 days of t herapy, t he chloride ion accumulat es and t he pat ient
develops a hyperchloremic met abolic acidosis. O ne met hod of prevent ing t his
occurrence is t o balance t he addit ion of ions in TPN bet w een chloride and
acet at e. This approach w ill bot h maint ain t he elect rical and chemical neut ralit y
and prevent
t he occurrence of t he acidosis w hile giving t he appropriat e elect rolyt e

const it uent s t o t he pat ient .
Suggested Readings
G aut hier PM, Szerlip HM. Met abolic acidosis in t he int ensive care unit . Crit
Care Clin. 2002; 18: 289–308.
Rodriguez Soriano J. Renal t ubular acidosis: t he clinical ent it y. J Am Soc

Nephrol. 2002; 13: 2160–2170.
> Table of C ontents > S hoc k/Fluids /E lec tr olytes > 180 - C ons ider Hyper c hlor em ic Metabolic Ac idos is
to be a R enal Tubular Ac idos is Until P r oven O ther w is e if O bvious S our c es of B ic ar bonate Los s es Like
D iar r hea, Ur inar y D iver s ions , and the Adm inis tr ation of C hlor ide ar e not P r es ent
180
Consider Hyperchloremic Metabolic Acidosis to
be a Renal Tubular Acidosis Until Proven
Otherwise if Obvious Sources of Bicarbonate
Losses Like Diarrhea, Urinary Diversions, and
the Administration of Chloride are not Present
I t is import ant t o consider t he diagnosis of renal t ubular acidosis (RTA) w hen a
non-anion-gap met abolic acidosis has a new onset . I n RTA, bicarbonat e is
excret ed t hrough t he kidneys and chloride ion is reabsorbed, leading t o t he t w o
major serum charact erist ics of t hese syndromes: met abolic acidosis and
hyperchloremia. These disorders may be associat ed w it h eit her hypokalemia
(t ypes I and I I RTA) or hyperkalemia (t ype I V RTA). The ot her causes of a
hyperchloremic met abolic acidosis include bicarbonat e losses t hrough t he
gast roint est inal t ract (e. g. , diarrhea or urinary diversions) and an excess
administ rat ion of chloride, w hich may occur f rom excess chloride ion
administ rat ion. The lat t er can occur secondary t o t he large-volume administ rat ion
of normal saline during resuscit at ion f rom t rauma or w it h t reat ment of diabet ic
ket oacidosis. I n addit ion, t he imbalance can occur w it h t he excess use of
chloride salt s rat her t han acet at e salt s f or crit ically ill pat ient s receiving t ot al
parent eral nut rit ion.
What to Do
Type I RTA is described by a hypokalemic, hyperchloremic met abolic acidosis
w it h a high urinary pH (>5. 5). The def ect result s f rom a reduced number of dist al
nephron ion exchangers t hat result f rom a genet ic def ect , t he presence of
cert ain t oxins or drugs, or diseases such as mult iple myeloma, cyst inosis, or
syst emic lupus. These pat ient s are suscept ible t o urinary st ones. The t reat ment
is t o provide alkali in a dose of 4 mmol/ kg/ day, w hich improves t he acidosis and
hypokalemia. Type 2 RTA is t he t ype of RTA t hat is associat ed w it h proximal
t ubular dysf unct ion and is associat ed w it h Fanconi syndrome. The present ing
sympt oms include a hyperchloremic, met abolic acidosis w it h a low urine pH. The
t reat ment includes alkali administ rat ion and at t ent ion t o grow t h if t he pat ient is a
child. Type I V RTA is charact erized by a hyperchloremic, met abolic acidosis w it h
hyperkalemia and a urine pH
of <5. 5. These pat ient s have a reduct ion in t heir ammonia excret ion and
bicarbonat e product ion. I t is common f or t hese pat ient s t o have some degree of
renal insuff iciency and t o have hypoaldost eronism (or aldost erone resist ance).
These pat ient s are usually t reat ed w it h mineral ocort icoid and a loop diuret ic,
w hich correct s t he aldost erone resist ance, hyperkalemia, and acidosis.
Suggested Readings
G aut hier PM, Szerlip HM. Met abolic acidosis in t he int ensive care unit . Crit
Care Clin. 2002; 18: 289–308.
Rodriguez Soriano J. Renal t ubular acidosis: t he clinical ent it y. J Am Soc

Nephrol. 2002; 13: 2160–2170.
> Table of C ontents > Neur o > 181 - B e Aler t for New - O ns et C auda E quina S yndr om e in P atients w ith
S ac r al or S pinal Fr ac tur es or S ur ger y and O btain E m er gent Neur os ur gic al C ons ultation if S us pec ted
181
Be Alert for New-Onset Cauda Equina Syndrome
in Patients with Sacral or Spinal Fractures or
Surgery and Obtain Emergent Neurosurgical
Consultation if Suspected
Michael J. Dorsi MD
Low back pain is t he f if t h most common reason f or all physician visit s and t he
second most commonly report ed sympt om. I n most cases, low back pain signals
a muscular disorder t hat can be managed conservat ively. How ever, physicians
should have height ened vigilance f or low back pain combined w it h bow el and/ or
bladder incont inence, mot or and/ or sensory loss in t he low er ext remit ies, and
saddle anest hesia, as t his may represent a neurosurgical emergency t ermed
cauda equina syndrome.
The cauda equina (f rom t he Lat in f or “horse's t ail”) is t he descript ive name f or
t he lumbar nerve root s emanat ing f rom t he dist al t ip of t he spinal cord. The
spinal cord t erminat es as t he conus medullaris t ypically at t he L1–L2 level. The
low er lumbar and sacral nerve root s cont inue caudally in t he cauda equina, exit
t heir respect ive neural f oramina, and ult imat ely provide mot or and sensory
innervat ion t o t he muscles and skin below t he w aist .
Cauda equina syndrome arises f rom compression of t he lumbosacral nerve root s.
Large herniat ed lumbar int ervert ebral discs are t he most common culprit . O t her
sources of compression include spinal or sacral f ract ures or surgery common in
I CU t rauma pat ient s, lumbosacral neoplasms, spinal st enosis, nonneoplast ic
masses such as cyst s, peripheral neuropat hy, and inf ect ious processes.
Signs and Symptoms

Cauda equina syndrome most of t en can be diagnosed clinically. Compression of
t he lumbar and S1 nerve root s result s in sensory and/ or mot or def icit s in t he
low er ext remit ies. St rengt h and sensat ion should be caref ully t est ed in a
dermat omal pat t ern. The pat ellar (L3 and L4) and Achilles (S1) deep t endon
ref lexes should be t est ed and may be decreased. The sacral nerve root s supply
mot or innervat ion t o t he uret hral and anal sphinct ers, sensat ion t o t he perineum,
and in a “bull's-eye” pat t ern t o t he skin surrounding t he anus. Thus, decreased
rest ing and/ or volit ional rect al t one are a clinical sign of sacral nerve root
compression, as is numbness in t he perineum. Anal
sphinct er ref lexes may be t est ed by st ret ching t he phallus (bulbocavernosus

ref lex) or by pinprick t o t he skin surrounding t he anus (anal w ink ref lex).
Emergent imaging should be obt ained in all pat ient s present ing w it h suspect ed
cauda equina syndrome. Magnet ic resonance imaging (MRI ) has emerged as t he
imaging modalit y of choice. Comput ed t omography myelography is indicat ed f or
pat ient s w it h a cont raindicat ion t o MRI or t hose w it h spinal inst rument at ion. I n
addit ion, neurosurgical consult at ion should be sought immediat ely in all cases of
cauda equina syndrome. Surgical int ervent ion aims t o decompress t he aff ect ed
lumbosacral nerve root s w it hin 48 hours of present at ion. The approach and
ext ent of surgery depends on t he nat ure and locat ion of t he compressing lesion.
I nst rument ed f usion is indicat ed w hen t here is evidence of spinal inst abilit y.
Suggested Readings
G reenberg MS, ed. Handbook of Neurosurgery. New York: Thieme; 2005: 298–
299.
St aat s P, ed. Pain: Just t he Fact s. Philadelphia: Lippincot t Williams & Wilkins;
2004: 141–146.
> Table of C ontents > Neur o > 182 - K eep P atients w ith D ur al Tear s Flat for 24–48 Hour s
182
Keep Patients with Dural Tears Flat for 24–48
Hours
Jayme E. Locke MD
Dural t ear is a know n complicat ion of spinal surgeries, such as laminect omies,
spinal f usions, and disc excisions. I t most commonly occurs in t he lumbar region.
I n general, t he t ear is not ed at t he t ime of operat ion w hen leakage of
cerebrospinal f luid (CSF) is not ed by t he surgeon. A t ear in t he dura result s in
decompression of t he t hecal sac and reduct ion of local pressure on t he epidural
veins, allow ing CSF t o leak in t o t he operat ive sit e. Dural t ears t hat are not
not ed at t he t ime of surgery of t en present in t he post operat ive period. Pat ient s
manif est severe headaches t hat are exacerbat ed by upright post ure. I n addit ion,
CSF can be seen leaking f rom t he w ound, or a subcut aneous collect ion may be
not ed. Conf irmat ion t hat t he f luid is CSF can be gained by t est ing t he f luid f or
bet a-2 t ransf errin.
What to Do
Whet her t he dural t ear is not ed int raoperat ively or st igmat a of a dural t ear are
not ed post operat ively (e. g. , headache, labile vit al signs post operat ively, f luid
collect ion), repair of t he dural t ear must be considered. O t herw ise, t he pat ient is
at risk f or t he development of pseudomeningocele and meningit is. O nce t he
condit ion is ident if ied, t he pat ient should be kept f lat t o minimize sympt oms. The
goal of t he repair is t o achieve a w at ert ight closure. The t ear should be repaired
using 4-0 or 6-0 dural sut ure w it h a t apered or reverse cut t ing needle. The
closure can be accomplished w it h simple int errupt ed or running locking sut ure
t echnique. I f t he t ear is t oo large f or primary repair, t hen a f ree graf t or f ascial
graf t can be used t o repair t he t ear. The repair should be t est ed by placing t he
pat ient in reverse Trendelenburg posit ion or via a Valsalva maneuver. I f t he
leakage of CSF persist s, t hen t he repair can be augment ed w it h f ibrin glue,
addit ional sut ure, gelat in sponge, or aut ologous f at . O nce complet e, t he
paraspinous muscles and t he f ascia should be closed in t w o layers. The f ascial
layer is responsible f or prevent ing durocut aneous f ist ulae. A drain should not be
lef t behind, as t he negat ive pressure creat ed may encourage a persist ent leak.
Af t er repair, all pat ient s should remain on bed rest in t he supine posit ion f or 24
t o 48 hours.
Suggested Readings
Brow ner B, ed. Skelet al Trauma: Basic Science, Management , and
Reconst ruct ion. 3rd ed. Philadelphia: WB Saunders; 2003: 937, 1026–1027.
Canale ST, ed. Campbell's O perat ive O rt hopaedics. 10t h ed. Philadelphia:
Mosby; 2003: 2011–2012.
> Table of C ontents > Neur o > 183 - K now The S tatus of C er vic al, Thor ac ic , and Lum bar S pine
S tability on All P os toper ative and Tr aum a P atients
183
Know The Status of Cervical, Thoracic, and
Lumbar Spine Stability on All Postoperative and
Trauma Patients
When a pat ient ret urns f rom t he operat ing room f ollow ing spinal surgery it is
import ant t o est ablish w it h t he surgical t eam t he st abilit y of t he spine as w ell as
t he int egrit y of t he dura. This inf ormat ion is crit ical should t he pat ient need t o be
reint ubat ed, vomit , or elevat e t he head of bed. There w ould be no great er
cat ast rophe t han f or a caregiver t o cause a spinal cord injury by incorrect ly
manipulat ing a pat ient w ho has an unst able spine, part icularly t he cervical spine
(Fig. 183. 1).
I n t he pat ient w it h an unst able cervical spine requiring reint ubat ion t here are
t hree opt ions: direct laryngoscopy w it h in-line immobilizat ion; f iberopt ic
int ubat ion; or an emergent surgical airw ay. Direct laryngoscopy w it h in-line
st abilizat ion (avoiding t ract ion on t he spine) is unquest ionably challenging f or t he
uninit iat ed but should be at t empt ed f irst . Fiberopt ic int ubat ion provides a more
opt imal view of t he airw ay but requires t hat t he equipment be readily available
and t hat t he pat ient is relat ively st able and noncombat ive. Emergency surgical
airw ay should be reserved f or circumst ances in w hich oral int ubat ion has f ailed.
I t should be st ressed t hat a pat ient does not need t o be int ubat ed t o be
vent ilat ed. I f t he bedside provider has know ledge of an unst able cervical spine
and can adequat ely oxygenat e a pat ient w it h a bag valve mask, it may be
prudent t o do so w hile obt aining t he services of an experienced airw ay
prof essional t hat may be more adept at t he procedures list ed earlier.
I n t he pat ient w ho vomit s and must lie f lat , most commonly due t o lumbar spine
inst abilit y or a dural t ear, t he pat ient should be log rolled t o prot ect t he airw ay.
While not unif ormly agreed upon, bed rest f ollow ing durot omy is t hought t o
decrease any pot ent ial cerebrospinal f luid (CSF) leak, allow t he dural t ears t o
seal, and reduce sympt oms (e. g. , nausea, vomit ing, dizziness). I t is incumbent
on t he I CU provider t o consider t his bed-rest prot ocol and if inst it ut ed, st art
appropriat e t hromboembolic (deep vein t hrombosis/ pulmonary embolism)
prophylaxis measures if t he query regarding t he presence of a dural t ear is
aff irmat ive.
FIG URE 183. 1. Sixt h cervical (C6) vert ebral compression f ract ure. (Court esy
of Robert Hendrickson, MD; reused w it h permission f rom Badaw y M.
Cervical vert ebral compression w edge f ract ure. I n: G reenberg MI .
G reenberg's Text -At las of Emergency Medicine. Philadelphia: Lippincot t
Williams & Wilkins; 2005: 629.)
Selected Readings
G haf oor AU, Mart in TW, G opalakrishnan S, et al. Caring f or t he pat ient s w it h
cervical spine injuries: w hat have w e learned? J Clin Anest h. 2005; 17: 640–
649.
Hodges SD, Humphreys SC, Eck JC, et al. Management of incident al

durot omy w it hout mandat ory bed rest . Spine. 1999; 24(19): 2062–2064.
> Table of C ontents > Neur o > 184 - Apply Appr opr iate D eep Vein Thr om bos is P r ophylaxis To P atients
W ith S pinal C or d Injur y
184
Apply Appropriate Deep Vein Thrombosis
Prophylaxis To Patients With Spinal Cord Injury
Jose I. Suarez MD
Acut e spinal cord injury (SCI ) is common w it h an est imat ed 8, 000 t o 10, 000 new
cases in t he Unit ed St at es each year. The mean age of present at ion is 31 years
and t he majorit y of pat ient s are male. SCI is associat ed w it h signif icant long-
t erm disabilit y and mort alit y.
Etiology
SCI has t radit ionally been classif ied as t raumat ic and nont raumat ic. Traumat ic
SCI is usually due t o mot or vehicle collisions (MVCs) (MVAs) (20% t o 25%),
mot orcycle collisions (MCCs) (25% t o 30%), or f alls (20% t o 25%). The most
common locat ions f or t raumat ic SCI are t he cervical spine (50% t o 55%),
f ollow ed by t he t horacolumbar region (15% t o 20%), t horacic spine (10% t o
15%), and t he lumbosacral region (10%).
Nont raumat ic causes of SCI are numerous. Such condit ions include: various
bact erial (including spinal epidural abscesses), viral, f ungal, or parasit ic
inf ect ions; neoplast ic lesions (usually ext ramedullary primary or met ast at ic
t umors); vascular event s (inf arct ions or hemorrhages f rom vascular
malf ormat ions); demyelinat ing lesions such as mult iple sclerosis; t oxins;
aut oimmune disorders; and nut rit ional abnormalit ies such as vit amin B12
def iciency.
Neurologic damage and t he ensuing clinical manif est at ions w ill depend on t he
ext ent and level of t he SCI . Physical examinat ion usually reveals muscle
w eakness and a sensory level below t he level of t he lesion in complet e lesions.
I n t hose pat ient s w it h incomplet e SCI t here is some preservat ion of mot or and
sensory f unct ions below t he level of compromise. I nit ially pat ient s w it h SCI w ill
present w it h at onia and arref lexia below t he level of t he lesion (spinal shock)
and w it hin a f ew days or w eeks w ill experience spast icit y and hyperref lexia.
Venous Thromboembolism in SCI

Pat ient s w it h SCI w ho do not receive prophylact ic t reat ment have t he highest
incidence of deep venous t hrombosis (DVT) and pulmonary embolism (PE) among
all pat ient s. A sympt omat ic DVT has been f ound in 60% t o 100% of SCI pat ient s
and PE represent s t he t hird leading cause of deat h. Addit ional risk f act ors f or
DVT in SCI pat ient s include age, concomit ant low er limb f ract ure, and delayed
init iat ion of prophylact ic t reat ment . I t has been show n t hat SCI pat ient s are at
great est risk of DVT and PE during t he acut e care phase. How ever, t he risk is
st ill present during t he rehabilit at ion period. The incidence of new DVT may be
as high as 30% bet w een 2 t o 6 w eeks f ollow ing SCI .
Thromboprophylaxis after SCI

Six randomized st udies (albeit small) have show n t hat t he use of low -dose
unf ract ionat ed heparin (LDUH) alone or sequent ial compression devices alone
are not suff icient as t hromboprophylaxis. How ever, adjust ed-dose unf ract ionat ed
heparin and low -molecular-w eight heparin (LMWH) have been f ound t o be more
eff icacious. This f inding has also been report ed in SCI pat ient s undergoing
rehabilit at ion (uncont rolled st udies). There is also evidence t hat oral
ant icoagulant s such as w arf arin are also eff ect ive t hromboprophylact ic
t reat ment s w hen init iat ed short ly af t er SCI . I t is import ant t o not e t hat
t hromboprophylaxis should be init iat ed once pract it ioners f eel secure about
having achieved eff ect ive hemost asis af t er SCI .
Rout ine insert ion of inf erior vena cava f ilt ers (I VCFs) in SCI pat ient s is st ill
cont roversial. There is st rong evidence t hat w hen appropriat e
t hromboprophylaxis is given I VCFs are not necessary. Even t hough I VCFs may
t heoret ically reduce t he incidence of PE, t hey are associat ed w it h complicat ions
and higher cost s. Pat ient s w ho have been show n t o benef it f rom I VCF insert ion
include t hose w it h concomit ant long bone f ract ures, DVT f ormat ion despit e
t hromboprophylaxis, and t hose w it h cont raindicat ions t o ant icoagulat ion.
The current recommendat ions f or t hromboprophylaxis af t er SCI include t he
f ollow ing:
1. Thromboprophylaxis f or all pat ient s w it h acut e SCI .

2. Single prophylaxis modalit ies (i. e. , low -dose unf ract ionat ed heparin,
graduat ed compression st ockings, or sequent ial compression devices alone)
should be avoided.
3. Thromboprophylaxis w it h LMWH is t he pref erred t reat ment once primary

hemost asis f rom t he injury is obt ained. Alt ernat ives t o LMWH include t he
combined use of sequent ial compression devices and eit her low -dose
unf ract ionat ed heparin or LMWH.
4. When ant icoagulat ion is cont raindicat ed early in SCI , sequent ial compression
devices and/ or graduat ed compression st ockings should be used w it h
considerat ion of an I VCF.
5. During t he rehabilit at ion phase, SCI pat ient s should be t reat ed w it h LMWH or
convert ed t o oral w arf arin (int ernat ional normalized rat io range 2. 0 t o 3. 0).
6. The insert ion of I VCFs as primary prophylaxis is not recommended.
Suggested Readings
G eert s WH, Pineo G F, Heit JA, et al. Prevent ion of t hromboembolism: t he
sevent h ACCP conf erence on ant it hrombot ic and t hrombolyt ic t herapy. Chest .
2004; 126: 338–400.
Maxw ell RA, Chavarria-Aguilar M, Cockerham WT, et al. Rout ine prophylact ic
vena cava f ilt rat ion is not indicat ed af t er acut e spinal cord injury. J Trauma.
2002; 52: 902–906.
> Table of C ontents > Neur o > 185 - B e Aler t For Autonom ic D ys r eflexia in Intens ive C ar e Unit P atients
W ith a S pinal C or d Injur y
185
Be Alert For Autonomic Dysreflexia in Intensive
Care Unit Patients With a Spinal Cord Injury
Jose I. Suarez MD
Aut onomic dysref lexia is an acut e syndrome t hat f requent ly occurs in spinal cord
injury (SCI ) pat ient s w it h a level usually above T6. I t is charact erized by
excessive unmodulat ed sympat het ic out f low in response t o noxious st imuli below
t he spinal cord level. This can lead t o dangerous elevat ions of blood pressure
w it h disast rous clinical sequelae.
Pathophysiology
I n bot h normal and SCI pat ient s, noxious st imulat ion of peripheral sensory
recept ors (i. e. , bladder dist ent ion) w ill act ivat e aff erent pat hw ays t o produce a
sympat het ic response. The sympat het ic out f low st imulat es peripheral
vasoconst rict ion, t hus elevat ing art erial blood pressure. Art erial vessel
barorecept ors are t hen act ivat ed and signal brain-st em inhibit ory pat hw ays t hat
descend t he spinal cord t o limit t he sympat het ic out f low and prevent excessive
art erial vasoconst rict ion. The brain-st em inhibit ory cent ers increase vagal
st imulat ion t o t he heart , result ing in bradycardia. I n SCI pat ient s, vagal out put t o
t he heart st ill occurs; how ever, brain-st em inhibit ory descending pat hw ays t hat
limit sympat het ic out f low are blocked in t he spinal cord at t he level of t he spinal
cord lesion. This result s in excessive vasoconst rict ion and markedly elevat ed
blood pressure (Fig. 185. 1).
Signs and Symptoms

The main clinical signs and sympt oms of aut onomic dysref lexia are relat ed t o
bot h t he excessive sympat het ic out put below t he spinal cord level and f rom
compensat ory parasympat het ic out put originat ing f rom above t he spinal cord
level. Prof ound hypert ension occurs because of excessive peripheral
vasoconst rict ion, and t he skin w ill appear cool and clammy below t he spinal cord
level. Above t he spinal cord level, parasympat het ic out put may result in
vasodilat at ion of blood vessels, result ing in a pounding headache, skin f lushing,
sw eat ing, and nasal congest ion. O t her sympt oms t he pat ient may report include
visual spot s, blurry vision, and anxiet y. The elevat ed blood pressure may lead t o
a host of clinical sequelae, including int racranial
hemorrhage, seizures, hypert ensive encephalopat hy, cerebral edema, at rial

f ibrillat ion, pulmonary edema, renal f ailure, coma, and deat h.
FIG URE 185. 1. I n bot h normal and spinal-cord-injury pat ient s, noxious
st imulat ion of peripheral sensory recept ors (i. e. , bladder dist ent ion) w ill
act ivat e aff erent pat hw ays t o produce a sympat het ic response. From
Blackmer J. Rehabilit at ion medicine: 1. Aut onomic dysref lexia. CMAJ.
2003; 169(9): 931–935.
Mult iple st imuli may t rigger aut onomic dysref lexia. The general cat egories
include bladder and urinary t ract , gast roint est inal, dermat ological, skelet al,
reproduct ive, and hemat ological sources. The most common specif ic t riggers f or
aut onomic dysref lexia are t hought t o be f rom bladder dist ension and f ecal
impact ion (Table 185. 1).
M anagement
I mmediat e assessment of blood pressure should be made in pat ient s w it h
sympt oms of aut onomic dysref lexia. Normal blood pressure in
quadriplegic pat ient s is about 90 t o 110 mm Hg syst olic in t he sit t ing posit ion.
G enerally, a blood pressure above 150 mm Hg in adult s w arrant s pharmacologic
t reat ment . The pat ient should be moved t o t he upright posit ion and legs low ered
t o allow f or pooling of blood in t he low er ext remit ies, w hich may reduce blood
pressure. Any const rict ive clot hing should be loosened. The blood pressure
should be checked every f ew minut es. Pharmacologic t reat ment of blood
pressure should be done by using rapid-onset short -act ing agent s. I mmediat e-
release nif edipine 10-mg capsule is a pref erred agent . This should be t aken by a
“bit e-and-sw allow ” met hod t o f acilit at e absorpt ion. How ever, sublingual
nif edipine is not advisable. O t her medicat ions t hat can be used include
nit ropast e, nit roglycerin, hydralazine, and sodium nit roprusside.
TABLE 185-1 PRECIPITATING FACTORS FOR

AUTONOM IC DYSREFLEXIA
SYST EM PRECIPITAT ING FACTORS
Bladder distention, infection, urethral

Urinary tract distension, instrumentation, catheter
traction
Gastrointestinal Fecal impaction, constipation,

tract instrumentation, infection, ulceration
Skin Pressure sore, ingrown toenail
Heterotopic ossification, fracture,

Skeletal joint dislocation
Labor and delivery, menstruation,
Reproductive
testicular torsion
Deep vein thrombosis, pulmonary

Hematological
embolism
Nasal decongestants,
Medications
sympathomimetics
While blood pressure is being st abilized, a prompt search f or t he precipit at ing

cause of aut onomic dysref lexia should be made. Since bladder dist ension is
t hought t o be t he most common t rigger, a t rial of bladder cat het erizat ion should
be done if no indw elling cat het er is already present . A solut ion of 2% lidocaine
jelly should be inst illed int o t he uret hra bef ore t his is done. I f t he pat ient already
has an indw elling cat het er, it should be checked f or obst ruct ions along t he ent ire
lengt h. I f blocked, t he cat het er should be irrigat ed w it h normal saline. I f no
problems are f ound w it h t he bladder, a rect al exam and abdominal radiograph
must be perf ormed. I f f ecal impact ion is ruled out , a syst emat ic search f or t he
causes list ed previously must be undert aken.
Suggested Readings
Acut e management of aut onomic dysref lexia: individuals w it h spinal cord injury
present ing t o healt h-care f acilit ies. J Spinal Cord Med. 2002; 25 (Suppl
1): S67–S88.
Blackmer J. Rehabilit at ion medicine: 1. Aut onomic dysref lexia. CMAJ.

2003; 169(9): 931–935.
> Table of C ontents > Neur o > 186 - C ons ider the Us e of S ter oids for Neur ologic al Tr aum a in B lunt
S pinal C or d Injur y O nly
186
Consider the Use of Steroids for Neurological
Trauma in Blunt Spinal Cord Injury Only
Jacob T. G utsche MD
Patrick K. Kim MD
Cort icost eroids have been st udied f or decades as a t reat ment of t raumat ic
cent ral nervous syst em injury. The proposed mechanism of benef it is t he
reduct ion of inf lammat ion and edema, leading t o decreased ischemic injury and
neuronal apopt osis. High-dose met hylprednisolone w as w idely adopt ed as t he
st andard of care f or pat ient s w it h blunt spinal cord injury f ollow ing t he
publicat ion of t he Nat ional Acut e Spinal Cord I njury St udies (NASCI S) I I and I I I .
These st udies suggest ed t hat st eroid t reat ment result s in slight improvement in
mot or f unct ion at long-t erm f ollow -up. Subsequent clinical st udies and crit ical
review s have challenged t he validit y of t he result s of t he NASCI S I I st udy.
Despit e it s w ide adopt ion, st eroid t herapy f or spinal cord injury is st ill
cont roversial.
Met hylprednisolone (Solu-Medrol) is t he st eroid of choice based on NASCI S.
Large doses are required, st art ing w it h a bolus int ravenous inf usion of 30 mg/ kg
of body w eight over 15 minut es f ollow ed by an inf usion of 5. 4mg/ kg per hour. I f
st eroid t reat ment begins w it hin 3 hours of injury, t he durat ion of t he inf usion is
23 hours. I f st eroid t reat ment begins bet w een 3 and 8 hours af t er injury, t he
durat ion of inf usion is 47 hours. I f t he 8-hour w indow is missed, st eroids should
not be given. The use of st eroids in blunt spinal cord injury should not be
ext rapolat ed t o pat ient s w it h penet rat ing spinal cord injury (e. g. , gunshot
w ounds, st abs). St udies have show n no improvement in neurological out come in
penet rat ing injury and should not be used in t his pat ient populat ion.
The use of st eroids in acut e t raumat ic brain injury (TBI ) has also been a source
of debat e and has been even more ext ensively st udied. A 1997 syst emat ic
review suggest ed t hat st eroids may reduce mort alit y af t er TBI . How ever, t he
dat a f rom a 10, 000-pat ient , mult icent er, randomized, placebo-cont rolled t rial
proved t he opposit e. The Cort icost eroid Randomisat ion af t er Signif icant Head
I njury (CRASH) st udy proved a st at ist ically signif icant increase in t he risk of
deat h f or pat ient s randomized t o receive cort icost eroids. The st udy w as st opped
early because of more deat hs at 2 w eeks in t he st eroid group. Final
analysis of 6-mont h f ollow -up conf irmed t he higher mort alit y in pat ient s t reat ed
w it h st eroids.
I t must be remembered t hat st eroids can have severe side eff ect s on various
organ syst ems. These include higher rat es of inf ect ion, gast roint est inal bleeding,
and pancreat it is. Pat ient s t reat ed w it h st eroids also have delayed w ound healing
and diff icult y w it h blood sugar cont rol. Considerat ion must be given t o t hese
risks, and st eroids should be given only w hen t heir proven benef it out w eighs
t hese signif icant risks. St eroids may be considered f or blunt spinal cord injury,
alt hough t he dat a are not as persuasive as once t hought . How ever, t he dat a are
quit e convincing t hat st eroids should not be used f or penet rat ing spinal cord
injury or in pat ient s w it h t raumat ic brain injury.
Suggested Readings
Bracken MB, Shepard MJ, Collins WF, et al. A randomized, cont rolled t rial of
met hylprednisolone or naloxone in t he t reat ment of acut e spinal-cord injury.
Result s of t he Second Nat ional Acut e Spinal Cord I njury St udy. N Engl J Med.
1990; 322(20): 1405–1411.
Bracken MB, Shepard MJ, Holf ord TR, et al. Administ rat ion of
met hylprednisolone f or 24 or 48 hours or t irilazad mesylat e f or 48 hours in t he
t reat ment of acut e spinal cord injury. Result s of t he Third Nat ional Acut e
Spinal Cord I njury Randomized Cont rolled Trial. Nat ional Acut e Spinal Cord
I njury St udy. JAMA. 1997; 277(20): 1597– 1604.
Coleman WP, Benzel D, Cahill DW, et al. A crit ical appraisal of t he report ing
of t he Nat ional Acut e Spinal Cord I njury St udies (I I and I I I ) of
met hylprednisolone in acut e spinal cord injury. J Spinal Disord.
2000; 13(3): 185–199.
Edw ards P, Arango M, Balica L, et al. CRASH t rial collaborat ors. Final result s
of MRC CRASH, a randomised placebo-cont rolled t rial of int ravenous
cort icost eroid in adult s w it h head injury-out comes at 6 mont hs. Lancet .
2005; 365(9475): 1957–1959.
Heary RF, Vaccaro AR, Mesa JJ, et al. St eroids and gunshot w ounds t o t he
spine. Neurosurgery. 1997; 41(3): 576–583; discussion 583–584.
> Table of C ontents > Neur o > 187 - S tar t an E ar ly B ow el R egim en in P atients After S pinal C or d Injur y
187
Start an Early Bowel Regimen in Patients After
Spinal Cord Injury
Jose I. Suarez MD
Acut e spinal cord injury (SCI ) is a devast at ing problem result ing in more t han
10, 000 permanent ly disabled pat ient s in t he Unit ed St at es each year. The vast
majorit y of t raumat ic SCI s occur in ot herw ise healt hy young adult s bet w een 16
and 30 years of age w it h a long lif e expect ancy. Because of t hat SCI is
associat ed w it h a signif icant burden on healt h care resources.
Gastrointestinal Consequences of SCI

SCI is associat ed w it h common gast roint est inal abnormalit ies including
gast roesophageal ref lux (G ERD); delayed gast ric empt ying t ime; alt ered colonic
mot ilit y w it h increased t ransit t ime; severe const ipat ion; prolonged bow el
evacuat ion t ime; abdominal dist ent ion; and hemorrhoids. Such alt erat ions w ill
become evident in isolat ion or in combinat ion depending on t he ext ent and level
of t he SCI . G ERD and delayed gast ric empt ying t ime are more commonly seen in
t et raplegic pat ient s. I leus can be seen a f ew days af t er SCI and t ypically in
pat ient s w it h complet e lesions. Pat ient s w it h lesions above T12 w ill have a
spast ic anal sphinct er, t hus ret aining ref lex bow el empt ying. How ever, pat ient s
w it h lesions below T12 w ill have a f laccid anal sphinct er w it h accompanying loss
of bot h ref lexive and volunt ary ref lex bow el empt ying.
M anagement of Gastrointestinal Complications of SCI

All SCI pat ient s should receive adequat e pept ic ulcer prophylaxis and a bow el
care regimen f rom admission (Tabl e 187. 1). G ast roint est inal care st art s w it h
adequat e f luid administ rat ion and diet ary int ake. Fluids are very import ant f or
st ool consist ency. I t has been recommended t hat SCI pat ient s should receive at
least 2 t o 3 lit ers of f luids daily. A diet high in f iber (at least 15 g/ day) may be
benef icial t o increase st ool bulk. I f adequat e diet ary management does not
improve bow el f unct ion, t hen pharmacologic t reat ment s should be inst it ut ed.
Agent s t hat have been used include st ool sof t eners such as docusat e; st ool bulk
f ormers such as calcium polycarbophil, met hylcellulose, psyllium, or
lact ulose; st imulant s of perist alsis and prokinet ic agent s such as senna; and
cont act irrit ant s such as bisacodyl or glycerin supposit ories. Such programs
should be consist ent but individualized depending upon t he SCI severit y, level of
injury, pat ient 's lif est yle, and plans t o ret urn t o w ork. During t he acut e phase of
spinal shock t he rect um is usually f laccid. Theref ore pat ient s may require manual
removal of st ool f rom t he rect um daily. Pat ient s w it h cervical or t horacic spine
lesions may ret ain t heir ref lexive bow el responses. Such pat ient s can be
managed w it h digit al st imulat ion of t he rect um, w hich result s in ref lexive bow el
empt ying. Pat ient s w it h incomplet e SCI may not require any specif ic t reat ment
beyond t he diet ary recommendat ions
TABLE 187-1 GASTROINTESTINAL M ANIFESTATIONS

AFTER SCI AND SOM E TREATM ENT
RECOM M ENDATIONS
CONDIT ION RECOMMENDED MANAGEMENT
Gastroesophageal Head elevation, motility, or

reflux prokinetic agents
Prophylaxis with H2-blockers or

Peptic ulcer
proton pump inhibitors
Bowel rest, correct electrolytes

Ileus (serum potassium >4 mEq/dL,
consider total parenteral nutrition
Bowel rest, correct electrolytes,

decompression with nasogastric
Bowel obstruction tube decompression. If medical
management does not work, then
consider surgery
Fecal impaction Bowel care program, scheduled

laxative
Suggested Readings
Chen D, Nussbaum SB. The gast roint est inal syst em and bow el management
f ollow ing spinal cord injury. Phys Med Rehab Clin Nort h Am. 2000; 11: 45–56.
Derw enskus J, Zaidat O O . Spinal cord injury and relat ed disorders. I n Suarez
JI , ed. Crit ical Care Neurology and Neurosurgery. Tot ow a, NJ: Humana Press;
2004: 417–432.
> Table of C ontents > Neur o > 188 - C ons ider Moder ate Hypother m ia After C ar diac Ar r es t
188
Consider Moderate Hypothermia After Cardiac
Arrest
Moderat e hypot hermia has been demonst rat ed t o improve neurologic out come
and mort alit y in pat ient s w ho remain comat ose despit e ret urn of spont aneous
circulat ion (RO SC) af t er vent ricular f ibrillat ion (VF) or pulseless vent ricular
t achycardia (VT)–induced cardiac arrest . This t ype of event leads t o global
cerebral ischemia w it h t he pot ent ial f or severe anoxic or hypoxic brain injury and
progression t o brain deat h. The pat ient populat ions st udied in t he original t rials,
w hich demonst rat ed t his benef it , w ere t hose w ho had out -of -hospit al event s and
as such t here has been considerable debat e as t o w het her t his t reat ment w ould
be usef ul f or in-hospit al cardiac arrest s. I n-hospit al cardiac arrest s are of t en
secondary t o long-developing det eriorat ions and are of t en t he end result of a
t erminal illness. This is in cont rast t o out -of -hospit al arrest s, w hich are usually
secondary t o an acut e coronary event . While t he dat a are imperf ect , t here is
some suggest ion t hat inpat ient s w it h w it nessed arrest s f rom VF and VT may also
benef it f rom t his t reat ment .
The t ime bet w een t he w it nessed arrest and RO SC is import ant . RO SC may not
occur prompt ly and t here is an out er limit t hat should obviat e init iat ing
hypot hermia. The t ime limit adhered t o f or t he out -of -hospit al arrest s w as 60
minut es f rom t ime of collapse. Most in-hospit al resuscit at ions do not cont inue
t his long and it is unclear w hat t he out er limit f or RO SC should be f or init iat ing of
cooling f or in-hospit al arrest s.
The earlier t he cooling is init iat ed af t er ret urn of spont aneous circulat ion, t he
bet t er. O ne of t he out -of -hospit al st udies init iat ed cooling in t he f ield. The t arget
t emperat ure should be bet w een 32° and 34° Celsius core t emperat ure (based on
a bladder pressure pulmonary art ery cat het er or ot her “core” t emperat ure
measurement ). I t is desirable t o achieve t he t arget t emperat ure w it hin a 2-hour
t ime f rame. How long t o maint ain t he hypot hermia is unclear. O ne of t he t w o
st udies maint ained it f or 12 hours and t he ot her f or 24 hours and bot h show ed
improved out come. Many w ould advocat e t hat unless t here is a cont raindicat ion
t o maint aining t he hypot hermia f or 24 hours, t his w ould be t he pref erred t ime.
Af t er t his point t he pat ient is act ively rew armed t o normot hermia (37° C).
During t he period of hypot hermia, very close at t ent ion t o a w ide variet y of
paramet ers needs t o be maint ained. Elect rolyt es need t o be checked f requent ly
and volume needs t o be appropriat ely resuscit at ed since hypot hermia can induce
diuresis and elect rolyt e abnormalit ies. Shivering must be cont rolled, usually w it h
sedat ives and/ or paralyt ics, as it increases myocardial oxygen demand
dramat ically. Cardiac dysrhyt hmias may occur secondary t o t he hypot hermia as
w ell as a result of t he original cardiac insult . Some prot ocols use prophylact ic
lidocaine inf usion t o cont rol t his. Coagulopat hy may occur, t hough t his is
generally not a problem unt il w ell below t he t arget t emperat ure range.
Nevert heless, t he prot hrombin t ime/ part ial t hromboplast in t ime should be
monit ored f requent ly. Severe vasoconst rict ion occurs as a result of t he
hypot hermia and t his may place addit ional w ork on t he heart in t erms of
increased af t erload. I t is prudent t o place a pulmonary art ery cat het er t o monit or
cardiac perf ormance as w ell as volume st at us during t he t reat ment and f or a
period of t ime af t er rew arming.
O nce t he pat ient is rew armed, t he primary issue is neurologic assessment
including f requent neurologic exams, head comput ed t omography scan, magnet ic
resonance imaging, and possibly ot her t est s such as evoked pot ent ials t o assist
in prognost icat ion f or neurologic recovery. Assessment and t reat ment of
underlying causes of t he original event also need t o be perf ormed in parallel.
Suggested Readings
Bernard SA, G ray TW, Buist MD, et al. Treat ment of comat ose survivors of
out -of -hospit al cardiac arrest s w it h induced hypot hermia. N Engl J Med.
2002; 346: 557– 563.
The Hypot hermia af t er Cardiac Arrest St udy G roup. Mild t herapeut ic

hypot hermia t o improve t he neurologic out come af t er cardiac arrest . N Engl J
Med. 2002; 346: 549– 556.
> Table of C ontents > Neur o > 189 - B e Vigilant for B lunt C er ebr ovas c ular Injur y After Tr aum a
189
Be Vigilant for Blunt Cerebrovascular Injury
After Trauma
Michael D. G rossman MD
Underst anding of t he spect rum of blunt cerebrovascular injury has changed
markedly during t he past decade. I n t he past , t hese injuries w ere likely t o be
det ect ed only af t er development of f ocal neurologic signs in pat ient s w it h normal
comput ed t omography (CT) scans of t he head. I mproved access t o diagnost ic
t echnology and improved survival of t he most severely injured t rauma pat ient s
have f ocused at t ent ion upon earlier diagnosis and t reat ment of t his condit ion.
Est imat ion of incidence is diff icult and dependent upon specif icat ion of t he
populat ion t hat should undergo screening, t he met hod used t o screen, sensit ivit y
of t he t est , and clear agreement as t o w hat f indings const it ut e an injury.
Published f igures range bet w een 0. 1% t o more t han 1. 0%. I ncidence seems t o
rise as eff ort s at screening are int ensif ied, suggest ing t he t rue incidence
probably exceeds 1%; how ever, many injuries are “minimal” and t heir clinical
consequences indet erminat e.
Watch Out For

Blunt cerebrovascular injury may be caused by a variet y of injury mechanisms.
Relat ively minor injury mechanisms may produce signif icant injury. Examples
include hyperf lexion or ext ension injuries of t he neck producing vert ebral art ery
occlusion and subsequent embolizat ion int o t he post erior cerebral circulat ion. For
t he purposes of screening large populat ions of t rauma pat ient s, pot ent ial f or
blunt cerebrovascular injury is most of t en associat ed w it h severe closed head
injury, cervical spine f ract ure (especially if t he vert ebral f oramina are aff ect ed),
skull base f ract ures, severe midf ace f ract ures, and presence of a “seat -belt
sign” across t he base of t he neck. Cranial nerve f indings associat ed w it h blunt
f orce t rauma t o t he head or neck are also suggest ive of proximit y t o carot id and
vert ebral art eries.
Blunt cerebrovascular injury may aff ect t he major t runks of t he carot id or
vert ebral art erial syst ems eit her in t he neck or skull base. Anat omic injury
includes lacerat ion or t ransact ion w it h bleeding, lacerat ion or t ransact ion w it h
t hrombosis, development of int imal f laps, dissect ions, pseudoaneurysms, and
art eriovenous f ist ulas. O ccurrence of st roke or cerebrovascular accident , t he
most f eared pat hophysiologic consequences of blunt cerebrovascular injury,
depends upon t he
presence or absence of t hrombosis, embolic phenomenon, and collat eral f low.

Decisions regarding ant icipat ed risks and benef it s of t herapy should t ake t hese
f act ors int o considerat ion.
What to Do
The gold st andard f or ident if icat ion of blunt cerebrovascular injury is
convent ional f our-vessel cerebral angiography. How ever, because t his t est is
invasive and resource int ensive, w idespread screening of populat ions at risk has
been inf requent . Recent at t ent ion has t urned t o t he use of magnet ic resonance
imaging/ art eriography (MRI , MRA) and CT angiography f or screening purposes.
Comparison of CT angiography t o angiography has been f avorable in recent
publicat ions provided high-speed, mult idet ect or devices are used (“high speed”
or “light speed” scanners). CT angiography has good sensit ivit y and excellent
specif icit y f or carot id injuries; sensit ivit y f or vert ebral art eries may not be as
good. Sixt een-slice scanners appear t o provide bet t er sensit ivit y f or vert ebral
injuries. Current lit erat ure support s using CT angiography t o screen f or blunt
cerebrovascular injury f ollow ed by angiography t o est ablish t he anat omic
diagnosis.
MRI , MRA can det ect vascular injury w it h excellent sensit ivit y and specif icit y and
can ident if y end-organ eff ect s of embolizat ion or t hrombosis if t hese have
occurred. Even in t he absence of ident if ied vascular injury on CT angiography,
t he presence of cerebral inf arct ion on MRI might indicat e t he need f or t reat ment
or at t he very least more aggressive diagnost ic w orkup using angiography. CT
angiography has no ut ilit y in det ect ion of end-organ eff ect s of injury. MRI / A is
more cumbersome t o obt ain t han CT angiography but w ould seem t o have
great er overall ut ilit y part icularly if clinical suspicion is very high.
Treat ment depends upon t he injury ident if ied. While t hrombosis might w arrant
ant icoagulat ion, associat ed injuries or condit ions might preclude it . There is no
clear consensus on w het her heparin, w arf arin, or ant iplat elet agent s are t he best
t herapy. Cert ain f indings (including bleeding, pseudoaneurysm, and dissect ion)
might w arrant int ervent ional neuroradiology procedures.
Suggested Readings
Biff l WL, Egglin T, Benedet t o B, et al. Sixt een slice comput ed t omographic
angiography is a reliable noninvasive screening t est f or clinically signif icant
blunt cerebrovascular injuries. J Trauma. 2006; 60(4): 745–751.
Biff l WL, Ray CE, Moore EE, et al. Non-invasive diagnosis of blunt
cerebrovascular injuries: a preliminary report . J Trauma. 2002; 850–856.
Bub L, Hollingsw ort h W, Jarvik JG , et al. Screening f or blunt cerebrovascular

injury: evaluat ing t he eff icacy of mult idet ect or comput ed t omographic
angiography. J Trauma. 2005; 59: 697.
Mayberry JC, Brow n CV, Mullins RJ, et al. Blunt carot id art ery injury: t he
f ut ilit y of screening and diagnosis. Arch Surg. 2004; 139: 609–613.
> Table of C ontents > Neur o > 190 - Have an E xtr em ely High Thr es hold in Giving Antihyper tens ives in
Head Tr aum a
190
Have an Extremely High Threshold in Giving
Antihypertensives in Head Trauma
Ronald F. Sing DO
Blood pressure cont rol in t he int ensive care unit (I CU) can result in cont roversial
and diff icult management decisions because of t he paucit y of dat a. This seems
especially apparent in pat ient s w it h t raumat ic brain injury (TBI ). Hypert ension
can be t he body's nat ural response t o TBI in an at t empt t o aut oregulat e cerebral
blood f low as t he body may require an elevat ed blood pressure t o maint ain
adequat e cerebral blood f low. Hypert ension af t er TBI may also be t he result of a
nonspecif ic ref lex pressor response or t riggers of a cat echolamine response,
such as pain and agit at ion.
O nce t he primary t rauma t o t he brain has occurred, minimizing secondary insult s
is t he primary object ive t o achieve maximal neurologic recovery. Avoiding
hypot ension (syst olic blood pressure <90mm Hg) and hypoxia (oxygen sat urat ion
<90% or PaO2 <60 mm Hg) in TBI pat ient s is essent ial t o prevent cerebral
ischemia, since bot h independent ly predict a w orse out come. The Brain Trauma
Foundat ion recommends t he mean art erial blood pressure (MAP) be maint ained
above 90 mm Hg in an at t empt t o maint ain a cerebral perf usion pressure (CPP)
great er t han 70 mm Hg.
What to Do
I n t he event t hat t he MAP is excessively elevat ed beyond t hat required t o
maint ain an adequat e CPP, sedat ion and analgesia should be opt imized prior t o
considerat ion of ant ihypert ensive administ rat ion. No ant ihypert ensive t herapy
should be inst it ut ed w it hout consult at ion w it h a senior member of t he t reat ment
t eam. I n t he very unlikely event t hat blood pressure reduct ion is w arrant ed, t he
most desirable agent w ould have a smoot h dose-response relat ionship providing
a predict able and cont rollable onset . The drug w ould also be a short -act ing,
t it rat able inf usion t o allow t he specif ic blood pressure t arget s t o be achieved
w hile avoiding hypot ension. Addit ionally, t he drug w ould have f ew adverse
eff ect s and w ould not cause int racranial pressure (I CP) elevat ion.
Esmolol is an accept able choice as it is a short -act ing bet a-recept or ant agonist
administ ered as a const ant inf usion w it h f ew adverse eff ect s. Labet alol, a bet a-
recept or ant agonist w it h alpha-blocking act ivit y, can
also be used and is administ ered as a const ant inf usion or can be administ ered
int ermit t ent ly f or acut e elevat ions. Eit her agent is pref erred in pat ient s w it h
baseline t achycardia. Nicardipine, a calcium-channel blocker, w orks exclusively
as a peripheral vasodilat or. Administ ered as a const ant inf usion, it s smoot h
dose-response relat ionship is pref erable in pat ient s w it h bradycardia, heart
f ailure, bronchospasm, or chronic obst ruct ive pulmonary disease. These agent s
have no direct eff ect on I CP or cerebrovascular blood volume, making t hem
accept able t reat ment opt ions.
What not to Do
Sodium nit roprusside should be avoided in TBI pat ient s since it can cause I CP
elevat ions secondary t o dilat at ion of cerebral vasculat ure. I t s pot ent eff ect s and
lack of smoot h dose-response relat ionship can result in hypot ension during
inf usion and rebound hypert ension af t er discont inuat ion of t he agent . I n addit ion,
it s use put s pat ient s at risk f or it s severe pot ent ial side eff ect of cyanide or
t hiocyanat e t oxicit y w it h prolonged use, high doses, or in pat ient s w it h liver or
renal dysf unct ion.
Suggested Readings
Brain Trauma Foundat ion. G uidelines. ht t p: / / w w w 2. braint rauma. org/ guidelines
Chobanian AV, Bakris G L, Black HR, et al. Nat ional Heart , Lung, and Blood
I nst it ut e Joint Nat ional Commit t ee on Prevent ion, Det ect ion, Evaluat ion, and
Treat ment of High Blood Pressure; Nat ional High Blood Pressure Educat ion
Program Coordinat ing Commit t ee. The sevent h report of t he Joint Nat ional
Commit t ee on Prevent ion, Det ect ion, Evaluat ion, and Treat ment of High Blood
Pressure: t he JNC 7 report . JAMA. 2003; 289(19): 2560–2572.
Marik PE. Management of t raumat ic brain injury. I n Handbook of Evidence-

Based Crit ical Care. 1st ed. New York: Springer-Verlag New York; 2001: 321–
330.
Rose JC, Mayer SA. O pt imizing blood pressure in neurological emergencies.

Neurocrit ical Care. 2004; 1(3): 287–300.
> Table of C ontents > Neur o > 191 - D o Not Give Mor e than 7 D ays of Antis eiz ur e Medic ation in Head
Tr aum a
191
Do Not Give More than 7 Days of Antiseizure
Medication in Head Trauma
Lauren Paton MD
Ronald F. Sing DO
Among all pat ient s w it h severe head t rauma w ho receive medical at t ent ion, about
12% develop post t raumat ic seizures; t his rat e is more t han 50% f or t hose w it h
penet rat ing head injuries. Post t raumat ic seizures are def ined as early (occurring
in t he f irst w eek af t er injury) or lat e (longer t han 7 days af t er injury) (Tabl e
191. 1). Seizure act ivit y in t he early post t raumat ic period f ollow ing head injury
may cause secondary brain damage because of increased met abolic demands,
raised int racranial pressure, and excess neurot ransmit t er release. The
development of seizures complicat es t he acut e management and rehabilit at ion of
head-injured pat ient s. These seizures can be physically and psychologically
debilit at ing. How ever, early seizure development does not aff ect long-t erm
pat ient out comes.
Watch Out For

The use of ant iseizure prophylaxis must be balanced against t he medicat ion's
demonst rat ed t oxicit y and side eff ect s, w hich may be especially disabling in t his
populat ion. Valproat e, phenyt oin, and carbamazepine can cause a rash,
hepat ot oxicit y, blood dyscrasias, at axia, dizziness, and nausea. I n addit ion,
phenyt oin can cause hypot ension and lead t o a decrease in t he cerebral
perf usion pressure. The impact of being on seizure medicat ion in regard t o t he
impact on act ivit ies of daily living should not be minimized; in some st at es
pat ient s on ant iseizure medicat ions are not allow ed t o drive a mot or vehicle.
The prophylact ic use of ant iseizure medicat ions (usually phenyt oin or
carbamazepine) af t er t raumat ic brain injury has become common, alt hough t he
available dat a are not suff icient t o support a pract ice st andard. Class I evidence
demonst rat es a signif icant ly low er rat e of early post t raumat ic seizures among
pat ient s t reat ed w it h prophylact ic ant iseizure medicat ions. I mport ant ly, t hese
st udies had f ew adverse eff ect s occurring w it hin t he f irst w eek of t herapy f or
penet rat ing and nonpenet rat ing brain injury pat ient s. I n t hese pat ient s, t he
benef it s of seizure prophylaxis out w eighed t he low risk rat e.
TABLE 191-1 RISK FACTORS FOR POSTTRAUM ATIC

SEIZURES
Glasgow Coma Scale (GCS) score less than 10
Cortical contusion
Depressed skull fracture
Subdural hematoma
Epidural hematoma
Intracerebral hematoma
Penetrating head wound
Seizure within 24 hours of injury
What not to Do
A common I CU error is prolonging t he durat ion of ant iseizure medicat ion in t he
head-injured pat ient beyond 1 w eek. The Brain Trauma Foundat ion guidelines
st at e t hat prophylact ic use of ant iseizure medicat ions should not be used t o
prevent lat e post t raumat ic seizures. Class I evidence exist s support ing t he
st at ement t hat t he use of phenyt oin, carbamazepine, or valproat e is not
recommended in t he reduct ion of lat e post t raumat ic seizures. Prophylact ic
t reat ment should st op 7 days af t er t raumat ic brain injury w hen caring f or
crit ically ill neurot rauma pat ient s.
O ne f inal not e is t hat serum levels of ant iseizure medicat ions should not be
t est ed; t he half -lif e of t hese drugs is suff icient ly long t hat st eady st at e is not
reached w it hin 7 days.
Suggested Readings
Brain Trauma Foundat ion. G uidelines. ht t p: / / w w w 2. braint rauma. org/ guidelines
Chang BS, Low enst ein DH; Q ualit y St andards Subcommit t ee of t he American
Academy of Neurology. Pract ice paramet er: ant iepilept ic drug prophylaxis in
severe t raumat ic brain injury: report of t he Q ualit y St andards Subcommit t ee
of t he American Academy of Neurology. Neurology. 2003; 60(1): 10â 1 6.
> Table of C ontents > Neur o > 192 - C alc ulate the Glas gow C om a S c ale Us ing the B es t Motor
R es pons e
192
Calculate the Glasgow Coma Scale Using the
Best Motor Response
Rajan G upta MD
The G lasgow Coma Scale (G CS) w as init ially developed t o assess t he level of
brain f unct ion f ollow ing injury. Recent ly it s applicat ion has expanded t o assess
brain f unct ion f rom ot her neurological et iologies, especially in t he crit ical care
set t ing. The score is calculat ed by adding assigned values t o diff erent levels of
f unct ion in t hree main cat egories: mot or, verbal, and eye response. The G CS
(Tabl e 192. 1) allow s f or a common language bet w een healt h care providers
during t he assessment of neurologically impaired pat ient s. Using t his score t o
st rat if y t he severit y of brain injury assist s in det ermining f urt her diagnost ic and
t herapeut ic int ervent ions. I t s predict ive value is limit ed w hen t he score is used
alone; how ever, w hen combined w it h ot her physiologic and anat omic crit eria, it
may be more usef ul in evaluat ing out comes.
The severit y of t raumat ic brain injury is of t en def ined by t he G CS. A score ≤8 is
considered severe injury, 9 t o 13 is moderat e, and 14 t o 15 is mild. The mot or
score may be t he most reliable and accurat e component and appears t o
correlat e w ell w it h t he t ot al G CS score. The use of t he mot or score alone avoids
problems of t en associat ed w it h t he collect ion of t he verbal and eye component s.
The eye port ion of t he score adds lit t le t o t he overall predict ive abilit y of t he
G CS score and can easily be aff ect ed by ot her f act ors. The verbal response is
impossible t o det ermine in int ubat ed pat ient s and has led t o t he adopt ion of
annot at ing t he G CS score w it h a T in t hese pat ient s (e. g. , 3T). The G CS is most
reliable w hen used in assessing pat ient s w it h isolat ed head injuries; how ever,
t here remains variabilit y in calculat ing t he score among diff erent pract it ioners as
w ell as diff erent inst it ut ions.
TABLE 192-1 THE GLASGOW COM A SCALE

EYE
MOTOR VERBAL
OPENING
6 Obeys 4
5 Oriented
commands Spontaneous
5 Localizes pain 4 Confused 3 To voice
4 W ithdraws to 3 Inappropriate
2 To pain
pain words
2
3 Flexion to pain
Incomprehensible 1 None
(decorticate)
sounds
2 Extension to pain
1 None
(decerebrate)
1 None
There are several pit f alls t hat can of t en render t he G CS score inaccurat e and
unreliable. I nt oxicat ion and subst ance abuse, t he use of sedat ives and narcot ics,
t he need f or int ubat ion, t he presence of f acial injuries, hearing loss, and
dement ia can all be conf ounding f act ors. I n addit ion, t he mot or response can be
inaccurat e in t he set t ing of a spinal cord or peripheral nerve injury. How ever,
despit e it s inexact ness, most providers agree t hat pat ient s w it h a G CS of <<9
require emergent int ubat ion.
Suggested Readings
Bast os PG , Sun X, Wagner DP, et al. G lasgow Coma Scale score in t he
evaluat ion of out come in t he int ensive care unit : f indings f rom t he Acut e
Physiology and Chronic Healt h Evaluat ion I I I st udy. Crit Care Med.
1993; 21(10): 1459–1465.
Healey C, O sler TM, Rogers FB, et al. I mproving t he G lasgow Coma Scale
score: mot or score alone is a bet t er predict or. J Trauma. 2003; 54(4): 671–
680.
> Table of C ontents > Neur o > 193 - Us e Magnetic R es onanc e Im aging ( Not Head C om puted
Tom ogr aphy) as the Gold S tandar d Tes t for D iffus e Axonal Injur y
193
Use Magnetic Resonance Imaging (Not Head
Computed Tomography) as the Gold Standard
Test for Diffuse Axonal Injury
Eliahu S. Feen MD
Jose I. Suarez MD
Diff use axonal injury develops most commonly in t he set t ing of angular
accelerat ion of t he brain in head injury. This movement causes a shear injury
bet w een t he gray and w hit e mat t er w it h microscopic hemorrhages developing in
part icular areas of t he brain. The shear also disrupt s axons, w hich sw ell and
w hose severed ends event ually become t he globular st ruct ures seen
pat hologically as t he “ret ract ion balls of Cajal. ” The earliest changes occur hours
af t er injury and t he pat hologic changes cont inue f or several days. The clinical
manif est at ions of diff use axonal injury include a period of unconsciousness (and
amnesia) last ing more t han 6 hours.
Prognosis f or head-injury pat ient s is diff icult t o predict . The severit y of diff use
axonal injury can vary w it h t he severit y of head injury. Prognosis f or t raumat ic
brain injury also varies w it h t he severit y of t he injury. I n t he most severe cases
(in w hich pat ient s w ho do not die remain in a veget at ive st at e), about one-t hird
of pat ient s recover consciousness by 3 mont hs. How ever, recovery cont inues
signif icant ly af t er t hat , w it h 46% of pat ient s recovering consciousness by 6
mont hs, and 52% by 1 year. Clinicians must t heref ore t ake ext reme caut ion in
giving a prognosis f or such pat ient s f or at least 6 mont hs. The great majorit y of
pat ient s in a veget at ive st at e have diff use axonal injury on pat hological
examinat ion of t heir brains. Thus, w hen diff use axonal injury is ident if ied
premort em (via head imaging), t he same caut ion must be used in giving a
prognosis f or diff use axonal injury.
Watch Out For

Head comput ed t omography (CT) in t raumat ic brain injury is obviously absolut ely
essent ial. Many of t he various t ypes of t raumat ic brain injury are easily seen on
head CT and include subdural hemat oma, epidural hemat oma, int racranial
hemorrhage, and brain cont usion. Head CT is also usef ul in ident if ying acut e
hydrocephalus, massive cerebral edema, and midline shif t or herniat ion. Head
CT, how ever, is not usef ul in ident if ying diff use axonal injury. Blood w it hin t he
brain t hat is seen on CT imaging can be a marker f or diff use axonal injury.
How ever, t he head CT can be normal in head-injury pat ient s f or w hom diff use
axonal injury is observed on MRI . For t his reason, even some pat ient s w it h a
normal head CT need int racranial pressure monit oring af t er t raumat ic brain
injury.
The diff erent modes of magnet ic resonance imaging (MRI ) sequences have
bet t er sensit ivit y f or det ect ing diff use axonal injury. Fast f luid-at t enuat ed
inversion-recovery (FLAI R), diff usion-w eight ed imaging (DWI ), and gradient echo
sequences have all been report ed t o have higher yield in t he det ect ion of diff use
axonal injury t han regular (T2-w eight ed) MRI sequences. MRI can also det ect
microhemorrhages w it hin t he brain not visible on CT (t hrough t he use of gradient
echo sequences). DWI sequences of an MRI are very sensit ive f or ischemia.
Thus, MRI is usef ul in ident if ying diff use axonal injury, w hich may not become
visible unt il 6 t o 12 days af t er t he insult . As w ell, MRI is much more sensit ive
t han CT f or ident if ying ischemic, hypoxic, or hypot ensive injuries t hat may be
delayed or develop as secondary brain injury. Many clinicians use CT acut ely and
MRI in t he subacut e and chronic phases of t raumat ic brain injury f or t hese
reasons.
Suggested Readings
Adams JH, Jennet t B, McLellan DR, et al. The neuropat hology of t he
veget at ive st at e af t er head injury. J Clin Pat h. 1999; 52: 804–806.
Huisman TAG M, Sorensen AG , Hergan K, et al. Diff usion-w eight ed imaging f or

t he evaluat ion of diff use axonal injury in closed head injury. J Comp Asst
Tomography. 2003; 27: 5–11.
McArt hur DL, Chut e DJ, Villablanca JP. Moderat e and severe t raumat ic brain
injury: epidemiologic, imaging and neuropat hologic perspect ives. Brain Pat hol.
2004; 14: 185–194.
Mit t l RL, G rossman RI , Hiehle JF, et al. Prevalence of MR evidence of diff use
axonal injury in pat ient s w it h mild head injury and normal head CT f indings.
AJNR. 1994; 15: 1583–1589.
Mult i-societ y Task Force on PVS. Medical aspect s of t he persist ent

veget at ive st at e (in t w o part s). N Engl J Med. 1994; 330: 1499–1508, 1572–
1579.
> Table of C ontents > Neur o > 194 - R em em ber that ther e is Us ually an Upw ar d D r ift in S om e
Intr ac r anial- P r es s ur e Monitor R eadings The Longer they Have B een in P lac e
194
Remember that there is Usually an Upward Drift
in Some Intracranial-Pressure Monitor Readings
The Longer they Have Been in Place
Eliahu S. Feen MD
Jose I. Suarez MD
The normal int racranial pressure (I CP) ranges bet w een 5 and 15 mm Hg (7 t o 20
cm H2 O ). I CP can be monit ored invasively as w ell as noninvasively. O ne of t he
most sensit ive indicat ors of elevat ed I CP is t he neurologic exam—as t he I CP
rises above normal, t he neurologic exam declines. A drop in t he level of
consciousness serves as one of t he f irst signs of elevat ed I CP. I t is w ell know n
t hat as I CP rises, cert ain neurologic signs develop. For example, pressure on
t he t hird cranial nerve causes a f ixed, dilat ed pupil, and diff usely increased I CP
can cause bilat eral sixt h cranial nerve palsies, w it h impaired lat eral gaze. When
I CP rises high enough t o cause t ranslocat ion of brain t issues f rom t heir normal
neuroanat omic locat ions, classic herniat ion syndromes can develop. Neurologic
signs, how ever, lack suff icient accuracy and specif icit y. Most import ant ly, t he
goal of neurologic monit oring is t o ident if y and t reat elevat ed I CP bef ore
neurologic damage, such as herniat ion occurs. I n addit ion, in pat ient s w it h
t raumat ic brain injury, a delet erious rise in I CP is observed short ly af t er t he
injury f or an ext ended period of t ime. As I CP rises, cerebral perf usion pressure
(CPP) drops, since CPP is t he diff erence bet w een mean art erial pressure (MAP)
and I CP. When CPP declines suff icient ly, ischemia can result (secondary brain
injury). Management in t his area is cont roversial, but some guidelines suggest
maint aining a CPP bet w een 60 and 70 mm Hg in order t o avoid secondary brain
injury as a result of decreased cerebral blood f low.
Noninvasive met hods of monit oring I CP include neuroimaging (especially
comput ed t omography and magnet ic resonance imaging) and neurovascular
ult rasound. Neit her of t hese met hods provides good accuracy. Transcranial
Doppler st udies have not been f ound t o provide precise predict ions of w hen an
I CP w ill become acut ely and dangerously elevat ed. Theref ore, direct (i. e. ,
invasive) met hods remain t he pref erred t echnique. These include int ravent ricular
cat het ers, parenchymal cat het ers, subarachnoid bolt s, epidural cat het ers, and
lumbar drains (t o measure t he cerebrospinal f luid [ CSF] pressure in t he lumbar

spinal space).
When a pat ient has an ext ernal int ravent ricular cat het er placed int o eit her t he
lat eral vent ricles or anot her part of t he vent ricular syst em of t he brain, a
pressure t ransducer can be connect ed t o measure t he f luid pressure of t he CSF
t hat drains. I nt ravent ricular cat het ers are considered t he gold st andard of direct
I CP monit oring, because f luid conveys pressure so w ell. Placement of
int ravent ricular cat het ers carries w it h it up t o a 6% risk of hemorrhage and up t o
about a 20% risk of inf ect ion, w it h t he low est report ed inf ect ion rat es of about
5%. Clinicians commonly use prophylact ic ant ibiot ics in conjunct ion w it h t he
placement of int ravent ricular cat het ers, but dat a are lacking about w het her t his
pract ice reduces t he inf ect ious complicat ions. O f t he inf ect ions t hat do occur,
most occur af t er at least 5 days of cat het er placement . Some clinicians have
recommended on t his basis t o replace cat het ers t hat are more t han 5 days old.
Parenchymal cat het ers, such as t he Camino or Codman Micro Sensor cat het ers,
are insert ed t hrough a burr hole in t he skull w it h dept h cat het ers going t hrough
t he dura int o brain parenchyma. Measurement of t he t issue pressure represent s
I CP measurement . How ever, many st udies show t hat I CP is not unif orm
t hroughout t he brain. So, t he value of t he I CP measured may ref lect only t he
localized pressure in t he area of t he cat het er. This is not a problem unique t o
parenchymal cat het ers, as int ravent ricular cat het ers also suff er f rom t his issue,
w hich may even be exacerbat ed if t here is not f reely circulat ing CSF t o equally
dist ribut e pressure. Hemorrhagic and inf ect ious complicat ions appear t o be
about t he same as w it h int ravent ricular cat het ers. O ne of t he problems report ed
w it h parenchymal cat het ers is t he phenomenon of “zero drif t , ” in w hich t he zero
pressure baseline (t o w hich t he cat het er is calibrat ed during init ial placement )
drif t s. I n Camino cat het ers one st udy indicat ed an average drif t of up t o 3 mm
Hg of pressure. The longer t he cat het er remains in place, t he great er t he drif t .
This must be considered w hen t rends in I CP are observed.
Subarachnoid bolt s are f iberopt ic pressure t ransducers placed t hrough a burr
hole but w it h only a saline-f illed lumen exposed t o t he subarachnoid space
t hrough a t iny incision in t he dura. Subarachnoid bolt s carry a much low er risk of
hemorrhagic complicat ions and a low er inf ect ion rat e, w it h report ed rat es of
inf ect ion ranging only as high as about 7%. How ever, w it h subarachnoid bolt s
CSF cannot be analyzed, and t he bolt s are not as mechanically st able w it h
respect t o accurat e I CP measurement as compared w it h int ravent ricular
cat het ers
or parenchymal cat het ers. Recommendat ions about t he rout ine use of
prophylact ic ant ibiot ics w it h t he parenchymal cat het ers and subarachnoid bolt s
are lacking, but prophylact ic ant ibiot ics are commonly administ ered.
Epidural cat het ers are I CP-measuring of t he epidural space devices insert ed int o
t he epidural space. G iven t he devices anat omic locat ion bet w een t he dura and
calvaria, inf ect ion and hemorrhagic complicat ions are low er. How ever, t he
accuracy of t he I CP measurement is compromised, since it is not direct ly
measured. Lumbar drains can measure I CP and do provide CSF analysis w hile
avoiding signif icant hemorrhagic complicat ions and having a low inf ect ion rat e.
How ever, t he accuracy of t he I CP measurement s is uncert ain, because it is not
clear in many brain-injured pat ient s t hat t he lumbar spinal cist ern w ill accurat ely
ref lect t he int ravent ricular pressure.
Suggested Readings
Czosnyka M, Pickard JD. Monit oring and int erpret at ion of int racranial
pressure. J Neurol Neurosurg Psychiat ry. 2004; 75: 813–821.
Hollow ay KL, Barnet T, Choi S, et al. Vent riculost omy inf ect ions: t he eff ect of
monit oring durat ion and cat het er exchange in 584 pat ient s. J Neurosurg.
1996; 85: 419–424.
Mayhall CG , Archer NH, Lamb VA, et al. Vent riculost omy-relat ed inf ect ions: a
prospect ive epidemiologic st udy. N Engl J Med. 1984; 310: 553–559.
Paramore CG , Turner DA. Relat ive risks of vent riculost omy inf ect ion and
morbidit y. Act a Neurochir. 1994; 127: 79–84.
Piper I , Barnes A, Smit h D, et al. The Camino int racranial pressure sensor: is
it opt imal t echnology? An int ernal audit w it h a review of current int racranial
pressure monit oring t echnologies. Neurosurgery. 2001; 49: 1158–1165.
Suarez JI , ed. Crit ical Care Neurology and Neurosurgery. Tot ow a, New Jersey:
Humana Press; 2004.
> Table of C ontents > Neur o > 195 - B e Aw ar e that Inc r eas ing P os itive E nd- E xpir ator y P r es s ur e m ay
R es ult in Inc r eas ing Intr ac r anial P r es s ur e
195
Be Aware that Increasing Positive End-
Expiratory Pressure may Result in Increasing
Intracranial Pressure
Eliahu S. Feen MD
Jose I. Suarez MD
Watch Out For

Posit ive end-expirat ory pressure (PEEP) can cause numerous physiologic
changes. These changes may t ake place in t he lungs, w it h respect t o
int rat horacic pressures and lung compliance, and in cardiac out put . Clinical
observat ion has also demonst rat ed f or more t han 30 years t hat PEEP can
increase int racranial pressure (I CP). Animal and some human dat a have
support ed t w o t heories about how t his increase in I CP t akes place. By
increasing int rat horacic pressure, venous ret urn is compromised. As a result ,
cerebral venous blood volume increases and causes elevat ion in I CP.
Alt ernat ively, in a case w here increased PEEP reduces cardiac out put , t he mean
art erial pressure (MAP) may go dow n and necessarily low er t he cerebral
perf usion pressure (CPP) and if t he injured brain maint ains aut oregulat ion, a
decreased CPP w ill result in art eriodilat at ion, w hich increases t ot al cerebral
blood volume (CBV). Wit h increased CBV comes increased I CP.
I t is import ant t o not e t hat increased PEEP does not alw ays t ranslat e int o
increased I CP. Body posit ion can int erf ere w it h t he t ransmission of cent ral
venous pressure. For example, in pat ient s sit t ing upright (at an angle of close t o
90° degrees), t he jugular vein w ill collapse in syst ole and f ail t o t ransmit
increased int rat horacic pressure. Alt ernat ively, f or cases of brain injuries in
w hich aut oregulat ion is lost , decreased CPP may not alt er I CP; how ever,
decreased CPP may lead t o ot her problems, such as ischemia. Some evidence
has show n t hat in lungs w it h normal compliance, increased PEEP t ransmit s t o
increased I CP, but in lungs w it h low compliance, it does not .
Making recommendat ions about PEEP in brain injury pat ient s is diff icult . No
specif ic level of PEEP has yet been ident if ied as ideal f or brain injury pat ient s.
Addit ionally, dat a f rom t he Adult Respirat ory Dist ress Syndrome Net w ork
(ARDSNet ) t rial and ot her animal and human st udies have demonst rat ed t hat
some level of PEEP carries overall benef it f or a w ide variet y of crit ically ill
pat ient s (w it h respect t o t he development of ARDS). We can saf ely say t hat
clinicians must be aw are t hat PEEP can increase I CP. Thus, in pat ient s w it h
know n int racranial hypert ension or even pat ient s w it h any brain injury (w ho are
at risk f or low int racranial compliance), care must be t aken t o monit or pat ient s
closely. I CP monit oring and f requent neurologic examinat ions w ill represent
ext remely import ant element s of t he care of t hese pat ient s. I f pract it ioners
ident if y a correlat ion bet w een PEEP and I CP elevat ions, t hey should exert t heir
judgment and adjust t he f ormer t o ameliorat e t he lat t er w it hout compromising
syst emic oxygenat ion.
Suggested Readings
Burchiel KJ, St eege TD, Wyler AR. I nt racranial pressure changes in brain-
injured pat ient s requiring posit ive end-expirat ory pressure vent ilat ion.
Neurosurgery. 1981; 8: 443–439.
Caricat o A, Cont i G , Della Cort e F, et al. Eff ect s of PEEP on t he int racranial
syst em of pat ient s w it h head injury and subarachnoid hemorrhage: t he role of
respirat ory syst em compliance. J Trauma. 2005; 58: 571–576.
Cooper KR, Bosw ell PA, Choi SC. Saf e use of PEEP in pat ient s w it h severe
head injury. J Neurosurg. 1993; 63: 552–555.
McG uire G , Crossley D, Richards J, et al. Eff ect s of varying levels of posit ive
end-expirat ory pressure on int racranial pressure and cerebral perf usion
pressure. Crit Care Med. 1997; 25: 1059–1062.
Shapiro HM, Marshall LF. I nt racranial pressure responses t o PEEP in head-

injured pat ient s. J Trauma. 1978; 18: 254–256.
> Table of C ontents > Neur o > 196 - O btain a C om puted Tom ogr aphy S c an of the Head Im m ediately
after any C r aniotom y or Intr ac r anial P r oc edur e if P atient' s Neur ologic E xam ination is D iffer ent fr om
P r eoper ative As s es s m ents
196
Obtain a Computed Tomography Scan of the
Head Immediately after any Craniotomy or
Intracranial Procedure if Patient's Neurologic
Examination is Different from Preoperative
Assessments
Jose I. Suarez MD
Management of t he crit ically ill pat ient af t er a craniot omy poses special
challenges f or t he int ensivist . Not only do pract it ioners have t o manage t he
syst emic alt erat ions, but t hey must also underst and how such alt erat ions int eract
and aff ect brain f unct ions. For inst ance, pat ient s w ho develop int racerebral
hemorrhage (I CH) af t er a craniot omy are more likely t o experience hypert ensive
episodes bot h int raoperat ively and in t he f irst f ew hours post operat ively.
Pat ient s w ho undergo craniot omy f or subdural hemat oma (SDH) evacuat ion are
more likely t o experience recurrent SDH if t hey w ere coagulopat hic bef ore t he
surgical procedure w as perf ormed. Also t he t iming of neurological det eriorat ion
may vary depending on w hen t hey experience t he int racranial complicat ion. For
inst ance, post operat ive I CH w ill most ly present w it hin 6 hours af t er surgery.
Neurologic Examination
Whenever possible, pat ient s undergoing craniot omy should have a baseline
neurologic examinat ion prior t o surgery. Such evaluat ion should include
assessment of t he level of consciousness; cranial nerves; mot or st rengt h;
ref lexes; coordinat ion (i. e. , f inger-t o-nose and heel-t o-shin maneuvers); sensory
percept ion (light t ouch and pain and t emperat ure); and meningeal signs.
I mmediat ely af t er craniot omy t he pat ient 's neurological st at us should be
reassessed and compared w it h t he baseline. The t w o major limit at ions t o t he
immediat e post operat ive neurologic examinat ion are sedat ion and mechanical
vent ilat ion. Perf orming t he G lasgow Coma Scale (G CS) bef ore and af t er
craniot omy is a very reliable w ay of assessing level of consciousness. As
discussed in Chapt er 192, G CS has been validat ed w idely across medical and
surgical specialt ies and can be perf ormed in int ubat ed pat ient s using a predict ive
verbal score f rom t he eye and t he mot or scores. I t is of t he ut most import ance
t o act upon observed neurologic changes. The airw ay should be assessed and
blood pressure evaluat ed. Pat ient s w it h
G CS ≤8 should be int ubat ed (if not done already). Ext reme blood pressure levels
(hypo- or hypert ension) should be correct ed. Seizures should be t reat ed if
present . O nce t he init ial st abilizat ion of t he pat ient has been achieved, a head
comput ed t omography (CT) scan should be t he next st ep.
Head CT and Neurosurgical Evaluation

Many of t he immediat e post craniot omy complicat ions are due t o cerebral
ischemia, I CH, seizures, and cerebral edema w it h or w it hout elevat ed
int racranial pressure (I CP). Theref ore, head CT should be t he f irst imaging st udy
since it w ill ident if y most of t hese complicat ions. I f t he head CT scan reveals no
abnormalit ies, t hen an elect roen-cephalogram (EEG ) should be perf ormed. EEG
w ill be helpf ul t o exclude nonconvulsive seizures t hat w ould be undet ect able
ot herw ise. Neurosurgical evaluat ion should be sought immediat ely af t er
neurologic det eriorat ion is det ect ed (and bef ore t he head CT is obt ained). This
approach w ill guarant ee t hat pat ient s w ill be t aken t o t he operat ing room
prompt ly f or repeat craniot omy if t here is a neurosurgical t arget .
Suggested Readings
Basali A, Mascha EJ, Kalf as I , et al. Relat ion bet w een perioperat ive
hypert ension and int racranial hemorrhage af t er craniot omy. Anest hesiology.
2000; 93: 48–54.
Higuchi Y, I acono RP. Surgical complicat ions in pat ient s w it h Parkinson's

disease af t er post erovent ral pallidot omy. Neurosurgery. 2003; 52: 558–571.
Konig SA, Schick U, Dohnert J, et al. Coagulopat hy and out come in pat ient s
w it h chronic subdural hemat oma. Act a Neurol Scand. 2003; 107: 110–116.
Rodrigue T, Selman WR. Post operat ive management in t he neurosciences

crit ical care unit . I n Suarez JI , ed. Crit ical Care Neurology and Neurosurgery.
Tot ow a, NJ: Humana Press; 2004: 433–448.
Taylor WAS, Thomas NWM, Wellings JA, et al. Timing of post operat ive
int racranial hemat oma development and implicat ions f or t he best use of
neurosurgical int ensive care. J Neurosurg. 1995; 82: 48–50.
> Table of C ontents > Neur o > 197 - C ons ider Las t- D itc h Maneuver s to Low er Intr ac r anial Hyper tens ion
in Im pending Her niation
197
Consider Last-Ditch Maneuvers to Lower
Intracranial Hypertension in Impending
Herniation
Eliahu S. Feen MD
Jose I. Suarez MD
What to Do
Elevat ed int racranial pressure (I CP), or int racranial hypert ension, is a common
occurrence in crit ically ill neurologic pat ient s and is associat ed w it h poor
out come. While much disagreement remains about t he opt imal t reat ment of
elevat ed I CP, t here is recognit ion of t he most commonly employed t echniques f or
low ering crit ically elevat ed I CP; hypervent ilat ion f ollow ed by osmot herapy
(administ rat ion of an osmot ic diuret ic, most commonly mannit ol). Recent
research support s t he use of hypert onic saline (e. g. , 3% saline solut ion) as an
alt ernat ive or adjunct ive agent t o mannit ol. These medical measures w ill
hopef ully t emporize t he elevat ed I CP w hile invest igat ions as t o t he part icular
cause ensue. O f course, depending upon t he cause, surgery may be necessary
t o def init ively t reat t he elevat ed I CP.
What to Do Next
I f t hese medical measures are not successf ul and a def init ive surgical procedure
is not an opt ion, t hen f or many pract it ioners t he next st ep is usually general
anest hesia. For many years barbit urat e-induced coma w as t he accept ed
approach. Current ly, propof ol is recommended because of it s short half -lif e,
rapid w ashout t ime, and abilit y t o allow f or more rapid assessment of neurologic
f unct ion. Keeping pat ient s under generalized anest hesia f or several days
reduces cerebral met abolism and consequent ly reduces cerebral blood f low. I t is
t hrough t his mechanism t hat generalized anest hesia is t hought t o reduce t he I CP.
O ne import ant caveat t o general anest hesia is t he development of clinically
signif icant hypot ension. The lat t er can occur in up t o 25% of pat ient s t reat ed
w it h barbit urat es and is cert ainly seen w it h propof ol. Hypot ension is associat ed
w it h poor out come in pat ient s w it h neurologic injury.
Should met abolic coma not w ork, t w o more drast ic measures have been
advocat ed—hemicraniect omy and decompressive laparot omy. A subst ant ial
amount of evidence over t he past 30 years has lent support t o t he use of
decompressive hemicraniect omy f or t he t reat ment of int ract ably elevat ed I CP. I n
t he case of unilat eral cerebral edema (as det ermined by head comput ed
t omography scanning),
unilat eral decompressive hemicraniect omy can be perf ormed. I n t he case of

bilat eral cerebral edema, bilat eral hemicraniect omy is of t en necessary
(f ront ot emporal-pariet al). An alt ernat ive surgical approach f or t he t reat ment of
diff use bilat eral edema is bif ront al decompressive craniect omy. Case series and
case report s document success w it h perf ormance of early craniect omy (w it hin 24
t o 48 hours) as w ell as lat e craniect omy (af t er 24 t o 48 hours). Some pat ient s
may have had unilat eral mass lesions t hat w ere resect ed surgically and t hen
developed localized react ive edema, w hich became ref ract ory t o st andard
t reat ment s f or int racranial hypert ension. I n some of t hese cases (w here
ipsilat eral surgical resect ion had already been perf ormed), success has been
report ed w it h a decompressive hemicraniect omy. Dat a f rom randomized,
cont rolled t rials are lacking. As a result , pat ient -select ion prot ocols do not exist
and t he ident if icat ion of risk f act ors predict ing poor out come remains uncert ain.
Nevert heless, t he approach has become w idespread, especially in cert ain
specif ic circumst ances, such as malignant cerebral edema due t o middle
cerebral art ery occlusion. O f int erest , t he management of t he bone f lap varies
bet w een inst it ut ions. Some surgeons place t he bone f lap in t he pat ient 's
abdomen f or several w eeks prior t o reat t achment , w hereas ot hers place t he
bone f lap in (ext ernal) st erile environment s at subzero t emperat ures f or several
w eeks t o mont hs.
From t he t rauma lit erat ure, it has emerged t hat decompression of t he abdomen
result s in low ering of t he I CP. Not uncommonly, t rauma pat ient s f requent ly have
bot h abdominal injuries and head injuries (and as a result have I CP monit oring).
I f such a pat ient developed t he abdominal compart ment syndrome, w it h elevat ion
of int ra-abdominal pressure (I AP), and t hen t he pat ient w as t aken t o surgery f or
abdominal decompression, it w as coincident ally not ed in some pat ient s t hat t he
I CP w as reduced af t erw ard. From such cases t rauma surgeons developed an
approach t o perf orm abdominal decompression as a met hod of low ering
ref ract ory int racranial hypert ension, even in pat ient s w it h normal int ra-abdominal
pressures. At t his point only case series and animal dat a exist . I n many of t he
pat ient s f rom t he case series, t he I CP cont inued t o remain elevat ed despit e
craniect omy. Consequent ly, abdominal decompression t ruly remains a last -dit ch
at t empt t o low er I CP—but one t hat has been show n t o produce result s in t he
most ref ract ory of cases.
Suggested Readings
Albanese J, Leone M, Alliez J-R, et al. Decompressive craniect omy f or severe
t raumat ic brain injury: evaluat ion of t he eff ect s at one year. Crit Care Med.
2003; 31: 2535–2538.
G uerra WKW, G aab MR, Diet z H, et al. Surgical decompression f or t raumat ic

brain sw elling: indicat ions and result s. J Neurosurg. 1999; 90: 187–196.
Joseph DAK, Dut t on RP, Aarabi B, et al. Decompressive laparot omy t o t reat
int ract able int racranial hypert ension af t er t raumat ic brain injury. J Trauma.
2004; 57: 687–695.
Migliet t a MA, Salzano LJ, Chiu WC, et al. Decompressive laparot omy: a novel
approach in t he management of severe int racranial hypert ension. J Trauma.
2003; 55: 551–555.
Polin RS, Shaff rey ME, Bogaev CA, et al. Decompressive bif ront al
craniect omy in t he t reat ment of severe ref ract ory post t raumat ic cerebral
edema. Neurosurgery. 1997; 41: 84–94.
Robert s I . Barbit urat es f or acut e t raumat ic brain injury. Cochrane Dat abase
Syst Rev. 2000; CD000033.
Saggi BH, Bloomf ield G L, Sugerman HJ, et al. Treat ment of int racranial
hypert ension using nonsurgical abdominal decompression. J Trauma.
1999; 46: 646–651.
> Table of C ontents > Neur o > 198 - R em em ber that P atients Under going B ar bitur ate C om a m us t have
Adequate E lec tr ophys iologic al Monitor ing
198
Remember that Patients Undergoing Barbiturate
Coma must have Adequate Electrophysiological
Monitoring
Jose I. Suarez MD
High-dose barbit urat es have been used f or more t han 60 years in crit ically ill
neurologic pat ient s. The most common applicat ion has been t o low er elevat ed
int racranial pressure (I CP) in pat ient s w it h severe t raumat ic brain injury (TBI ).
O t her uses include ref ract ory st at us epilept icus, cerebral vasospasm af t er
subarachnoid hemorrhage (SAH), comat ose survivors of cardiac arrest , and
pat ient s w it h cerebral edema and elevat ed I CP f rom any et iology ot her t han TBI .
The most commonly used barbit urat e is pent obarbit al.
High-dose barbit urat es have various eff ect s on t he brain t hat may explain t heir
apparent ly neuroprot ect ive propert ies. I t has been suggest ed t hat t he most
import ant propert y is t hat of coupling cerebral blood f low t o regional cerebral
met abolic demands. I t is post ulat ed t hat low ering cerebral met abolic
requirement s w ould low er cerebral blood volume, w hich in t urn w ill result in low er
I CP. O t her mechanisms may include inhibit ion of f ree-radical-mediat ed lipid
peroxidat ion and ameliorat ion of glut amat e release. High-dose barbit urat es also
suppress bot h clinical and elect rical seizure act ivit y.
Dosage, M onitoring, and Adverse Effects

A number of t herapeut ic regimens using pent obarbit al have been used. O ne
common st rat egy uses an init ial loading dose of 10 mg/ kg over 30 t o 40 minut es
f ollow ed by a maint enance drip of 0. 5 t o 2. 0 mg/ kg/ h. The most reliable
monit oring t echnique is cont inuous elect roencephalogram (EEG ) monit oring.
Pat ient s w ho undergo pent obarbit al coma have no usef ul clinical examinat ion;
t heref ore, EEG is most helpf ul at t it rat ing t he t herapy. Usually, t he inf usion rat e
of pent obarbit al is t it rat ed t o a burst suppression pat t ern on t he EEG record.
There is no agreement regarding t he dept h of t he EEG suppression t hat must be
accomplished. How ever, many expert s recommend an EEG burst suppression
pat t ern of 5 t o 10 seconds. There is also no general agreement on t he durat ion
of pent obarbit al coma. Cert ainly
t he durat ion w ill depend on t he condit ion being t reat ed. For inst ance, pat ient s
w it h ref ract ory I CP elevat ions w ill have t o be t reat ed as long as t he abnormal
int racranial compliance is present , w hich may be f or several days. O n t he ot her
hand, many pat ient s w it h st at us epilept icus may need bet w een 24 and 96 hours
of t reat ment t o cont rol seizures. How ever, some pat ient s w it h st at us epilept icus
may also require days or w eeks of t reat ment .
Pent obarbit al coma is associat ed w it h a host of serious adverse eff ect s. The
most common side eff ect s include respirat ory depression and hypot ension. The
lat t er is an import ant issue since t he benef icial eff ect s of t he t herapy may be
negat ed by t he hypot ensive eff ect . I n f act , many t reat ed pat ient s w ill need ext ra
cardiovascular support including vasopressors and pulmonary art ery cat het er
insert ion t o monit or cardiac f unct ion and volume st at us. O t her adverse eff ect s of
pent obarbit al coma include ileus, decreased clearance of bronchial secret ion
leading t o pneumonia, and liver dysf unct ion. All of t hese f act ors and t he f act t hat
t here is no good evidence f rom randomized, cont rolled clinical t rials t hat
pent obarbit al coma improves out come in pat ient s w it h various neurological
pat hologies have led many pract it ioners t o consider t his t herapy as t he last
resort w hen ot her t reat ment s have been exhaust ed. As such, pent obarbit al coma
should be used only in pat ient s w it h elevat ed I CP ref ract ory t o medical and
surgical t herapy and st at us epilept icus ref ract ory t o ot her t reat ment s including
propof ol and midazolam drips.
Suggested Readings
Brain Resuscit at ion Clinical Trial I St udy G roup. Randomized clinical st udy of
t hiopent al loading in comat ose survivors of cardiac arrest . N Engl J Med.
1986; 314: 397–403.
Eisenberg HM, Frankow ski RF, Cont ant CF, et al. High-dose barbit urat e
cont rol of elevat ed int racranial pressure in pat ient s w it h severe head injury. J
Neurosurg. 1988; 69: 15–23.
Finf er SR, Ferch R, Morgan MK. Barbit urat e coma f or severe, ref ract ory
vasospasm f ollow ing subarachnoid hemorrhage. I nt ensive Care Med.
1999; 25: 406–409.
G oodman JC, Valadka AB, G opinat h SP, et al. Lact at e and excit at ory amino
acids measured by microdialysis are decreased by pent obarbit al coma in
head-injured pat ient s. J Neurot rauma. 1996; 13: 549–556.
Robert s I . Barbit urat es f or acut e t raumat ic brain injury. Cochrane Dat abase
Syst Rev. 2000; (2): CD000033.
Surgeons. The Joint Sect ion on Neurot rauma and Crit ical Care. Use of
barbit urat es in t he cont rol of int racranial hypert ension. J Neurot rauma.
2000; 17: 527–530.
Ward JD, Becker DP, Miller JD, et al. Failure of prophylact ic barbit urat e coma
in t he t reat ment of severe head injury. J Neurosurg. 1985; 62: 383–388.
> Table of C ontents > Neur o > 199 - B e Aler t for C onver s ion of Nonhem or r hagic ( Is c hem ic ) S tr oke to
Hem or r hagic S tr oke
199
Be Alert for Conversion of Nonhemorrhagic
(Ischemic) Stroke to Hemorrhagic Stroke
Jose I. Suarez MD
I schemic st roke account s f or about 80% of all st rokes. Many pat ient s w it h an
init ial ischemic st roke w ill experience a hemorrhagic t ransf ormat ion. This may
result in f urt her neurological det eriorat ion. As observed on aut opsy st udies,
many pat ient s w ill have some degree of hemorrhagic t ransf ormat ion, w it h 51% t o
71% of cardioembolic and 2% t o 21% of noncardioembolic ischemic st rokes
having hemorrhagic t ransf ormat ion. The incidence of hemorrhagic t ransf ormat ion
seen on comput ed t omography (CT) scan ranges f rom a f ew percent t o 40% of
pat ient s, depending on t he crit eria used.
Watch Out For

Alt hough hemorrhagic t ransf ormat ion occurs f requent ly af t er ischemic st roke, it
is of t en asympt omat ic, part of t he nat ural course of ischemic inf arct ion, and may
not alt er t he clinical out come beyond w hat is t o be expect ed based on t he
int rinsic charact erist ics of t he ischemic inf arct . O nly sympt omat ic hemorrhagic
t ransf ormat ion has clinical relevance. Diff erent st roke t rials have varying
def init ions of w hat const it ut es sympt omat ic hemorrhagic t ransf ormat ion. I n order
t o clarif y t his conf usion, Berger et al. analyzed dat a f rom pat ient s previously
enrolled in t he European Cooperat ive Acut e St roke St udy (ECASS I I ) t rial and
f ound t hat clinically signif icant hemorrhagic t ransf ormat ion occurred mainly if
great er t han 30% of t he inf arct ed area had dense homogenous-appearing
hemorrhage w it h mass eff ect . I n t his case, pat ient s had signif icant ly increased
risk of early neurologic det eriorat ion and w orse long-t erm out come. I n pat ient s
w it h smaller (<30%) homogenous hemorrhagic t ransf ormat ion w it h minimal mass
eff ect , a w orse short -t erm out come w as observed, but no eff ect on long-t erm
clinical out come w as seen. I n cont rast , pat ient s w it h small het erogeneous
pet echial hemorrhages did not have w orse early or long-t erm out come.
The mechanisms f or hemorrhagic t ransf ormat ion vary depending on t he clinical
scenario; how ever, one of t he essent ial requirement s is breakdow n of t he blood–
brain barrier (BBB). Cert ain f act ors at baseline may predict a higher rat e of
hemorrhagic t ransf ormat ion. These f act ors include hyperglycemia, increased
pat ient age, embolic st roke
mechanism, increased inf arct size, edema f ormat ion, excessive hypert ension,
increased severit y of st roke sympt oms, and receiving ant icoagulant or t issue-
t ype plasminogen act ivat or (t PA). Pat ient s receiving t PA had about a 6. 4% rat e
of sympt omat ic hemorrhagic t ransf ormat ion, as seen in t he Nat ional I nst it ut e of
Neurological Disorders and St roke (NI NDS) t rial. O ne group of invest igat ors
f ound t hat t PA use f acilit at ed dysregulat ion of ext racellular prot eolysis, result ing
in BBB breakdow n.
What to Do
Management of hemorrhagic t ransf ormat ion requires a mult if acet ed approach.
First , t he precipit at ing cause needs t o be quickly ident if ied and t reat ed. The
pat ient should have normoglycemia maint ained, excessively elevat ed blood
pressure should be gent ly low ered, and coagulopat hy should be correct ed
prompt ly w it h administ rat ion of f resh f rozen plasma. G enerally, ant iplat elet
agent s and ant icoagulant s should be held f or at least a w eek depending on t he
size of t he hemorrhagic t ransf ormat ion; how ever, solid dat a guiding t his decision
are lacking. Management of signif icant mass eff ect and edema associat ed w it h
sympt omat ic hemorrhagic t ransf ormat ion should be reduced by inst it ut ing
neuroprot ect ive, pharmacologic, and surgical measures. Neuroprot ect ive
measures include maint aining normoglycemia and normot hermia, as
hyperglycemia and hypert hermia may w orsen cerebral edema. I f t he pat ient
becomes obt unded, a noncont rast ed head CT should be perf ormed immediat ely
t o exclude hydrocephalus or brain herniat ion.
I f sympt omat ic hemorrhagic t ransf ormat ion occurs w it hin t he cerebellum,
compression of t he f ourt h vent ricle can rapidly lead t o obst ruct ive hydrocephalus
and deat h. Direct brain-st em compression w it h subsequent brain deat h may also
occur. Thus, immediat e neurosurgical int ervent ion w it h a vent riculost omy and
plan f or possible post erior f ossa decompression or cerebellect omy must occur.
I n addit ion, in pat ient s w it h neurological det eriorat ion and signs of impending
herniat ion, immediat e int ubat ion should t ake place t o prot ect t he airw ay, and
alt hough very cont roversial, mild hypervent ilat ion t o keep t he PCO2 bet w een
about 30 and 34 mm Hg can be considered (t his is a t emporizing measure only).
Addit ionally, 1 g/ kg of mannit ol, an osmot ic agent , must be given immediat ely t o
help low er int racranial pressure. The serum osmolalit y should be maint ained at a
range of 310 t o 320 mO sm/ L. A hemicraniect omy should be considered in
younger pat ient s (<50 years) w it h large hemispheric st roke if conservat ive
measures f ail.
Anot her point t hat is very import ant is t he correct ion of exist ing coagulopat hy.
Acut e ischemic st roke pat ient s are of t en t reat ed w it h ant it hrombot ic and
ant iplat elet agent s. Thus, pract it ioners need t o consider blood product
t ransf usion (i. e. , plat elet s, f resh f rozen plasma, and cryoprecipit at e) t o correct
pot ent ial coagulat ion abnormalit ies. I n pat ient s receiving int ravenous
unf ract ionat ed heparin, prot amine sulf at e (1mg per 100 unit s of heparin given
over t he previous 4 hours) should be given, af t er considerat ion of prot amine's
risks. I n addit ion, some early dat a on f act or VI I a f or t reat ment of coagulopat hy
has show n some benef it .
Suggested Readings
Berger C, Fiorelli M, St einer T, et al. Hemorrhagic t ransf ormat ion of ischemic
brain t issue: asympt omat ic or sympt omat ic? St roke. 2001; 32: 1330–1335.
Feen ES, Suarez JI . Raised int racranial pressure. Curr Treat O pt ions Neurol.
2005; 7(2): 109–117.
O kada Y, Yamaguchi T, Minemat su K, et al. Hemorrhagic t ransf ormat ion in

cerebral embolism. St roke. 1989; 20(5): 598–603.
Wang X, Tsuji K, Lee SR, et al. Mechanisms of hemorrhagic t ransf ormat ion
af t er t issue plasminogen act ivat or reperf usion t herapy f or ischemic st roke.
St roke. 2004; 35: 2726.
> Table of C ontents > Neur o > 200 - O btain a Head C om puted Tom ogr aphy S c an and Lum bar P unc tur e
in P atients w ith Hum an Im m unodefic ienc y Vir us or Ac quir ed Im m unodefic ienc y S yndr om e and New -
O ns et Mental S tatus C hanges
200
Obtain a Head Computed Tomography Scan and
Lumbar Puncture in Patients with Human
Immunodeficiency Virus or Acquired
Immunodeficiency Syndrome and New-Onset
Mental Status Changes
Eric M. Bershad MD
Jose I. Suarez MD
Human immunodef iciency virus (HI V) and acquired immunodef iciency syndrome
(AI DS) pat ient s may host a myriad of inf ect ious, inf lammat ory, t oxic, and
neoplast ic diseases t hat are not seen commonly in t he immune-compet ent
pat ient . Thus, it is import ant f or t he clinician t o keep in mind t he expanded
diff erent ial diagnosis of ment al st at us changes in t he HI V pat ient . The et iology
f or ment al st at us changes in general includes a vast diff erent ial of diseases in
t he cat egories of vascular, inf ect ious, neoplast ic, drugs, inf lammat ory,
aut oimmune, t rauma, and endocrine or met abolic causes. I n HI V/ AI DS pat ient s
t he diff erent ial must be expanded t o include diseases and condit ions relat ed t o
immune dysf unct ion. The clinician should have a much low er t hreshold f or
suspect ing a neurologic cause f or ment al st at us changes in t he HI V/ AI DS pat ient
as compared w it h t he immune-compet ent pat ient .
The init ial w orkup f or HI V/ AI DS pat ient s w it h new -onset ment al st at us changes
must include a cont rast ed head comput ed t omography (CT) scan t o rule out
int racranial mass lesions. A magnet ic resonance imaging (MRI ) w it h gadolinium is
more sensit ive f or small int racranial lesions; how ever, t he limit ed availabilit y in
some cent ers and t he increased t ime t o obt ain images make CT scan a more
pract ical init ial imaging modalit y (Tabl e 200. 1).
In the HIV/ AIDS pati ent i t i s i mportant to obtai n the head CT bef ore perf ormi ng
a l umbar puncture ( Fi g. 200. 1). A f ocal mass lesion may produce locally
elevat ed int racranial pressure. Alumbar punct ure done in t his set t ing can result in
a high pressure gradient bet w een t he lesion and lumbar cist ern t hat f orces t he
brain dow nw ard, result ing in f at al herniat ion. The most common mass lesions t o
consider are t oxoplasmosis and primary cent ral nervous syst em (CNS)
lymphoma. The incidence of cerebral t oxoplasmosis in HI V/ AI DS pat ient s ranges
f rom 3% t o 10% in t he Unit ed St at es, w hile primary CNS lymphoma occurs in
about 2% of pat ient s.
TABLE 200-1 ETIOLOGIES OF M ASS LESIONS IN

HIV/AIDS PATIENTS
Parasites: Toxoplasm osis,

cysticercosis
Fungi: Cryptococcus, Candida,
aspergillosis, mucormycosis,
Infectious
coccidioidomycosis
Bacteria: Mycobacterium tuberculosis,
Mycobacterium avium, Nocardia,
Listeria, Treponema pallidum
Prim ary CNS lym phom a, glioma,

Neoplastic
metastatic neoplasm
Progressive multifocal
Inflammatory
leukoencephalopathy (PML)
Ischemic stroke, intracerebral

Vascular
hemorrhage
FIG URE 200. 1. Head CT w it h cont rast show ing a solit ary enhancing mass
lesion (arrow ) in t he right basal ganglia lat er ident if ied as a t oxoplasmosis
abscess. Primary CNS lymphoma may have a similar appearance.
Reproduced w it h permission f rom a neuroradiology t eaching f ile at Universit y
Hospit als of Cleveland. Court esy of David Prest on.
What to Do
O nce an int racranial mass lesion has been f ound on imaging, it is import ant t o
det ermine t he et iology. Based on t he American Academy of Neurology (AAN)
pract ice paramet er on evaluat ion and management of int racranial mass lesions in
AI DS, large lesions w it h mass eff ect t hreat ening impending herniat ion require
open biopsy w it h decompression. I n all ot her cases, empiric t herapy f or
t oxoplasmosis should be inst it ut ed, unless t here is only a single lesion seen and
also negat ive t oxoplasmosis serology. Pat ient s t reat ed f or t oxoplasmosis
inf ect ions presumpt ively need t o be caref ully monit ored clinically and
radiographically over t he f irst 10 t o 14 days. The median t ime t o response is 5
days. By day 14, 91% of pat ient s w it h t oxoplasmosis cent ral nervous syst em
inf ect ion exhibit a response. I f no response is observed af t er empiric
t oxoplasmosis t herapy, one should proceed w it h st ereot act ic brain biopsy.
I n t he absence of a brain mass lesion, a lumbar punct ure should be done t o
exclude meningit is and encephalit is. Cryptococcus is t he most common cause of
meningit is in t he HI V pat ient . HI V may direct ly result in asept ic meningit is. O t her
et iologies include Cyt omegalovirus (CMV), varicella-zost er virus (VZV), Epst ein-
Barr virus (EBV), herpes simplex virus (HSV), Li steri a monocytogenes,
coccidioidomycosis, hist oplasmosis, syphilis, t uberculosis, and lymphomat ous
meningit is.
When perf orming lumbar punct ure, cerebrospinal f luid (CSF) should be sent f or
cell count , cell diff erent ial, G ram st ain, prot ein, glucose, bact erial and f ungal
cult ures, I ndia ink st ain, crypt ococcus, and VDRL t est (f or syphilis). Polymerase
chain react ion (PCR) should be sent f or VZV, CMV, HSV, EBV, and t uberculosis.
Furt hermore, it is import ant t o send CSF f or cyt ology t o assess f or at ypical
lymphocyt es as seen in lymphomat ous meningit is.
Suggested Readings
Bradley W, ed. Neurology in Clinical Pract ice. New York: But t erw ort h-
Heinemann; 2004: 1581–1602.
Evaluat ion and management of int racranial mass lesions in AI DS. Report of
t he Q ualit y St andards Subcommit t ee of t he American Academy of Neurology.
Neurology. 1998; 50(1): 21–26.
> Table of C ontents > Neur o > 201 - Have a Low Thr es hold for O btaining an Initial and R epeat Head
C om puted Tom ogr aphy S c an after S ubar ac hnoid Hem or r hage
201
Have a Low Threshold for Obtaining an Initial
and Repeat Head Computed Tomography Scan
after Subarachnoid Hemorrhage
Eric M. Bershad MD
Jose I. Suarez MD
Subarachnoid hemorrhage (SAH) is a neurologic emergency w it h a high mort alit y
rat e (50%). Morbidit y result s f rom ext ravasat ion of blood int o t he subarachnoid
space, most commonly f rom a rupt ured aneurysm. The result ing blood can lead
t o vasospasm and cerebral inf arct ion, hydrocephalus, cerebral edema w it h brain
herniat ion, seizures, and ot her complicat ions. The t ypical present at ion of SAH is
a sudden-onset severe headache w it h nausea, vomit ing, neck pain, phot ophobia,
and brief loss of consciousness.
An immediat e noncont rast ed head comput ed t omography (CT) scan should be t he
f irst diagnost ic t est w hen suspect ing SAH. The sensit ivit y of head CT
approaches 100% w it hin t he f irst 12 hours and 93% w it hin 24 hours af t er onset
of sympt oms. The amount of subarachnoid hemorrhage seen on head CT helps
predict t he subsequent chance of vasospasm and also t he long-t erm out come of
t he pat ient . A head CT can also assess f or int ravent ricular hemorrhage, w hich
port ends a higher risk f or hydrocephalus and cerebral edema (Fi g. 201. 1).
What Not to Do
I n t he absence of t he classic signs and sympt oms, SAH may be misdiagnosed.
The f requency of misdiagnosis may be up t o 50% in pat ient s present ing f or t heir
f irst visit t o a physician. The main reasons f or misdiagnosis include f ailure t o
obt ain t he appropriat e imaging st udy (i. e. , head CT) in 73% of cases and f ailure
t o perf orm, or correct ly int erpret t he result s of , a lumbar punct ure in 23% of
cases. Pat ient s w ho are misdiagnosed are usually f ound t o be less ill and have
normal neurologic examinat ion. How ever, it is import ant t o realize t hat neurologic
complicat ions event ually occur in up t o 50% of pat ient s w it h delayed diagnosis
and are associat ed w it h an increased risk of deat h and disabilit y. Every pat ient
present ing w it h a severe and unusual headache should be evaluat ed f or SAH.
Headache may be t he only present ing complaint in up t o 40% of pat ient s and
may abat e complet ely
w it hin minut es or hours (so-called sent inel or t hunderclap headaches or “w arning

leaks”). Emergent evaluat ion of sent inel headaches is required since pat ient s
may experience a major SAH w it hin 3 w eeks.
Based on several f act ors, t he clinician can cat egorize t he severit y of SAH t o help
predict out come. The World Federat ion of Neurological Surgeons scale combines
t he score on t he G lasgow Coma Scale and mot or def icit s t o help prognost icat e
out come, w hile t he Head CT G rading Scale helps predict out come based on t he
amount of SAH and presence or absence of int ravent ricular hemorrhage.
FIG URE 201. 1. Head CT of subarachnoid bleed. This noncont rast ed head CT
show s a large amount of hyperint ense signal, indicat ing acut e subarachnoid
blood in t he basal cist erns. This head CT is a normal st udy f or comparison.
Reproduced w it h permission, court esy of David Prest on, neuroradiology
t eaching f ile at Universit y Hospit als of Cleveland.
What to Do
I n pat ient s w it h a clinical hist ory suspicious f or SAH and a negat ive head CT, a
lumbar punct ure should be done t o assess f or xant hochromia, a result of
breakdow n of red blood cells in t he cerebrospinal f luid. Tubes 1 and 4 should be
analyzed f or cell count . This enables diff erent iat ion bet w een an iat rogenic
t raumat ic lumbar punct ure, w hich produces elevat ed red blood cells in t he f irst
t ube but not t he last t ube, versus a sust ained red blood cell elevat ion, w hich is
seen in t rue SAH.
O nce t he diagnosis of SAH is made, immediat e st eps should be t aken t o ident if y
and secure rupt ured aneurysms. The best init ial diagnost ic t est is CT
angiography, w hich is less invasive t han st andard digit al subt ract ion angiography
t hat requires invasive cat het erizat ion.
O nce an aneurysm has been ident if ied, it should be secured by eit her
int ervent ional neuroradiologic coiling or surgical clipping based on it s int rinsic
charact erist ics. This w ill help prevent rebleeding of t he aneurysm, w hich carries
a high mort alit y rat e of 50%.
O nce t he aneurysm has been secured, t he next st ep in t he management of SAH
pat ient s is t o monit or and t reat t he pat ient s in t he int ensive care unit (I CU)
set t ing. Common complicat ions f ollow ing aneurysmal SAH include sympt omat ic
vasospasm (46%), hydrocephalus (20%), and rebleeding (7%). O t her
complicat ions include seizures, deep venous t hrombosis, pulmonary embolism,
hyponat remia, neurogenic pulmonary edema, and met abolic and cardiac
complicat ions.
An immediat e noncont rast ed head CT should be done in any pat ient w hodevelops
ment al st at us changes f ollow ing SAH. This st udy w ill prompt ly diagnose
rebleeding of t he aneurysm, hydrocephalus, int raparenchymal hemorrhage,
and/ or cerebral edema.
The quest ion may arise as t o w het her it is saf e t o use heparin or enoxaparin t o
prevent t hromboembolic complicat ions such as deep venous t hrombosis and
pulmonary embolism f ollow ing SAH. A recent randomized double-blind st udy w it h
120 pat ient s w it h aneurysmal SAH compared enoxaparin 20 mg subcut aneously
daily versus placebo f or 3 w eeks af t er securing t he aneurysm. The pat ient s in
t he enoxaparin group had signif icant ly low er incidence of cerebral vasospasm,
cerebral inf arct ion, and severe hydrocephalus. Furt hermore, at 1-year f ollow -up,
t he pat ient s w ho w ere t reat ed w it h enoxaparin had improved overall out come
compared w it h placebo-t reat ed pat ient s. How ever, t his conf lict s w it h a previous
randomized double-blind st udy of 170 pat ient s w it h SAH t hat w ere t reat ed w it h
eit her enoxaparin 40 mg daily or placebo. I n t his st udy, t here w as no signif icant
diff erence in long-t erm out come (3 mont hs) bet w een t he groups. Thus, it is st ill
unclear w het her or not ant icoagulant agent s should be used in SAH pat ient s in
t he acut e set t ing.
Suggested Readings
Siironen, J, Juvela S, Varis J, et al. No eff ect of enoxaparin on out come of
aneurysmal subarachnoid hemorrhage: a randomized, double-blind, placebo-
cont rolled clinical t rial. J Neurosurg. 2003; 99(6): 953–959.
Suarez JI , Tarr RW, Selman WR. Aneurysmal subarachnoid hemorrhage. N

Engl J Med. 2006; 354(4): 387–396.
Wurm G , Tomancok B, Nussbaumer K, et al. Reduct ion of ischemic sequelae

f ollow ing spont aneous subarachnoid hemorrhage: a double-blind, randomized
comparison of enoxaparin versus placebo. Clin Neurol Neurosurg.
2004; 106(2): 97–103.
> Table of C ontents > Neur o > 202 - B e Aler t for R ebleeding in P atients w ith S ubar ac hnoid Hem or r hage
202
Be Alert for Rebleeding in Patients with
Subarachnoid Hemorrhage
Nirav G . Shah MD
Subarachnoid hemorrhage (SAH) is a commonly misdiagnosed problem. Alt hough
SAH may produce minor sympt oms (see Tabl e 202. 1), it is of t en f at al, and early
surgical int ervent ion can improve out comes.
Watch Out For

The t ypical pat ient w ho develops a subarachnoid hemorrhage f requent ly has a
posit ive personal or f amily hist ory f or SAH, a hist ory of polycyst ic kidney
disease, hypert ension, alcohol abuse, or cigaret t e smoking. I n addit ion, t he
incidence is more common in pat ient s w it h herit able connect ive-t issue disease,
especially f ibromuscular dysplasia. A rupt ured saccular aneurysm is t he most
common et iology of subarachnoid hemorrhage, f ollow ed by t rauma, art eriovenous
malf ormat ions, illicit drug use (especially cocaine), and vasculit ides.
The most common present ing sympt om of subarachnoid hemorrhage is headache
of unusual severit y. Pat ient s t ypically describe t heir pain as “t he w orst headache
of my lif e. ” Somet imes, t he lif e-t hreat ening bleed may be preceded by a small
bleed and headache, w hich is called a sent inel headache. O t her sympt oms may
include nuchal rigidit y, diminished level of consciousness, aphasia, and bilat eral
w eakness of t he low er ext remit ies. Signs t hat are apparent on physical
examinat ion include papilledema, t hird or sixt h nerve palsy, nyst agmus, lef t -side
visual neglect , or ret inal hemorrhage.
The most import ant st udy t o obt ain w hen suspect ing a subarachnoid hemorrhage
is a noncont rast ed comput ed t omography (CT) scan of t he head. Pref erably, t he
st udy should be done w it h t hin cut s about 3 mm in t hickness in order t o ident if y
small collect ions of blood. The CT scan is most sensit ive if obt ained w it hin 24
hours of t he bleed. I n one st udy, t he percent age of posit ive scans w ent f rom
92% t o 50% f rom day one t o day seven. O f not e is t hat if t he underlying et iology
of t he subarachnoid bleed is a rupt ured aneurysm, t he locat ion of blood on t he
CT scan does not accurat ely predict t he locat ion of t he aneurysm. I f t he CT scan
is negat ive but t he clinical syndrome is st ill consist ent w it h SAH, a lumbar
punct ure should be perf ormed t o ident if y blood in t he cerebrospinal f luid.
TABLE 202-1 HUNT-HESS SCALE FOR GRADING

SUBARACHNOID HEM ORRHAGE
GRADE NEUROLOGICAL STAT US
Asymptomatic; or minimal headache and

1
slight nuchal rigidity
Moderate to severe headache; nuchal rigidity;

2 no neurologic deficit except cranial nerve
palsy
3 Drowsy; minimal neurologic deficit
Stuporous; moderate to severe hemiparesis;

4 possible early decerebrate rigidity and
vegetative disturbances
Deep coma; decerebrate rigidity; moribund

5
appearance
What to Do
The t reat ment of subarachnoid hemorrhage includes act ive management of
cerebral perf usion pressure, prophylaxis f or vasospasm, ant iconvulsive t herapy
t o prevent seizures, and def init ive t herapy via surgery or endovascular repair.
There is ongoing debat e w het her early surgery or endovascular repair port ends
a bet t er prognosis compared w it h w ait ing 10 t o 14 days f or t he improvement of
edema prior t o def init ive t herapy.
The t hree most common complicat ions of subarachnoid hemorrhage are
rebleeding, vasospasm, and hydrocephalus. Rebleeding is a signif icant risk in
pat ient s w it h an SAH and occurs w it h an incidence of approximat ely 3% t o 6%. I t
most f requent ly occurs in t he f irst 24 hours f ollow ing a bleed w it h t he highest
risk in t he f irst 6 hours. Rebleeding can present w it h alt ered ment al st at us. Any
changes in ment al st at us af t er SAH w arrant s an emergent repeat CT scan. The
f act ors t hat can most reliably predict t he possibilit y of rebleeding are size of
aneurysm and t he Hunt -Hess G rade of t he neurological st at us.
The second complicat ion, vasospasm, is t he leading cause of mort alit y in
pat ient s w it h rupt ured aneurysms. The t ime of occurrence is usually bet w een
days t hree and seven but can be seen as f ar out as 14 days f rom subarachnoid
hemorrhage. This should be suspect ed w it h any change in neurologic st at us or
w it h any f ocal def icit s. O ne met hod used t o monit or f or vasospasm is
t ranscranial Doppler, w hich looks f or changes in t he velocit y of blood f low as a
precursor t o act ual sympt omat ic vasospasm. The risk of vasospasm is based on
size of bleed, locat ion, age, and neurologic st at us. Nimodipine, a calcium
ant agonist , has been used t o signif icant ly reduce t he occurrence of vasospasm.
O nce vasospasm occurs af t er aneurysm clipping, t reat ment is opt imized by use
of t he t riple-H t herapy, w hich includes hemodilut ion, induced hypert ension, and
hypervolemia.
Acut e or chronic hydrocephalus is also a possible complicat ion associat ed w it h

subarachnoid hemorrhage. The et iology is t hought t o be secondary t o occlusion
of cerebrospinal f luid f low f rom debris or decreased cerebrospinal f luid
product ion. Risk f act ors f or t he development of hydrocephalus include age at
present at ion, t he locat ion of t he bleed, t he size of t he bleed, and concomit ant
int ravent ricular hemorrhage. The clinical present at ion t ypically includes alt ered
ment al st at us w it h radiologic conf irmat ion by CT scan. The t reat ment includes
drainage of cerebrospinal f luid via vent riculost omy. Typically, about 50% of t he
pat ient s w it h acut e hydrocephalus have resolut ion of sympt oms w hile t he
remaining pat ient s have poorer out comes w it h a high rat e of mort alit y.
Suggested Readings
Edlow JA, Caplan LR. Avoiding pit f alls in t he diagnosis of subarachnoid
hemorrhage. N Engl J Med. 2000; 342: 29–36.
van G ijn J. Subarachnoid hemorrhage. Lancet . 1992; 339: 653.

> Table of C ontents > Neur o > 203 - R em em ber that New E lec tr oc ar diogr am C hanges in a P atient w ith
a S ubar ac hnoid B leed m ay be a S ign of P r ogr es s ion of the B leed
203
Remember that New Electrocardiogram Changes
in a Patient with a Subarachnoid Bleed may be a
Sign of Progression of the Bleed
Amisha Barochia MD
Loss of consciousness is a common clinical occurrence in subarachnoid
hemorrhage. How ever, t he sudden deat h seen in t hese pat ient s is t hought t o be
t he result of cardiac rhyt hm dist urbances. Alt hough t he exact pat hophysiology is
unknow n, t here appears t o be an imbalance in t he aut onomic nervous syst em,
w hich leads t o a cat echolamine surge, changes in repolarizat ion and
depolarizat ion, and subsequent ect opic f oci. Hypokalemia secondary t o
increased circulat ing cat echolamines, vomit ing, and hypercort isolism may also
cont ribut e t o t he development of arrhyt hmias associat ed w it h a subarachnoid
bleed.
Watch Out For

The most f requent ly observed elect rocardiogram (ECG ) changes in pat ient s w it h
subarachnoid bleeds include T w ave inversions, ST-segment elevat ions or
depressions, Q w aves, U w aves ≥0. 1 mV, and Q T prolongat ion. ECG changes
secondary t o subarachnoid bleeds may even mimic t he changes seen in
acut emyocardial inf arct ion. How ever, even t hough t hese lat t er changes may be
associat ed w it h abnormal t hallium perf usion scans in up t o one-t hird of pat ient s,
t hey are not generally indicat ive of underlying myocardial ischemia. When
aut opsy specimens f rom pat ient s w it h t hese ECG changes are examined, t here is
no correlat ion w it h t he presence of coronary art ery disease and t he myocardium
is int act on microscopic examinat ion. ECG changes usually associat ed w it h a
poor prognosis include pat hologic Q w aves, raised ST segment s, P mit rale, Q T
prolongat ion ≥450 milliseconds, and t achycardia. How ever, one recent st udy did
not f ind a relat ionship bet w een t he t ype or locat ion of ECG change and mort alit y.
ECG changes are usually seen in t he f irst 72 hours af t er a subarachnoid
hemorrhage and usually resolve in a f ew days, but occasionally may persist f or 4
t o 8 w eeks. I f t he pat ient rebleeds, t he same ECG changes t hat occurred w it h
t he f irst episode are usually observed. Transient cardiac arrhyt hmias have been
not ed in ≥90% of pat ient s in t he acut e phase of subarachnoid hemorrhage on
Holt er monit oring. The arrhyt hmia t ypes include vent ricular t achycardia,
vent ricular f lut t er, and torsades de poi ntes. There is no associat ion bet w een t he
pat ient 's
clinical condit ion and t he occurrence of t hese arrhyt hmias, except t hat t hose
pat ient s w ho experienced arrhyt hmias t end t o be older and have low er pot assium
levels. The severit y, t ype, and lesion locat ion generally do not correlat e w it h t he
development of an arrhyt hmia.
I n addit ion t o ECG changes, echocardiographic changes have been not ed in
pat ient s w it h subarachnoid bleed t hat have no relat ion t o ECG changes in t he
same pat ient . G lobal or segment al w all mot ion abnormalit ies have been report ed
and are t hought t o be due t o st unned myocardium, since t he changes w ere f ound
t o be reversible. Unlike ECG abnormalit ies, echocardiogram f indings do
correspond t o t he degree of neurologic changes.
O ne f inal not e is t hat is advisable t o have pat ient s w it h subarachnoid
hemorrhage on cont inuous t elemet ry monit oring so t hat arrhyt hmias are det ect ed
and prompt ly t reat ed, w hich can be lif esaving f or t he pat ient .
Suggested Readings
De O liveira JJ, Silva SR. Signs of myocardial ischemia associat ed w it h
subarachnoid bleed. Arq Bras Cardio. 1996; 67(6): 403–406.
Manninen PH, Ayra B, G el B, et al. Associat ion bet w een elect rocardiographic
abnormalit ies and int racranial blood in pat ient s f ollow ing acut e subarachnoid
hemorrhage. J Neurosurg Anest hesiol. 1995; 7(1): 12–16.
> Table of C ontents > Neur o > 204 - C ons ider the Us e O f Thr om bolytic Agents for Tr eatm ent of Ac ute
Is c hem ic S tr oke
204
Consider the Use Of Thrombolytic Agents for
Treatment of Acute Ischemic Stroke
Nirav G . Shah MD
Acut e ischemic st roke is t he t hird leading cause of mort alit y in t he Unit ed St at es
and, w hen not f at al, can result in signif icant morbidit y. The most eff ect ive w ay t o
decrease morbidit y is t o rest ore t he blood f low t o t he ischemic area in a t imely
manner w it h t hrombolyt ic t herapy. Upon present at ion t o an emergency room,
st andard procedures such as est ablishing medical st abilit y, obt aining a t horough
hist ory, and perf orming a physical examinat ion and laborat ory t est ing should be
undert aken immediat ely. A quick diagnosis in t he set t ing of ischemic st roke is
associat ed w it h a reduct ion in morbidit y. Follow ing t hese measures, a comput ed
t omography (CT) scan must be perf ormed t o diff erent iat e bet w een subarachnoid
hemorrhage, int racerebral hemorrhage, and ischemic inf arct ion. O t her st udies
t hat may need t o be perf ormed based on t he hist ory of illness include lumbar
punct ure (encephalit is, subarachnoid hemorrhage), art erial blood gas
(hypoxemia), cervical spine x-rays (t rauma), magnet ic resonance imaging (bet t er
anat omic def init ion), and cerebral art eriography (if considering int ra-art erial
t hrombolysis).
What to Do
The indicat ion f or int ravenous t hrombolysis in t he set t ing of acut e ischemic
st roke is predicat ed on t he present at ion and diagnosis occurring w it hin 3 hours
of sympt om onset . The mort alit y benef it w as demonst rat ed by t he Nat ional
I nst it ut e of Neurological Disorders and St roke Recombinant Tissue Plasminogen
Act ivat or (t PA) St roke St udy. Tissue-t ype plasminogen act ivat or is
cont raindicat ed in many pat ient s including t hose w it h st roke or head t rauma in
t he previous 3 mont hs, a hist ory of myocardial inf arct ion in t he previous 3
mont hs, a hist ory of gast roint est inal bleeding in t he previous 3 w eeks, and a
major surgical procedure in t he previous 2 w eeks (alt hough surgery is not an
absolut e cont raindicat ion). I n addit ion, pat ient s w it h t hrombocyt openia or
coagulopat hy should be excluded f rom use of t hrombolyt ic agent s. Blood
pressure must also be assessed. Pat ient s w it h a blood pressure great er t han
185/ 110 mm Hg should not receive t PA. The pat ient can receive a dose of
int ravenous (I V) 20 mg of labet ot ol t imes 2 doses given 10 minut es apart . I f t he
blood pressure decreases below
185/ 110 mm Hg and remains, t he pat ient can be considered f or t PA. The overall
benef it demonst rat ed by use of t PA w as signif icant w it h close t o 50% of pat ient s
having complet e or near-complet e recovery at 3 mont hs and 1 year as compared
w it h less t han 30% in t he group t reat ed w it h placebo.
Watch Out For

While t he use of t PA demonst rat ed a mort alit y benef it , t here is also signif icant
risk associat ed w it h it . The most signif icant risk is t hat of int racerebral
hemorrhage, w hich occurred in approximat ely 6% of pat ient s enrolled in t he t PA
t rials. The risk of int racerebral hemorrhage is increased w it h CT-scan
demonst rat ion of mass eff ect prior t o t he use of t hrombolyt ics, severit y of
neurologic st at us, age, and comorbidit ies including diabet es, t hrombocyt openia,
and cardiac disease. O t her complicat ions of t PA include syst emic bleeding and
angioedema (usually mild).
Thrombolyt ic t herapy can also be provided t hrough an int raart erial met hod,
almost alw ays in conjunct ion w it h experienced neuroint ervent ional radiology
physicians. The agent s t hat have been invest igat ed are urokinase, t PA, and
prourokinase. These agent s demonst rat ed benef it w it h 75% of pat ient s achieving
t ot al or part ial recanalizat ion of t he occluded vessel. I nt ra-art erial t hrombolyt ic
t herapy is administ ered on an increment al basis unt il t he t hrombus has resolved.
The rat e of complicat ion of int racerebral hemorrhage is low er f or int ra-art erial
t han f or int ravenous administ rat ion of t he agent . The risk f act ors associat ed w it h
t his complicat ion w ere hyperglycemia, more severe neurologic st at us, low er
plat elet count s, and longer t ime t o recanalizat ion. Addit ional st udies are being
perf ormed t o evaluat e t he combined use of int ravenous and int ra-art erial
t hrombolyt ics in order t o opt imize t he benef it s of bot h.
Suggested Readings
Brot t T, Bogousslavsky J. Treat ment of acut e ischemic st roke. N Engl J Med.
2000; 343: 710â 7 22.
The Nat ional I nst it ut e of Neurological Disorders and St roke rt -PA St roke
St udy G roup. Tissue plasminogen act ivat or f or acut e ischemic st roke. N Engl
J Med. 1995; 333: 1581.
> Table of C ontents > Neur o > 205 - C ons ider Abs enc e of W ithdr aw al to P ain at 24 hour s and Abs enc e
of E ye R eflexes at 72 hour s P os t–C ar diac Ar r es t to be Highly C or r elated w ith P er m anent C om a
205
Consider Absence of Withdrawal to Pain at 24
hours and Absence of Eye Reflexes at 72 hours
Post–Cardiac Arrest to be Highly Correlated with
Permanent Coma
Amisha Barochia MD
Neurologic recovery af t er cardiac arrest occurs in only a small number of
pat ient s. O ne of t he biggest challenges in t he int ensive care unit (I CU) is
prognost icat ing out comes in comat ose pat ient s and opt imizing t reat ment
decisions w hen a pat ient survives a cardiac arrest . The physical examinat ion can
help great ly in predict ing deat h or poor out come in comat ose survivors of
cardiac arrest .
When a cardiac arrest occurs, t here is cessat ion of f low t o t he brain and ot her
organs. The oxygen st ores of t he brain last f or only approximat ely 20 seconds
af t er arrest and t he glucose and adenosine t riphosphat e (ATP) st ores in t he
brain last less t han 5 minut es. Neuronal deat h occurs af t er arrest and,
depending on t he ext ent , can result in a coma. During eff ect ive cardiopulmonary
resuscit at ion, f low is rest ored, but it is almost alw ays abnormal (e. g. , a low f low
st at e). Wit h a short er durat ion of arrest and resuscit at ion, t he damage due t o
ischemic injury and hypoperf usion is expect ed t o be less. When spont aneous
circulat ion is rest ored, t here is a brief period of cerebral hyperemia, f ollow ed by
vasospasm and prolonged hypoperf usion.
The overall rat e of a poor out come in pat ient s surviving a cardiac arrest w ho are
comat ose is approximat ely 77%. Poor out comes w ere def ined as severe
cerebral disabilit y, coma, veget at ive st at e, or deat h (G lasgow -Pit t sburgh
Cerebral Perf ormance Cat egories 3 t hrough 5). Clinical exam f indings can assist
in det ermining t he prognosis of comat ose survivors of cardiac arrest . Decisions
t o w it hdraw care should not be made on t he basis of t hese f indings alone, but
t his inf ormat ion is of t en usef ul in helping t he f amily make decisions about f urt her
care. The physical exam f indings t hat w ere f ound t o be most helpf ul in predict ing
poor out come w ere relat ed t o mot or and brain-st em f unct ion. The precision of
t he neurologic exam in predict ing out comes w as f ound t o be only moderat ely t o
subst ant ially helpf ul. The posit ive likelihood rat ios (LR) of various physical exam
f indings in predict ing poor out come are list ed in Tabl e 205. 1.
Theref ore, t he absence of t he pupillary and corneal ref lexes at 24 hours and t he
absence of a mot or response at 72 hours make it
ext remely likely t hat t he pat ient w ill have a poor out come. Levy et al. , in an older
st udy, f ound t hat none of t he 210 pat ient s w ho had all t hree of t hese f indings
ever regained an independent lif est yle. Alt hough many clinicians believe t hat
seizures or myoclonus port ends a poor prognosis, t his does not appear t o be
support ed by t he recent lit erat ure.
TABLE 205-1 POSITIVE LIKELIHOOD RATIOS
At Tim e of Com a Onset
Absent withdrawal to pain LR 1.7
At 24 Hours
Absent withdrawal to pain LR 4.7
Absent pupillary response LR 10.2
Absent motor response LR 4.9
Absent corneal reflex LR 12.9
At 72 Hours
Absent pupillary response LR 3.4
Absent motor response LR 9.2

Seizure or myoclonus LR 1.4
Finally, neuroprot ect ive st rat egies f or survivors of cardiac arrest hospit alized in
t he I CU include maint enance of cerebral perf usion and normoglycemia, cont rol of
seizures, and prevent ing hypert hermia. Wit h regard t o t he lat t er, an elevat ion of
1° C leads t o a 10% increase in brain oxygen consumpt ion. Theref ore, t he
avoidance of hypert hermia eit her by t reat ing f ever prompt ly or inducing mild
hypot hermia is f elt t o be benef icial.
Suggested Readings
Fischer C, Luaut e J, Nemoz C, et al. I mproved predict ion of aw akening or
nonaw akening f orm severe anoxic coma using t ree-based classif icat ion
analysis. Crit Care Med. 2006; 34(5): 1520–1524.
Koenig MA, Kaplan PW, Thakor NV. Clinical neurophysiological monit oring and
brain injury f rom cardiac arrest . Neurol Clin. 2006; 24(1): 89–106.
Levy DE, Coronna JJ, Singer BH, Lapinski RH, et al. Predict ing out come f rom
hypert oxic-ischamic coma. JAMA. 1985: 253(10): 1420–1426.
> Table of C ontents > Neur o > 206 - R em em ber that Failur e to R ec ogniz e P ituitar y Apoplexy C an
R es ult in a Neur ologic C atas tr ophe
206
Remember that Failure to Recognize Pituitary
Apoplexy Can Result in a Neurologic
Catastrophe
Eric M. Bershad MD
Jose I. Suarez MD
Pit uit ary apoplexy ref ers t o a clinical syndrome of abrupt severe headache,
visual loss, opht halmoplegia, and alt ered ment al st at us due t o hemorrhage or
inf arct ion w it hin t he pit uit ary gland, usually int o an underlying pit uit ary adenoma.
Clinical Signs and Symptoms

I n pit uit ary apoplexy, headache is universal and usually occurs abrupt ly and is
severe in int ensit y. Meningeal signs including st iff neck, nausea, and vomit ing
may be present because of blood ent ering t he subarachnoid space. Det eriorat ion
of visual f ields is common because of compression of t he proximat e opt ic
chiasm. The compression of crossed nasal ret inal f ibers, w hich represent s t he
t emporal visual f ields, produces a charact erist ic bit emporal hemianopia. The
proximit y of t he pit uit ary gland t o t he cavernous sinuses may result in
opt halmoparesis and diplopia due t o dysf unct ion of t he t hird, f ourt h, and sixt h
cranial nerves and a Horner syndrome, pt osis, and miosis due t o compression of
t he sympat het ic plexus w it hin t he cavernous sinus. Some pat ient s w ill complain
of f acial pain or alt ered sensat ion, w hich is relat ed t o t rigeminal nerve
compression. Hypot ension may occur because of def iciency of
adrenocort icot ropic hormone (ACTH).
Pit uit ary apoplexy usually occurs w it hout any specif ic provoking f act or. Most
pat ient s are f ound t o have an underlying pit uit ary adenoma. Several proposed
pat hophysiological mechanisms t hat may t rigger t he apoplect ic event include (a)
reduced blood f low in t he pit uit ary gland relat ed t o eit her syst emic hypoperf usion
or t ransient elevat ed int racranial pressure; (b) acut e increased blood f low ; (c)
st imulat ion of t he pit uit ary gland by exogenous or endogenous sources; and (d)
coagulopat hic condit ions. Some of t he common clinical scenarios in w hich t he
previous condit ions occur include surgical or procedural int ervent ions, pregnancy,
and acut e syst emic illnesses.
Diagnosis
A magnet ic resonance imaging (MRI ) of t he brain is t he diagnost ic modalit y of
choice (see Fi g. 206. 1). This w ill usually show a pit uit ary adenoma w it h
het erogeneous blood product s w it hin t he suprasellar mass. Ext ension of blood
and edema t o surrounding t issue may be visualized. Noncont rast ed comput ed
t omography (CT) head scan has poor resolut ion compared w it h MRI in
delineat ing t he suprasellar cont ent s but may be considered in some pat ient s. A
lumbar punct ure is not sensit ive or specif ic f or pit uit ary apoplexy but may be
usef ul in demonst rat ing subarachnoid blood if one suspect s aneurysmal
subarachnoid hemorrhage.
FIG URE 206. 1. MRI of t he brain show ing hemorrhage int o a pit uit ary
adenoma (arrow). Unenhanced sagit t al T1-w eight ed MRI demonst rat ing
het erogeneous areas of high-int ensit y signals mixed w it h regions of low er
signals, indicat ing subacut e hemorrhage w it hin a pit uit ary lesion. Reproduced
f rom Verrees, M, Araf ah B, Selman W. Pit uit ary t umor apoplexy:
charact erist ics, t reat ment and out comes. Neurosurg Focus. 2004; 16(4) E6.
Treatment
Emergent neurosurgical consult at ion f or def init ive surgical t reat ment is
mandat ory. This w ill allow f or prompt decompression of t he pit uit ary gland and
surrounding st ruct ures, t hus salvaging visual and oculomot or f unct ion. Surgical
resect ion via t he t ranssphenoidal rout e is usually suff icient , alt hough occasionally
an int racranial rout e may be necessary. Endocrinologic dist urbances occur
f requent ly af t er pit uit ary apoplexy; t hus comprehensive assessment of pit uit ary
axis should be perf ormed w it h t he consult at ion of an endocrinologist . I nt ravenous
hydrocort isone should be given immediat ely af t er t he diagnosis of pit uit ary
apoplexy has been made, as some pat ient s develop adrenal crisis because of
severe dysf unct ion of t he pit uit ary gland.
Prognosis
Wit h expedit ious neurosurgical int ervent ion, most pat ient s have excellent
out come. I n one series of 37 pat ient s, ocular paresis, visual f ield def ect s, and
visual acuit y resolved in 100%, 95%, and 88% of pat ient s, respect ively, af t er
surgical int ervent ion. I n t he same series, how ever, persist ent hormonal
dist urbances occurred in t he majorit y of pat ient s. The most common def iciencies
included grow t h hormone (88%), lut einizing hormone (76%), prolact in (67%),
adrenocort icot ropic hormone (66%), and est radiol (33%). Diabet es insipidus
occurred in only 2% t o 3% of t he pat ient s.
Suggested Readings
Biousse V, New man NJ, O yesiku NM. Precipit at ing f act ors in pit uit ary
apoplexy. J Neurol Neurosurg Psychiat ry. 2001; 71(4): 542–545.
Bradley W, ed. Neurology in Clinical Pract ice. Philadelphia: But t erw ort h-
Heinemann; 2004: 858–859.
Verrees M, Araf ah BM, Selman WR. Pit uit ary t umor apoplexy: charact erist ics,
t reat ment and out comes. Neurosurg Focus. 2004; 16(4): E6.
> Table of C ontents > Neur o > 207 - D o not P er for m a Lum bar P unc tur e on P atients w ith P os ter ior
Fos s a Mas s es
207
Do not Perform a Lumbar Puncture on Patients
with Posterior Fossa Masses
Jose I. Suarez MD
Anatomy
The post erior f ossa represent s t he area of t he cranial vault locat ed below t he
t ent orium cerebella. The lat t er arises f rom t he superior crest of t he pet rous
port ion of t he t emporal bone and roof s over t he post erior f ossa. The cerebellum
overlies t he post erior aspect of t he brain st em and ext ends lat erally under t he
t ent orium t o f ill most of t he post erior f ossa. The most caudal aspect of t he
post erior f ossa is t he f oramen magnum, t hrough w hich t he medulla exit s t he
cranium and becomes t he spinal cord.
Types of Lesions and Clinical Presentation

Mass lesions locat ed in t he post erior f ossa mainly represent neoplasms,
vascular lesions such as vascular malf ormat ions, and ischemic st rokes w it h
edema. Such mass lesions are classif ied according t o t heir locat ion w it hin t he
post erior f ossa (Tabl e 207. 1). The main areas w here such lesions are f ound
include t he parenchyma (cerebellum and brain st em), t he cerebellopont ine angle,
t he f ourt h vent ricle, and t he f oramen magnum. Because of t he small anat omical
space, t hese lesions can easily impinge upon t he vent ricular syst em (e. g. , t he
f ourt h vent ricle) and lead t o obst ruct ive hydrocephalus. Pat ient s may present
w it h headache, nausea, and vomit ing, f ollow ed by alt ered sensorium. Post erior
f ossa lesions can also compress t he brain st em, leading t o various syndromes
w it h mult iple cranial nerve involvement w it h serious respirat ory and
cardiovascular consequences. Anot her import ant issue t hat needs t o be
recognized is t hat any occupying space in t he post erior f ossa has t he pot ent ial
t o lead t o cerebellar t onsillar herniat ion. The cerebellar t onsils are locat ed in t he
most caudal and inf erior aspect of t he post erior lobe of t he cerebellum. I t can be
easily appreciat ed t hat should t here be t onsillar herniat ion, t he f oramen magnum
becomes occluded and t he cervicomedullary junct ion compressed. Respirat ory
f ailure and arrest w it h circulat ory collapse ensues. Last ly, lesions t hat compress
t he upper brain st em may lead t o t he so-called upw ard herniat ion of t he midbrain
and t he ant erior cerebellar lobe clinically manif est ed by coma, hypervent ilat ion,
f ixed pupils, and vert ical gaze paralysis.
TABLE 207-1 TYPES OF POSTERIOR FOSSA M ASS

LESIONS ACCORDING TO ANATOM ICAL LOCATION IN
ADULTS
Metastasis
Choroid plexus
papilloma
Intraventricular masses
Subependymoma
(fourth ventricle)
Dermoid/epidermoid
tumors
Hemangioblastoma
Acoustic neuroma
Meningioma
Cerebellopontine angle Vascular
masses ectasia/aneurysms
Epidermoid tumors
Metastasis
Ischemic stroke with

mass effect
Intracranial
Intra-axial masses (brain hemorrhage
stem and cerebellum) Vascular malformation
Metastasis
Hemangioblastoma
Exophytic glioma
Gliomas
Fourth ventricular
Foramen magnum masses tumors (see above)
Metastasis
Hemangioblastoma
Lumbar Puncture in Patients w ith Posterior Fossa

M asses
Perf orming a lumbar punct ure in pat ient s w it h know n or suspect ed post erior
f ossa masses is very risky. Lumbar punct ure w ill upset t he delicat e balance
bet w een parenchyma, mass, cerebrospinal f luid (CSF), and blood volume and
may precipit at e herniat ion w it h disast rous consequences. Pract it ioners should
ref rain f rom perf orming a lumbar punct ure under t hese circumst ances. I n t hose
inst ances w here obt aining CSF may be necessary (e. g. , meningit is), direct
sampling of t he vent ricular syst em is pref erred. Many of t hese pat ient s w ill
require ext ernal vent ricular drainage (most likely in t he lat eral vent ricles) t o
alleviat e hydrocephalus, f acilit at ing CSF recovery. I t is also import ant t o realize
t hat pat ient s w ho undergo such ext ernal vent ricular drainage may experience
upw ard herniat ion if t he post erior f ossa pressure is elevat ed. Theref ore, close
neurologic monit oring of t hese pat ient s is required.
Suggested Readings
Brazis PW, Masdeu JC, Biller J, eds. Localizat ion in Clinical Neurology. 3rd
ed. Bost on: Lit t le, Brow n and Company; 1996: 565–595.
Carpent er MB, ed. Core Text of Neuroanat omy. 4t h ed. Balt imore: Williams &
Wilkins; 1991: 1, 42–56.
O sborne AG , ed. Diagnost ic Neuroradiology. St . Louis: Mosby; 1994: 401–

528.
> Table of C ontents > Neur o > 208 - E valuate for Guillain- B ar r é S yndr om e in P atients w ith Ac ute
P ar alys is or R es pir ator y Failur e and Ar eflexia
208
Evaluate for Guillain-Barré Syndrome in Patients
with Acute Paralysis or Respiratory Failure and
Areflexia
Jose I. Suarez MD
G uillain-Barré syndrome (G BS) represent s t he most common cause of acut e
f laccid paralysis in ot herw ise healt hy adult s in t he Unit ed St at es. G BS is more
f requent in men as compared w it h w omen of all ages. There is a slight increase
in t he incidence of G BS in younger adult s and in t hose older t han 50 years of
age.
Most pat ient s w ill complain of a f lulike illness about 1 t o 3 w eeks prior t o
init iat ion of sympt oms. O t her possible f act ors t hat may be associat ed w it h
G uillain-Barré syndrome include inf ect ious agent s such as Campyl obacter jejuni,
Cyt omegalovirus, Epst ein-Barr virus, inf luenza virus, and human
immunodef iciency virus; recent surgery; immunizat ion; and syst emic illnesses
such as Hodgkin's disease and syst emic lupus eryt hemat osus. How t hese f act ors
t rigger or lead t o G uillain-Barré syndrome is st ill unclear. How ever, mount ing
basic and clinical evidence support s an immune-mediat ed at t ack on t he
peripheral nervous syst em. Pat hologic evaluat ion of peripheral nerves of G BS
pat ient s reveals perivascular inf ilt rat ion by mononuclear cells along w it h
demyelinat ion and axonal loss.
The major clinical sympt oms and signs include progressive limb w eakness and
aref lexia (Tabl e 208. 1). Such w eakness usually present s over a period of several
hours t o days. About half of pat ient s w ill have ascending involvement of low er
ext remit ies f ollow ed by upper ext remit ies. How ever, a variant present ing w it h
descending paralysis st art ing in cranial nerves or upper limbs can be seen in
about 14% of cases. Limb w eakness does not usually progress beyond 4 w eeks
and half of t he pat ient s experience a nadir w it hin 2 w eeks. Frequent ly pat ient s
complain of limb parest hesias and dysest hesias concomit ant ly or preceding limb
w eakness t hat is somet imes accompanied by low er back pain. O n physical
examinat ion pract it ioners encount er aref l exi a or severel y decreased myotati c
ref l exes and usually symmet ric limb involving bot h dist al and proximal muscles.
O n exam, t he most common cranial nerves involved are f acial, glossopharyngeal,
and vagus nerves. Thus pat ient s w ill have f acial w eakness (usually bilat eral) and
oropharyngeal
dysf unct ion manif est ed by dysphagia and dysart hria. Compromise of t he cranial
nerves innervat ing t he ext raocular muscles (i. e. , I I I , I V, and VI cranial nerves) is
less common and is usually part of a G uillain-Barré syndrome variant also know n
as Miller Fisher variant . The lat t er
can present w it h opht halmoparesis or opht halmoplegia and aref lexia f ollow ed by
respirat ory f ailure. The progression and severit y of G BS are variable w it h 33%
of pat ient s requiring mechanical vent ilat ion.
TABLE 208-1 DIAGNOSTIC CRITERIA FOR GUILLAIN-

BARRÉ SYNDROM E
Features
Progressive motor weakness of
Required for
more than one limb Areflexia
the Diagnosis
Features Strongly Supportive of the Diagnosis
Symptoms and signs of motor

weakness develop rapidly and
cease to progress 4 weeks into the
illness
Symmetry is seldom absolute
Mild sensory symptoms or signs
Clinical features Cranial nerve involvement (most
common is facial nerve)
Recovery usually begins 2 to 4
weeks after progression stops
Presence of autonomic dysfunction
Absence of fever at the onset of
symptoms
After the first week of symptoms,

cerebrospinal fluid protein is
Cerebrospinal
elevated or has been shown to rise
fluid features
on serial lumbar punctures
Counts of ≤10 leukocytes/mm3
Evidence of nerve conduction

slowing or block at some point
during the illness
Electrodiagnostic Distal latencies may be increased
features to as much as three times normal
Conduction studies may not
become abnormal until several
weeks into the illness
Marked, persistent asymmetry of

weakness
Persistent bladder or bowel
dysfunction
Features casting Bladder or bowel dysfunction at
doubt on the onset
diagnosis >50 mononuclear leukocytes/mm3
in cerebrospinal fluid
Presence of polymorphonuclear
cells in cerebrospinal fluid
Sharp sensory level
Diagnosis of botulism, myasthenia,

poliomyelitis, or toxic neuropathy
Features that
Abnormal porphyrin metabolism
rule out the
diagnosis Recent diphtheria infection
Occurrence of purely sensory
syndrome
From Asbury AK, Cornblath DR. Assessment of current

diagnostic criteria for Guillain-Barré syndrome. Ann
Neurol. 1990;27:S21–S24.
Diagnosis
Suspicion of G uillain-Barré syndrome should be based on t he clinical
present at ion and support ed by cerebrospinal f luid and elect romyographic
f indings (Tabl e 208. 1). All pat ient s w it h suspect ed G BS require a lumbar
punct ure and an elect romyogram (EMG ). Bot h of t hese t est s can help conf irm
t he diagnosis or make t he clinician suspect ot her condit ions t hat can mimic G BS
(Tabl es 208. 1 and 208. 2). Accurat e diagnosis is import ant f or eff ect ive
t reat ment t o be init iat ed prompt ly.
TABLE 208-2 DIFFERENTIAL DIAGNOSIS OF

GUILLAIN-BARRÉ SYNDROM E
KEY PRESENT ING CLINICAL

CONDIT ION
FEAT URES
Acute Asymmetric limb weakness

intermittent progressing to quadriplegia after
porphyria several attacks
Nausea and vomiting preceding

muscle weakness. Blurred vision,
dysphagia, dysarthria, descending
Botulism
muscle paralysis, dilated pupils,
dry mouth, constipation, and
urinary retention
Patient with chronic obstructive
pulmonary disease or asthma
Critical illness
requiring mechanical ventilation
myopathy
and use of neuromuscular blockers
and corticosteroids
Patient with sepsis and difficulty

Critical illness
weaning from the ventilator,
polyneuropathy
diminished or absent reflexes
Generalized muscle weakness,

Electrolyte
cardiac arrhythmias with or without
imbalance
rhabdomyolysis
Symmetric proximal muscle

Lambert-Eaton weakness, hypoactive or absent
myasthenic deep tendon reflexes, dry mouth,
syndrome blurred vision, orthostatic
hypotension
Pure motor weakness, initially of

extensors muscles, fasciculations,
Lead poisoning
abdominal pain, constipation,
anemia, renal failure
Motor neuron
W eakness, wasting, fasciculations
disease
Fatigue worse toward the end of

Myasthenia
the day, fluctuating symptoms and
gravis
signs, no sensory complaints
Exposure to insecticides,
petroleum additives, and modifiers
Organophosphate of plastics followed by acute
poisoning cholinergic crisis (muscle
weakness, miosis, abdominal
cramping)
Proximal, symmetrical muscle

Polymyositis weakness, elevated creatine
kinase
Patient with impaired renal or

Prolonged
hepatic failure who had been on
neuromuscular
continuous neuromuscular blocking
blocking
agents
Treatment
Management of t he G uillain-Barré syndrome pat ient w ill depend on t he severit y
and t he rat e of disease progression. I n t hose pat ient s w it h mild and
nonprogressing sympt oms, no specif ic t reat ment is necessary. How ever, in t hose
pat ient s w it h more severe compromise, admission t o t he int ensive care unit
(I CU) is indicat ed t o evaluat e respirat ory st at us and t o monit or f or
dysaut onomia. Several randomized cont rolled clinical t rials have support ed t he
use of plamapheresis (200 t o 250 mL of plasma/ kg body w eight f or a t ot al of 5
sessions) or immunoglobulin administ rat ion (400 mg/ kg/ d int ravenous
immunoglobulin [ I VI g] f or 5 days). Treat ment should be init iat ed w it hin 2 w eeks
of sympt om onset t o maximize response. Pat ient s should also receive ot her
support ive t reat ment s such as mechanical vent ilat ion (usually int ubat ion is
required w hen vit al capacit y is <15 cm3 / kg) and gast roint est inal and deep venous
t hrombosis prophylaxis. Dysaut onomia manif est ed as int ermit t ent episodes of
bradycardia, t achycardia, and hypo- or hyper-t ension may be severe and lead t o
deat h.
Prognosis
Most pat ient s recover sat isf act orily over a period of w eeks t o mont hs af t er init ial
present at ion. Approximat ely 10% of pat ient s are lef t w it h signif icant disabilit y
due t o mot or w eakness. I t should be not ed t hat pat ient s w ho require mechanical
vent ilat ion but are t reat ed w it h plasma-pheresis or I VI g w it hin 2 w eeks of
sympt om init iat ion are w eaned off mechanical vent ilat ion earlier t han pat ient s
w hose t reat ment w as delayed, t hus minimizing vent ilat or complicat ions in t he
early-t reat ment group.
Suggested Readings
Asbury AK, Cornblat h DR. Assessment of current diagnost ic crit eria f or
G uillain-Barré syndrome. Ann Neurol. 1990; 27: S21–S24.
Ropper AH. The G uillain-Barré syndrome. N Engl J Med. 1992; 326: 1130–
1136.
Suarez JI , ed. Crit ical Care Neurology and Neurosurgery. Tot ow a, NJ: Humana
Press; 2004: 481–492.
van der Merche FG , Schmit z PI . A randomized t rial comparing int ravenous

immunoglobulin and plasma exchange in G uillain-Barré syndrome. N Engl J
Med. 1992; 326: 1123–1129.
> Table of C ontents > Neur o > 209 - S tar t E ar ly P las m apher es is or Im m unoglobulins in Guillain- B ar r é
P atients
209
Start Early Plasmapheresis or Immunoglobulins
in Guillain-Barré Patients
Jose I. Suarez MD
G uillain-Barré syndrome (G BS) is an acut e inf lammat ory disease of t he
peripheral nervous syst em t hat is usually charact erized by ascending paralysis
w it h a paucit y of object ive sensory f indings. The progression and severit y of t he
disease vary. How ever, many pat ient s w ill require int ensive care w it hin t he f irst
f ew days of present at ion and about one-t hird of t hem w ill need mechanical
vent ilat ion.
What to Do
There are current ly t w o t herapeut ic modalit ies recommended f or G BS: plasma
exchange and immunoglobulin administ rat ion (I VI g). The decision t o t reat G BS
pat ient s depends on t he severit y of t he disease, rat e of progression of
sympt oms, and durat ion of sympt oms t o t ime of present at ion. Pat ient s w it h mild
sympt oms (e. g. , mild sensory complaint s and very dist al mot or involvement ) may
need t o be t reat ed w it h support ive care alone. Pat ient s w ho present w it hin 2
w eeks of onset of sympt oms may require t reat ment . Pat ient s present ing beyond
2 w eeks should probably also be t reat ed, but t he evidence is less convincing.
Plasma exchange has been available longer t han I VI g f or G BS and t he only one
compared w it h convent ional t herapy. St at ist ically signif icant diff erences, f avoring
t he plasma-exchange-t reat ed group, w ere f ound in t erms of improvement t o 4
w eeks, t ime t o independent w alking, and out come at 6 mont hs. Plasma exchange
is part icularly eff icacious in t hose pat ient s present ing w it hin 7 days of sympt om
onset and in t hose requiring mechanical vent ilat ion. O ne of t he main
disadvant ages of plasma exchange is it s availabilit y (Tabl e 209. 1). Not all
cent ers are w ell equipped and experienced t o perf orm plasma exchange in all
pat ient s w it h G BS. About 10% of pat ient s may experience recurrent sympt oms
w it hin 1 t o 2 w eeks af t er t reat ment w it h plasma exchange. This may be because
of a rebound product ion of ant ibodies. Furt her t reat ment s w it h plasma exchange
improve sympt oms. The recommended amount of plasma exchange is 200 t o 250
mL/ kg of body w eight every ot her day f or a t ot al of f ive t reat ment s.
I VI g has never been compared w it h placebo. How ever, I VI g has been compared
w it h plasma exchange in adequat e randomized,
cont rolled clinical t rials. Such t rials have show n t hat in pat ient s w it h severe G BS,
I VI g st art ed w it hin 2 w eeks of sympt om onset hast ens recovery as much as
plasma exchange. Also t reat ment w it h I VI g is more likely t o be complet ed t han
plasma exchange. O ne great advant age is it s ease of administ rat ion and
pot ent ially broader availabilit y. I VI g appears t o be superior t o plasma exchange
in t hose G BS pat ient s w ho are serum I gG ant i-G M1 posit ive. The recommended
dose of I VI g is 400 mg/ kg/ day f or 5 days.
TABLE 209-1 ADVANTAGES AND DISADVANTAGES OF

PLASM A EXCHANGE AND IVIG
T REAT MENT ADVANTAGES DISADVANTAGES
Available in
specialized centers;
Efficacious central-line-related
Plasmapheresis compared with complications;
placebo hypotension;
coagulopathy; mild
thrombocytopenia
More expensive;
headaches; fever;
myalgias;
anaphylactic
Immunoglobulin Easily given reactions
therapy (intravenously) (especially IgA-
deficient patients);
hypercoagulable
and hyperviscosity
syndrome
What not to Do
The use of combinat ion t herapy (i. e. , plasma exchange f ollow ed by I VI g) w as
advocat ed in t he past . How ever, one randomized, cont rolled t rial invest igat ed t he
eff icacy of plasma exchange alone, I VI g alone, and combinat ion t herapy w it hin 2
w eeks of sympt om onset . The st udy show ed t hat t he combinat ion of plasma
exchange w it h I VI g did not conf er signif icant advant age over each t reat ment
alone. Current ly combinat ion t herapy cannot be recommended.
Suggested Readings
Hughes RA, Raphael JC, Sw an AV, et al. I nt ravenous immunoglobulin f or
G uillain-Barré syndrome. Cochrane Dat abase Syst Rev. 2006; (1): CD002063.
Kuw abara S, Mori M, O gaw ara K, et al. I nt ravenous immunoglobulin t herapy

f or G uillain-Barré syndrome w it h I gG ant i-G M1 ant ibody. Muscle Nerve.
2001; 24: 54–58.
Plasma Exchange/ Sandoglobulin G uillain-Barré Syndrome Trial G roup.

Randomised t rial of plasma exchange, int ravenous immunoglobulin, and
combined t reat ment s in G uillain-Barré syndrome. Lancet . 1997; 349: 225–230.
The G uillain-Barré Syndrome St udy G roup. Plasmapheresis and acut e

G uillain-Barré syndrome. Neurology. 1985; 35: 1096–1104.
Van der Meche FG , Schmit z PI . A randomized t rial comparing int ravenous

immunoglobulin and plasma exchange in G uillain-Barré syndrome. Dut ch
G uillain-Barré St udy G roup. N Engl J Med. 1992; 326: 1123–1129.
> Table of C ontents > Neur o > 210 - R em em ber that P atients w ith Myas thenia Gr avis E xac er bation
Us ually Look W ell Until Jus t B efor e they R equir e Intubation
210
Remember that Patients with Myasthenia Gravis
Exacerbation Usually Look Well Until Just
Before they Require Intubation
Eliahu S. Feen MD
Jose I. Suarez MD
Myast henia gravis is an aut oimmune disease in w hich a pat ient primarily f orms
aut oant ibodies t o t he acet ylcholine recept ors on t he post synapt ic membrane in
t he neuromuscular junct ion. O t her kinds of aut oant ibodies also may play a role.
I n general f or myast henia gravis, t he post synapt ic muscle membrane is dist ort ed
and t he concent rat ion of acet ylcholine recept or prot eins is reduced. As a result ,
t he neurot ransmit t er acet ylcholine released f rom t he presynapt ic nerve t erminal
—alt hough it is released in normal concent rat ion—has a mut ed eff ect . Themuscle
w eakness t hat result s can t ake several f orms. I n ocularmyast henia gravis, t he
w eakness aff ect s t he ext raocular muscles, w it h f indings of diplopia and pt osis.
I n bot h bulbar myast henia gravis (aff ect ing primarily t he bulbar muscles) and in
generalized myast henia gravis (aff ect ing primarily t he limb muscles), respirat ory
muscle involvement is common. Myast henic crisis represent s t he exacerbat ion of
myast henia gravis t o t he point of requiring mechanical vent ilat ion.
Factors Contributing to M echanical Ventilation

Many myast henia gravis pat ient s w ill complain of dyspnea f requent ly, especially
w it h exert ion. Exacerbat ions of myast henia gravis present w it h t w o import ant
risk f act ors f or respirat ory compromise. The f irst is t hat t he muscle w eakness
impairs diaphragmat ic accessory respirat ory muscle f unct ion. As a result breat hs
are shallow. Consequent ly, t o maint ain adequat e oxygenat ion, t he respirat ory
rat e climbs, w hich f urt her exacerbat es muscle f at igue. The shallow breat hs also
limit t idal volume, w hich impairs gas exchange. The second mechanism is t hat
w it h pharyngeal dysf unct ion, t he risk of aspirat ion rises. The development of
aspirat ion pneumonia w ill f urt her compromise bot h oxygenat ion and gas
exchange because of shunt ing. The ominous danger of t his sit uat ion is emergent
respirat ory f ailure. O nce a pat ient develops ext reme muscle f at igue, respirat ions
w ill very rapidly become insuff icient t o maint ain oxygenat ion. The accumulat ion of
carbon dioxide
(CO 2 ) due t o progressively inadequat e gas exchange w ill cause hypercarbic

respirat ory f ailure. Caref ul monit oring of myast henia gravis pat ient s is necessary
in order t o avoid an emergent int ubat ion as w ell as t o avoid t he complicat ions of
emergent respirat ory f ailure and hypercarbia (such as hypoxia, hypot ension,
acidosis, and arrhyt hmias). As amyast henia gravis pat ient show s signs of
respirat ory compromise, closer monit oring becomes essent ial. I f such pat ient s
are not yet in an int ensive care set t ing, it is necessary t o place t hem in one.
Bedside Evaluation and Indications for M echanical

Ventilation
The most common bedside t est s of respirat ory st rengt h are t he measurement of
vit al capacit y (VC) and negat ive inspirat ory f orce (NI F). I n normal adult s VC is
great er t han 60 mL/ kg, and t he absolut e value of t he NI F is great er t han 70 cm
H2 O . Adult s on t he verge of int ubat ion due t o acut e respirat ory f ailure have a VC
of about 15 t o 20 mL/ kg and an absolut e value of NI F less t han 20 cm H2O .
Myast henia gravis pat ient s should have t heir VC and NI F checked at least every
6 hours. A progression of t hese bedside pulmonary t est s t ow ard dangerously low
levels may w arrant int ensive care unit (I CU) monit oring. I n addit ion t here is some
evidence t hat myast henia gravis pat ient s w it h an abnormalit y on chest x-ray
(such as an inf ilt rat e, eff usion, or at elect asis) are at higher risk t han ot her
pat ient s of requiring int ubat ion during t heir hospit al course and should t heref ore
be admit t ed direct ly t o an I CU f or close monit oring. I t is import ant t o remember
t hat art erial oxygenat ion may remain high up t o t he point of respirat ory collapse
because of t he pulmonary physiology of gravis. Theref ore, pulse oximet ry and
art erial blood gases may not be of help w hen deciding t he need f or int ubat ion.
Rat her, a VC ≤15 mL/ kg and NI F ≤20 cm H2 O should be used t o int ubat e
myast henia gravis pat ient s semielect ively.
M anagement
Treat ment of myast henia gravis exacerbat ions acut ely involves aggressive
immunot herapy in t he f orm of int ravenous immunoglobulin (I VI g) or
plasmapheresis. High-dose st eroids are used by many clinicians acut ely t o
improve t he condit ion of pat ient s. Cholinest erase inhibit ors, w hile usef ul as
chronic t herapy f or sympt omat ic relief , can be t roublesome in t he case of
exacerbat ions w here t here is a concern f or respirat ory compromise. Many
pat ient s suff er side eff ect s f rom t hese medicat ions in t he f orm of increased
secret ions. As a result
t his can f urt her exacerbat e respirat ory diff icult ies. Addit ionally, w hile pat ient s
cert ainly need pulmonary t oilet , incent ive spiromet ers are not appropriat e
because t his can f at igue pat ient s, w it h a risk of expedit ing respirat ory
compromise t hat out w eighs any benef it .
Suggested Readings
Mayer SA. I nt ensive care of t he myast henic pat ient . Neurology. 1997; 48
(Suppl): S70–S75.
Q ureshi AI , Choudry MA, Akbar MS, et al. Plasma exchange versus

int ravenous immunoglobulin t reat ment in myast henic crisis. Neurology.
1999; 52: 629–632.
Suarez JI , ed. Crit ical Care Neurology and Neurosurgery. Tot ow a, NJ: Humana
Press; 2004.
Thomas CE, Mayer SA, G ungor Y, et al. Myast henic crisis: clinical f eat ures,
mort alit y, complicat ions, and risk f act ors f or prolonged int ubat ion. Neurology.
1997; 48: 1253–1260.
> Table of C ontents > Neur o > 212 - R em em ber that not all S eiz ur es ar e C onvuls ive and O bvious
212
Remember that not all Seizures are Convulsive
and Obvious
Jose I. Suarez MD
Seizures are a common occurrence in t he int ensive care unit (I CU). The t erm
epi l epsy ref ers t o recurrent seizures. Seizures present ing in t he I CU can be
somew hat simplist ically classif ied int o part ial (or f ocal) and generalized. The
most common f ocal seizures are mot or and involve f ace or limb mot or seizure
act ivit y w it hout alt erat ion of sensorium. G eneralized seizures can be of various
t ypes: generalized t onic-clonic (generalized convulsions w it h loss of
consciousness); complex part ial (dist urbed sensorium w it h common
aut omat isms); and non-convulsive seizures (disordered sensorium or loss of
consciousness).
Watch Out For

Crit ically ill pat ient s experience seizures f requent ly regardless of w het her t hey
have underlying medical illnesses or have undergone a surgical procedure. Many
medical and surgical complicat ions increase t he likelihood of seizures including
hypoxia, cerebral ischemia, medicat ions, drug w it hdraw al, inf ect ion, surgical
injury, and met abolic derangement s. Most seizures in t he I CU occur in pat ient s
w it hout prior hist ory of seizures. I n f act it has been report ed t hat neurologic
complicat ions occur in about 12% of pat ient s admit t ed t o t he I CU w it hout prior
int racranial pat hology, and of t hese, 28% experience seizures. About 90% of
seizures in t he I CU are generalized t onic-clonic. How ever, about 10% of pat ient s
w ill have complex-part ial or ot her non-convulsive seizures including non-
convulsive st at us epilept icus (i. e. , ongoing elect rical seizures last ing more t han 5
minut es). Pat ient s w it h non-convulsive seizures represent a diff icult dilemma f or
int ensivist s. I t has been est imat ed t hat t he incidence of non-convulsive st at us
epilept icus may be 5% t o 50% in pat ient s in coma and t he incidence of non-
convulsive seizures in general could be as high as 34%. The key point t o learn
f rom t hese dat a is t hat w it hout monit oring (e. g. , elect roencephalography [ EEG ] ),
t he diagnosis of non-convulsive seizures and st at us epilept icus w ill be missed, t o
t he det riment of t he pat ient .
What to Do
The diagnost ic t est of choice t o det ect ongoing seizures in t he I CU is t he EEG
(ot her t han overt clinical manif est at ions). Even in pat ient s w ho appear t o
have been successf ully t reat ed f or overt st at us epilept icus, t he incidence of

subclinical or non-convulsive seizures can be as high as 20%. Current
recommendat ion is t hat unless pat ient s ret urn t o t heir baseline level of
consciousness af t er a seizure, an EEG , and pref erably cont inuous EEG
monit oring, should be perf ormed. Cont inuous EEG monit oring should be
perf ormed f or at least 24 hours t o exclude non-convulsive seizures. I n f act , some
invest igat ors recommend cont inuous EEG monit oring in most I CU pat ient s,
part icularly t hose w it h underlying neurologic injury or medical illnesses
predisposing t o seizures as ment ioned earlier. Availabilit y of cont inuous EEG
monit oring may be a problem in nonspecialized cent ers. I n such sit uat ions
int ensivist s may w ant t o obt ain an EEG immediat ely af t er t reat ment of a seizure
or in any pat ient w it h alt ered sensorium and repeat t he EEG daily f or t he
f ollow ing 24 t o 72 hours. How ever, it must be remembered t hat I nt ermit t ent
EEG s may f ail t o det ect many subclinical seizures and is not a replacement f or
cont inuous EEG .
Suggested Readings
Bleck TP, Smit h MC, Pierre-Louis SJC, et al. Neurologic complicat ions of
crit ical medical illness. Crit Care Med. 1993; 21: 98–103.
Jordan KG . Cont inuous EEG and evoked pot ent ial monit oring in t he
neurosciences int ensive care unit . J Clin Neurophysiol. 1993; 10: 445–475.
Tow ne AR, Wat erhouse EJ, Boggs JG , et al. Prevalence of non-convulsive

st at us epilept icus in comat ose pat ient s. Neurology. 2000; 54: 340–345.
Varelas PN, ed. Seizures in Crit ical Care: A G uide t o Diagnosis and
Therapeut ics. Tot ow a, NJ: Humana Press; 2005.
> Table of C ontents > Neur o > 213 - Tr eat S tatus E pileptic us as a Medic al E m er genc y
213
Treat Status Epilepticus as a Medical Emergency
St at us epilept icus is a medical emergency t hat requires immediat e recognit ion
and t reat ment . I t is def ined as a series of seizures t hat occur w it hout recovery
of underlying neurologic f unct ion. There are many causes of st at us epilept icus,
w hich may be relat ed t o an underlying epilept ic disorder or secondary t o an
acut e neurologic insult (Tabl e 213. 1). I n t he I CU set t ing, up t o 8% of pat ient s
experience t his condit ion w it h an associat ed mort alit y t hat can be as high as
40%.
Watch Out For

There are many diff erent present at ions of st at us epilept icus. I t can present w it h
generalized bilat eral involvement of skelet al muscles and loss of consciousness.
I t may begin as a f ocal seizure on one side, w hich t hen generalizes t o t he
cont ralat eral side, w it h loss of consciousness or, it may be nonconvulsive in
nat ure. St at us epilept icus can be conf irmed by an elect roencephalogram (EEG ),
but t reat ment should not be delayed pending an EEG result . The EEG can be
used t o conf irm bot h clinically apparent and subclinical st at us epilept icus.
What to Do
When a pat ient is experiencing a seizure, f undament al principles of advanced lif e
support need t o be applied. The pat ient should receive supplement al oxygen and
should be int ubat ed if t he airw ay is compromised. Blood pressure, heart rat e,
and oxygen sat urat ion need t o be assessed. I nt ravenous (I V) access should be
obt ained and serum chemist ry prof iles, complet e blood count , and serum glucose
level assessed. I f t he glucose level is low, dext rose 50% (D50) should be
administ ered. G lucose in t he absence of t hiamine should not be given unless
hypoglycemia is document ed since it can precipit at e Wernicke encephalopat hy in
t hiamine-def icient pat ient s. Alt ernat ively, pat ient s can be given 100 mg of
t hiamine bef ore or w it h glucose administ rat ion. O nce I V access is obt ained,
int ravenous benzodiazepines, pref erably lorazepam (0. 1 mg/ kg) should be
administ ered. I t is import ant t o not e t hat administ rat ion of benzodiazepines may
f urt her compromise t he pat ient 's ment al and respirat ory st at us. Theref ore,
preparat ion f or int ubat ion and mechanical vent ilat ion should be done. I f t he
seizure persist s,
phenyt oin 20 mg/ kg should be given and repeat doses up t o 30 mg/ kg should be
administ ered. The blood pressure, w hich may init ially be elevat ed, can decrease
w it h t he administ rat ion of ant iconvulsant t herapy or af t er 15 t o 30 minut es of
cont inuous seizure act ivit y due t o impaired cerebrovascular aut oregulat ion.
I nt ravenous f luids and vasopressor support should be administ ered t o maint ain
adequat e blood pressure.
TABLE 213-1 CAUSES OF STATUS EPILEPTICUS
Subtherapeutic
anticonvulsants
Prior seizure history
Ethanol related
Intractable epilepsy
Ethanol related
Drug toxicity
No prior seizure Central nervous system
history infection
Head trauma
Central nervous system tumor
Less common Metabolic aberration

etiologies Stroke
I n st at us epilept icus cerebral injury result s f rom prolonged elect rical discharge
t hat occurs. I f seizures persist in spit e of t he previously described t herapy, t hen
t he pat ient should receive int ravenous phenobarbit al, cont inuous inf usion of
int ravenous benzodiazepine, or propof ol. Cont inuing seizures t hat f ail t o respond
should prompt an evaluat ion of t he sodium level. Hyponat remia may cause st at us
epilept icus t hat f ails t o respond t o convent ional ant iconvulsant t herapy. For t hese
pat ient s, empiric 3% saline may be indicat ed t o st op t he seizures. I n addit ion,
hypomagnesia may be t he cause of seizures and st at us epilept icus in t ransplant
pat ient s because of t he magnesium-deplet ing eff ect of most immunosuppression
regimens. The experienced pract it ioner w ill realize t hat normal serum magnesium
is not indicat ive of t ot al body magnesium load. Many t ransplant cent ers
administ er I V magnesium empirically w hen any t ransplant pat ient seizures,
regardless of serum levels.
Suggested Readings
Fink MP, Abraham E, Vincent JL, et al. , eds. Text book of Crit ical Care. 5t h
ed. Philadelphia: Elsevier Saunders; 2005: 355–365.
Wyllie E, ed. The Treat ment of Epilepsy: Pract ice and Principles.
> Table of C ontents > Neur o > 214 - K now the P otential Adver s e E ffec ts of Valpr oic Ac id
214
Know the Potential Adverse Effects of Valproic
Acid
Jose I. Suarez MD
Valproic acid (Depakot e) is usef ul in t reat ing mult iple seizure t ypes including
complex part ial, generalized t onic-clonic, myoclonic, as w ell as st at us
epilept icus. Valproic acid has mult iple proposed mechanisms of act ion including
blocking sodium channels, augment ing t he act ion of glut amic acid decarboxylase,
a γ-aminobut yric acid (G ABA)–synt hesizing enzyme, rest rict ing G ABA
t ransaminase (G ABA-T), an enzyme t hat degrades G ABA, and blocking T-t ype
calcium current s. Unlike some ot her ant iconvulsant medicat ions, it can be rapidly
loaded, has lit t le eff ect on blood pressure, and is available as an int ravenous
f ormulat ion. Thus, valproic acid may be especially usef ul in unst able int ensive
care unit (I CU) pat ient s. Alt hough valproic acid is very usef ul in t reat ing st at us
epilept icus, t he clinician using it needs a t horough underst anding of it s pot ent ial
adverse eff ect s.
Adverse Effects
The most import ant adverse eff ect s of valproic acid include hepat ot oxicit y,
pancreat it is, t erat ogenicit y, t hrombocyt openia, and hyperammonemia. Valproic
acid carries a black-box w arning f or hepat ot oxicit y, pancreat it is, and
t erat ogenicit y. Serious hepat ot oxicit y usually occurs in young children rat her t han
adult s and occurs w it hin t he f irst 6 mont hs of t herapy; how ever, cases of
hepat ot oxicit y have been report ed in t he older populat ion. Preceding sympt oms
of malaise, w eakness, anorexia, or vomit ing may occur bef ore serious or f at al
hepat ot oxicit y is det ect ed. Pancreat it is occurs rarely and appears t o be an
idiosyncrat ic eff ect . I t may occur years af t er init iat ion of valproat e t herapy.
Cases of rapidly f at al hemorrhagic pancreat it is have been report ed in t he
lit erat ure. I n addit ion, valproic acid has a st rong t erat ogenic eff ect . I t s use in
pregnant pat ient s is associat ed w it h a 1% t o 2% rat e of neural t ube def ect s in
off spring, w hich is much higher t han t he general populat ion. High-dose f olic acid
supplement at ion should be given rout inely t o w omen of childbearing age w ho
require valproat e t herapy. Thrombocyt openia may occur w it h valproat e use and
appears t o be dose relat ed. I n a clinical t rial of valproat e monot herapy f or
epilepsy, 27% of pat ient s had at least one value below 75x109 / L w hen t aking a
dose of 50 mg/ kg/ day average. The incidence of t hrombocyt openia
increases signif icant ly w hen plasma valproic acid levels exceeded about 110 t o
135 mcg/ mL. Finally, hyperammonemia may occur in pat ient s t aking valproat e,
even in t he absence of abnormal liver-f unct ion t est s. Pat ient s w ho develop
unexplained let hargy, nausea, vomit ing, or ment al st at us changes w hile t aking
valproat e should have plasma ammonia levels checked. Underlying urea cycle
disorders should be evaluat ed in pat ient s f ound t o have hyperammonemia in t he
set t ing of valproat e use. Asympt omat ic elevat ions of ammonia may occur.
M onitoring
Laborat ory monit oring of pat ient s init iat ed on valproat e should include baseline
complet e blood count (CBC), liver-f unct ion t est s (LFTs) and t ransaminases, and
bet a–human chorionic gonadot ropin. Some clinicians repeat t he CBC and liver
t est s 1 w eek af t er st art ing t he drug. Subsequent ly, LFTs and t ransaminases
should be perf ormed severalw eeks af t er init iat ion of t herapy and at mont hly
int ervals at least f or t he f irst 6 mont hs. CBC should also be checked periodically.
I t is not necessary t o regularly check pancreat ic enzyme levels or ammonia
levels unless clinical sympt oms develop.
Suggested Readings
BradleyW, ed. Neurology in Clinical Pract ice. Philadelphia: But t erw ort h-
Heinemann; 2004: 1953–1992.
Physician's Desk Ref erence. ht t p: / / w w w. pdr. net / . New Jersey: Thomson

Healt h; 2006.
> Table of C ontents > Neur o > 215 - Lear n the C r anial Ner ve E xam ination as O ne C an O btain a Lot of
Infor m ation E ven in C om atos e and P oor ly C ooper ative P atients
215
Learn the Cranial Nerve Examination as One Can
Obtain a Lot of Information Even in Comatose
and Poorly Cooperative Patients
Eliahu S. Feen MD
Jose I. Suarez MD
First Cranial Nerve: The Olfactory Nerve

The olf act ory nerve conveys t he sense of smell t o appropriat e cent ers of t he
cerebral cort ex via olf act ory recept ors locat ed in t he superior nasal sept um.
Smell is properly t est ed t hrough t he use of nonirrit at ing volat ile oils or liquids,
f or example, oil of w int ergreen, oil of cloves, eucalypt us, oil of cinnamon, vanilla,
or anise. O ne nost ril is occluded w hile t he pat ient inhales vigorously w it h t he
st imulant held close t o t he open nost ril and w it h t he pat ient 's eyes closed. The
f act t hat t he pat ient can sense an odor is more import ant t han being able t o
ident if y t he part icular st imulant . Even some bedside subst ances can be usef ul f or
rough, qualit at ive t est ing of smell, such as coff ee grounds, lemon oil, or f lavored
t oot hpast e. O bviously a conscious pat ient is necessary f or t est ing of t he
olf act ory nerve. Some irrit at ive subst ances may act ually st imulat e t he t rigeminal
nerve and conf use t he examiner, even t hough t hey are at t ract ive t o use in
semiconscious or obt unded pat ient s. Examples of t he subst ances t o be avoided
are chlorof orm, ment hol, camphor, ammonia, st rong acet ic acid, alcohol, and
f ormaldehyde.
Second Cranial Nerve: The Optic Nerve

The opt ic nerve conveys ret inal f ibers t o t he lat eral geniculat e bodies of t he
t halamus t hrough t he opt ic chiasm. From t he lat eral geniculat e bodies, t he opt ic
radiat ion f ibers t ravel t o t he primary visual cort ex t o begin t he process of
const ruct ing vision based upon t he visual st imulat ion of t he ret inal f ibers. For t he
scope of t his t ext , only t he basic bedside clinical t est s of vision are of concern.
These f undament ally consist of conf ront at ion t est ing and visual acuit y t est ing.
Visual acuit y is t est ed in t he conscious, cooperat ive pat ient by means of one of
t he easily available reading cards, such as t he Snellen, Jaeger, or Rosen-baum
reading cards. Conf ront at ion t est ing consist s most commonly of having a pat ient
st are at a f ixed point in t he cent er of his or her vision (such as t he examiner's
nose, if t he examiner is in f ront of t he pat ient ) and count ing f ingers of t he
examiner's hand, held in f ront of
t he pat ient . I deally, each eye is t est ed independent ly by covering t he

cont ralat eral eye w it h an opaque object . The examiner should hold up f ingers of
his or her ow n hand in each of f our quadrant s of t he pat ient 's scope of vision f or
each eye. I t is import ant t o remember t hat f or t he purpose of count ing f ingers,
t he f ingers should be held up st eadily and not moved or w iggled. Moving t he
f ingers w hile asking t he pat ient t o count t hem w ill act ually t est t he sense of
visual movement and not object percept ion. I n semiconscious, uncooperat ive or
poorly communicat ive pat ient s, visual conf ront at ion t est ing may be perf ormed at
t he bedside by f licking f ingers close t o each eye but w it hout t ouching t he eye. A
blinking response indicat es visual f ield percept ion. There are t w o caveat s. First ,
t he f licking should not be so vigorous as t o induce air movement st rong enough
t o st imulat e t he eyelashes or cornea, w hich w ould induce a blinking ref lex.
Second, t his met hod can only grossly t est visual percept ion w it hout giving any
def init ive sense of quadrant anopsias. I t can usually be done only in t he f ield of
t he nasal ret ina.
The opt ic nerve, because it conveys light percept ion, is also involved in t he
pupillary light ref lex (consensual pupillary const rict ion in response t o light
st imulat ion), but because of t he involvement of t he t hird cranial nerve in t his
ref lex arc, t his is t reat ed in t he next sect ion.
Third Cranial Nerve (The Oculomotor Nerve), Fourth

Cranial Nerve (The Trochlear Nerve), and Sixth Cranial
Nerve (The Abducens Nerve)
The oculomot or nerve subserves f our of t he six oculomot or muscles; t he superior
rect us, medial rect us, inf erior rect us, and inf erior oblique muscles. I t also
innervat es t he levat or palpebrae superioris, w hich cont ribut es t o t he elevat ion of
t he upper eyelid. There is a parasympat het ic component t o t he innervat ion of t he
oculomot or nerve, w hich innervat es t he sphinct er pupillae muscle and ciliary
muscles. These lat t er muscles cont rol pupillary size and are int imat ely involved in
t he pupillary light ref lex and in accommodat ion t o near vision. The t rochlear
nerve innervat es t he superior oblique muscle. The superior oblique muscle
depresses t he eye w hen it is adduct ed and has some rot at ory eff ect . The
abducens nerve innervat es t he lat eral rect usmuscle, w hich abduct s t he eye.
I n a conscious pat ient t est ing of t he oculomot or, t rochlear, and abducens nerves
consist s basically of having t he pat ient t rack a moving t arget , such as t he
examiner's f inger. Each eye may be t est ed
independent ly. I n poorly cooperat ive pat ient s t he examiner may observe
spont aneous eye movement s and use gross st imulat ion, such as moving f rom one
side of a pat ient 's bed t o t he ot her w hile calling or clapping, t o observe f or
t racking eye movement s. Some comat ose pat ient s may have cert ain kinds of
roving eye movement s, w hich can give inf ormat ion about w hich ext raocular
muscles are int act . Test ing t rochlear nerve f unct ion independent ly is diff icult at
t he bedside, most especially in an int ensive care set t ing. The most common
clinical present at ion of acut e isolat ed t rochlear nerve palsy in an ambulat ory
pat ient is a head t ilt —t he pat ient present s w it h his or her head t ilt ed f orw ard and
t o t he cont ralat eral side of t he aff ect ed muscle. I ncreasing t he head t ilt in t he
direct ion of t he palsied side may induce vert ical diplopia, w hich is part of t he
t est ing maneuver f or t rochlear muscle w eakness (Bielschow sky t est ). O bviously,
t his is somew hat diff icult in many int ensive care unit (I CU) pat ient s. Examinat ion
of eye movement s must include observat ion of nyst agmus and ot her eye-
movement abnormalit ies such as f orced gaze in a part icular direct ion or
opsoclonus (irregular oscillat ions of t he eyes in bot h vert ical and horizont al
direct ions), as t hese may help t o localize neurologic lesions.
For pat ient s w ho cannot cooperat e (f or example, comat ose pat ient s or t hose on
sedat ion), t he perf ormance of oculocephalic maneuvers can elicit eye
movement s. The oculocephalic ref lex depends upon an int act vest ibular syst em
f or it s reliabilit y. I n a pat ient w it h an int act oculocephalic ref lex, t urning t he
pat ient 's head in a given horizont al direct ion w ill drive t he eyes t o t he opposit e
side. This t est s t he ipsilat eral medial rect us (t hird cranial nerve) and t he
cont ralat eral lat eral rect us (sixt h cranial nerve). Vert ical eye movement s can be
t est ed w it h a vert ical mot ion of t he head (w it h t he head held in t he midline
posit ion). I solat ion of a part icular ext raocular muscle abnormalit y is diff icult ,
t hough, because of t he combinat ion of ext raocular muscle f unct ion in t he
product ion of vert ical eye movement s. I n addit ion t o eye movement s, ot her
clinical f eat ures w ill indicat e oculomot or dysf unct ion. The presence of pt osis may
suggest a lesion in t he oculomot or nerve.
Finally, t est ing of t he pupillary light ref lex is an absolut ely essent ial component
of bedside cranial nerve t est ing. Light is conveyed t hrough t he opt ic nerves t o
t he midbrain and int o t he Edinger-West phal nucleus. Eff erent mot or and
parasympat het ic f ibers t o t he pupil are conveyed by t he oculomot or nerve. There
are mult iple crossing point s along t he pat hw ay, so t he pupillary light ref lex is
bilat eral in nat ure. I t is t est ed w it h t he sw inging f lashlight t est , in w hich a bright
light is direct ed int o each eye individually. The examiner must observe f or
bot h t he direct response and t he consensual response. The pupil in a normal eye
const rict s in response t o light show n direct ly int o it ; t his is t he direct response.
The cont ralat eral pupil should simult aneously (and equally) const rict if it is also
normal, w hich represent s t he consensual response. An abnormalit y in one of t he
responses suggest s a lesion somew here along t he pat hw ay, w hich can involve
bot h t he opt ic nerve, t he oculomot or nerve, or midbrain connect ions. I n a
conscious, cooperat ive pat ient , t he accommodat ion ref lex can also be t est ed. I n
t he normal accommodat ion ref lex, if a pat ient looks t o a f ar object and t hen
shif t s his or her gaze t o a near object , such as a hand held in f ront of t he
pat ient 's f ace, t he pupils const rict . The eff erent limb of t he ref lex is conveyed by
t he oculomot or nerve.
Fifth Cranial Nerve: The Trigeminal Nerve

The t rigeminal nerve has bot h sensory and mot or component s. The sensory
component conveys pain, t emperat ure, and light t ouch sensat ion f rom t he f ace
region. I n addit ion it carries t ast e f rom t he oral mucosa and t o some ext ent t he
t ongue and propriocept ion f rom cert ain muscles of t he f ace. The mot or
component primarily innervat es several muscles coordinat ing mast icat ion
(chew ing).
Clinical examinat ion of t he t rigeminal nerve involves t est ing bot h it s mot or and
sensory component s. I n an aw ake, cooperat ive pat ient , t he mot or limb may be
examined by t est ing t he st rengt h of jaw movement s. The examiner can t ry t o
open t he pat ient 's jaw against resist ance, close t he jaw against resist ance, or
resist t he pat ient 's eff ort t o lat erally displace his or her mandible (t o t he lef t or
right ). I n a unilat eral lesion of t he mot or component of t he t rigeminal nerve
aff ect ing t he branches t hat innervat e muscles of jaw opening or closing, an
at t empt t o do t his w ill result in def lect ion of t he jaw t o t he side cont ralat eral t o
t he lesion.
Sensory examinat ion is st raight f orw ard in an aw ake, cooperat ive pat ient . There
are t hree branches of t he t rigeminal nerve innervat ing sensat ion of t he f ace—t he
opht halmic, maxillary, and mandibular branches. These t hree branches roughly
divide t he ant erior, cent ral f ace int o t hree sensory zones, f rom t op t o bot t om,
respect ively. Touching t he pat ient 's f ace w it h eit her a sof t or sharp object w ill
indicat e int act sensat ion t hrough t he t rigeminal nerve on each respect ive side.
Tradit ionally t he pat ient is st imulat ed on t he f orehead f or t he opht halmic branch,
on t he cheek lat eral t o t he nose f or t he maxillary branch, and on t he chin or jaw
f or t he mandibular branch (on each side of t he f ace).
Uncooperat ive or unconscious pat ient s may require ref lex t est ing of t he
t rigeminal nerve, as caref ul observat ion f or lat eral def lect ion may be inadequat e
and a sensory exam unreliable. The jaw ref lex (also called masset er ref lex or
mandibular ref lex) is elicit ed by placing t he examiner's f inger on t he middle of
t he chin and t apping w it h a ref lex hammer over t hat f inger. A relaxed or slight ly
open jaw is necessary. For an int act ref lex t he jaw w ill suddenly and moment arily
close in response t o t he t apping. The corneal ref lex is arguably t he most f amous
t rigeminal ref lex. Corneal sensat ion is conveyed by t he t rigeminal nerve. St roking
t he cornea gent ly w it h a nonabrasive object (such as a w isp of cot t on) or
w ashing t he cornea w it h st erile, saline f lush w ill st imulat e t he aff erent limb of t he
ref lex. The sevent h cranial nerve (f acial nerve) conveys t he eff erent limb of t he
corneal ref lex. The eff erent limb consist s of t w o part s in most pat ient s—a direct
corneal ref lex and a consensual ref lex. The direct ref lex consist s of t he
ipsilat eral closing of t he eye (i. e. , ipsilat eral t o t he side of t he st imulus). The
consensual ref lex involves simult aneous closure of t he cont ralat eral eye. Not all
pat ient s w ill normally evince t he consensual response.
O t her ref lexes may be especially usef ul in comat ose pat ient s. I n t hese cases t he
sensory limb of t he ref lex arc is conveyed by t he t rigeminal nerve. St imulat ion of
t he cornea may cause pupillary const rict ion or dilat at ion f ollow ed by const rict ion
(bilat eral due t o t he nat ure of pupillary innervat ion, as described earlier).
Tickling t he nasal mucosa w it h, f or example, a cot t on sw ab may cause sneezing
or f acial muscle cont ract ion, especially in t he vicinit y of t he nose (t he
st ernut at ory ref lex).
Seventh Cranial Nerve: The Facial Nerve

The f acial nerve has primarily mot or f unct ion t o innervat e t he muscles of t he f ace
(except ing t he of mast icat ion, w hich are innervat ed by t he t rigeminal nerve) and
plat ysma of t he upper, ant erior neck t o give f acial expression. There is some
sensory f unct ion, in part icular t ast e, over t he ant erior t w o-t hirds of t he t ongue.
Variable port ions of t he ext ernal audit ory meat us have superf icial sensat ion
conveyed by t he f acial nerve in some individuals. Parasympat het ic input t o t he
lacrimal (t ear) glands, t he several salivary glands, and mucous-secret ing glands
is also conveyed by t he f acial nerve.
Clinical t est ing of t he f acial nerve in aw ake, cooperat ive pat ient s involves asking
t hem t o raise t heir eyebrow s, squeeze t heir eyes shut , smile or f row n, blow out
t heir cheeks, w hist le, and cont ract t he chin muscles by ret ract ing t he angles of
t he mout h dow nw ard (f or t est ing of
t he plat ysma). Uncooperat ive or unconscious pat ient s may be observed caref ully
f or t he perf ormance of t hese act ions spont aneously. As w ell, a f acial droop or an
asymmet ry of t he nasolabial f olds and size of t he palpebral f issures w ill give
some clue t o asymmet ric f acial t one. I f a pat ient is lying w it h his or her eyelids
closed, asymmet ry in t he ease w it h w hich t he eyelids are passively pulled open
by an examiner may give some indicat ion of a lesion in one of t he f acial nerves
(opht halmic branch).
The most reliable ref lexes t est ing t he f acial nerve are t hose w it h a sensory limb
along t he t rigeminal nerve, namely t he corneal and st ernut at ory ref lexes. There
are various ref lexes not ed in some pat ient s t hat involve st imulat ion of eye
closure, puckering of t he mout h, or low er f acial muscle f lexion, but t hese are t oo
variable t o be generally reliable.
Eighth Cranial Nerve: The Vestibulocochlear Nerve

The eight h cranial nerve consist s of t w o separat e nerve bundles t hat combine
int o a single nerve t runk, namely t he acoust ic nerve and t he vest ibular nerve. The
acoust ic nerve conveys t he sensat ion of sound, as derived f rom t he recept or
cells in t he organ of Cort i. The vest ibular nerve conveys t he sensat ion of balance
and head posit ion in space, as derived f rom t he recept or cells in t he semicircular
canals.
Bedside t est ing of t he acoust ic nerve in an int ensive care set t ing is necessarily
inexact . For aw ake, cooperat ive pat ient s, t est ing air and bone conduct ion of
sound t hrough t he Rinne and Weber t est s may be appropriat e. The det ails are
beyond t he scope of t his t ext . I t is import ant t o not e, t hough, t hat almost
immediat ely af t er each acoust ic nerve ent ers t he brain st em, t here is f iber
crossing. Thus, sound conduct ion f rom t he brain st em up t o t he cort ex is
bilat eral, and unilat eral acoust ic nerve lesions may be missed on simple sound
t est ing.
The most common sympt om of vest ibular dysf unct ion is vert igo and one of t he
most common signs on exam is posit ional nyst agmus. Caref ul t est ing of aw ake,
cooperat ive pat ient s in t erms of peripheral mot or f unct ions, muscle t one,
post ural ref lexes, and eye movement s can cont ribut e t o t he delineat ion of a
vest ibular problem, but again t his is w ell beyond t he scope of t his t ext . For our
purposes, t he most import ant clinical t est of vest ibular f unct ion is in comat ose
pat ient s f or w hom caloric t est ing may be necessary. As discussed in t he sect ion
on t he cranial nerves cont rolling eye movement s, t he oculovest ibular ref lex t est s
eye movement s based upon t he st imulat ion of t he vest ibular
syst em. I n t he clinic, pat ient s can be placed on rot at ing chairs t o st imulat e t he
vest ibular syst em. Aw ake, cooperat ive pat ient s can undergo caloric t est ing as
w ell, w hich is of t en uncomf ort able f or t hem.
At t he bedside of comat ose pat ient s, in preparat ion f or caloric t est ing, each
ext ernal audit ory meat us should be examined t o be sure it is f ree of cerumen or
ot her subst ances. The t ympanic membrane should be examined w it h t he use of
an ot oscope f or any evidence of perf orat ion. Should perf orat ion be f ound, it is
advisable t o avoid t est ing t hat side. How ever, w here t est ing is absolut ely
necessary, an ant isept ic solut ion may be used. For a supine pat ient t he head is
placed at a 30-degree angle above t he horizont al plane. This head posit ion w ill
put t he lat eral semicircular canal in t he vert ical plane. I ce w at er is recommended
(at 0 t o 5C). The examiner should inst ill a minimum of 10 mL over a maximum of
30 t o 40 seconds. O ne common pract ice, t hough, is t o inst ill about 60 mL over
30 seconds and t hen observe f or anot her 30 seconds. The examiner observes f or
any eye movement s. The reason t he ext ernal audit ory canal is f illed w it h cold (or
in some variat ions of t he t est , w arm) w at er is t o st imulat e endolymph f low w it hin
t he semicircular canals. A response t o cold-w at er caloric t est ing in a normal
pat ient consist s of a rot at ory nyst agmus of t he eyes, w it h a slow phase t o t he
side ipsilat eral t o t he cold-w at er st imulat ion. (Warm w at er produces a slow
phase t o t he side cont ralat eral t o t he st imulat ion. ) Caloric t est ing is a very
st rong st imulus, and f ailure t o produce eye movement s suggest s a lesion in t he
vest ibular syst em (on t he appropriat e side), t he ocular mot or syst em, or in t he
brain-st em connect ions t hat link t hese t oget her. Absence of t his ref lex is a
common t est in conf irming brain deat h.
Ninth Cranial Nerve (The Glossopharyngeal Nerve) and

Tenth Cranial Nerve (The Vagus Nerve)
The glossopharyngeal nerve supplies some minimal muscular f unct ion t o t he arch
of t he sof t palat e, t he aff erent limb of part of t he ref lex of salivat ion, and t ast e
sensat ion t o t he post erior t hird of t he t ongue and part of t he pharynx. For clinical
purposes t he most import ant innervat ion of t he glossopharyngeal nerve is t he
aff erent limb of t he gag ref lex. St roking t he post erior pharyngeal t issue,
including t he area of t he t onsils, w it h a sof t or blunt device, such as a cot t on
sw ab or t ongue blade, w ill elicit elevat ion and const rict ion of t he muscles of t he
pharynx as w ell as ret ract ion of t he t ongue. This mot or response, how ever, is
carried t hrough t he vagus nerve. A lesion of t he glossopharyngeal nerve
w ill cause unilat eral impairment in t he gag ref lex, so t hat st imulat ion t o t he side
ipsilat eral t o t he lesion w ill f ail t o elicit a response.
The vagus nerve has mult iple f unct ions t hroughout t he body, as it s anat omic
course and name (“w andering”) suggest . The mot or cont rol of t he vagus involves
t he muscles of t he sof t palat e, pharynx, and larynx. Coordinat ion of sw allow ing
and phonat ion are under vagal cont rol primarily. For clinical purposes, sensory
f unct ions of t he vagus include a small area of skin sensat ion in t he ext ernal
audit ory meat us and pinna in some individuals and t ast e over cert ain regions of
t he deep pharynx. (Some ot her sensory f unct ions of t he vagus lack clinical ut ilit y
since t hey are diff icult t o t est , such as sensory innervat ion of t he meninges and
t ast e over t he epiglot t is. ) The vagus also conveys a large amount of
parasympat het ic innervat ion t o t he viscera, as w ell as carrying some general
sensory inf ormat ion f rom t he viscera.
When a pat ient has diff icult y w it h sw allow ing and slurred speech, t his may be an
indicat ion of vagal dysf unct ion (f or a low er mot or neuron lesion). A hoarse or
raspy voice, as opposed t o ot her phonat ion abnormalit ies, represent s t he most
t ypical manif est at ion of vagal nerve lesions t hrough t he laryngeal branches. To
def init ively prove a vocal cord palsy requires laryngoscopy. The gag ref lex is
also necessary in t he examinat ion of vagal f unct ion. A unilat eral vagus lesion t hat
causes palat al and pharyngeal w eakness w ill produce an ipsilat eral droopy sof t
palat e. Upon cont ract ion of t he palat al and pharyngeal muscles (eit her volunt arily
or t hrough st imulat ion of t he gag ref lex in comat ose pat ient s), t he uvula and sof t
palat e w ill be pulled t o t he side cont ralat eral t o t he lesion.
Eleventh Cranial Nerve: The Spinal Accessory Nerve

The spinal accessory nerve, ent irely a mot or nerve f or all pract ical purposes,
innervat es most of t he f ibers of t he st ernocleidomast oid muscle and a minorit y of
t he f ibers of t he t rapezius muscles. The rest of t he innervat ion f rom t hese
muscles comes f rom cervical spinal nerves. As a result , complet e lesions of t he
spinal accessory nerve produce w eakness as opposed t o complet e paralysis.
Clinical examinat ion of t he st ernocleidomast oid involves having t he pat ient t urn
his or her head f ully t o one side against resist ance. This requires cont ract ion of
t he st ernocleidomast oid muscle cont ralat eral t o t he side t o w hich t he head is
t urned. Weakness of head t urning w ill suggest a lesion. The t rapezius muscle
elevat es t he shoulder, ret ract s t he head, and assist s in t he elevat ion of t he arm
above t he horizont al posit ion. Clinical
examinat ion of t hese f unct ions requires aw ake, cooperat ive pat ient s. The
examiner has t he pat ient shrug his or her shoulders against resist ance. Again,
w eakness may suggest a low er mot or neuron lesion of t he spinal accessory
nerve.
Tw elfth Cranial Nerve: The Hypoglossal Nerve

The hypoglossal nerve innervat es almost all of t he muscles t hat cont rol t ongue
movement . O bservat ions of t ongue movement , even in uncooperat ive or
unconscious pat ient s, w ill give insight int o lingual paresis. Unilat eral lesions of
t he hypoglossal nerve w ill produce ipsilat eral w eakness in t he t ongue. The f ibers
of t he lingual muscles t hat prot rude t he t ongue are angled in such a w ay t hat
unilat eral lesions of t he nerve cause t he t ongue t o deviat e t o t he ipsilat eral side.
When t he t ongue cannot be examined in t his w ay or t ongue movement s are
rest rict ed by endot racheal t ubes or t he like, some clinical f eat ures suggest
hypoglossal nerve lesions if t he lesions are chroni c. Chronic hypoglossal nerve
lesions produce at rophy of t he lingual muscles and f asciculat ions of t he muscles,
w hich are visible upon inspect ion of t he t ongue.
O f not e, w hile t he f ull spect rum of cranial nerve abnormalit ies is w ell beyond t he
scope of t his t ext , some vit amin def iciencies produce t ypical changes in t he
mucosa of t he t ongue. For example, t he t ongue of Wernicke (t hiamine def iciency)
appears smoot h and shiny but reddened w it h some at rophy present .
Suggested Readings
Brazis PW, Maseu JC, Biller J. Localizat ion in Clinical Neurology. 4t h ed.
Philadelphia: Lippincot t Williams & Wilkins; 2001.
Campbell WW, ed. DeJong's t he Neurologic Examinat ion. 6t h ed. Philadelphia:

Lippincot t Williams & Wilkins; 2005.
> Table of C ontents > Neur o > 216 - D o not S queez e the Toenails or Adm inis ter a P ainful S tim ulus to
the foot to Tes t R es pons e to P ain in C om atos e P atients as this m ay Tr igger a S pinal R eflex that m ay
be s een even in B r ain- D ead P atients
216
Do not Squeeze the Toenails or Administer a
Painful Stimulus to the foot to Test Response to
Pain in Comatose Patients as this may Trigger a
Spinal Reflex that may be seen even in Brain-
Dead Patients
Jose I. Suarez MD
The evaluat ion of t he comat ose pat ient using t he G lasgow Coma Scale (G CS)
requires t est ing t he respect ive modalit ies of eye opening, verbal response, and
mot or response. I n t est ing mot or response, it is import ant t o precisely
charact erize t he degree of mot or responsiveness. This helps t o charact erize t he
degree of coma and it may also have prognost ic value w hen assessing a pat ient
af t er hypoxic or ischemic injury.
Watch Out For

O ne specif ic element in t he mot or exam t hat of t en creat es conf usion f or t he
examiner (and t he f amily) is t he int erpret at ion of t he pat ient 's response t o a
painf ul st imulus in t he low er ext remit y. The conf usion relat es t o t he issue of
diff erent iat ing bet w een a purposef ul w it hdraw al response t o pain versus a simple
spinal ref lex w it hdraw al t o pain. The import ance of t his diff erent iat ion is t hat a
purposef ul w it hdraw al response t o pain may be a good prognost ic sign in
pat ient s w it h hypoxic-ischemic injury, w hereas a spinal ref lex w it hdraw al may be
seen in brain-dead pat ient s. I t is import ant t o realize t hat brain-dead pat ient s
may exhibit a variet y of movement s t hat are spinally mediat ed and do not imply
int act brain-st em or cort ical f unct ion. Among t he possible movement s are
ext ensor plant ar responses, jerks of t he f ingers, t oe f lexion, and a Lazarus sign,
w hich involves f lexion of t he arms at t he elbow, adduct ion of t he shoulders, lif t ing
of t he arms, dyst onic post uring of t he hands, and crossing of t he hands.
Furt hermore, a t riple-f lexion response w it h dorsif lexion at t he ankle, knee, and
hip may be seen w it h dist al low er-ext remit y pain st imulat ion, w hich may appear
at f irst glance t o be a purposef ul w it hdraw al response but is act ually a spinal
cord ref lex.
The w ays t o diff erent iat e bet w een a spinal cord ref lex and a t rue w it hdraw al
response involve observing t he charact er of t he w it hdraw al response t o pain. A
spinal cord ref lex w ill usually have a st ereot yped amplit ude and durat ion. The
amplit ude of ref lexic movement is usually proport ional t o t he st imulus int ensit y.
Furt hermore, a spinal ref lex such as t he t riple-f lexion response is usually more
easily elicit ed by
dist al low er-ext remit y st imulat ion, especially t he f oot or t oenail beds. Thus, it is
bet t er t o t est f or w it hdraw al response in t he low er ext remit y by using a more
proximal st imulat ion such as squeezing t he calf or t high.
Selected Readings
Saposnik G , Maurino J, Saizar R, et al. Spont aneous and ref lex movement s in
107 pat ient s w it h brain deat h. Am J Med. 2005; 118(3): 311–314.
Teasdale G , Jennet t B. Assessment and prognosis of coma af t er head injury.

Act a Neurochir (Wien). 1976; 34(1–4): 45–55.
> Table of C ontents > Labor ator y > 217 - D o not As c r ibe an Inc r eas ed S er um Lac tate Level to R enal
Ins uffic ienc y
217
Do not Ascribe an Increased Serum Lactate
Level to Renal Insufficiency
Lact ic acid levels are a usef ul adjunct in assessing perf usion in t he int ensive
care unit (I CU). Lact at e levels >2 mEq/ L may ref lect t issue ischemia. How ever,
lact at e levels may be conf ounded by various circumst ances and are also prone
t o misint erpret at ion. When using lact at e levels in clinical management , it is
import ant t o underst and t he physiology behind lact at e product ion and t he
commonly held misconcept ions about lact ic acidosis.
When cells f ace insuff icient oxygen t o carry out t heir met abolic f unct ions, t hey
convert f rom aerobic t o anaerobic met abolism. Wit hout an available oxygen
molecule, pyruvat e is inst ead convert ed t o lact at e, cont ribut ing t o an anion-gap
acidosis. I n ot her w ords, rat her t han yielding t he complet e 36 adenosine
t riphosphat e (ATP) molecules f or each mole of glucose, met abolism is halt ed at
lact at e, rendering only 2 ATP. The clearance of lact at e (in t he set t ing of
adequat e t issue perf usion) occurs in t he liver, t hrough one of t w o mechanisms. I t
may combine w it h oxygen w it h t he end result being carbon dioxide, w at er, and
bicarbonat e:
CH3 CHO HCO O - + 3O2 → 2CO2 + 2H2 O + HCO- 3
Alt ernat ively, lact at e may be processed in t he Cori cycle in t he liver, w here
lact at e is convert ed t o pyruvat e, t hen t o glucose in t he process of
gluconeogenesis.
There are t w o t ypes of lact ic acidosis, A and B. Type A w as described earlier:
lact at e result ing f rom inadequat e oxygen t o complet e aerobic met abolism. Type
B, on t he ot her hand, st ems f rom insuff icient clearance. I f t he pat ient suff ers
f rom liver f ailure, he or she may not be able t o clear t he lact at e by convert ing it
t o pyruvat e. Thus, a persist ent ly high lact at e in t he set t ing of liver f ailure may
not be indicat ive of ongoing ischemia, but simply of t he inabilit y t o resolve t he
lact ic acidosis met abolically.
O f not e, lact ic acidosis should not be ascribed t o renal insuff iciency. The kidney
f ilt ers lact ic acid only at a t hreshold of 6 t o 8 mEq/ L; at low er levels lact at e is
cleared hepat ically as described previously. I n addit ion, Lact at ed Ringer solut ion
does not cont ribut e t o lact ic acidosis and may be given in t he set t ing of ischemia
w it hout concern
f or conf ounding t he pict ure. I sot onic w it h plasma, Lact at ed Ringer cont ains a
balanced solut ion of sodium (130 mEq/ L), chloride (109 mEq/ L), pot assium (4
mEq/ L), and calcium (3 mEq/ L) w it h a lact at e buff er t o render a pH of 6. 4.
Alt hough a small amount of lact at e is cont ained w it hin t his f luid, it s isot onicit y
renders t his lact at e clinically inconsequent ial. I n f act , Lact at ed Ringer is
pref erred over 5% dext rose solut ion in t he set t ing of crit ical illness: t he lat t er
has been show n t o cont ribut e t o higher post operat ive lact at e levels as t he added
glucose is met abolized t o lact at e during periods of ischemia.
I nt erpret at ion of lact ic acid levels may be f urt her conf ounded by lag t ime in
laborat ory processing. I f a sample sit s f or a period of t ime prior t o processing,
t he measured lact at e may be spuriously elevat ed. As t he blood remains w it hin
t he t ube, t he red cells cont inue t heir met abolism, using t he oxygen dissolved
w it hin t he serum. As t his is deplet ed, t he red cells convert t o anaerobic
met abolism, yielding a f urt her elevat ion in lact at e.
Suggested Readings
Marino PL, ed. The I CU Book. 2nd ed. Philadelphia: Lippincot t Williams &
Wilkins; 1998: 196–198, 230–231, 592–600.
Souba WW, Fink MP, Jurkovich G J, eds. American College of Surgeons

Principles and Pract ice. New York: WebMD; 2004: 1171–1178.
> Table of C ontents > Labor ator y > 218 - D o not Tr eat Lac tic Ac idos is w ith B ic ar bonate
218
Do not Treat Lactic Acidosis with Bicarbonate
Lact ic acidosis is one t ype of anion-gap met abolic acidosis. I t occurs w hen
aerobic t issue met abolism convert s t o anaerobic met abolism, w hich produces a
signif icant amount of lact ic acid.
Lact ic acidosis occurs at t he cellular level (mit ochondria) w hen t he oxygen
delivery t o t he t issue is inadequat e or t he cells cannot ut ilize t he oxygen. Many
causes of lact ic acidosis are recognized, w it h t he most common being due t o low
oxygen delivery f rom shock (sept ic, hypovolemic, cardiogenic, neurogenic),
severe anemia, and severe hypoxemia. I t also can occur f rom exposure t o drugs
(e. g. , met f ormin, ant iret roviral t herapy, acet aminophen, salicylat es, cocaine,
valproic acid, salf asalazine, isonicot inic acid hydrazide, f luorouracil), t oxins, and
sugar alcohols (sorbit ol, xylit ol, f ruct ose).
Watch Out For

Many syst emic diseases can also be complicat ed w it h lact ic acidosis including
diabet es mellit us, alcoholism, pancreat it is, cancer, inf ect ions, vit amin B1
def iciency, short gut syndrome, and malabsorpt ion syndrome. Cert ain rare
congenit al inborn errors in met abolism (e. g. , Von G ierke disease, met hylmalonic
aciduria, pyruvat e dehydrogenase def iciency) can also be responsible f or lact ic
acidosis.
The best t reat ment of lact ic acidosis is t o t reat t he underlying causes. This of t en
includes volume expansion, correct ing anemia, increasing cardiac perf ormance,
and correct ing hypoxemia. I f t he cause is det ect ed and t reat ed prompt ly, t he
lact ic acidosis should gradually subside. The severit y of acidemia (art erial pH)
can be used t o f ollow t he eff ect iveness of t he int ervent ion.
Alkali t herapy (especially sodium bicarbonat e) should not be given in lact ic
acidosis. Alt hough it may init ially correct t he deranged lab value, it act ually
w orsens t he clinical condit ion. The exogenous bicarbonat e w ill result in a
signif icant amount of CO2 , w hich can move f reely bet w een t he compart ment s
(including across t he blood–brain barrier) w hile t he bicarbonat e does not . This
creat es disequilibrium of st rong ions. Many st udies bot h in animal and human
cells have conf irmed t he increase in int racellular acidosis by bicarbonat e
t herapy. Bicarbonat e administ rat ion is w idely accept ed in acidosis associat ed
w it h chronic loss of bicarbonat e (e. g. , renal t ubular acidosis, pancreat ic

t ransplant at ion, diarrhea) but not in lact ic acidosis.
O ne f inal t hought is t hat mild t o moderat e syst emic acidosis is not as delet erious
as once t hought . The grow ing use of permissive hypercapnia in pat ient s w it h
acut e respirat ory dist ress syndrome has show n t hat many adult pat ient s have no
change in cardiac cont ract ilit y even w hen syst emic pH is as low as 7. 15.
Suggested Readings
Forsyt he SM. Sodium bicarbonat e f or t he t reat ment of lact ic acidosis. Clinical
invest igat ions in crit ical care. Chest . 2000; 117: 260–267.
> Table of C ontents > Labor ator y > 219 - R em em ber that P os toper ative Hyper tens ion c an be a R es ult
of Inc r eas ed P C O 2
219
Remember that Postoperative Hypertension can
be a Result of Increased PCO2
Post operat ive hypert ension is common and can be diff icult t o manage. When
ot her vit al signs are normal, including normal pulse oximet ry and normal
respirat ory rat e, it is common t o assume t hat t he hypert ension is due t o poorly
t reat ed pain and should be t reat ed w it h opioids. This may be a dangerous
approach.
Hypovent ilat ion is a not uncommon post operat ive occurrence. Frequent causes of
hypovent ilat ion include residual eff ect s of inhaled anest het ics and opioids, as
w ell as persist ent neuromuscular blockade, subopt imal vent ilat ion due t o pain or
surgical sit e f act ors, and coexist ing chronic obst ruct ive pulmonary disease
(CO PD). As hypovent ilat ion w orsens, it can lead t o carbon dioxide ret ent ion and
hypercarbia. As hypercarbia w orsens, direct cardiac and vascular depression
occurs, and in order t o compensat e f or t his depression, hypercarbia can also
st imulat e cat echolamine release. There is a linear correlat ion bet w een increases
in PaCO2 and increases in heart rat e and cardiac out put . The overall eff ect of
t he cat echolamine release is hypert ension, along w it h increased pulse pressure,
st roke volume, myocardial cont ract ilit y, and heart rat e. Treat ing hypert ension
w it h opioids, t heref ore, may w orsen t he sit uat ion, as opioids may lead t o f urt her
hypovent ilat ion, hypercarbia, hypert ension, and event ually respirat ory
insuff iciency or f ailure.
What not to Do
I t cannot be overly st ressed t hat t he O 2 sat urat ion on t he pulse oximet er is not
an adequat e measure of vent ilat ory st at us. While t he pulse oximet er is
considered t he st andard of care in t he post operat ive set t ing and is able t o
noninvasively measure peripheral oxygen sat urat ion, it is an est imat e only of
art erial oxygenat ion. This is a nonlinear relat ionship, and as illust rat ed in t he
oxyhemoglobin dissociat ion curve, an oxygen sat urat ion of 90% t o 100% can
ref lect an ext remely w ide range of values f or t he PaO2 , f rom 60 mm Hg t o >100
mm Hg. The pulse oximet er is t heref ore a poor indicat or of vent ilat ory st at us, as
it can be misleading about t he PaO2 value and can be poorly ref lect ive of t he
underlying respirat ory eff ort , including t he PaCO2 . Furt her, as t he PaCO2
increases, it shif t s t he oxyhemoglobin dissociat ion curve t o t he right , ref lect ive
of increased oxygen unloading. I n managing
a post operat ive pat ient w it h hypert ension, t heref ore, it is import ant t o assess
t he f ull clinical pict ure f or ot her signs of respirat ory depression, including
dyspnea, shallow respirat ions, use of respirat ory muscles, paradoxical mot ions
of t he chest and abdomen, cyanosis, and obt undat ion, t o det ermine adequacy of
vent ilat ion, likelihood of hypovent ilat ion/ hypercarbia, and t he f ull diff erent ial f or
t he hypert ension.
Suggested Readings
Dubbink DA. Physiologic Eff ect s of Hyper- and Hypocarbia. I n Faust RJ, ed.
Anest hesiology Review. 3rd ed. Philadelphia: Churchill Livingst one, 2002: 26–
27.
Wilson WC and Benum of JL. Respirat ory Physiology and Respirat ory
Funct ion during Anest hesia. I n Miller RD, ed. Anest hesia. 6t h ed. Philadelphia:
Churchill Livingst one, 2005: 679–722.
Brian JE. Carbon dioxide and t he cerebral circulat ion. Anest hesiology.
1998: 88: 1365– 1386.
> Table of C ontents > Labor ator y > 220 - D o not us e a Nor m al Ar ter ial B lood Gas to r ule out a
P ulm onar y E m bolis m
220
Do not use a Normal Arterial Blood Gas to rule
out a Pulmonary Embolism
David N. Hager MD
Venous t hromboembolic disease aff ect s 200, 000 persons in t he Unit ed St at es
annually. Among aff ect ed persons, almost half suff er sympt omat ic pulmonary
embolism (PE). Wit hout t reat ment , 30% of pat ient s w ill die w it hin 1 year.
How ever, despit e it s high incidence, PE is a diff icult diagnosis t o make. Clinical
present at ions vary and t he sympt oms are of t en nonspecif ic. Though an abnormal
art erial blood gas can be inf ormat ive, normal indices of oxygenat ion and
vent ilat ion do not rule out PE.
Risk Factors
The major risk f act ors f or PE are t he same as t hose f or venous t hrombosis.
Broadly def ined, t hese include a prior t hromboembolic event , neoplast ic disease,
immobilit y, recent surgery or t rauma, est rogen replacement t herapy (especially
in t he cont ext of t obacco use), and a f amily hist ory of hypercoaguabilit y such as
act ivat ed prot ein C resist ance, hyperhomocyst einemia, prot eins C and S
def iciency, ant it hrombin I I I def iciency, and f act or V Leiden.
Signs and Symptoms

PE most commonly present s as a t riad of dyspnea (70% t o 80%), pleurisy, and
t achypnea (RR >20). Among t hose pat ient s enrolled and proven t o have PE in
t he Prospect ive I nvest igat ion of Pulmonary Embolism Diagnosis (PI O PED) st udy,
97% exhibit ed one of t hese t hree sympt oms. A less common present at ion is
hemo-dynamic compromise, w hich may be severe enough t o cause syncope. The
elderly of t en present w it h sympt oms consist ent w it h unresolving pneumonia or
heart f ailure despit e appropriat e t herapy. O n clinical exam an enhanced second
heart sound, t achycardia, and jugular venous dist ent ion may be appreciat ed. A
normal chest radiograph is consist ent w it h PE, as are ot her f indings such as
pleural eff usion, West ermark sign (a f ocal loss of peripheral vascular markings),
and even f ocal inf ilt rat es. Tachycardia and nonspecif ic ST-T w ave changes are
t he most common abnormal f inding on an elect rocardiogram (ECG ). The
f requent ly discussed S1 Q 3 T 3 pat t ern is present in less t han 12% of pat ient s w it h
PE.
Hypoxia (PaO2 <80 mm Hg), hypocapnia (PaCO2 <35 mm Hg), and an elevat ed
alveolar-art erial oxygen diff erence (P[ A–a] O 2 >20mm Hg) are t he most common
art erial blood gas abnormalit ies in pat ient s w it h PE. How ever, in a group of
subject s suspect ed of having PE, t hese indices did not eff ect ively discriminat e
bet w een t hose ult imat ely proven t o have PE and t hose w ho did not . Furt her,
among individuals w it hout prior lung disease, t hese indices w ill of t en f all w it hin
normal limit s even in t he presence of PE. For t hese reasons, art erial blood gas
dat a cont ribut e lit t le t o t he diagnosis or exclusion of PE.
Diagnosis
Because of t he nonspecif ic nat ure of t he clinical f indings associat ed w it h PE and
t he risks and limit ed availabilit y of invasive pulmonary angiography (t he def init ive
t est ), approaches t o diagnosis should include a combinat ion of (a) a clinical
est imat ion of t he likelihood of PE prior t o f urt her radiographic t est ing, and (b)
t he result s of f urt her radiographic t est ing. As demonst rat ed in t he PI O PED
st udy, t his combinat ion approach is more successf ul at diagnosing and excluding
PE t han clinical assessment or radiographic assessment alone. Though t he
int erpret at ion of vent ilat ion/ perf usion (V/ Q ) imaging and comput ed t omographic
(CT) angiography is beyond t he scope of t his chapt er, all clinicians should be
able t o det ermine if a pat ient 's clinical present at ion is consist ent w it h a high,
int ermediat e, or low probabilit y of PE.
Hi gh probabi l i ty is def ined as an 80% t o 100% chance of having had a PE.

Such pat ient s have (a) unexplained dyspnea, t achypnea, and/ or pleurisy, (b)
an unexplained gas exchange abnormalit y or x-ray abnormalit y, and (c) a risk
f act or f or hypercoaguabilit y.
Intermedi ate probabi l i ty is def ined as a 20% t o 79% chance of having had a
PE. Most simply, such pat ient s are def ined as not meet ing crit eria f or high or
low probabilit y.
Low probabi l i ty is def ined as a 1% t o 19% chance of having had a PE. Such
pat ient s may have dyspnea, t achypnea, pleurisy, x-ray abnormalit ies, or gas
exchange abnormalit ies. How ever, in t hese pat ient s, such f indings can be
at t ribut ed t o a diff erent condit ion. Typical risk f act ors f or hypercoaguabilit y
are not present .
The clinical impression can t hen be combined w it h t he radiologist 's int erpret at ion
of a V/ Q scan, CT angiogram, or bot h. Each radiographic t est has limit at ions.
V/ Q st udies are very diff icult t o
int erpret in t he cont ext of coexist ing lung disease such as pneumonia,
emphysema, or f ibrosis and t ake a long t ime t o complet e compared w it h a CT
angiogram. How ever, V/ Q has been more t horoughly evaluat ed and, in ot herw ise
normal lungs, is more sensit ive t o peripheral emboli. CT angiography can be
perf ormed quickly and usually det ect emboli in t he large cent ral and segment al
pulmonary art eries if present . How ever, some report s suggest t hat t he sensit ivit y
of CT angiography may be as low as 70%. Furt her, t he resolut ion of CT scanner
at diff erent cent ers varies w idely, int ravenous cont rast is required (w hich may
pose risks) and t he expert ise of radiologist s also varies.
These imaging limit at ions emphasize t he import ant role of t he clinical impression
in t he w orkup of a suspect ed PE. I f t he clinical impression and radiographic
assessment bot h suggest a low probabilit y of PE, t he evaluat ion can be st opped.
I f bot h suggest a high probabilit y of PE, ant icoagulat ion should be cont inued or
init iat ed. How ever, if bot h t he clinical impression and radiographic assessment
suggest int ermediat e probabilit y, or if one or t he ot her suggest s high probabilit y,
invasive angiography should be pursued. For t he pat ient in w hom t he clinical
probabilit y of PE is low and t he radiographic assessment suggest s int ermediat e
probabilit y (or vice versa), some invest igat ors have advocat ed t he use of low er-
ext remit y duplex ult rasound imaging t o rule out venous t hrombosis. I f such
t est ing is negat ive, ant icoagulat ion t herapy may be w it hheld w it hout f urt her
evaluat ion provided close f ollow -up is available.
D-Dimer
There has been much emphasis on t he use of d-dimer assays in t he evaluat ion of
deep venous t hrombosis and PE in t he last several years. I t is f requent ly
suggest ed t hat a posit ive d-dimer is nondiagnost ic, but t hat a negat ive result
rules out venous t hrombosis and/ or PE. How ever, st udies assessing t he ut ilit y of
d-dimer t est ing in t he diagnosis of t hromboembolic disease show t hat t he result s
of d-dimer t est s are subject t o not only t he t ype of t est used (lat ex agglut inat ion
or enzyme-linked immunosorbent assay [ ELI SA] ) but also t he prevalence of
t hromboembolic disease in t he st udy populat ion, pat ient age, lengt h of
hospit alizat ion, and even locat ion in t he hospit al (f loor vs. emergency
depart ment ). For t hese reasons, d-dimer t est ing remains of lit t le ut ilit y,
especially w hen t he clinical impression suggest s a high probabilit y of PE.
Suggested Readings
Hyers TM. Venous t hromboembolism. Am J Respir Crit Care Med.
1999; 159: 1–14.
Tapson VF, Carroll BA, Davidson BL, et al. The diagnost ic approach t o acut e
venous t hromboembolism. Am J Respir Crit Care Med. 1999; 160: 1043–1066.
Value of t he Vent ilat ion/ Perf usion Scan in Acut e Pulmonary Embolism. Result s
of t he Prospect ive I nvest igat ion of Pulmonary Embolism Diagnosis (PI O PED).
JAMA. 1990; 263: 2753–2759.
> Table of C ontents > Labor ator y > 221 - R em em ber that Ar gatr oban Inc r eas es Inter national
Nor m aliz ed R atio B ut does Not Affec t C oagulation in the E xtr ins ic S ys tem
221
Remember that Argatroban Increases
International Normalized Ratio But does Not
Affect Coagulation in the Extrinsic System
Kelly G rogan MD
Heparin-induced t hrombocyt openia (HI T) is an adverse drug react ion
charact erized by t hrombocyt openia and a high risk f or venous and art erial
t hrombosis. I t is def ined as a decrease in plat elet count during or short ly
f ollow ing exposure t o heparin. There are t w o classif icat ions of HI T: a benign
f orm, called t ype I , and an immune-mediat ed f orm, called t ype I I , w hich is
associat ed w it h an increased risk f or pot ent ially cat ast rophic t hrombosis. Type I
aff ect s 10% of pat ient s exposed t o heparin and is classif ied by a rapid, mild
decrease in plat elet count (100 t o 150, 000/ µL) in t he f irst 1 t o 4 days of
exposure t hat resolves despit e cont inued heparin use and is t hought t o be due t o
a direct int eract ion bet w een heparin and plat elet s. I t does not require any
medical t reat ment and does not appear t o have any major clinical consequences.
Type I I HI T, in pat ient s receiving heparin f or t he f irst t ime, present s as a more
severe t hrombocyt openia (80% w it h nadirs bet w een 20 and 100, 000/ µL) t hat
occurs bet w een 4 and 14 days af t er t herapy is init iat ed. Pat ient s w ho have
received heparin in t he past and have been sensit ized may show a decrease in
plat elet count t hat may occur w it hin hours and up t o 3 days af t er exposure. Af t er
discont inuat ion of t he heparin, t he plat elet count st art s t o rise af t er 2 t o 3 days
and usually ret urns t o normal w it hin 4 t o 10 days. The ant ibody disappears w it hin
2 t o 3 mont hs af t er cessat ion of heparin t herapy. Plat elet s of sensit ized pat ient s
are capable of aggregat ion and release react ions at 2 mont hs af t er recovery
w it hout evidence of circulat ing ant ibody in t he serum, suggest ing t hat a major
port ion of t he ant iplat elet ant ibody is membrane bound. Alt hough HI T does not
invariably recur during subsequent re-exposure t o heparin, f ut ure exposure t o
heparin must be approached w it h ext reme caut ion.
Treatment of HIT
I n pat ient s w ho have t est ed serologically posit ive or w ho are st rongly suspect ed
of having HI T based on clinical dat a, heparin should be st opped immediat ely and
an alt ernat ive f orm of ant it hrombot ic t herapy commenced. Tw o direct t hrombin
inhibit ors (DTI s), lepirudin
and argat roban, are approved in t he Unit ed St at es f or heparin-induced

t hrombocyt openia.
Di rect Thrombi n Inhi bi tors. Four parent eral DTI s have been approved by t he
Food and Drug Administ rat ion in Nort h America: lepirudin and argat roban f or t he
t reat ment of heparin-induced t hrombocyt openia, bivalirudin as an alt ernat ive t o
heparin in percut aneous coronary int ervent ion, and desirudin as prophylaxis
against venous t hromboembolism in hip replacement .
To brief ly review, t hrombin is cent ral in t he clot t ing process: it covert s soluble
f ibrinogen t o f ibrin; act ivat es f act ors V, VI I I , and XI , w hich generat es more
t hrombin; and st imulat es plat elet s. Furt hermore, by act ivat ing f act or XI I I ,
t hrombin f avors t he f ormat ion of cross-linked bonds among t he f ibrin molecules,
st abilizing t he clot . The coagulat ion cascade is regulat ed by nat ural
ant icoagulant s, such as t issue f act or pat hw ay inhibit or, t he prot ein C and prot ein
S syst em, and ant it hrombin.
The DTI s also have an ant iplat elet eff ect secondary t o t heir eff ect of reducing
t he t hrombin-mediat ed act ivat ion of plat elet s. Since DTI s do not bind t o plasma
prot eins, t hese agent s should produce a more predict able response t han does
unf ract ionat ed heparin and should be more eff ect ive t han low -molecular-w eight
heparins because t hey inhibit f ibrin-bound t hrombin.
Argatroban. Ant ihirudin ant ibodies develop in 40% t o 74% of pat ient s receiving
lepirudin af t er 4 days or more of t reat ment . O f not e, f at al anaphylaxis has been
described w it h lepirudin, part icularly in pat ient s w ho are t reat ed again w it hin 3
mont hs of a previous exposure t o t his agent . I n cont rast , argat roban does not
appear t o be immunogenic.
Argat roban is a small-molecule direct t hrombin inhibit or t hat , unlike lepirudin, is
not immunogenic. I t is a univalent compet it ive inhibit or of t hrombin and binds only
t o t he cat alyt ic sit e of t hrombin via a noncovalent bond t o f orm a reversible
complex. Argat roban has a plasma half -lif e of 39 t o 52 minut es and is ext ensively
met abolized by t he liver int o f our most ly inact ive met abolit es. Renal dysf unct ion,
age, and gender do not alt er t he eliminat ion half -lif e of t he drug.
Current ly argat roban is approved in t he Unit ed St at es f or t he prophylaxis and
t reat ment of t hrombosis in pat ient s w it h HI T. The recent recommended dose of
argat roban in t he t reat ment of HI T f or pat ient s in t he int ensive care unit (I CU) is
0. 5 µg/ kg/ min and t he dose should be adjust ed t o maint ain t he act ivat ed part ial
t hromboplast in t ime (aPTT) at 1. 5 t o 3. 0 t imes t he pat ient 's baseline, w it h t he
maximum inf usion rat e at 10 µg/ kg/ min. A general guideline is t o
increase t he inf usion rat e by 20% and recheck t he aPTT in 4 hours. I f
suprat herapeut ic, it is recommended t o decrease t he inf usion rat e by 50% and
recheck t he aPTT in 4 hours. Argat roban increases t he a PTT and act ivat ed
clot t ing t ime (ACT) in a dose-dependent manner.
Moni tori ng Argatroban Cl i ni cal Acti vi ty. The best met hod t o monit or t herapy w it h
DTI s has not been clearly est ablished. The ant icoagulat ion eff ect of direct
t hrombin inhibit ion is t ypically monit ored using t he act ivat ed part ial
t hromboplast in t ime. Alt hough dose-dependent increases also occur in t he
prot hrombin t ime (PT) and int ernat ional normalized rat io (I NR), t hese are not
dependable markers of act ivit y. I t must be st ressed t hat an elevat ed PT does not
mean t here is a clinically signif icant alt erat ion in t he ext rinsic pat hw ay. The
ecarin clot t ing t ime bet t er ref lect s t he act ual plasma concent rat ion of DTI s, but
t his t est is not w idely available. Recombinant hirudins and argat roban can be
monit ored w it h t he use of t he aPTT and bivalirudin w it h t he ACT.
Since direct t hrombin inhibit ors also increase t he PT and I NR, a reliable means
t o monit or w arf arin t herapy during coadminist rat ion of a direct t hrombin inhibit or
is import ant . Eff ort s t o charact erize argat roban's eff ect on t he I NR during
concurrent w arf arin t herapy have been report ed. I n one st udy, argat roban 1 t o 2
µg/ kg/ min and w arf arin 2. 5 t o 5. 0 mg/ day w ere administ ered concurrent ly t o
healt hy subject s over 6 days. Because of it s short half -lif e, argat roban w as
t emporarily discont inued daily t o simulat e w arf arin monot herapy. Wit hin 4 hours
of st opping t he argat roban, mean I NRs decreased almost t w of old alt hough mean
f unct ional f act or X levels remained unchanged. Subsequent in vit ro st udies using
plasma f rom w arf arin-t reat ed pat ient s demonst rat ed a linear relat ionship
bet w een t he I NR in t he presence versus absence of exogenous argat roban.
This underscores t he import ance of post poning w arf arin init iat ion pending
subst ant ial resolut ion of HI T (plat elet s great er t han 100, 000/ dL) t o avoid
w arf arin-induced microvascular t hrombosis. This complicat ion appears t o be
caused by deplet ion of t he vit amin K–dependent nat ural ant icoagulant , prot ein C.
This complicat ion is usually seen w hen t he I NR rises above 3. 5 as t his
represent s a surrogat e marker f or prot ein C deplet ion.
A chromogenic-based met hod can be used t o det ermine f act or X levels as a
monit or of t he oral ant icoagulat ion f rom w arf arin w it hout eff ect f rom argat roban.
Also, pat ient s receiving argat roban cannot be evaluat ed f or pot ent ial coagulat ion
abnormalit ies w it h rout ine f unct ional (clot -based) assays f or f ibrinogen, f act or
levels, or prot ein C. Argat roban act s as an inhibit or in t hese assays, causing a
dose-dependent
f alse decrease of f ibrinogen and f act or levels, and a f alse increase of prot ein C.
Suggested Readings
Di Nisio M, Middeldorp S, Buller HR. Drug t herapy: direct t hrombin inhibit ors.
N Engl J Med. 2005; 353: 1028–1040.
Schmit t BP, Adelman B. Heparin-associat ed t hrombocyt openia: a crit ical

review and pooled analysis. Am J Med Sci. 1993; 305: 208–215.
Warkent in TE, Heddle NM. Laborat ory diagnosis of immune heparin-induced

t hrombocyt openia. Curr Hemat ol Rep. 2003; 2: 148–157.
> Table of C ontents > Labor ator y > 222 - C hec k S er ial Methem oglobin Levels in P atients on Inhaled
Nitr ic O xide
222
Check Serial Methemoglobin Levels in Patients
on Inhaled Nitric Oxide
T imothy M. Moore MD, PHD
I nhaled nit ric oxide acut ely and locally relaxes const rict ed pulmonary vascular
smoot h muscle, result ing in decreased pulmonary vascular resist ance, but it
generally does not cause hemodynamic changes out side t he lung. Addit ionally,
inhaled nit ric oxide may improve art erial oxygenat ion in hypoxemic pat ient s by
reducing int rapulmonary shunt and improving vent ilat ion-perf usion mat ching. I n
bot h pediat ric and adult pat ient s, inhaled nit ric oxide has been st udied and used
as a t reat ment modalit y in t he int ensive care unit (I CU) set t ing f or a large
number of clinical condit ions in w hich reduct ion of pulmonary vascular resist ance
and/ or improvement in art erial oxygenat ion is desired, including persist ent
pulmonary hypert ension of t he new born; congenit al diaphragmat ic hernia;
premat ure inf ant respirat ory f ailure; respirat ory dist ress syndrome (adult s and
children); cyst ic f ibrosis; sickle cell disease; primary pulmonary hypert ension;
post operat ive congenit al heart disease; one-lung vent ilat ion during t horacic
surgery; post pneumonect omy; lung t ransplant at ion ischemia/ reperf usion injury;
and perioperat ive pulmonary hypert ension associat ed w it h cardiac
t ransplant at ion and lef t vent ricular assist device implant at ion. Mult iple out come
paramet ers have been measured t o show inhaled nit ric oxide t reat ment eff icacy
f or t he various clinical condit ions list ed above, but result s have varied, yielding
bot h posit ive and negat ive f indings. How ever, it is saf e t o say t hat t here is
current ly no evidence t hat inhaled nit ric oxide t herapy improves long-t erm
morbidit y or mort alit y rat es f or any clinical condit ion st udied, and any short -t erm
benef it s of inst it ut ing inhaled nit ric oxide t herapy must be w eighed against t he
know n acut e and pot ent ial long-t erm t oxic eff ect s of inhaled nit ric oxide. The
current Food and Drug Administ rat ion (FDA)–approved use f or inhaled nit ric
oxide (market ed as I NO max) is f or t he t reat ment of t erm and near-t erm (>34
w eeks) neonat es w it h hypoxic respirat ory f ailure associat ed w it h clinical or
echocardiographic evidence of pulmonary hypert ension, w here it improves
oxygenat ion and reduces t he need f or ext racorporeal membrane oxygenat ion.
Thus, all ot her off -label indicat ions f or inhaled nit ric oxide are regulat ed at t he
inst it ut ional level.
Because of t he short half -lif e of nit ric oxide, sust ained pulmonary vasodilat at ion
requires cont inuous delivery of inhaled nit ric oxide via
commercially available syst ems t hat accurat ely deliver inspired concent rat ions
bet w een 1 and 80 part s per million (ppm). These syst ems must be able t o
deliver a const ant concent rat ion of inhaled nit ric oxide t o pat ient s w hile
minimizing t he generat ion of nit rogen dioxide (NO2 ), a direct t oxin and
environment al pollut ant , as w ell as cont inuously monit or inspired NO , NO2 , and
O 2 concent rat ions. Clinically signif icant levels of NO2 are unlikely t o occur w hen
inhaled nit ric oxide is delivered by an eff icient delivery syst em at concent rat ions
of 20 ppm or less. I nhaled NO is usually delivered during mechanical vent ilat ion
int o t he inspirat ory limb of t he vent ilat ory circuit by eit her pulse or cont inuous
modes, alt hough it may be administ ered t o spont aneously breat hing pat ient s w it h
a close-f it t ing mask.
What to Do
FDA-recommended dosing f or inhaled nit ric oxide is 20 ppm f or neonat es, w it h
t reat ment durat ion as clinically indicat ed f or up t o 14 days. For ot her uses,
dosing is usually t it rat ed t o t he desired pulmonary syst olic and diast olic
pressures (assessed w it h a pulmonary art erial cat het er) and/ or art erial blood
oxygenat ion (measured by cont inuous O 2 sat urat ion and/ or art erial blood gases).
There is lit t le evidence t o suggest t hat concent rat ions of 40 ppm or less are
t oxic even f or ext ended periods, but it is import ant t o not e t hat nont it rat ed
w eaning off or sudden w it hdraw al of inhaled nit ric oxide t herapy has been show n
t o produce severe, lif e-t hreat ening rebound hypoxemia and pulmonary
hypert ension. Addit ional complicat ions or side eff ect s of inhaled nit ric oxide
t herapy may include peroxynit rit e (O NO O- ) f ormat ion leading t o cellular
membrane lipid peroxidat ion, alt ered pulmonary inf lammat ory responses, alt ered
surf act ant f unct ion, genot oxic alt erat ions, inhibit ion of plat elet f unct ion, and
alt erat ion of iron- and heme-based prot ein f unct ion such as cyclo-oxygenase and
cyt ochromes.
Watch Out For

Met hemoglobinemia is a pot ent ially very dangerous complicat ion of inhaled nit ric
oxide t reat ment , and inhaled nit ric oxide is cont raindicat ed in pat ient s w it h
congenit al or acquired met hemoglobin reduct ase def iciency. The danger w it h
met hemoglobinemia lies in it s insidious nat ure of development , w it h development
and progression occurring despit e normal art erial blood gas measures. I t is a
f unct ional anemia w here nit rat es accumulat e during inhaled nit ric oxide t herapy
and lead t o met hemoglobin f ormat ion, charact erized by oxidat ion of f errous iron
t o f erric iron, w hich in t urn prevent s O2 release int o t he t issues (i. e. , shif t s t he
oxyhemoglobin dissociat ion curve t o t he lef t ).
Hypoxia secondary t o met hemoglobinemia is t heref ore ref ract ory t o correct ion by
O 2 t herapy and if not det ect ed and t reat ed can be rapidly f at al. Normal
met hemoglobin, report ed as a f ract ion of normal hemoglobin, is <2%. Cyanosis
is appreciat ed w hen met hemoglobin levels approach 15% t o 20%, w it h
progressive, mult iorgan hypoxic sympt oms developing w it h increasing
met hemoglobin f ract ions. Deat h occurs w hen met hemoglobin levels reach 70%.
How ever, in t he crit ically ill, vent ilat ed pat ient receiving inhaled nit ric oxide, it is
diff icult t o say w hat level of met hemoglobin correlat es w it h w hat degree of
t issue hypoxia, since t he pat ient may be also be in a hypoperf used st at e and
more severe t issue hypoxia could result at low er met hemoglobin f ract ions.
Thus, inst it ut ion of inhaled nit ric oxide t herapy requires concomit ant
met hemoglobin monit oring. Clinically signif icant levels of met hemoglobin are
unlikely t o result unless inhaled nit ric oxide concent rat ions over 20 ppm are
administ ered, but met hemoglobin should be measured every 4 hours af t er
inst it ut ing inhaled nit ric oxide delivery and daily t hereaf t er, since inadvert ent ly
high NO concent rat ions may be delivered. High met hemoglobin levels (>5%) may
provoke changes in dosing, but t he overall risks and benef it s of alt ering dosing
must be considered given t he overall clinical pict ure, part icularly in light of t he
know n possible det riment al eff ect of rapid off w eaning of inhaled nit ric oxide.
Decreasing dosing alone w ill not rapidly correct met hemoglobinemia. High levels
alone may w arrant t reat ment , w hereas increased levels plus addit ional clinical
f indings of t issue hypoxia such as elect rocardiogram (ECG ) changes, lact ic
acidosis, change in neuro exam, and so on mandat e t reat ment . The st andard
t reat ment f or met hemoglobinemia is int ravenous (I V) inf usion of met hylene blue
at 1 t o 2 mg/ kg of a 1% solut ion over 5 minut es. The dose may need t o be
repeat ed but should not exceed 1. 5 t imes t he init ial dose. Met hylene blue w ill
rapidly convert f erric iron back t o it s f errous st at e, w hich w ill ref acilit at e O2 off
loading int o t he t issues. Ascorbic acid t reat ment may be used f or pat ient s w ho
are resist ant t o met hylene blue.
Suggested Readings
G ermann P, Braschi A, Della Rocca G , et al. I nhaled nit ric oxide t herapy in
adult s: European expert recommendat ions. I nt ens Care Med. 2005; 31: 1029–
1041.
G riff it hs MJD, Evans TW. I nhaled nit ric oxide t herapy in adult s. N Engl J Med.
2005; 353(25): 2683–2695.
Haddad E, Low son SM, Johns RA, et al. Use of inhaled nit ric oxide
perioperat ively and in int ensive care pat ient s. Anest hesiology.
2000; 92(6): 1821–1825.
Maimo G , Redick E. Recognizing and t reat ing met hemoglobinemia. A rare but
dangerous complicat ion of t opical anest het ic or nit rat e overdose. Dimens Crit
Care Nurs. 2004; 23(3): 116–118.
Weinberger B, Laskin DL, Heck DE, et al. The t oxicology of inhaled nit ric
oxide. Toxicol Sci. 2001; 59: 5–16.
> Table of C ontents > Labor ator y > 223 - R em em ber that Tr oponin Levels ar e Inac c ur ate as a Meas ur e
of C ar diac D am age in R enal Ins uffic ienc y
223
Remember that Troponin Levels are Inaccurate
as a Measure of Cardiac Damage in Renal
Insufficiency
Bradford D. Winters MD, PHD
The measurement of cardiac t roponins has become a commonly used
biochemical marker f or myocardial injury and inf arct ion. Unf ort unat ely, t here are
several f act ors t hat may make t he int erpret at ion of t roponin values diff icult ,
including spurious elevat ions t hat are not indicat ive of myocardial injury af t er
f urt her evaluat ion. O ne of t he major clinical conf ounders, part icularly in crit ically
ill pat ient s, is renal insuff iciency. Pat ient s w it h chronic renal insuff iciency are
know n t o be at higher risk f or developing accelerat ed at herosclerot ic coronary
art ery disease (CAD), so t he abilit y t o int erpret biochemical markers f or
myocardial disease in t his populat ion is of great import ance. For pat ient s w it h
acut e renal insuff iciency or f ailure, t here is no similar increment al risk of CAD
direct ly at t ribut able t o t he renal f ailure, but myocardial event s are nevert heless
common in t hese pat ient s.
Troponins are prot ein molecules t hat part icipat e in t he regulat ion of cont ract ile
f unct ion of t he myocardial sarcomeres and are released int o t he general
circulat ion w hen myocardial injury occurs much as creat inine kinase–MB (CK-MB)
is under similar condit ions. There are t w o t roponins: I and T. Elevat ed t roponin T
has been show n t o correlat e w it h an increased risk of f ut ure acut e coronary
syndromes. For t roponin I , t his relat ionship is not as clear, w it h t he st udy result s
being cont radict ory. Some hospit als measure and report t roponin I , but many
hospit al laborat ories report t roponins as a combined value. The problem w it h t his
is t hat t roponin T has been show n t o lack specif icit y f or making t he diagnosis of
an acut e coronary event in pat ient s w it h end-st age renal disease. The exact
values f or t he upper normal limit of t roponins may vary f rom hospit al lab t o
hospit al lab based on t he assay used. Where t he cut off is set f or any part icular
lab t est w ill det ermine it s specif icit y and sensit ivit y and should be opt imized
using receiver operat or curve (RO C) analysis. The preponderance of dat a
suggest s t hat higher t hresholds are required f or using cardiac t roponins t o def ine
out comes f or renal pat ient s suspect ed of having an acut e coronary syndrome.
RO C analysis suggest s t hat f or nonrenal pat ient s, t he upper limit should be
approximat ely 0. 1 µg/ L w hile f or renal pat ient s t he upper limit should be
approximat ely 0. 5 µg/ L, t hough t hese values are not absolut e. I n comparison,
most of t he dat a on CK-MB suggest
t hat values f or t his analyt e do not need t o be signif icant ly adjust ed f or renal
pat ient s.
Since t roponin levels rise over a 6- t o 8-hour period post -myocardial injury and
remain elevat ed f or up t o 10 days, it is imperat ive t o f ollow t he values
sequent ially. I f t here is a def init e rise and t hen f all over t his t ime f rame, t hen an
acut e myocardial event has likely occurred especially if in cont ext w it h a
suspicious clinical scenario. CK-MB values generally rise more slow ly but f all
more quickly. These are usually measured on a q8h basis f rom t he t ime of t he
suspect ed event f or a period of 24 hours w it h t he second and t hird values being
most predict ive. The CK-MB f ract ion is usually report ed as a percent of t he t ot al
CK, w it h values of great er t han 3% being indicat ive of myocardial injury
assuming t he t ot al CK is great er t han approximat ely 200. Percent ages higher
t han 3 in f ace of low t ot al CK are generally considered spurious.
The evidence in support of requiring higher t roponin t hresholds f or pat ient s w it h
renal disease f or diagnosing and prognost icat ing acut e coronary syndromes is
primarily based on pat ient s w it h end-st age renal disease on chronic dialysis
t herapy. While t here are limit ed dat a on crit ically ill pat ient s, it seems
reasonable t o consider int ensive care unit (I CU) pat ient s w ho are on dialysis f or
renal f ailure similar t o t he chronic renal f ailure pat ient in t erms of t he use of
t roponins as biomarkers t hough t here is evidence t hat t he longer a pat ient is on
dialysis, t he higher his or her t roponins may be. The great er diff icult y comes in
int erpret ing t roponins in crit ically ill pat ient s w it h lesser degrees of renal
insuff iciency, some nat ive glomerular f ilt rat ion, and w ho are not on dialysis. I n
t his sit uat ion, t here are virt ually no dat a and even less consensus on t hresholds
despit e t he concern t hat t roponins may be f alsely elevat ed even w hen t he renal
f ailure is incomplet e. Reliance on t he clinical scenario and ot her t est s including
CK-MB and echocardiography are necessary in t his sit uat ion. G iven t hat
t hresholds and assays vary f rom inst it ut ion t o inst it ut ion and some use t roponin I
and/ or t roponin T as t he analyt e in t heir lab t est , clinicians should f amiliarize
t hemselves w it h t he ref erence values at t heir hospit al and w het her allow ances
are made f or renal insuff iciency.
A f inal not e is t hat a myriad of commercial assays are available t o measure
t roponin level. Up t o a 20-f old variat ion in t roponin values may be seen
depending on t he assay used.
Suggested Readings
Buet i J, Krahn J, Karpinski M, et al. Troponin I t est ing in dialysis pat ient s
present ing t o t he emergency room: does t roponin I predict t he 30 day
out come? Nephron Clin Pract . 2006; 103: 129–136.
Kanderian AS, Francis G S. Cardiac t roponins and chronic kidney disease.

Kidney I nt . 2006; 69: 1112–1114.
Van Lent e F, McErlean ES, DeLuca SA, et al. Abilit y of t roponins t o predict
adverse out comes in pat ient s w it h renal insuff iciency and suspect ed acut e
coronary syndromes: a case-mat ched st udy. J Am Coll Cardiol. 1999; 33: 471–
478.
> Table of C ontents > Labor ator y > 224 - As k the Labor ator y for a S yner gy P anel in R es is tant
P s eudom onas Infec tions
224
Ask the Laboratory for a Synergy Panel in
Resistant Pseudomonas Infections
Harjot K. Singh MD
Sara E. Cosgrove MD
The use of combinat ion t herapy (double coverage) in t he management of G ram-
negat ive inf ect ions is a cont roversial pract ice. Reasons t o consider combinat ion
t herapy include (a) broadening empiric coverage in t he event t hat t he causat ive
organism is resist ant t o one agent , (b) prevent ion of emergence of resist ance
during t herapy, and (c) obt aining synergist ic act ivit y bet w een t w o agent s t hat is
great er t han w ould be expect ed f rom t he sum of t he act ivit ies of t he individual
agent s.
Broadening empiric coverage should be considered in pat ient s w ho are crit ically
ill as dat a suggest t hat inappropriat e empiric t herapy can lead t o adverse pat ient
out comes. Agent s f rom t w o diff erent classes of ant ibiot ics should be used.
Review of t he local ant ibiogramis import ant t o assess w het her combinat ion
t herapy is needed and w hat agent s should be used. For example, if t he usual
empiric t herapy regimen f or a penicillin-allergic pat ient is a f luoroquinolone, but
40% of Pseudomonas isolat es in a unit are know n t o be resist ant t o
f luoroquinolones, addit ion of a second agent such as an aminoglycoside may
eff ect ively broaden empiric coverage if Pseudomonas is suspect ed. I n all cases
in w hich combinat ion t herapy is chosen f or empiric coverage, t herapy should be
narrow ed on t he basis of microbiologic dat a.
What to Do
Prevent ion of emergence of resist ance is of t en cit ed as a reason f or combinat ion
t herapy in t he management of def init ive inf ect ions w it h G ram-negat ive
organisms; how ever, emergence of resist ance on t herapy is uncommon,
occurring in 5% t o 10% of G ram-negat ive inf ect ions. I n most circumst ances,
long-t erm combinat ion t herapy is not needed and may be harmf ul as it increases
t oxicit y f rom adverse drug react ions. I t can be considered in t he t reat ment of a
deep-seat ed inf ect ion (such as pneumonia or severe ost eomyelit is) w hen t he
organism being t reat ed is already resist ant t o one or more classes of ant ibiot ics.
Discont inuing one of t he agent s af t er 5 t o 7 days of t herapy w hen t he bact erial
burden has decreased is an opt ion. I n all cases, at t ent ion should be given t o
appropriat e dosing of ant ibiot ics and t o obt aining f ollow -up cult ures if t he pat ient
is f ailing t herapy. O f not e is t hat w hen organisms t hat produce inducible
bet a-lact amases (e. g. , Enterobacter, Serrati a, Ci trobacter) are t reat ed w it h

bet a-lact am agent s, emergence of resist ance on t herapy can occur in ~20% of
cases. Monot herapy w it h bet a-lact am agent s is best avoided f or t reat ment of
serious inf ect ions in t hese pat ient s if ot her agent s are available.
Alt hough in vit ro synergy exist s bet w een ant ipseudomonal penicillins and
aminoglycosides f or Pseudomonas aerugi nosa and ot her G ram-negat ive rods, no
def init ive evidence exist s t hat combining t hese agent s in humans leads t o
enhanced out comes. A st udy by Hilf et al. in 1989 suggest ed t hat combinat ion
t herapy w as superior t o monot herapy in pat ient s w it h Pseudomonas bact eremia,
but 84% of pat ient s receiving monot herapy received only an aminoglycoside.
Aminoglycosides should not be used as monot herapy f or inf ect ions caused by
Pseudomonas except f or uncomplicat ed urinary t ract inf ect ions. More recent
st udies have not show n a diff erence in mort alit y w hen pat ient s received
appropriat e monot herapy f or bact eremia due t o G ram-negat ive rods, including
Pseudomonas; t hese st udies have been summarized in a recent met a-analysis
examining pat ient s w it h sepsis due t o G ram-negat ive bact eria.
Synergy bet w een drug combinat ions ot her t han bet a-lact am and aminoglycosides
is less predict able and has unclear clinical signif icance. Synergy t est ing w it h t w o
ant ibiot ics is of signif icant value only w hen t reat ing a highly resist ant G ram-
negat ive organism t hat is resist ant t o one or bot h of t he drugs in t he
combinat ion. Formal synergy t est ing is recommended if using combinat ion
t herapy in t his set t ing, alt hough such t est ing is not w it hin t he purview of many
microbiology laborat ories.
Suggested Readings
Comber KR, Basker MJ, O sborne CD, et al. Synergy bet w een t icarcillin and
t obramycin against Pseudomonas aerugi nosa and Ent erobact eriaceae in vit ro
and in vivo. Ant imicrob Agent s Chemot her. 1977; 11: 956–964.
Fish DN, Piscit elli SC, Danziger LH. Development of resist ance during
ant imicrobial t herapy: a review of ant ibiot ic classes and pat ient
charact erist ics in 173 st udies. Pharmacot herapy. 1995; 15: 279–291.
Hilf M, Yu VL, Sharp J, et al. Ant ibiot ic t herapy f or Pseudomonas aerugi nosa
bact eremia: out come correlat ions in a prospect ive st udy of 200 pat ient s. AmJ
Med. 1989; 87: 540–546.
I brahim EH, Sherman G , Ward S, et al. The inf luence of inadequat e
ant imicrobial t reat ment of bloodst ream inf ect ions on pat ient out comes in t he
I CU set t ing. Chest . 2000; 118: 146–155.
Kaye KS, Cosgrove S, Harris A, et al. Risk f act ors f or emergence of

resist ance t o broadspect rum cephalosporins among Enterobacter spp.
Ant imicrob Agent s Chemot her. 2001; 45: 2628–2630.
Paul M, Benuri-Silbiger I , Soares-Weiser K, et al. Bet a lact am monot herapy

versus bet a lact am-aminoglycoside combinat ion t herapy f or sepsis in
immunocompet ent pat ient s: syst emat ic review and met a-analysis of
randomised t rials. BMJ. 2004; 328: 668–672.
> Table of C ontents > Labor ator y > 225 - O btain B lood Us ed for Mixed Venous O xygen Tes ting Fr om
the D is tal P ulm onar y Ar ter y C atheter P or t
225
Obtain Blood Used for Mixed Venous Oxygen
Testing From the Distal Pulmonary Artery
Catheter Port
Analysis of mixed venous blood provides valuable inf ormat ion in evaluat ing t he
oxygen supply-demand axis. Mixed venous oxygen sat urat ion (SvO2 ) is t he O2
sat urat ion of blood in t he pulmonary art ery af t er t he venous eff luent f rom various
organs has mixed t horoughly in t he right vent ricle (normal SvO2 = 70% t o 80%,
normal mixed venous oxygen pressure [ PvO2 ] =35 t o 45 mm Hg). Normally, blood
f rom t he inf erior vena cava is more f ully sat urat ed (5% t o 7%) t han blood f rom
t he superior vena cava due t o highly sat urat ed blood f rom t he renal vein ent ering
t he inf erior vena cava, but t he reverse is t rue in shock st at es because of
redist ribut ion of f low aw ay f rom t he splanchnic, renal, and mesent eric beds.
The SvO2 can be measured int ermit t ent ly by slow ly w it hdraw ing a sample of
blood f rom t he dist al port of t he unw edged pulmonary art ery cat het er or
cont inuously w it h a f iberopt ic pulmonary art ery cat het er t hat measures O2
sat urat ion by ref lect ance oximet ry. I nt ermit t ent sampling of SvO2 is
accomplished by discarding t he init ial 3 mL of blood and t hen w it hdraw ing a
sample very sl owl y in order t o avoid cont aminat ion w it h capillary blood, w hich
may art if act ually increase t he oxygen cont ent .
The value of mixed venous blood analysis is best underst ood in t he f ramew ork of
t issue O2 delivery-supply dynamics. O xygen delivery is t he product of cardiac
out put and art erial O2 cont ent , t he lat t er being det ermined by t he hemoglobin
and art erial O2 sat urat ion (SaO2 ). Each organ receives a variable percent age of
t he t ot al amount , a f low t hat may be luxuriant , just adequat e, or insuff icient t o
sat isf y t he aerobic met abolism demand. The O2 t ension (PvO2 ) and sat urat ion of
t he venous eff luent ref lect t he balance bet w een supply and demand. Normally,
peripheral oxygen consumpt ion is independent of t he oxygen delivery. Theref ore,
as cardiac out put and oxygen delivery decline, peripheral ext ract ion increases t o
keep consumpt ion const ant . This result s in decreased mixed venous oxygen
sat urat ion. Under normal condit ions, t he SaO2 -SvO 2 diff erence is 20% t o 25%,
yielding an SvO2 of 65% t o 75% w hen art erial blood is w ell oxygenat ed.
I ncreased SvO2 is seen in a variet y of condit ions. I n sepsis, SvO2 is of t en
normal, but in some cases t here is ext reme peripheral vasodilat at ion, and
cardiac
out put increases disproport ionat ely t o met abolic demands, result ing in increased
SvO 2 . Cirrhosis is one of t he more common causes of marked increase in SvO2 ,
w it h values usually >85%. A number of vasodilat ing agent s and lef t -t o-right
shunt s (eit her int racardiac or peripheral) also result in increased SvO2 . Wedging
of t he cat het er and mit ral regurgit at ion (w hich t end t o bring t he cat het er t ip int o
cont act w it h art erialized blood) are addit ional f act ors t hat raise t he SvO2 .
Finally, agent s t hat int erf ere w it h mit ochondrial cyt ochrome act ivit y (e. g. ,
cyanide) may produce an elevat ed SvO2 value ow ing t o inabilit y of t he t issues t o
ut ilize oxygen.
The cont inuous measurement of O2 sat urat ion in mixed venous blood is
perf ormed w it h specialized pulmonary art ery cat het ers (more expensive t han
regular ones) equipped w it h f iberopt ic bundles t hat can t ransmit light t o and f rom
t he cat het er t ip. The opt ical det ect ion of SvO2 is perf ormed by ref lect ion
spect rophot omet ry. Wavelengt hs of light similar t o t hose used in pulse oximet ry
are passed along t he f iberopt ic bundles in t he pulmonary art ery cat het er and out
f rom t he cat het er t ip. The light beam is t ransmit t ed t hrough t he circulat ing blood
and t he light t hat comes in cont act w it h hemoglobin in t he circulat ing
eryt hrocyt es is ref lect ed back t o t he cat het er t ip. This light is t hen t ransmit t ed
back t hrough t he cat het er t o a phot odet ect or and microprocessor t hat record t he
average SvO2 at 5-second int ervals. A >5% variat ion in SvO2 t hat persist s f or
longer t han 10 minut es is considered a signif icant change. Perhaps t he most
rat ional use of t he f iberopt ic pulmonary art ery cat het er is in t he management of
pat ient s w it h cardiac dysf unct ion, in w hom rapid changes in t he cardiac out put
maybe expect ed t o occur on t he basis of t he underlying disease process or in
response t o a t herapeut ic int ervent ion.
Watch Out For

I t must be st ressed t hat a normal or high SvO2 does not necessarily mean
adequat e oxygen delivery (as in sepsis), and t he need t o f urt her augment
cardiac out put in t his set t ing is cont roversial. A normal SvO2 may be associat ed
w it h diff erent levels of cardiac out put depending on t he underlying aerobic
met abolism. I n essence, t he SvO2 helps us def ine t he appropriat eness of
measured cardiac out put . I n one recent st udy, t he mixed venous oxygen
sat urat ion correlat ed w ell w it h decreased morbidit y af t er cardiac surgery w hen
t herapy w as t arget ed t o keep SvO2 >70% and a lact at e level <2 mg/ dL.
How ever, no benef it w as show n in anot her st udy t hat involved a more
het erogeneous group of crit ically ill pat ient s w hen goal-direct ed hemodynamic
t herapy w as based on SvO2 .
Determinants of Svo2
S[ v w it h bar above] O2 = SaO2 - (Vo2 / Q × Hb × 13)
VO 2 is oxygen consumpt ion; Q is cardiac out put ; Hb is hemoglobin. (Low SvO2
<60% indicat es an abnormalit y of one or more of t he f act ors on t he right -hand
side of t he above equat ion. )
Suggested Readings
G at t inoni L, Brazzi L, Pelosi P, et al. A t rial of goal-orient ed hemodynamic
t herapy in crit ically-ill pat ient s: SvO2 Collaborat ive G roup. N Engl J Med.
1995; 333: 1025.
Polonen P, Ruokonen E, Hippelainen M, et al. A prospect ive, randomized

st udy of goal-orient ed hemodynamic t herapy in cardiac surgical pat ient s.
Anest h Analg. 2000; 90: 1052.
Vaugh S, Puri VK. Cardiac out put changes and cont inuous mixed venous
sat urat ion measurement in t he crit ically ill. Crit Care Med. 1988; 16: 495.
> Table of C ontents > Labor ator y > 226 - C hec k Thyr oid Func tion in C r itic ally Ill P atients
226
Check Thyroid Function in Critically Ill Patients
Meghan C. Tadel MD
Bot h hypot hyroidism and hypert hyroidism can cont ribut e t o lif e-t hreat ening
sit uat ions during t he care of crit ically ill pat ient s. Alt hough t he most severe f orms
of t hese diseases, myxedema coma and t hyroid st orm, generally occur in
pat ient s w it h previously diagnosed t hyroid disorders, t hey can rarely be t he
present ing episode of t hyroid abnormalit y in a pat ient undergoing t he ext reme
st resses of t rauma, inf ect ion, perioperat ive periods, or recovery f rom
anest hesia.
Hypothyroidism
Alt hough up t o 90% of int ensive care unit (I CU) pat ient s have been f ound t o have
abnormal t hyroid f unct ion t est s (most commonly low t riiodot hyronine [ T3 ] ), t he
vast majorit y of t hem are f ound t o have sick eut hyroid syndrome, also know n as
nont hyroidal illness syndrome. Because of t his complicat ing f act or it has been
very diff icult t o quant it at e t he t rue incidence of hypot hyroidism in I CU pat ient s,
but est imat es range f rom 5% t o 30% of pat ient s admit t ed t o medical I CUs.
Signs and sympt oms of hypot hyroidism exist along a cont inuum. Pat ient s may
exhibit some sympt oms w hile lacking ot hers, and t he most severe sympt oms are
present only w hen a pat ient has progressed t o myxedemacoma. Possible
manif est at ions of hypot hyroidism include hypoglycemia, hypot hermia (or cold
int olerance), hypot ension, hypovent ilat ion, hyponat remia, bradycardia,
elect rocardiogram changes (T-w ave inversions and event ually J w aves),
gast roint est inal at ony, bladder at ony, skelet al muscle myopat hy, doughy
nonpit t ing edema, f at igue, let hargy, and alt ered level of consciousness. Perhaps
t he most ubiquit ous sign of hypot hyroidism is a pericardial eff usion, w hich is
present in up t o 30% of hypot hyroid pat ient s; pleural and perit oneal eff usions
can be seen as w ell. Medicat ions t hat have been associat ed w it h hypot hyroidism
include lit hium, amiodarone, aminoglut et himide, int erf eron alpha, t halidomide,
bet aroxine, and st avudine.
Crit ically ill pat ient s w ho demonst rat e any number of t he af orement ioned
st igmat a of hypot hyroidism should be t est ed f or t hyroid f unct ion as early as
possible as t his minimizes t he likelihood of values being abnormal because of
sick eut hyroid syndrome. There are many t est s available f or screening and
diagnosis of t hyroid dysf unct ion and
some cont roversy surrounds t he order in w hich t hese t est s should be complet ed.
The most common met hod, w hich provides >90% sensit ivit y f or det ect ion of
hypot hyroidism, is a combinat ion of t hyrot ropin and eit her f ree t hyroxine (T4 ) or
t ot al T4 and f ree T4 index. I n primary hypot hyroidism (w hich account s f or 95% of
hypot hyroidism), t hyrot ropin levels w ill be elevat ed out side t he normal range of
0. 5 t o 5. 0 mU/ L and in most cases of clinical hypot hyroidism w ill act ually be
elevat ed above 20 mU/ L. I n secondary, or cent ral, hypot hyroidism, w hich is due
t o adenohypophyseal or pit uit ary dysf unct ion, t he t hyrot ropin w ill be very low or
undet ect able. I n eit her sit uat ion, f ree T4 w ill be below t he normal range. Tot al T4
levels are not helpf ul because only f ree hormone is act ive and t ot al T4 levels may
be low in t he f ace of normal f ree T4 in sit uat ions of alt ered prot ein binding or
decreased prot eins available f or binding. The f ree T4 index t akes int o account
binding and prot ein availabilit y by mult iplying t he t ot al T4 by a f act or det ermined
by T3 resin upt ake.
What to Do
Treat ment of hypot hyroidism must encompass support ive measures, in addit ion
t o t hyroid hormone replacement . Hormone replacement can be accomplished via
oral or parent eral supplement at ion of t hyroid hormone in t he f orm of eit her T3 or
T 4 . Because int ravenous supplement at ion of t hyroid hormone, in part icular T3 ,
has been associat ed w it h lif e-t hreat ening arrhyt hmias, specif ically vent ricular
t achyarrhyt hmias, it has been argued t hat t his be reserved f or pat ient s
demonst rat ing only t he most severe sympt oms of hypot hyroidism—myxedema
coma or hypot ension and bradycardia not responsive t o ot her t herapies.
Because of t his concern, t here is no consensus on t herapy but several generally
accept ed regimens of t hyroid hormone replet ion are most commonly
recommended. I n mild cases, w here a pat ient can t olerat e oral (PO )
medicat ions, t hey can be st art ed at a dose of 50 µg/ day (elderly should st art at
25 µg/ day) w it h t it rat ions of 50 µg/ day of T4 every 3 t o 4 w eeks w it h monit oring
of t hyrot ropin level t o det ermine w hen t herapeut ic doses have been reached. The
usual ending dose is 50 t o 200 µg/ day. Pat ient s w ho cannot t olerat e oral
medicat ions or w ho are demonst rat ing signif icant sympt oms f rom hypot hyroidism
can be given 200 t o 500 µg int ravenously (I V) as a single bolus t o help replet e
t he t hyroxine circulat ing volume, t hen 100 µg 24 hours lat er and t hen 50 µg daily,
w it h f urt her t it rat ions every 3 t o 4 w eeks based on t hyrot ropin levels. I f I V T4 is
given, it can be supplement ed w it h I V T3 at t he dose of 25 µg f or a young,
noncardiac pat ient , 12. 5 µg f or an elderly or cardiac pat ient ; t he single dose of
T 3 may be repeat ed once
at 12 hours. The rat ionale f or supplement ing T3 in addit ion t o T4 is t hat t here are
concerns given t hat T3 is t he more act ive f orm of t hyroid hormone, alt hough 90%
of T3 is made via peripheral conversion f rom T4 and in a hypot hyroid st at e,
peripheral conversion of T4 t o T3 is reduced. I t should be not ed w hen convert ing
f rom an I V t o a PO , t he dose should be doubled as t he bioavailabilit y is 50%.
Hyperthyroidism
Hypert hyroidism, or as it is commonly called, t hyrot oxicosis, is rarely diagnosed
in t he I CU set t ing, but decompensat ion of chronic hypert hyroidism in t hese
crit ically ill pat ient s can be lif e t hreat ening. The most common cause of
t hyrot oxicosis, account ing f or up t o 90% of cases in t he Unit ed St at es, is G raves
disease, w hich occurs in anyw here f rom 0. 02% t o 0. 4% of t he populat ion. Signs
and sympt oms of t hyrot oxicosis include supravent ricular t achyarrhyt hmias (sinus
t achycardia and at rial f ibrillat ion most commonly), agit at ion (alt hough let hargy
can be seen in t he elderly), f ine t remors, anxiet y, heat int olerance, increased
perspirat ion, and w eight loss, of t en in spit e of an increased appet it e. When
t hyrot oxicosis progresses t o it s most severe f orm, t hyroid st orm, t he previous
sympt oms are accompanied by f ever, severe agit at ion or coma, and high-out put
cardiac f ailure. The medicat ion most commonly associat ed w it h t hyrot oxicosis is
chronic amiodarone t herapy because of it s high iodine cont ent . The ot her
diagnosis t o consider, part icularly in a pat ient w it h no know n hist ory of
t hyrot oxicosis, is t hat of t hyrot oxicosis f act it ia, w hich is caused by ingest ion of
suprat herapeut ic doses of t hyroxine.
The laborat ory t est most usef ul in diagnosing t hyrot oxicosis is f ree T4 or t ot al T4
and T4 index. By def init ion, f ree T4 or T4 index w ill be elevat ed in t hyrot oxicosis.
I t has been t radit ionally argued t hat t hyrot ropin levels provide lit t le or no insight
int o hypert hyroidism. How ever, now t hat sensit ive t hyrot ropin t est s are available
t hat can det ect levels of t hyrot ropin as low as 0. 01 mU/ L (normal range 0. 5 t o
5. 0 mU/ L), it is argued t hat a value below 0. 1 mU/ L is indeed diagnost ic of
primary t hyrot oxicosis w it h appropriat e negat ive f eedback of t hyrot ropin release.
What to Do
Thyrot oxicosis should be managed w it h support ive measures as w ell as
ant it hyroid medicat ions. The t w o available medicat ions t hat suppress t hyroxine
product ion are met himazole and propylt hiouracil (PTU). There is some
disagreement as t o w hich t herapy is pref erred because met himazole is
t hought t o have a more rapid onset of act ion and has a bet t er side-eff ect prof ile;
how ever, PTU has been t radit ionally pref erred because it demonst rat es inhibit ion
of peripheral conversion of T4 t o T3 , w hich is t he more act ive f orm of t hyroid
hormone. There is also lit t le consensus as t o t he opt imal dosing of t hese
medicat ions but most prot ocols are w it hin t he range of PTU 100 t o 200 mg t w o
or t hree t imes a day or met himazole 10 t o 60 mg per day. Bot h medicat ions are
available only in ent eral f ormulat ions and if a pat ient cannot have PO
medicat ions t hey can be given via nasogast ric t ube.
These medicat ions primarily aff ect product ion of t hyroxine, w hich means t hey
can t ake w eeks t o aff ect t he available pool of t hyroid hormone. I n t he meant ime,
addit ional t herapies can be of benef it . Propranolol in eit her I V or PO f ormulat ion
can be t it rat ed t o eff ect f or improvement of t achyarrhyt hmias and agit at ion.
Alt hough ot her select ive bet a-blockers have been show n t o be helpf ul,
part icularly in pat ient s w it h bronchoconst rict ive lung disease, propranolol is
generally f avored because it demonst rat es some minimal inhibit ion of peripheral
T 4 -t o-T 3 conversion. I odide, given as Lugol's solut ion eit her PO or as an enema,
has an inst ant aneous inhibit ion of release of t hyroid hormone f rom t he gland.
How ever, t his t herapy should be inst it ut ed only af t er eit her met himazole or PTU
t herapy has begun t o prevent an increase in t hyroxine product ion. Via t he same
mechanism, t he gallbladder dyes iopanoic acid and ipodat e, w hich bot h cont ain
high levels of iodide, can also be given. These dyes also act t o decrease
peripheral T4 -t o-T 3 conversion. I n t he short t erm amiodarone has also been used
because of it s high iodine cont ent and inhibit ion of T4 -t o-T 3 conversion, but as
not ed earlier in chronic t herapy t his medicat ion can cause t hyrot oxicosis. Finally,
in lif e-t hreat ening cases, plasmapheresis and ot her f orms of direct removal of
t hyroxine f rom circulat ion have been used, including cholest yramine f or
gast roint est inal binding of t hyroxine, alt hough no t est ed prot ocol exist s f or any of
t hese met hods.
Suggested Readings
G oldman L, ed. Cecil Text book of Medicine. 22nd ed. Philadelphia: WB
Saunders; 2004: 1392–1405.
Larsen P, ed. Williams Text book of Endocrinology. 10t h ed. Philadelphia: WB

Saunders; 2003: 346–365.
> Table of C ontents > Labor ator y > 227 - Us e an E m pty Lab Tube to C hec k the Viability of a S tom a
227
Use an Empty Lab Tube to Check the Viability of
a Stoma
While st oma placement is t ypically considered a minor procedure, complicat ions
f ollow ing it are f requent and pot ent ially devast at ing. The complicat ion rat e has
been report ed t o vary f rom 25% f ollow ing colost omy t o 75% f ollow ing end
ileost omy. These complicat ions include ret ract ion, prolapse, hernia, f ist ula,
ulcerat ion, st enosis, bleeding, and necrosis.
St omal ischemia is more common w it h end st omas t han w it h loop st omas.
Causes of compromised blood f low are varied and usually relat ed t o surgical
t echnique around t he t ime of st oma f ormat ion. A t oo-t ight abdominal st oma sit e
may cause ext rinsic compression of t he blood vessels f eeding t he dist al bow el.
Tension on t he vascular pedicle due t o progressive abdominal dist ent ion or
inadequat e mobilizat ion may lead t o progressive ischemia of t he st oma.
Excessive int raoperat ive dissect ion of t he bow el mesent ery can lead t o
compromised dist al art erial blood supply. Blood vessels may also become kinked
as t he bow el mesent ery t raverses t he abdominal w all.
I deally, a clear st oma appliance w ill be in place prior t o pat ient arrival in t he
int ensive care unit (I CU). This not only allow s f or collect ion and measurement of
st oma eff luent but also allow s f or bedside visual inspect ion of t he bow el mucosa
w it hout necessit at ing a dressing or appliance change. I mmediat ely
post operat ively, it is common f or bow el edema t o be present at t he st oma sit e.
While a f requent cause of concern f or pat ient s and st aff , some duskiness is t o
be expect ed. Typically, t he st oma appearance improves and t he duskiness
resolves w it hin 48 hours.
I f t here is no improvement in t he st omal appearance w it hin 48 hours, or if t he
st oma becomes black, t he dept h of st omal necrosis must be assessed t o rule
out an impending abdominal cat ast rophe. A simple t echnique f or bedside
examinat ion of t he st omal mucosa is as f ollow s: Remove t he rubber t op and
adhesive paper label of a glass blood collect ion t ube and af t er lubricat ion gent ly
place t he bot t om-closed end int o t he st oma. Then, shine a f lashlight t hrough t he
open mout h of t he t ube t o illuminat e t he mucosa. I n a noncompromised st oma,
t he reassuring pink color of viable bow el is t ypically apparent just below skin
level.
I f t here is f urt her concern regarding necrosis of t he st oma, a f lexible endoscope
insert ed gent ly int o t he st oma can bet t er def ine t he limit s of necrosis. Necrosis
t hat does not involve a large, cont inuous segment of bow el or is conf ined t o t he
mucosa may be managed expect ant ly. Typically t his t ype of lesion heals by
secondary int ent ion w it hout perf orat ion. I f t his is t he case, t he clinician must be
aw are of an increased risk of f ist ula f ormat ion or st enosis as t he segment of
necrosis heals. Necrosis t hat ext ends below t he level of t he f ascia or is
circumf erent ial requires emergency surgery.
Suggested Readings
Deit ch EA, ed. Tools of t he Trade and Rules of t he Road: A Surgical G uide.
Philadelphia: Lippincot t -Raven Publishers; 1997: 186–202.
Nyhus LM, Baker RJ, Fischer JE, eds. Mast ery of Surgery. 3rd ed. Bost on:
Lit t le, Brow n and Company; 1997: 1366–1377.
> Table of C ontents > Labor ator y > 228 - Never R etes t Low S er um Gluc os e; Tr eat Im m ediately and
Then P r event Fur ther E pis odes
228
Never Retest Low Serum Glucose; Treat
Immediately and Then Prevent Further Episodes
Kathleen A. Williams MSN, RN, CRNP
Maint aining euglycemia in t he inpat ient set t ing has been show n in mult iple
st udies t o decrease morbidit y and mort alit y across mult iple t ypes of pat ient
populat ions. O ne of t he leading barriers t o achieving opt imal glucose cont rol is
t he f ear of hypoglycemia on t he part of bot h pat ient s as w ell as care providers.
This f ear is underst andable. Hypoglycemia can cause behavioral changes,
seizures, permanent neurological injury, and even deat h. Prompt and appropriat e
recognit ion, t reat ment , and prevent ion of hypoglycemia is t heref ore of crit ical
import ance.
Current met hods f or inpat ient glucose monit oring include laborat ory analysis of
serum, point -of -care t est ing w it h port able bedside glucomet ers, and ongoing
commercial invest igat ions of cont inuous subcut aneous and serum glucose
monit oring syst ems f or inpat ient use. The most accurat e is laborat ory analysis of
a serum value. How ever, t hese values of t en t ake considerable t ime t o obt ain.
Bedside glucomet ers, w hen used w it hin t he manuf act urer's guidelines, can
provide a rapid assessment of t he pat ient 's glycemic cont rol. Appropriat e
t raining of st aff using and int erpret ing blood glucose values obt ained by bedside
monit oring is crit ical t o pat ient saf et y. Some condit ions t hat alt er t he accuracy of
glucomet er readings may include t he f ollow ing:
1. The skin sit e may be inadequat ely prepared.

2. The art erial or venous lines may not be adequat ely cleared bef ore blood is
draw n and placed on t he glucomet er st rip.
3. Condit ions of decreased peripheral blood f low may not ref lect t he t rue
physiological st at e. Examples of decreased blood f low include severe
dehydrat ion, hypot ension, shock, and peripheral vascular disease or
peripheral vasoconst rict ion due t o vasoact ive drugs.
4. Acet aminophen levels great er t han 8 mg/ dL may produce elevat ed glucose
result s
5. At glucose values above 200 mg/ dL, low hemat ocrit s (<20%) may cause
elevat ed glucose readings compared w it h a w hole blood ref erence.
6. At glucose values above 200 mg/ dL, high hemat ocrit s may cause reduced
glucose value compared w it h a w hole blood ref erence.
What to Do
G lucose monit oring orders should be w rit t en w it h not if icat ion paramet ers f or
bot h hypo- and hyper-glycemia. I f not if icat ion is given t hat a pat ient has a
glucose value of <60 mg/ dL, immediat e t reat ment is required. I f t he pat ient is
aw ake and able t o sw allow w it hout risk f or aspirat ion, t he pat ient should be
given 15 g of f ast -act ing carbohydrat e (4 oz of juice or nondiet soda, 8 oz of
milk, or a packet of graham crackers). I f t he pat ient is not able t o t ake anyt hing
by mout h, he or she should be given 25 t o 50 mL of 50% dext rose (D50)
int ravenously (I V). For t hose pat ient s w ho are unable t o sw allow and lack I V
access, init ial t reat ment should consist of glucagon 1 mg int ramuscularly and
w it h est ablishment of I V access. G lucose values should be rechecked w it hin 20
minut es of t he init ial administ rat ion of glucose or glucagon t o assess response.
Cont inued f requent monit oring is required f or t he durat ion of act ion of any
hypoglycemic agent given. This is especially import ant if t he causat ive agent is
eit her int ermediat e or long-act ing insulin or a sulf onylurea.
What not to Do
I t is import ant t o not e t hat some diabet ic pat ient s may demonst rat e
hypoglycemic unaw areness. Hypoglycemic unaw areness is def ined as t he loss of
hypoglycemic w arning sympt oms such as diaphoresis, t achycardia, and shaking
due t o def ect s in hormonal count erregulat ion. These are pat ient s w ho w ill have
low glucose readings but present w it hout obvious sympt oms. This lack of
sympt oms has caused t he nursing and medical st aff t o disbelieve t he glucose
result and delay t reat ment , w ait ing f or a repeat result t o verif y. This delay in
t reat ment can result in a prolonged hypoglycemic event w it h subsequent
neurologic injury. The t ransient increase in glucose f rom t he administ rat ion of a
single dose of D50 or oral int ake of 15 g of f ast ing-act ing carbohydrat es is
signif icant ly less likely t o cause harm t o t he pat ient as compared w it h t he risk
associat ed w it h prolonged hypoglycemia.
Pat ient s at part icular risk f or hypoglycemic unaw areness include t hose w ho have
been maint ained w it h int ensive insulin t herapies in t he out pat ient set t ing, t hose
w it h near-normal glycosylat ed hemoglobin (HgbA1c), and t hose w it h aut onomic
neuropat hy f rom t he diabet es. Pat ient s w ho are int ubat ed, sedat ed, or ot herw ise
lack t he abilit y t o self -report hypoglycemic sympt oms (e. g. , st roke) are also at a
very high risk f or unrecognized prolonged hypoglycemia.
Hypoglycemia is considered a prevent able complicat ion of ant i-hyperglycemic
t herapy. I n a recent st udy of 24 inpat ient hypoglycemic event s, 22 w ere
considered prevent able. I nadequat e management
response t o a change in t he course of care w as t he most common cause (82%)

and inappropriat e disease management , def ined as inappropriat e medicat ion
select ion and inadequat e monit oring select ion, occurred in 50% of t he event s. O f
not e, exclusive use of t he sliding scale regimen (w hich is st ill f requent ly being
ordered as a hyperglycemia t reat ment modalit y) account ed f or 26% of t he
hyperglycemic event s in t he same st udy.
To prevent hypoglycemia, t he care provider should f ollow t he American
Associat ion of Clinical Endocrinologist s (AACE) I npat ient G lycemic Cont rol
G uidelines recommendat ions, w hich include t he f ollow ing:
1. Appropriat e select ion of pat ient s f or t he use of I V insulin

2. Provision of I V insulin or addit ion of insulin in t he t ot al parent eral nut rit ion
(TPN) bag t o meet t he requirement s f or insulin creat ed by t he use of TPN.
This w ill avoid having a long-act ing subcut aneous insulin cont inuing t o have
an eff ect if t he TPN is st opped.
3. Use of a programmed subcut aneous insulin regimen t hat includes a basal–
nut rit ional–correct ion component
4. Daily revision of t he scheduled insulin regimen based on clinical condit ion and
t he response t o previous t herapy
5. Adjust ment of insulin t herapy w it h any change in renal f unct ion
6. Recognit ion of t he risk f or hypoglycemia w it h sudden changes in nut rit ion and
t he subsequent need f or insulin adjust ment and hypoglycemia prevent ion—
including t he use of D10 t o mit igat e t he loss of eit her TPN or ent eral f eeding
7. I mport ance of close and accurat e glucose monit oring at t he bedside
Suggested Readings
Furnary AP, G ao G , G runkemeier G L, et al. Cont inuous insulin inf usion
reduces mort alit y in pat ient s w it h diabet es undergoing coronary art ery bypass
graf t ing. J Thorac Cardiovasc Surg. 2003; 125: 1007–1021.
Krinsley JS. Associat ion bet w een hyperglycemia and increased hospit al
mort alit y in a het erogeneous populat ion of crit ically ill pat ient s. Mayo Clin
Proc. 2003; 78: 1471–1478.
van den Berghe G , Wout ers P, Weekers F, et al. I nt ensive insulin t herapy in
t he crit ically ill pat ient . N Engl J Med. 2001; 345: 1359–1367.
> Table of C ontents > Labor ator y > 229 - D - D im er Levels C an R ule O ut B ut Not R ule in P ulm onar y
E m bolis m in P os toper ative P atients
229
D-Dimer Levels Can Rule Out But Not Rule in
Pulmonary Embolism in Postoperative Patients
Hari Nathan MD
D-dimer is a degradat ion product of cross-linked f ibrin. A normal level of d-dimer
is t ypically less t han 500 ng/ mL. I n a pat ient w it h pulmonary embolism (PE),
deep venous t hrombosis (DVT), or ot her clot burden, elevat ed d-dimer levels
ref lect t he body's endogenous f ibrinolysis. I naddit ion, d-dimer levels may also be
raised in myocardial inf arct ion, pneumonia, sepsis, and cancer; during t he
second and t hird t rimest ers of pregnancy; and af t er surgery. Elevat ed d-dimer
levels may persist up t o 3 mont hs af t er surgery.
When DVT or PE is suspect ed in a post operat ive pat ient , normal d-dimer levels
argue against t he diagnosis. The negat ive predict ive value of t he t est varies f rom
85% t o 99%, depending on t he assay used and t he pret est probabilit y of PE.
Elevat ed d-dimer levels, w hile consist ent w it h PE, by no means conf irm t he
diagnosis in post operat ive pat ient s. Thrombus in locat ions ot her t han t he
pulmonary art erial bed (as in DVT or clot at t he sit e of surgery) can explain d-
dimer elevat ions, as can any of t he f act ors list ed previously. Theref ore, if t he
clinical suspicion of PE is low, a normal d-dimer level is helpf ul in ruling out t he
diagnosis. When t he clinical suspicion of PE is high, f urt her invest igat ion is
mandat ed regardless of t he d-dimer level.
Suggested Readings
G oldhaber SZ. Pulmonary embolism. Lancet . 2004; 363: 1295â 1 305.
Thompson BT, Hales CA. Clinical manif est at ions of and diagnost ic st rat egies
f or acut e pulmonary embolism. I n: Rose BD, ed. UpToDat e. Walt ham, MA:
UpToDat e; 2006.
> Table of C ontents > Labor ator y > 230 - R em em ber That D iabetic K etoac idos is O ften B egins W ith an
Anion- Gap Metabolic Ac idos is
230
Remember That Diabetic Ketoacidosis Often
Begins With an Anion-Gap Metabolic Acidosis
Nirav G . Shah MD
Diabet ic ket oacidosis (DKA) is a common diagnosis encount ered in t he hospit al
set t ing w it h an est imat ed incidence of 4. 6 t o 8 episodes per 1, 000 pat ient s w it h
diabet es. I t is a disorder t hat t ypically occurs af t er a precipit at ing event t hat
causes an increase in cat echolamines and is t ypif ied by t he t riad of
hyperglycemia, met abolic acidosis, and ket osis. The import ance of prompt
diagnosis and t reat ment is evident by t he severit y of lab derangement s and
clinical det eriorat ion t hat occurs if it is lef t unt reat ed.
Watch Out For

The t ypical int ensive care unit (I CU) pat ient w it h DKA is a know n diabet ic w it h
insulin def iciency w ho has an event t hat causes an increase in grow t h hormone,
glucagon, cort isol, or cat echolamines, or an omission or def iciency of insulin.
O ccasionally, it may also be t he present ing diagnosis f or a new -onset diabet ic
pat ient . An underlying cause should alw ays be excluded in t he pat ient w it h DKA.
The most common underlying et iologies include inf ect ion, myocardial inf arct ion,
st roke, and pancreat it is.
The physiology of DKA is f airly st raight f orw ard. The hyperglycemia in DKA
result s f rom gluconeogenesis and glycogenolysis in t he liver and a reduced
ut ilizat ion of glucose by peripheral t issues. This, in t urn, releases t he inhibit ory
eff ect on glucagon, w hich w hen combined w it h low insulin levels result s in more
hepat ic gluconeogenesis and glycogenolysis. G lucose levels can rout inely run as
high as 800 mg/ dL. This level of hyperglycemia leads t o dehydrat ion and ot her
elect rolyt e abnormalit ies because of t he ensuing osmot ic diuresis. Tot al-body
st ores of sodium, pot assium, chloride, phosphorus, and magnesium are reduced
by t he polyuria t hat accompanies t he diuresis. The serum pot assium may be
art if icially elevat ed secondary t o t he prof ound acidosis. I n addit ion, a f act it ious
hyponat remia may occur in t he set t ing of hyperglycemia. Theref ore, t he serum
sodium must be correct ed by increasing t he sodium level by 1. 6 mEq f or each
100-mg/ dL increase in serum glucose.
What to Do
The t reat ment f or DKA addresses t w o major goals: correct ing t he insulin
def iciency and elect rolyt e abnormalit ies and t reat ing t he underlying cause.
Theref ore, t he init ial management is int ravenous insulin t herapy and aggressive
int ravenous f luid resuscit at ion. I nsulin t herapy should begin w it h a 10- t o 20-unit
bolus of regular insulin int ravenously f ollow ed by a cont inuous inf usion of 5 t o 10
U/ h t hat can be t it rat ed t o response. I nt ravenous f luid t herapy should st art w it h 2
t o 3 lit ers of normal saline f ollow ed by 0. 45% normal saline. O nce t he serum
glucose reaches approximat ely 240 mg/ dL, a combinat ion of 5% glucose w it h
0. 45% normal saline should be init iat ed. Throughout t his course, caref ul at t ent ion
should be paid t o t he level of serum pot assium. O nce t he level is normal, or less
t han 5. 5 mEq/ L, replacement of pot assium should begin at a rat e of 10 mEq/ h
and should be increased t o 40 t o 80 mEq/ h if t he serum pot assium is less t han
3. 5 mEq/ L or if bicarbonat e t herapy is administ ered, t hough replacement of
bicarbonat e is not rout inely suggest ed.
O f not e, t he met abolic acidosis of DKA is f requent ly severe w it h serum
bicarbonat e measurement s less t han 10 t o 15 mEq/ L and art erial pH ranging
bet w een 6. 8 and 7. 3. The met abolic acidosis is an anion-gap t ype and is present
secondary t o t he inabilit y t o measure ket ones in t he blood. Bet a-hydroxybut yric
acid is t he main ket one present in DKA. The anion-gap acidosis resolves as
ket one product ion slow s w it h t he init iat ion of insulin and int ravenous f luid
t herapy. I n f act , w it h t he administ rat ion of normal saline (154 mEq of bot h
sodium and chloride) a hyperchloremic non-anion-gap met abolic acidosis may
develop. This has no adverse eff ect s and w ill gradually improve as renal acid
excret ion is enhanced in t he days f ollow ing admission w it h caref ul observat ion of
chloride load.
Complicat ions of DKA and/ or inadequat e t herapy can be devast at ing. These
include hypoglycemia, hypokalemia, cerebral edema (primarily in t he pediat ric
populat ion), and adult respirat ory dist ress syndrome.
Suggested Readings
Kit abchi AE, Umpierrez G E, Murphy MB, et al. Management of hyperglycemic
crises in pat ient s w it h diabet es. Diabet es Care. 2000; 24: 131–153.
Rose BD, Post TW. Clinical Physiology of Acid-Base and Elect rolyt e
Disorders. 5t h ed. New York: McG raw -Hill; 2001: 809–815.
> Table of C ontents > Labor ator y > 231 - C ons ider O btaining A S er um B - Type Natr iur etic P eptide Level
in the Ac utely D ys pneic P atient
231
Consider Obtaining A Serum B-Type Natriuretic
Peptide Level in the Acutely Dyspneic Patient
Lawrence O sei MD
B-t ype nat riuret ic pept ide (BNP) is a 32-amino-acid polypept ide cardiac
neurohormone specif ically secret ed f rom t he vent ricles in response t o volume
expansion and pressure overload. BNP w as originally ident if ied in ext ract s of
porcine brain. I t is present in minut e amount s in human brain, but t he clinically
import ant source originat es in t he cardiac vent ricles. The levels of B-t ype
nat riuret ic pept ide are elevat ed in pat ient s w it h lef t vent ricular dysf unct ion and
t he levels correlat e w it h bot h t he severit y of sympt oms and t he prognosis.
The present ing f eat ures of acut e cardiogenic and noncardiogenic pulmonary
edema are similar and can be clinically diff icult t o dist inguish. Common causes of
cardiogenic pulmonary edema include cardiac ischemia w it h or w it hout
myocardial inf arct ion, exacerbat ion of chronic syst olic or diast olic heart f ailure,
and dysf unct ion of t he mit ral or aort ic valve. Volume overload should also be
considered. A t ypical hist ory of paroxysmal noct urnal dyspnea or ort hopnea
suggest s cardiogenic pulmonary edema. How ever, a silent myocardial inf arct ion
or occult diast olic dysf unct ion may also manif est as acut e pulmonary edema,
w it h f ew clues provided by t he hist ory.
I n cont rast , noncardiogenic pulmonary edema is associat ed primarily w it h ot her
clinical disorders, including pneumonia, sepsis, aspirat ion of gast ric cont ent s,
and major t rauma associat ed w it h t he administ rat ion of mult iple blood-product
t ransf usions. I n a pat ient present ing w it h acut e dyspnea, measuring serum BNP
levels may be able t o help diff erent iat e bet w een cardiogenic and noncardiogenic
pulmonary edema.
I n one recent st udy by Maisel et al. , BNP levels by t hemselves w ere more
accurat e t han hist orical or physical f indings or laborat ory values in ident if ying
congest ive heart f ailure as t he cause of dyspnea. The diagnost ic accuracy of
BNP at a cut off of 100 pg/ mL w as 83. 4%. The negat ive predict ive value of BNP
at levels of less t han 50 pg/ mL w as 96%. The aut hors concluded t hat used in
conjunct ion w it h ot her clinical inf ormat ion, BNP levels w ere usef ul in excluding
t he diagnosis of congest ive heart f ailure.
Suggested Readings
Maisel A, Krishnasw amy P, Now ak RM, et al. Rapid measurement of B-t ype
nat riuret ic pept ide in t he emergency diagnosis of heart f ailure. NEngl J Med.
2002; 347(3): 161– 167.
Mueller C, Scholer A, Laule-Kilian K, et al. Use of B-t ype nat riuret ic pept ide in
t he evaluat ion and management of acut e dyspnea. N Engl J Med.
2004; 350(7): 647–654.
> Table of C ontents > Labor ator y > 232 - C ons ider Hypom agnes em ia as a C aus e of R efr ac tor y
Hypokalem ia
232
Consider Hypomagnesemia as a Cause of
Refractory Hypokalemia
Hypomagnesemia and hypokalemia are bot h common elect rolyt e dist urbances in
t he int ensive care unit (I CU) set t ing. Magnesium and pot assium have import ant
int eract ions t o consider w hen t reat ing def iciencies of eit her ion. Magnesium is
int imat ely involved in a number of biochemical react ions in t he body, part icularly
t hose involving adenosine t riphosphat e (ATP). The ionized f orm of magnesium in
t he serum is met abolically act ive but represent s only a small component of t he
t ot al-body magnesium. Furt hermore, t he relat ionship bet w een t he ionized and
serum concent rat ions, and t he serum and int racellular (cyt oplasmic)
concent rat ions are not clear. Theref ore, t ot al serum concent rat ions may not be
represent at ive of act ive ion.
Magnesium int ake t hrough t he diet is usually suff icient f or maint aining
appropriat e serum concent rat ions unless t here is insuff icient absorpt ion or
increased losses eit her t hrough t he gast roint est inal t ract or urine. When
hypomagnesemia occurs, t he st orage pool in t he bone, muscle, and sof t t issues
is of t en suff icient f or maint aining appropriat e serum concent rat ions.
The sympt oms of hypomagnesemia include arrhyt hmias (at rial f ibrillat ion and
f lut t er, vent ricular t achycardia, and supravent ricular t achycardia),
elect rocardiogram changes (prolonged Q T, increased PR, and w idening Q RS),
neurologic changes (seizures, alt ered ment al st at us), and ot her elect rolyt e
problems (hypophosphat emia, hypokalemia).
The t reat ment of hypomagnesemia is dependent upon t he underlying cause.
When magnesium is administ ered parent erally, t he reabsorpt ion at t he level of
t he kidney is reduced and a large component of t he administ ered dose w ill be
excret ed in t he urine. Hypokalemia coexist s w it h hypomagnesemia approximat ely
50% of t he t ime. The mechanism f or t his is relat ed t o t he ATP-dependent
inhibit ion of luminal pot assium channels in t he dist al nephron. When magnesium
is def icient , t he ATP is less eff ect ive and pot assium is secret ed int o t he t ubular
f luid, leading t o a ref ract ory hypokalemia. Theref ore,
magnesium (even in t he set t ing of a normal serum value and considering it s large
t herapeut ic w indow ) should be administ ered empirically in t he set t ing of
hypocalcemia or hypokalemia.
Suggested Readings
Dacey MJ. Hypomagnesemic disorders. Crit Care Clin. 2001; 17: 155â 1 73.
> Table of C ontents > Nutr ition > 233 - B e Aw ar e that E nter al Feeds c an Low er P henytoin Levels
233
Be Aware that Enteral Feeds can Lower
Phenytoin Levels
T imothy M. Moore MD, PHD
Faramarz Zarfeshanfard RPH
Phenyt oin (Dilant in) is an ant iconvulsant drug, w hich can be usef ul in t he
t reat ment of seizures and epilepsy. The primary sit e of act ion appears t o be t he
mot or cort ex w here spread of seizure act ivit y is inhibit ed, possibly by promot ing
sodium eff lux f rom neurons and st abilizing t he t hreshold against hyperexcit abilit y
caused by excessive st imulat ion or environment al changes capable of reducing
membrane sodium gradient . Loss of post t et anic pot ent iat ion prevent s cort ical
seizure f oci f rom det onat ing adjacent cort ical areas. Phenyt oin reduces t he
maximal act ivit y of brain-st em cent ers responsible f or t he t onic phase of t onic-
clonic (grand mal) seizures.
Phenyt oin exhibit s nonlinear, dose-dependent , Michaelis-Ment en
pharmacokinet ics. At t herapeut ic levels, phenyt oin enzyme hepat ic met abolism
reaches it s capacit y and it s pharmacokinet ics changes f rom f irst order t o zero
order, at w hich t ime a small increase in t he daily dose w ill result in a
disproport ionat ely higher level. Because of it s capacit y-limit ed met abolism,
phenyt oin half -lif e and t ime t o reach st eady st at e are concent rat ion dependent .
The higher t he concent rat ion, t he longer t he half -lif e and t he longer t he t ime it
t akes t o reach st eady st at e. The plasma half -lif e in humans af t er oral
administ rat ion of phenyt oin averages 22 hours, w it h a range of 7 t o 42 hours.
Af t er int ravenous (I V) administ rat ion, t he half -lif e ranges bet w een 11 and 15
hours. The diff erence in half -lif e is due t o slow er absorpt ion and varying
bioavailabilit y of phenyt oin f ormulat ions. Drug int eract ions may increase or
decrease phenyt oin half -lif e. Est imat ion of t ime t o reach st eady st at e cannot be
based on half -lif e. Serum level det erminat ions should be obt ained af t er t reat ment
init iat ion, dosage change, or addit ion or subt ract ion of an int eract ing f ood or
drug t o t he regimen. Trough levels provide inf ormat ion about clinically eff ect ive
serum level range and are obt ained just prior t o t he pat ient 's next scheduled
dose. Peak levels indicat e an individual's t hreshold f or emergence of dose-
relat ed side eff ect s and are obt ained at t he t ime of expect ed peak
concent rat ion. For Dilant in capsules, peak serum levels occur 4 t o 12 hours af t er
administ rat ion. For Dilant in I nf at abs and Dilant in suspension, peak serum levels
occur 1. 5 t o 3 hours af t er administ rat ion. The t iming of levels is dependent on
t he clinical sit uat ion as w ell
as rout e of administ rat ion and dosage f orm. I n t he acut e set t ing, it is advisable
t o check non-st eady-st at e phenyt oin levels t o avoid subt herapeut ic or t oxic
levels prior t o reaching st eady st at e. Suggest ed sampling t ime af t er oral load is
in 24 hours. Because of slow absorpt ion of oral f ormulat ions, t iming is not
crit ical f or level monit oring, but t rough levels are suggest ed. Suggest ed sampling
t ime f or I V maint enance is also a t rough level (or just prior t o t he next dose).
Though beyond t he scope of t his review, det ailed pharmacokinet ic equat ions are
available t o help dose and monit or phenyt oin t reat ment . (See suggest ed
readings. )
O pt imum cont rol w it hout clinical signs of t oxicit y occurs more of t en w it h t ot al
serum levels bet w een 10 and 20 µg/ mL and f ree levels of bet w een 1 and 2
µg/ mL, alt hough some mild cases of t onic-clonic (grand mal) epilepsy may be
cont rolled w it h low er serum levels of phenyt oin. I n most pat ient s maint ained at a
st eady dosage, st able phenyt oin serum levels are achieved. There may be w ide
int er-pat ient variabilit y in phenyt oin serum levels w it h equivalent dosages. The
pat ient w it h large variat ions in phenyt oin plasma levels, despit e st andard doses,
present s a diff icult clinical problem. Unusually high levels result f rom liver
disease, congenit al enzyme def iciency, or drug int eract ions result ing in met abolic
int erf erence. Plasma prot ein binding is decreased in renal f ailure and
hypoalbuminemia. Pat ient s w it h unusually low levels may be noncompliant or
hypermet abolizers of phenyt oin. How ever, in t he I CU set t ing in pat ient s receiving
ent eral f eeds, unusually or persist ent ly low serum phenyt oin levels may act ually
be at t ribut able t o a know n int eract ion w it h ent eral f eeding.
Alt hough it has been w idely report ed t hat serum phenyt oin levels are low ered in
pat ient s receiving concomit ant ent eral f eeding, t he exact mechanism underlying
t his int eract ion is unknow n. The majorit y of st udies have suggest ed t hat t here is
some physical incompat ibilit y bet w een phenyt oin and cert ain component s in
ent eral f eeding f ormulas causing decreased bioavailabilit y of t he drug. O t her
proposed mechanisms include binding of phenyt oin t o t he f eeding t ube lumen, pH
int eract ion of f eeds and drug, increased met abolism or clearance of t he drug
af t er prolonged use of ent eral f eeds, and int eract ion of phenyt oin and ot her
drugs being administ ered via t he ent eric t ube. A small number of st udies have
ref ut ed t he exist ence of an int eract ion bet w een phenyt oin and ent eral f eeding
alt oget her. I nt erest ingly, t hese st udies w ere prospect ive but w ere also
perf ormed in healt hy volunt eers rat her t han in an I CU pat ient populat ion. Et hical
and logist ic considerat ions limit t he abilit y t o perf orm prospect ive, cont rolled,
randomized t rials in appropriat e pat ient s.
The predominant ly recommended met hod f or reducing t his int eract ion bet w een
ent eral f eedings and phenyt oin is t o st agger phenyt oin administ rat ion w it h
f eeding t ime. When cont inuous f eeds are required, it has been recommended t o
hold f eeds 1 t o 2 hours prior t o phenyt oin dosing, f lush t he f eeding t ube w it h at
least 20 cm3 f ree w at er or saline bef ore and af t er administ rat ion, and resume
f eeding 1 t o 2 hours af t er dosing. I t should be not ed t hat t here is no clear
evidence t hat t his st rat egy increases absorpt ion. O t her st rat egies t o achieve and
maint ain seizure cont rol in t he t ube-f ed pat ient include simply adjust ing t he
phenyt oin dosage w hile on cont inuous f eeds w it hout int errupt ing t he f eeds
regimen, changing phenyt oin administ rat ion t o int ravenous, changing t he
ant iseizure medicat ion regimen, and changing t he ent eral f eeding f ormula. An
alt ernat ive st rat egy is t o use t he inject able product dow n t he f eeding t ube. This
st rat egy has been show n t o increase t he rat e of absorpt ion but not t he ext ent of
absorpt ion. I t is import ant t o not e t hat implement at ion of any or all of t hese
recommendat ions should be made only upon caref ul considerat ion of t he overall
clinical condit ion of t he crit ically ill pat ient . I t is also especially import ant t o
remember t hat if a correct ion is made t o increase serum phenyt oin levels in t he
pat ient receiving ent eral f eeds, a correct ion must likely also be made during
w eaning off f eeds t o limit t he possibilit y of phenyt oin t oxicit y secondary t o
overdose.
Suggested Readings
Au Yeung SCS, Ensom MH. Phenyt oin and ent eral f eedings: does evidence
support an int eract ion? Ann Pharmacot her. 2000; 34: 896–905.
Dipiro JT, Spruill WJ, Blouin RA, et al. Concept s in Clinical Pharmacokinet ics.
4t h ed. Bet hesda, MD: The American Societ y of Healt h-Syst em Pharmacist s;
2005.
Doak KK, Haas CE, Dunnigan KJ, et al. Bioavailabilit y of phenyt oin sodium
w it h ent eral f eedings. Pharmacot herapy. 1998; 18(3): 637–645.
G ilbert SJ, Hat t on J, Magnuson B. How t o minimize int eract ion bet w een
phenyt oin and ent eral f eedings: t w o approaches—a st rat egic approach. Nut r
Clin Pract . 1996; 11: 28–30.
Kit chen D, Smit h D. Problems w it h phenyt oin administ rat ion in neurology/
neurosurgery I TU pat ient s receiving ent eral f eeding. Seizure. 2001; 10: 265–
268.
Levy RH, Mat t son RH, Meldrum BS, Perucca E, eds. Ant iepilept ic Drugs. 5t h
ed. Philadelphia: Lippincot t Williams & Wilkins; 2002.
Murphy JE, ed. Clinical Pharmacokinet ics Pocket Ref erence. 2nd ed.
Bet hesda, MD: The American Societ y of Healt h-Syst em Pharmacist s; 2000.
Wint ers ME. Basic Clinical Pharmacokinet ics. 4t h ed. Philadelphia: Lippincot t
Williams & Wilkins; 2003.
> Table of C ontents > Nutr ition > 234 - C ons ider E ar ly E nter al Feeding
234
Consider Early Enteral Feeding
Bryan A. Cotton MD
I f t he gut w orks, use it !
As w it h ot her basic principles in pat ient care, t his simple maxim of nut rit ion has
somet imes been inexplicably ignored as our t echnological and pharmacological
advances have exponent ially increased over t he past several decades. How ever,
t o ignore t his simple idea is of t en t o t he det riment of t he pat ient . Numerous
st udies have demonst rat ed a correlat ion w it h poor nut rit ional st at us and poor
post operat ive out come. The current lit erat ure support s t he pref erent ial use of
ent eral f eeding over parent eral nut rit ion (t ot al parent eral nut rit ion, TPN) in
int ensive care unit (I CU) pat ient s w henever possible. The reasons f or t his include
not only t he benef icial eff ect s of ent eral support but also t he det riment al eff ect s
of TPN.
Beneficial Effects of Enteral Feeding

Several recent st udies have demonst rat ed t hat gut mucosal dysf unct ion, in t he
f orm of increased permeabilit y and villous sloughing, occurs early in t he absence
of ent eral f eedings. I n t he crit ically injured pat ient , several aut hors have
demonst rat ed improvement s in t he cat abolic st at e, specif ically t hrough improved
nit rogen balance, w hen ent eral nut rit ion is ut ilized inst ead of TPN. Physiological
advant ages of ent eral nut rit ion over TPN include it s st imulat ion of gallbladder
empt ying and release of pancreat ic secret ions, as w ell as maint enance of gut -
associat ed lymphoid t issue (G ALT) and mucosal immune f unct ion. The improved
gut mucosal int egrit y not ed w it h ent eral f eedings is likely responsible f or t he
decreased bow el perf orat ion rat e, improved int est inal anast omot ic healing, and
decreased sept ic complicat ions. I n addit ion, ent eral f eeding is signif icant ly less
expensive (even w hen excluding t he cost s associat ed w it h TPN complicat ions)
t han parent eral f ormulat ions.
Detrimental Effects of Parenteral Nutrition

The int est inal mucosa and submucosa is an area of int ense met abolic and
immunologic act ivit y, especially in t he crit ically ill and severely injured pat ient .
Ut ilizat ion of TPN in t hese pat ient s f urt her
compromises an already t enuous sit uat ion, w it h loss of mucosal mass and
w eight , increased villous sloughing, and dist urbed mucosal enzyme act ivit y. TPN
use has been show n t o decrease I gA in t he gut , as w ell as in upper respirat ory
secret ions. From a met abolic st andpoint , TPN causes met abolic acidosis,
hyperglycemia, hyperlipidemia, and signif icant elect rolyt e dist urbances. I n
addit ion, TPN has been associat ed w it h hepat ic st eat osis and cellular injury
leading t o liver dysf unct ion and f ailure. Syst emically, t he eff ect s of TPN include
impairment of leukocyt e chemot axis, impaired phagocyt osis, and an at t enuat ed
inf lammat ory response. O t her aut hors, how ever, have show n TPN associat ed
alt erat ions may act ually pot ent iat e t he syst emic inf lammat ory st at e by allow ing
increased bact erial t ranslocat ion and increasing f ree-radical f ormat ion. Some
st udies have demonst rat ed higher mort alit y, especially among t he crit ically ill, in
t hose receiving parent eral nut rit ion compared w it h ent eral f eeding, w it h TPN
almost doubling t he risk of dying. O f not e, t he risks of TPN t oxicit y can be
reduced by t he addit ion of low -rat e (somet imes ref erred t o as t rophic) t ube
f eedings (10 t o 30 mL/ h).
When and How to Give Enteral Feedings

Several aut hors have invest igat ed t he impact and t iming of early ent eral
nut rit ion. I n t he t rauma and burn set t ing, delays of as lit t le as 24 hours have
been demonst rat ed t o impact morbidit ies and out comes. I n f act , no evidence
exist s t o support w it hholding ent eral f eedings in t hose pat ient s w it h an open
abdomen. Among t he emergency surgery populat ion, ent eral f eedings should be
ut ilized early in t he post operat ive period. Alt hough t ube f eedings are rout inely
held because of concerns of bow el-w all edema and nonperist alsis, t his is not
support ed by t he lit erat ure. Wit h t he except ion of bow el obst ruct ion and proximal
ent eric f ist ulae, early ent eral nut rit ion has been demonst rat ed t o be t olerat ed
and of benef it even among t hose present ing w it h signif icant perit onit is and
premorbid malnut rit ion. Early ent eral support is also recommended f ollow ing
surgery f or gast roint est inal malignancies and even in cases of severe
pancreat it is, w it h evidence of at t enuat ed organ dysf unct ion and improved
out comes.
When possible, t he post pyloric posit ion should be ut ilized f or ent eral f eeding,
w it h t he nasojejunal locat ion pref erred in pancreat it is. Alt hough many pat ient s
experience gast roparesis and some evidence suggest s increased risk f or
aspirat ion, numerous st udies support saf et y and t olerance of t he gast ric rout e.
Most import ant ly, aggressive at t ent ion t o placing ent eral access (w het her
surgical or nasal rout e) should
be considered prior t o complet ion of t he operat ive procedure. I t is at t his t ime

t hat placement is most likely t o be successf ul f rom a t echnical st andpoint and
best t olerat ed w it h regard t o pat ient comf ort .
O ne f inal not e is t hat some experienced clinicians w ho believe in early and
aggressive ent eral f eedings do not support t he use of high volume ent eral
f eedings during t he init ial phase of act ive resuscit at ion f rom shock or sepsis or
in high-dose pressor use.
Suggested Readings
Jabbar A, McClave SA. Pre-pyloric versus post -pyloric f eeding. Clin Nut r.
2005; 24: 719–726.
Kaur N, G upt a MK, Minocha VR. Early ent eral f eeding by nasoent eric t ubes in
pat ient s w it h perf orat ion perit onit is. World J Surg. 2005; 29: 1023–1027.
Marik PE, Pinsky M. Deat h by parent eral nut rit ion. I nt ens Care Med.
2003; 29: 867–869.
Simpson F, Doig G S. Parent eral vs. ent eral nut rit ion in t he crit ically ill pat ient :
a met a-analysis of t rials using t he int ent ion t o t reat principle. I nt ens Care
Med. 2005; 31: 12– 23.
> Table of C ontents > Nutr ition > 235 - B e Aler t for O ver feeding
235
Be Alert for Overfeeding
Jason Sperry MD
Heidi L. Frankel MD
Providing inadequat e caloric supplement at ion during t imes of crit ical illness is
associat ed w it h negat ive eff ect s. Equally as det riment al is t he administ rat ion of
excessive calories, or overf eeding. O verf eeding is associat ed w it h signif icant
met abolic disorders including hyperglycemia, elevat ed serum t riglycerides, and
subsequent hepat ic st eat osis. I n addit ion, overf eeding may cause a signif icant
increase in CO2 product ion t hat can be delet erious t o t hose w it h respirat ory
insuff iciency and may make vent ilat or w eaning challenging (or impossible).
What to Do
To avoid overf eeding, accurat e est imat es of energy and caloric requirement s of
crit ically ill pat ient s are required. Crit ically ill pat ient s t ypically undergo a period
of cat abolism, w hich can be associat ed w it h signif icant body prot ein loss,
depending on t he severit y of crit ical illness and t he lengt h of t he cat abolic
process. Calculat ion of nit rogen balance can quant if y t he ext ent of cat abolism
and evaluat e t he eff icacy of supplement al nut rit ion in t hese pat ient s. Nit rogen
balance is calculat ed by subt ract ing t ot al nit rogen losses (urine, st ool, insensible
losses) f rom nit rogen int ake (1 g nit rogen per 6. 25 g of prot ein). The primary
mode of nit rogen excret ion is urinary, and 24-hour urine collect ion f or urinary
urea nit rogen (UUN) is t he most common means of measurement . Fecal and
insensible losses are t ypically small but est imat ed. UUN can be a poor est imat e
of overall nit rogen losses in burn pat ient s, w hen urine out put is low (<1L/ d), in
pat ient s w it h acut e or chronic renal f ailure and in pat ient s w it h ent eric f ist ulas
w here exceedingly high losses occur in t he f ist ulae out put .
Alt ernat ively, nut rit ional assessment can be accomplished via indirect
calorimet ry, w here O2 consumpt ion and CO2 product ion are measured and a
respirat ory quot ient (RQ ) is calculat ed at t he pat ient 's bedside. An RQ of 0. 7 is
t ypical of f at oxidat ion and an RQ of <0. 7 suggest s ket osis, lipolysis, and
underf eeding. An RQ ≥1. 0 exemplif ies primary carbohydrat e met abolism and
possible overf eeding. Because of t he accuracy of measurement s required f or
indirect calorimet ry, it is generally limit ed t o t hose w ho are on vent ilat ory
support . Since changes in minut e vent ilat ion, cardiac out put , f ract ion of inspired
oxygen, and acid-base st at us can aff ect CO2 product ion and O2 upt ake, indirect
calorimet ry must also be perf ormed, w it h t he pat ient at a relat ive st eady st at e
f or maximal accuracy.
Suggested Readings
Brandi LS, Bert olini R, Sant ini L, et al. Eff ect s of vent ilat or reset t ing on
indirect calorimet ry measurement in t he crit ically ill surgical pat ient . Crit Care
Med. 1999; 27(3): 531–539.
Dickerson RN, Tidw ell AC, Minard G , et al. Predict ing t ot al urinary nit rogen
excret ion f rom urinary urea nit rogen excret ion in mult iple-t rauma pat ient s
receiving specialized nut rit ional support . Nut rit ion. 2005; 21(3): 332–338.
Hunt er DC, Jaksic T, Lew is D, et al. Rest ing energy expendit ure in t he
crit ically ill: est imat ions versus measurement . Br J Surg. 1988; 75(9): 875–878.
Long CL, Schaff el N, G eiger JW, et al. Met abolic response t o injury and
illness: est imat ion of energy and prot ein needs f rom indirect calorimet ry and
nit rogen balance. J Parent er Ent eral Nut r. 1979; 3(6): 452–456.
Mann S, West enskow DR, Hout chens BA. Measured and predict ed caloric
expendit ure in t he acut ely ill. Crit Care Med. 1985; 13(3): 173–177.
> Table of C ontents > Nutr ition > 236 - B e Aler t for the D evelopm ent of R efeeding S yndr om e
236
Be Alert for the Development of Refeeding
Syndrome
William S. Hoff MD
Phosphat e (PO4 2- ) is t he most abundant int racellular anion, t he majorit y of

w hich is f ound in bones. As a result , t he measured serum PO4 2- does not
accurat ely ref lect t he t ot al-body cont ent of PO4 2- . Biologically, PO4 2- is an
import ant component of cell membranes (e. g. , phospholipids), enzyme syst ems,
and bone mat rix (e. g. , hydroxyapat it e). Phosphat e is also a component of
adenosine t riphosphat e (ATP), an import ant source of cellular energy. Synt hesis
of 2, 3-diphosphoglycerat e (2, 3-DPG ) also depends on PO4 2- .
Watch Out For

Hypophosphat emia is t he hallmark elect rolyt e abnormalit y associat ed w it h t he
ref eeding syndrome. Concomit ant serum abnormalit ies include hypokalemia,
hypomagnesemia, and hyperglycemia. Ref eeding syndrome is rare, occurring in
0. 8% of hospit alized adult s. How ever, t he mort alit y in hospit alized pat ient s w ho
develop severe hypophosphat emia (serum PO4 2- <<1 mg/ dL) is 30%. Risk
f act ors f or ref eeding syndrome include prolonged malnut rit ion, gast roint est inal
losses (e. g. , vomit ing, nasogast ric suct ioning, diarrhea), chronic alcohol abuse,
abdominal surgery, and met ast at ic cancer.
Ref eeding syndrome t ypically occurs w hen eit her ent eral or parent eral nut rit ion
is init iat ed in a pat ient w it h one of t he high-risk f act ors list ed. The sudden
int roduct ion of carbohydrat e st imulat es a release of insulin, w hich result s in an
int racellular shif t of PO4 2- , pot assium, and magnesium. This t ranscellular shif t
f requent ly occurs in t he presence of reduced body st ores of t hese key
elect rolyt es. Ref eeding syndrome should be suspect ed in t he presence of any of
t he elect rolyt e disorders not ed earlier. A common manif est at ion in t he int ensive
care unit (I CU) pat ient is respirat ory f ailure secondary t o diaphragm and
int ercost al muscle w eakness as a result of ATP deplet ion. This clinical
present at ion may be acut e respirat ory f ailure in a spont aneously breat hing
pat ient or inabilit y t o w ean a pat ient f rom mechanical vent ilat ion. Def iciency of
2, 3-DPG result ing in impaired O2 off -loading at t he cellular level may result in
dyspnea and f at igue, f urt her compromising respirat ion. Alt hough less specif ic in
t he I CU populat ion, pat ient s may also present w it h generalized skelet al muscle
w eakness
secondary t o deplet ed ATP st ores. O t her less common clinical f indings include
vent ricular dysrhyt hmias, lef t vent ricular dysf unct ion, conf usion, let hargy, and
gait dist urbances.
The key t o t reat ment of ref eeding syndrome is ant icipat ing development of t he
condit ion in pat ient s at risk. I n t his group of pat ient s, nut rit ion should be
int roduced gradually, f or example, beginning w it h a one-half nut rit ion goal of
5, 040 kilocalories and increasing by 840 kilocalories every 24 t o 48 hours, w it h
caref ul at t ent ion t o serum elect rolyt es (1 calorie = 4. 186 kJ). Act ive t reat ment
depends on t he degree of hypophosphat emia. Moderat e hypophosphat emia
(PO 4 2- = 1 t o 2. 5 mg/ dL) may be managed w it h oral or ent eral replacement .
Severe or sympt omat ic hypophosphat emia mandat es int ravenous replacement ; a
t ypical int ravenous replacement regimen is sodium phosphat e Na2 PO 4 0. 16 t o
0. 32 mmol/ kg administ ered over 4 t o 6 hours. I nt ravenous replacement should be
done caut iously in hypercalcemia. Serum elect rolyt es should be monit ored
closely during int ravenous replacement t herapy. I onized calcium levels should
also be monit ored as overzealous dosing of PO4 2- may result in hypocalcemia.
Suggested Readings
Fink MP, Abraham E, Vincent JL, eds. Text book of Crit ical Care. 5t h ed.
Khardori R. Ref eeding syndrome and hypophosphat emia. J I nt Care Med.

2005; 20: 174– 175.
Marinella MA. Ref eeding syndrome and hypophosphat emia. J I nt Care Med.
2005; 20: 155–159.
> Table of C ontents > Nutr ition > 237 - C ons ider Us ing E lem ental or S em ielem ental Feeds in P atients
w ith Album in Les s than 2.5 Gr am s P er D ec iliter
237
Consider Using Elemental or Semielemental
Feeds in Patients with Albumin Less than 2.5
Grams Per Deciliter
Watch Out For

Ent eral f eeding f ormulas can be cat egorized based on t heir prot ein source.
I nt act prot ein f eeds cont ain prot ein t hat has not been hydrolyzed and include
Ensure, O smolit e, Jevit y, and Promot e. Semielement al f eeds cont ain hydrolyzed
prot eins (small pept ides) as t heir source of nit rogen and include Pept amen,
Alit raq, and Perat ive. Element al f ormulas ut ilize f ree amino acids (FAA) as t heir
source of nit rogen and include Vivonex and F. A. A. Not all pat ient s can ut ilize
int act prot ein f eeds. I n pat ient s w it h hypoalbuminemia (albumin <2. 5) t here is
signif icant gut edema and an impairment of gut cell-w all prot eins necessary f or
nut rient absorpt ion. Use of element al or semielement al f ormulas is associat ed
w it h bet t er amino acid absorpt ion and insulin responses compared w it h int act
prot ein f ormulas. Semielement al f ormulas may be bet t er t olerat ed t han eit her
int act or element al f ormulas w it h regard t o diarrhea. Because element al f ormulas
are hypert onic t hey may also have a higher incidence of nausea, vomit ing, and
abdominal dist ension.
I n general, crit ically ill pat ient s are hypercat abolic and mildly hypermet abolic.
The goals of met abolic support st rat egies are t o preserve organ f unct ion, body
cell mass, and prot ein synt hesis. O f not e, providing high levels of nonprot ein
calories does not prevent prot ein cat abolism. O verf eeding can lead t o increased
product ion of carbon dioxide and int erf ere w it h w eaning f rom mechanical
vent ilat ion. Current recommendat ions suggest init iat ing nut rit ional support w it h
t ot al calories of approximat ely 84 t o 105 kilocalories/ kg/ d w it h prot ein cont ent of
1. 2 t o 1. 5 g/ kg/ d. Adjust ment of nonprot ein calories is based upon indirect
calorimet ry (met abolic cart ) t o ensure pat ient s are not over- or underf ed.
Assessment of t ot al prot ein needs and ut ilizat ion should be based on laborat ory
paramet ers (e. g. , prealbumin, t ransf errin) and nit rogen balance (urine urea
nit rogen [ UUN] ).
Crit ically ill pat ient s w it h ongoing syst emic inf lammat ory response syndrome,
w it h sepsis, or requiring vasopressors may be at part icular risk of
gast roint est inal complicat ions w it h ent eral f eeds. I n t hese pat ient s, parent eral
nut rit ion should be considered, alt hough w hen gut f unct ion is assessed t o be
present , ent eral f eeds are much
pref erable t o t he parent eral opt ion. Complicat ions of ent eral nut rit ion include
high gast ric residuals, diarrhea, const ipat ion, abdominal dist ension, vomit ing w it h
aspirat ion, and pneumonia. Assessment of gast ric residuals can be used t o
assess t olerance of ent eral f eeds. These can be checked at 1 and 4 hours and
t hen every 8 hours t o evaluat e w het her or not a plat eau has been reached.
Usually, residuals w ill plat eau at 20% t o 50% of t he volume inf used by 3 t o 13
hours af t er init iat ion of f eeds. Failure t o reach a plat eau suggest s t hat t he
inf usion should be st opped or slow ed dow n.
What to Do
Pat ient s receiving ent eral f eeds are at risk f or developing eit her diarrhea or
const ipat ion. Diarrhea can be addressed by met hods including t ransit ioning t o
semielement al f ormulas, decreasing osmolalit y by dilut ion of t he f ormula and
addit ion of ant idiarrheal agent s including kaolin-pect in, bismut h or salicylat e, or
deiodinized t inct ure of opium (DTO ). The import ance of f irst ruling out an
inf ect ious cause of diarrhea cannot be overst at ed. Cl ostri di um di f f i ci l e is a
commonly missed diagnosis w hen new onset diarrhea is erroneously at t ribut ed t o
t ube f eeds. Const ipat ion can be t reat ed by removal of ant idiarrheal t herapy (if
present ), correct ing dehydrat ion by t he addit ion of f ree-w at er f lushes, adding
f iber if using int act f ormulat ions, or adding laxat ives or enemas. Clinical
assessment should alw ays include caref ul at t ent ion t o signs and sympt oms
suggest ive of int olerance t o ent eral f eeds including abdominal dist ension, pain,
nausea, and/ or vomit ing.
Suggested Readings
G riff it hs RD, Bongers T. Nut rit ion support f or pat ient s in t he int ensive care
unit . Post grad Med J. 2005; 81: 629–636.
Mechanick JI , Bret t EM. Nut rit ion support of t he chronically crit ically ill
pat ient . Crit Care Clin. 2002; 18: 597–618.
> Table of C ontents > Nutr ition > 238 - C ons ider E nter al Feedings in P anc r eatitis and E nter oc utaneous
Fis tulae
238
Consider Enteral Feedings in Pancreatitis and
Enterocutaneous Fistulae
Christopher J. Sonnenday MD, MHS
Classic surgical t eaching has hist orically dict at ed t hat minimizat ion of ent eric
f low by w it hholding ent eral f eeds is essent ial t o t he management of pat ient s w it h
disrupt ion of t he normal cont inuit y of t he gast roint est inal t ract (i. e. ,
ent erocut aneous f ist ula). The use of parent eral nut rit ion t hus became
f undament al t o t he t reat ment of ent erocut aneous f ist ulae, as w ell as t o t he
management of severe acut e pancreat it is, w here ent eral f eeds t hat st imulat e
pancreat ic exocrine w ere believed t o cont ribut e t o exacerbat ion of t he complex
course of t hat disease. How ever, w it h a preponderance of evidence in f avor of
t he benef it s of ent eral nut rit ion in crit ical illness, clinicians have sought w ays t o
provide ent eral nut rit ion even in t hese disease st at es w here it w as previously
t hought t o be delet erious.
What to Do
I n t he case of proximal ent erocut aneous f ist ulae (e. g. , esophageal, gast ric,
duodenal, proximal jejunum), dist al jejunal f eeds are alw ays pref erable t o
parent eral nut rit ion and should be t he mainst ay of t herapy. The nut rit ional
requirement s f or pat ient s w it h ent erocut aneous f ist ula may be excessive because
of t he signif icant gast roint est inal (G I ) prot ein losses and cat abolism associat ed
w it h t hese condit ions. I t may be prudent t o maint ain t he pat ient on parent eral
nut rit ion unt il ent eral f eeds are advanced and opt imized. O pt imal caloric support
is f requent ly not possible f or 4 t o 5 days f ollow ing init iat ion of ent eral f eeds, and
parent eral nut rit ion can bridge t hat gap in crit ically ill pat ient s. I n t he case of a
more proximal f ist ula, it is of t en possible t o cannulat e t he dist al limb of t he
f ist ula it self t o f eed t he dist al small bow el. Such procedures of t en require expert
int ervent ional radiologist s and should not be perf ormed w it hout t he guidance of
an experienced G I surgeon.
Ent erocut aneous f ist ulae involving t he dist al small bow el may present more
diff icult challenges in est ablishing adequat e ent eral nut rit ion. I f at least 4 f eet of
small bow el f rom t he ligament of Treit z exist prior t o t he f ist ula sit e, t he st omach
or duodenum may be f ed w it h low -residue element al or semielement al f ormulas
t hat are most likely t o be absorbed in t he proximal gut (i. e. , bef ore it reaches
t he
f ist ula). Est ablishing a means t o supply ent eral nut rit ion w henever it is saf e and
f easible should be part of t he comprehensive management of an ent erocut aneous
f ist ula.
A similar paradigm shif t aw ay f rom parent eral nut rit ion has also occurred in t he
management of severe acut e pancreat it is. Tradit ional t eaching has argued t hat
ent eral f eeding w ould st imulat e t he pancreas, exacerbat ing t he ret roperit oneal
inf lammat ion t hat drives t he disease st at e. How ever, t w o recent prospect ive
t rials and at least t w o met a-analyses have show n pat ient out comes (including
mort alit y, need f or surgical int ervent ion, and hospit al lengt h of st ay) t o be at
least equivalent bet w een pat ient s f ed via ent eral versus parent eral rout es. I n
addit ion, inf ect ious complicat ions and cost are low er among pat ient s t reat ed w it h
ent eral nut rit ion. While all prospect ive t rials t o dat e have been small and based
at a single cent er, t he evidence is grow ing t hat ent eral nut rit ion is not harmf ul
and limit s t he risk of cent ral-line inf ect ions associat ed w it h parent eral nut rit ion.
Each of t he prospect ive t rials f or ent eral f eeding in pancreat it is have used
nasojejunal t ubes placed eit her endoscopically or under f luoroscopic guidance as
t heir delivery met hod f or ent eral f eeds. Early percut aneous gast rojejunost omy
t ubes should be considered in pat ient s project ed t o have a lengt hy clinical
course. While most expert s recommend f eeding dist al t o t he ligament of Treit z
as a w ay t o minimize pancreat ic st imulat ion, bot h animal st udies and anecdot al
report s in humans suggest t hat gast ric f eedings may be possible w it hout adverse
eff ect s. How ever, t his has not been st udied prospect ively and a jejunal rout e f or
ent eral nut rit ion should be achieved w henever possible. I n pat ient s w it h severe
acut e pancreat it is w ho require percut aneous or surgical int ervent ions, at t ent ion
should be paid t o int errupt ions in ent eral nut rit ion f or procedures and diagnost ic
st udies. I f ent eral nut rit ion is f requent ly int errupt ed, a t ransit ion period of
parent eral nut rit ion should be necessary. Any pat ient w ho requires a laparot omy
f or debridement and drainage should have a f eeding jejunost omy surgically
placed, if possible.
Watch Out For

Bot h of t hese groups of pat ient s require vigilant monit oring t o ensure t hat
nut rit ional supplement at ion is adequat e. Frequent biochemical monit oring of
serum markers such as albumin, prealbumin, t ransf errin, and lipid prof iles is
essent ial. Measurement of urine urea nit rogen excret ion should be perf ormed
w eekly in t he most crit ically ill pat ient s as t hey navigat e t he cat abolic phase of
t heir illness. Met abolic cart st udies may be necessary t o assist in t he est imat es
of energy requirement s in
t hese pat ient s w hose clinical condit ion, and t heref ore met abolic needs, change
rapidly. St udies have show n t hat a dedicat ed nut rit ion service can improve t he
out comes of t hese pat ient s and serious considerat ion should be given t o
obt aining consult at ion f rom t hem.
Suggested Readings
Dudrick SJ, Maharaj AR, McKelvey AA. Art if icial nut rit ional support in pat ient s
w it h gast roint est inal f ist ulas. World J Surg. 1999; 23: 570–576.
Heinrich S, Schaf er M, Rousson V, et al. Evidence-based t reat ment of acut e

pancreat it is: a look at est ablished paradigms. Ann Surg. 2006; 243: 154–168.
> Table of C ontents > Nutr ition > 239 - Us e The Metabolic C ar t only w hen P atients ar e on Low Vent
S ettings
239
Use The Metabolic Cart only when Patients are
on Low Vent Settings
Appropriat e nut rit ional support improves t he care of t he crit ically ill pat ient by
improving w ound healing, and by reducing inf lammat ion, inf ect ions, and t he rat e
of complicat ions. I n addit ion, it has been show n t o reduce overall vent ilat or days
and int ensive care unit (I CU) lengt h of st ay. Tradit ional approaches t o t he
assessment of nut rit ion in t he I CU have included a number of met abolic f ormulas.
The most w ell know n of t hese is t he Harris-Benedict equat ion, w hich measures
t he rest ing energy expendit ure (REE) and t hen “adjust s” t hat value based upon
t he overall st ressors of t he pat ient . The Harris-Benedict equat ion is represent ed
by t he f ollow ing f ormulas:
REE (men) = 66 + (13. 7 × w eight in kilos) + (5 × height in cent imet ers) - (6. 8 ×
age)
REE (w omen) = 655 + (9. 6 × w eight in kilos) + (1. 8 × height in cent imet ers) -
(4. 7 × age)
I n t he presence of invasive monit oring such as t he pulmonary art ery cat het er,
w here accurat e values f or oxygen consumpt ion are available, t he Weir equat ion
may allow a bet t er approximat ion of energy expendit ure.
energy expendit ure = (3. 94 × oxygen consumpt ion) + (1. 11 × carbon dioxide
product ion)
These values direct ly use t he component s of t he respirat ory quot ient (CO2
product ion/ O 2 consumpt ion) t o imput e t he energy expendit ure.
Apart f rom t he use of a calculat ed assessment of nut rit ional value, t here may be
opport unit ies t o improve upon t he assessment of t he met abolic st at us f or
pat ient s hospit alized in t he I CU. I t is likely t hat t he met abolic st at us of t he I CU
pat ient changes as his or her condit ion changes. The abilit y t o keep t rack of
t hese changes and t o modif y t herapy may be import ant in improving t he overall
out come of t he pat ient . I ndirect calorimet ry provides an import ant opport unit y t o
object ively measure t he met abolic needs of t he pat ient and t o mat ch t he
provision of nut rit ional support t o t he pat ient 's demand. This
is import ant part icularly w it h complex pat ient s w ho may need bot h ent eral and
parent eral support t o achieve nut rit ional goals.
Watch Out For

I ndirect calorimet ry is a simple bedside procedure t hat provides accurat e
measurement s f or oxygen consumpt ion, carbon dioxide product ion, and rest ing
energy expendit ure w hen t he abilit y t o det ect t hese measurement s f rom
pulmonary art ery cat het erizat ion is not available. While simple, t here are a
number of import ant caveat s t hat can aff ect t he obt ained measurement s. First ,
t here cannot be any air leaks f rom t he syst em since t hese leaks w ill cause
inaccuracies in t he measurement s of oxygen consumpt ion and carbon dioxide
product ion and w ill adversely aff ect t he calculat ions f or respirat ory quot ient s and
rest ing energy expendit ure. Second, a f ract ion of inspired oxygen (FI O2 ) of >0. 6
can also make t hese measurement s inaccurat e. Third, pat ient s need t o be in a
st eady st at e w it h lit t le f luct uat ion (≤10%) in oxygen consumpt ion or carbon
dioxide product ion. This limit at ion is of t en overcome by experienced clinicians by
increasing t he int erval of measurement f or t he pat ient t o 30 minut es.
Suggested Readings
I rw in RS, Rippe JM. I nt ensive Care Medicine. 5t h ed. Philadelphia: Lippincot t

Williams & Wilkins; 2003.
> Table of C ontents > Nutr ition > 240 - Us e a D edic ated, Upper - B ody, S ingle- Lum en C entr al Venous
C atheter for Adm inis tr ation of P ar enter al Nutr ition
240
Use a Dedicated, Upper-Body, Single-Lumen
Central Venous Catheter for Administration of
Parenteral Nutrition
Lisa Marcucci MD
Alt hough t he use of parent eral nut rit ion w as one of t he milest one breakt hroughs
in t he care of crit ically ill pat ient s, serious morbidit ies are associat ed w it h it s
use. O ne t roublesome complicat ion is t ot al parent eral nut rit ion (TPN) cat het er-
relat ed sepsis, w hich is associat ed w it h poorer out comes in pat ient s, including
longer int ensive care unit (I CU) st ays, longer hospit al st ays, and higher
mort alit y. To t his end, some hospit als have developed ent ire care t eams
dedicat ed t o t he management of cent ral venous lines used f or TPN, including
insert ion procedure, skin sit e inspect ion, dressing changes, and evaluat ion of
possible inf ect ion. G iven t he pot ent ial morbidit ies, numerous st udies have been
done t o det ermine t he best prot ocols f or cat het er care used f or TPN.
From an accumulat ing body of research regarding how t o reduce risks of
inf ect ion, including TPN cat het er t ip inf ect ion, sit e inf ect ion, bact eremia, and
cat het er colonizat ion, t he f ollow ing f act s seems clear: (a) TPN should be
init iat ed t hrough a new cat het er insert ed via a clean st ick, not a cat het er
changed over a w ire; (b) TPN cat het ers should be insert ed via subclavian
(pref erably) or int ernal jugular veins, not via f emoral veins; (c) TPN should have
it s ow n dedicat ed lumen used f or not hing else; and (d) a t eam dedicat ed t o TPN
cent ral-line care should be assembled t o low er cat het er inf ect ion rat es. A 2003
st udy by Dimick et al. show ed t hat single-lumen subclavian cat het ers dedicat ed
t o TPN use and cared f or by a dedicat ed t eam had a f ivef old decrease in
inf ect ion compared w it h ot her cat het ers t hat w ere used f or mult iple reasons.
O ne f inal not e is t hat because t he realit ies of care in seriously ill pat ient s do not
alw ays allow f or t he placement of a dedicat ed TPN line, one port of a new ly
placed mult iple-lumen cat het er can be designat ed f or exclusive administ rat ion of
TPN if no ot her opt ions exist .
Suggested Readings
Clark-Christ off N, Wat t ers VA, Sparks W, et al. Use of t riple-lumen subclavian
cat het ers f or administ rat ion of t ot al parent eral nut rit ion. J Parent er Ent eral
Nut r. 1992; 16: 403–407.
Dimick JB, Sw oboda S, Talamini MA, et al. Risk of colonizat ion of cent ral
venous cat het ers: cat het ers f or t ot al parent eral nut rit ion vs ot her cat het ers.
Am J Crit Care. 2003; 12: 328–335.
Kemp L, Burge J, Choban P, et al. The eff ect of cat het er t ype and sit e on
inf ect ion rat es in t ot al parent eral nut rit ion pat ient s. J Parent er Ent eral Nut r.
1994; 18: 71–74.
> Table of C ontents > R enal > 241 - B e Aler t for Hypophos phatem ia in the Intens ive C ar e Unit P atient
on D ialys is
241
Be Alert for Hypophosphatemia in the Intensive
Care Unit Patient on Dialysis
Adam R. Berliner MD
Derek M. Fine MD
O ut pat ient s on chronic dialysis usually have t rouble w it h hyperphosphat emia,
requiring high doses of oral phosphorus binders t o block absorpt ion of diet ary
phosphat e. I n t he int ensive care set t ing, t he opposit e sit uat ion—
hypophosphat emia—is of t en encount ered in pat ient s on dialysis f or several
reasons including higher f requency of dialysis and relat ive malnut rit ion. I t is
crit ical t o check t he serum phosphat e level daily in t he int ensive care unit (I CU)
pat ient on hemodialysis. A level t hat may have been quit e high prior t o dialysis
can plummet t o crit ical values if unmonit ored. Severe (<1. 5 mg/ dL)
hypophosphat emia can cause diaphragmat ic w eakness, apnea, cardiac
inst abilit y, rhabdomyolysis, or hemolysis.
Convent ional int ermit t ent hemodialysis is int rinsically very eff ect ive at phosphat e
removal, but t he overall t reat ment t ime (usually 2 t o 4 hours) is of t en not enough
t o cause prof ound hypophosphat emia, since ext racellular st ores are rapidly
replet ed f rom int racellular reserve. I n t he I CU set t ing it is common t o use
sust ained, high-int ensit y dialysis, usually in t he f orm of a cont inuous renal
replacement t herapy such as cont inuous venovenous hemof ilt rat ion (CVVH) or
cont inuous venovenous hemodiaf ilt rat ion (CVVHDF). A ret rospect ive st udy
show ed t hat CVVHDF curbs hyperphosphat emia bet t er t han int ermit t ent
hemodialysis, but at an increased risk of hypophosphat emia. I n addit ion t o
cont inuous renal replacement t herapy, ot her risk f act ors f or hypophosphat emia in
t he I CU include parent eral nut rit ion, diarrhea, and prof ound renal phosphat e
w ast ing due t o t ubular cell injury.
What to Do
I nt ermit t ent int ravenous replet ion w it h sodium phosphat e is recommended f or
serum phosphat e <1. 5 mg/ dL. Dosing regimens are not st andardized. O ne small
st udy of seven pat ient s not ed good result s f rom 15 mmol int ravenous sodium
phosphat e over 2 hours f or serumphosphat es of 1. 56 t o 2. 1 mg/ dL (0. 5 t o 0. 7
mmol/ L). The aut hors of t his st udy calculat ed a volume of dist ribut ion f or
phosphat e ranging f rom 0. 21 t o 0. 87 L/ kg and w ent on t o post ulat e t hat doses
closer t o 25 mmol should be given f or values closer t o 1. 09mg/ dL (0. 35 mmol/ L)
or less. O nce correct ed, chronic maint enance phosphat e supplement at ion, eit her
orally or in
parent eral nut rit ion, should be given w hen phosphat e is chronically low in order
t o prevent reoccurrence of crit ical hypophosphat emia.
O ne f inal not e is t hat st andard dialysat e cont ains no phosphat e. Phosphat e-
enriched dialysat e can be considered in ref ract ory cases, t hough it is usually not
required.
Suggested Readings
French C, Bellomo R. Arapid int ravenous phosphat e replacement prot ocol f or
crit ically ill pat ient s. Crit Care Resusc. 2004; 6: 175–179.
Subramanian R, Khardor R. Severe hypophosphat emia: pat hophysiologic

implicat ions, clinical present at ions, and t reat ment . Medicine. 2000; 79: 1–8.
Tan HK, Bellomo R, et al. Phosphat emic cont rol during acut e renal f ailure:
int ermit t ent hemodialysis versus cont inuous hemodiaf ilt rat ion. I nt J Art if
O rgans. 2001; 24(4): 186–191.
> Table of C ontents > R enal > 242 - K now the D r ugs that Mus t be R edos ed after D ialys is
242
Know the Drugs that Must be Redosed after
Dialysis
Acut e renal f ailure is a common disease encount ered in crit ically ill pat ient s, w it h
report ed incidences of 5% t o 10%. Hemodialysis and cont inuous venovenous
hemodialysis (CVVHD) are commonly used in int ensive care unit (I CU) pat ient s t o
t reat acut e renal f ailure, and t here are dat a t o suggest t hat increasing t he dose
of dialysis can improve mort alit y out comes in t hese pat ient s. I t is common sense
t hat as t he dose of dialysis is increased, clearance w ill increase, and t heref ore
medicat ions w ill need t o be dosed more f requent ly. When considering t his issue,
one must evaluat e various drug charact erist ics t hat w ill require more f requent
dosing and close monit oring f or t herapeut ic f ailure.
What to Do
When evaluat ing drug regimen adjust ment s in acut e renal f ailure pat ient s on
hemodialysis or CVVHD, one must f irst consider w het her or not t he drug is
signif icant ly cleared t hrough t he kidneys. I f t he drug has less t han 30% of it s
dose cleared renally, t hen dialysis is unlikely t o play a signif icant role in it s
eliminat ion. O t her drug charact erist ics t hat play a signif icant role in det ermining
t he eff ect of dialysis on t he need t o redose a medicat ion include prot ein binding,
volume of dist ribut ion (Vd), molecular size, and molecular charge.
Wit h respect t o prot ein binding, only t he unbound, or f ree, concent rat ion is
available t o be dialyzed out of t he blood. Drugs t hat are >80% bound are unlikely
t o be cleared by dialysis. Many physiologic derangement s can inf luence drug
binding, such as acidosis, uremia, hypoalbuminemia, and hyperbilirubinemia.
Drug int eract ions can also cause displacement of bound drug t hrough compet it ion
f or binding sit es (sulf amet hoxazole and t rimet hoprim are classic culprit s of t his).
While f ree drug is more available t o exert it s pharmacologic act ivit y, it is also
available t o be cleared as w ell.
Volume of dist ribut ion can be considered synonymous w it h plasma prot ein
binding in t hat only f ree drug available in t he blood can be cleared by
hemodialysis or CVVHD. Drugs w it h large volumes of dist ribut ion or increased
lipophilicit y or t issue binding w ill have relat ively small concent rat ions in t he
serum compared w it h t he
t ot al-body concent rat ion. I n general, drugs w it h Vd >0. 7 L/ kg w ill be less likely
t o be eff ect ively dialyzed w it h hemodialysis. This is because of t he usual short
durat ion of hemodialysis, w hich allow s f or a redist ribut ion phenomenon t o occur
w here drug f rom t he t issues w ill dist ribut e back t o t he serum af t er t he cessat ion
of hemodialysis. Clearance of t hese t ypes of drugs w ill be increased w it h
CVVHD because of t he slow, cont inuous nat ure of t his modalit y t hat allow s f or a
cont inuous removal of drug f rom t he serum. An example t hat illust rat es t his is t he
vancomycin dosing diff erences bet w een hemodialysis and CVVHD. Vancomycin
has a Vd of 0. 7 L/ kg and usually requires dosing once per w eek w it h regular
hemodialysis schedules (4 hours per day, 3 t imes per w eek). Wit h CVVHD,
vancomycin doses are usually required every 36 t o 48 hours.
TABLE 242-1 M EDICATIONS AND ADJUSTED DOSA
CRCL
<10ML/MIN
MEDICAT ION HEMODIALYSIS C
(ESRD
NOT HD)
16 mg/kg
IV ×1;
16 mg/kg IV ×1;
then start
Amikacin then dose based
regimen
on levels
based on
levels
Ampicillin 2 g q8h 2 g q8h
1 g q24h
Cefazolin 1 g q24h immediately after
dialysis
500 mg q24h 1 g
Cefepime 0.5 g q24h q24h
(Pseudomonas)
No dosage No dosage
Ceftriaxone
adjustment adjustment
IV: 400 mg q24h

IV: 400 mg
after dialysis PO:
q24h PO:
Ciprofloxacin 500 mg q24h
500 mg
immediately after
q24h
dialysis
No dosage No dosage
Clindamycin
400 mg
200 mg q24h
once; then
Gatifloxacin immediately after
200 mg
dialysis
q24h
4 mg/kg IV
×1; then
4 mg/kg IV ×1;
start
Gentamicin then dose based
regimen
on levels
based on
levels
500 mg q24h
Meropenem 0.5 g q24h
after dialysis
Oxacillin No dosage No dosage
3–4 g 3 g q12h after

Piperacillin
q12h dialysis a
Piperacillin/tazobactam 2.25 g q8h 2.25 g q8h
See
Tobramycin See gentamicin
gentamicin
10–15
Loading dose
mg/kg for
15–20 mg/kg;
one dose;
Vancomycin then check
then check
levels
level in 24
posthemodialysis
hours
IV, intravenously; PO, by mouth

a Dose to be given immediately after dialysis and then every
Molecular w eight also needs t o be considered w hen examining t he inf luence of

dialysis on drug clearance. Solut es, including drugs, t hat have a molecular
w eight less t han 500 d are readily cleared by hemodialysis, w here as CVVHD
can clear molecules t hat are much larger. This is because of t he high-f lux f ilt ers
t hat CVVHD machines ut ilize and t hat have t he capacit y t o remove solut es t hat
have a molecular size up t o 30 kd. Again, vancomycin illust rat es t his diff erence
nicely. I t has a molecular w eight of about 1, 450 d; t hus it is not removed w ell by
hemodialysis but is removed f airly w ell by CVVHD.
Taking all of t his int o account , it is clear t hat many of t he drugs t hat require
close monit oring f or dosing in dialysis pat ient s are ant ibiot ics. Unf ort unat ely, t he
only agent s in t his class t hat are able t o be monit ored adequat ely t hrough
obt aining serum concent rat ions are t he aminoglycosides (gent amicin, t obramycin,
and amikacin) and vancomycin. Tabl e 242. 1 det ails some import ant ant ibiot ic
dosing inf ormat ion in hemodialysis and CVVHD. When inf ormat ion is not available
in t ext books or handbooks, a search of primary lit erat ure is necessary as t he
dat a around t his issue are const ant ly evolving.
Suggested Readings
Bugge JF. Pharmacokinet ics and drug dosing adjust ment s during cont inuous
venovenous hemof ilt rat ion or hemodiaf ilt rat ion in crit ically ill pat ient s. Act a
Anaest hesiol Scand. 2001; 45: 929–934.
Joy MS, Mat zke G R, Armst rong DK, et al. A primer on cont inuous renal
replacement t herapy f or crit ically ill pat ient s. Annal Pharmacot her.
1998; 32(3): 362–375.
Schiff l H, Lang SM, Fischer R. Daily hemodialysis and t he out come of acut e
renal f ailure. N Engl J Med. 2002; 346: 305–310.
Uchino S, Kellum JA, Bellomo R, et al. Acut e renal f ailure in crit ically ill
pat ient s: a mult inat ional, mult icent er st udy. JAMA. 2005; 294(7): 813–818.
Velt ri MA, Neu AM, Fivush BA, et al. Drug dosing during int ermit t ent
hemodialysis and cont inuous renal replacement t herapy: special
considerat ions in pediat ric pat ient s. Pediat r Drugs. 2004; 6(1): 45–65.
> Table of C ontents > R enal > 243 - R em em ber that C ontinuous Venovenous Hem odialys is c an
O bs c ur e a Tem per atur e S pike
243
Remember that Continuous Venovenous
Hemodialysis can Obscure a Temperature Spike
Frank Rosemeier MD
I n t he normal person, body core t emperat ure is t ight ly regulat ed around a set
value of 37° C ± 0. 2° by hypot halamic t hermorecept ors, despit e f luct uat ion in
heat upt ake, heat product ion, and heat loss. Diurnal variat ions of ± 0. 5° C occur
w it h t he minimum value in t he morning and maximum in t he evening.
I n crit ical care pat ient s, t his normal t hermoregulat ion can be compromised by
several f act ors and can cause pert urbat ions of bot h hyper-and hypot hermia. For
inst ance, not every f ever is caused by inf ect ion. Bot h exogenous (e. g. , bact eria)
and endogenous pyrogens can set t he “t hermost at ” at a higher value. I nf ect ion,
inf lammat ion, and necrosis act ivat e macrophages, w hich in t urn release
int erleukin 1 and int erleukin 6 and increase prost aglandin met abolism. An init ial
but t emporary t emperat ure rise f ollow ing surgery is classically t hought t o be part
of an inf lammat ory response. Suspicion f or an inf ect ive cause is raised by
sust ained, recurrent , and/ or prolonged period of t emperat ure t ypically >38. 3° C.
I n addit ion, t hermoregulat ion is compromised by a combinat ion of cent ral nervous
syst em processes (dysregulat ion of t he aut onomic nervous syst em, inabilit y t o
engage skelet al muscles in shivering and heat product ion) and administ rat ion of
drugs such as acet aminophen, nonst eriodal ant i-inf lammat ory drugs (NSAI Ds),
opioids, propof ol, and neuromuscular blockers. Room-t emperat ure int ravenous
f luids and cold blood or blood product s also cont ribut e t o induced hypot hermia.
Environment al f act ors such as exposure t o laminar airf low and cold room
t emperat ure (w hich is t hermoneut ral and comf ort able t o w ork in f or t he clot hed
st aff , yet t oo low f or t he pat ient covered by a t hin blanket ) cont ribut e f urt her t o
heat loss.
Watch Out For

O ne part icular environment al f act or t hat is not orious f or low ering core
t emperat ure is cont inuous venovenous hemodialysis (CVVHD). This is a f orm of
renal replacement t herapy of t en encount ered in crit ically ill pat ient s w it h acut e
renal f ailure in w hich solut e is separat ed f rom blood across a semipermeable
membrane as a result of a diff usion gradient . Bot h t he aff erent limb and eff erent
limb are connect ed t o
a large vessel such as t he pat ient 's f emoral, subclavian, or int ernal jugular vein
using a large-bore, double-lumen cat het er (e. g. , Shiley or Davol cat het er). Most
of t he CVVHD machines in use t oday do not have an ext ernal heat ing source.
Hence, blood is cooled t hrough convect ive heat loss in t he ext racorporeal circuit
as t he blood t ubing is exposed. Room-t emperat ure dialysat e in count ercurrent t o
t he blood f low increases conduct ive heat loss. Large t emperat ure drops of >1° C
can be observed, rendering t he normot hermic pat ient hypot hermic and masking a
f ever in t he inf ect ed pat ient .
Suggested Readings
Fink MP, Abrahams E, Vincent J-L, et al. , eds. Text book of Crit ical Care. 5t h
ed. Philadelphia: Elsevier Saunders; 2005: 1151–1158.
Miller RD, ed. Miller's Anest hesia. 6t h ed. Philadelphia: Elsevier-Churchill

Livingst one; 2005: 1571–1592.
> Table of C ontents > R enal > 244 - C lam p the D ialys ate Line in C ontinuous Venovenous Hem odialys is
Im m ediately if it B ec om es P ink Tinged
244
Clamp the Dialysate Line in Continuous
Venovenous Hemodialysis Immediately if it
Becomes Pink Tinged
Acut e renal f ailure is a commonly occurring problem in t he care of int ensive care
unit (I CU) pat ient s. O ne met hod of improving t he care of t hese complex pat ient s
is t o use cont inuous renal replacement t herapies (CRRTs), of w hich cont inuous
venovenous hemof ilt rat ion w it h dialysis (CVVHD) is one of t he more common
f orms. This t echnique has t he same indicat ions as t radit ional dialysis
(hyperkalemia, ref ract ory acidosis, f luid overload, and sympt omat ic azot emia),
but CVVHD supposedly is a gent ler, more physiologic procedure t o achieve
solut e and f luid clearance and may be part icularly usef ul f or t hose pat ient s w it h
hemodynamic inst abilit y. While t his is t he commonly accept ed dogma in crit ical
care, it is import ant t o recognize t hat t here is current ly no lit erat ure t hat
demonst rat es a survival advant age in CRRT w hen compared w it h int ermit t ent
hemodialysis.
A simplif ied, schemat ized version of t he mechanics of CVVHD is show n in Fi gure
244. 1. There are several st eps t o be complet ed w hen init iat ing CVVHD. First ,
t he placement of a double-lumen, large-bore cent ral venous cat het er (e. g. ,
Shiley) f or access is accomplished. Next , blood f rom t his cat het er is w it hdraw n
at approximat ely 150 t o 200 mL/ min by a pump and passed t hrough a f ilt er w it h
mult iple layers of a semipermeable membrane. This membrane allow s f or t he
diff usion of f luid and solut e across t he membrane w it h t he result ing f ormat ion of
a f ilt rat e t hat is normally clear in appearance. Finally, t he blood f low s back t o
t he pat ient f rom t he ot her end of t he f ilt er by a venous ret urn line t hat in most
syst ems has a t rap f or clot s and an air det ect or. This line is reconst it ut ed w it h a
cryst alloid solut ion t hat mimics t he composit ion of plasma. I t is a necessit y t hat
t he blood f low in t he dialysis circuit cont inues w it hout st agnat ion so t hat clot s do
not f orm. Ant icoagulat ion is usually administ ered eit her t hrough t he use of
heparin or cit rat e-cont aining replacement f luids t o ensure an adequat e blood
f low.
Watch Out For
I t is import ant t o not e t hat t he f ilt er used in CVVHD maint ains a barrier bet w een
t he blood column and t he dialysis column. This f ilt er has
a number of charact erist ics t hat allow it t o meet t hreshold crit eria f or f ilt rat ion.
Several complicat ions can occur relat ed t o t he f ilt er, and it requires appropriat e
monit oring of t he device by t he provider. Filt ers t ypically last at least 3 days.
Dialysis f ilt ers have an int rinsic f ailure rat e. The longer a f ilt er is in use, t he
higher t he likelihood it w ill f ail over t ime. O ne met hod of monit oring a f ilt er's
f unct ion is t hrough t he use of plasma-f ree hemoglobin. Plasma-f ree hemoglobin
increases w hen eryt hrocyt e breakdow n product s begin t o accumulat e and signals
t he provider t hat a f ilt er change may be necessary. Changing t he circuit bef ore
t he f ilt er becomes dysf unct ional is import ant . Anot her f ilt er complicat ion t hat can
occur is a breech in t he cont inuit y of t he membrane. This complicat ion w ill
usually manif est it self as a pink-t inged or f rank blood in t he f ilt rat e. I f t his
occurs, it is import ant t o immediat ely clamp t he dialysis cat het er f low ing f rom
t he pat ient and assess t he int egrit y of t he f ilt er. I f t he membrane in t he f ilt er
rupt ures, t he semipermeable membrane w ill no longer separat e t he blood column
f rom t he f ilt rat e column and t he pat ient can receive a high concent rat ion of
delet erious solut ion t hrough t he venous line. For t his reason,
it is imperat ive t o clamp t he t ubing immediat ely if t he f luid becomes pink t inged.
FIG URE 244. 1 Schemat ic of CVVHD

Suggested Readings
Crit ical Care Medicine Tut orials. Renal Replacement Therapy.
ht t p: / / w w w. ccmt ut orials. com/ renal/ RRT/ index. ht m
Philadelphia: Elsevier Saunders; 2005: 1151â 1158.
> Table of C ontents > R enal > 245 - D o not us e C ontinuous Venovenous Hem odialys is in the S etting of
Angiotens in- C onver ting E nz ym e 2 and Vic e Ver s a
245
Do not use Continuous Venovenous
Hemodialysis in the Setting of Angiotensin-
Converting Enzyme 2 and Vice Versa
Kinins are a group of pept ides t hat have st rong vasodilat ing propert ies. They are
f ormed f rom large precursor molecules named kininogens by t he act ion of
prot eolyt ic enzymes called kallikreins, w hich are f ormed in t he liver. Kallikreins
exist peripherally as prekallikreins, w hich become act ivat ed t hrough a number of
diff erent mechanisms including exposure t o Hageman f act or (f act or XI I ) f rom t he
coagulat ion syst em and art if icial surf aces including dialysis membranes. Rapid
deact ivat ion occurs f or most kinins af t er f ormat ion t o prevent t he body f rom
having an overw helming response t o t hese pow erf ul subst ances.
Bradykinin, how ever, is an example of a kinin t hat has a longer half -lif e. I t is
f ormed f rom it s precursor kallidin, w hich is a product of low -molecular-w eight
kininogen in t he t issues. Bradykinin is a physiologically act ive product t hat has
several f unct ions. I t can act on t he vascular endot helium t o reduce smoot h
muscle t one and cause vasodilat at ion. I t is also act ive during acut e inf lammat ion
as a signal t o t he generat ion of arachidonic acid–derived product s such as
prost aglandins and leukot rienes. I n addit ion, bradykinin becomes act ivat ed
during act ivat ion of t he clot t ing cascade and t issue repair mechanisms. Finally,
t here is a pot ent ial role f or bradykinin in hypert ension since it is involved in t he
conversion of prorenin t o renin and levels appear t o be reduced in pat ient s w it h
hypert ension.
What to Do
I t is import ant f or t he clinician t o not e t hat hypersensit ivit y react ions t hat cause
alt erat ions in blood pressure have been described in pat ient s t reat ed w it h
polyacrylonit rile dialysis membranes (commonly used in cont inuous dialysis
t reat ment ) w ho are concurrent ly t aking angiot ensin-convert ing enzyme (ACE)
inhibit ors f or t heir hypert ension. The suspect ed hypersensit ivit y react ion occurs
because bradykinin is produced and it s degradat ion is inhibit ed in t he set t ing of
ACE administ rat ion, leading t o a longer half -lif e and t hereby magnif ying it s
eff ect s on t he cardiovascular syst em and leading t o hypot ension. Several
st rat egies exist f or reducing t his bradykinin response t o t he dialysis membrane
including changing t he pH of t he rinsing solut ion, t hereby alt ering it s charge
and act ivat ion. Anot her approach is t o use alt ernat ive dialyzers since not all
membranes evoke t he same level of response. I n animal st udies, it appears as
t hough t he response is relat ed t o t he dose of t he ACE inhibit or. So, an addit ional
one opt ion may be t o reduce or eliminat e t he ACE inhibit or dose and use
alt ernat ive met hods of blood pressure cont rol f or t hese pat ient s.
Suggested Readings
Amore A, G uarnieri G , At t i M, et al. Use of alkaline rinsing solut ion t o prevent
hypersensit ivit y react ions during hemodialysis: dat a f rom a mult icent re
ret rospect ive analysis. J Nephrol. 1999; 12: 383–389.
Kriet er DH, G rude M, Lemke HD, et al. Anaphylact oid react ions during
hemodialysis in sheep are ACE inhibit or dose-dependent and mediat ed by
bradykinin. Kidney I nt . 1998; 53: 1026–1035.
Schulman G , I kizler TA, Hakim R. Angiot ensin-convert ing inhibit ors and
hemodialysis membranes. Semin Dial. 1999; 12: S88–S91.
Wakasa M, Akizaw a T, Kinugasa E, et al. Plasma bradykinin levels during

hemodialysis w it h PAN DX and polysulf one membranes w it h and w it hout
concurrent ACE inhibit or. Clin Nephrol. 1995; 44: S29–S32.
> Table of C ontents > R enal > 246 - D o Not Give Fludr oc or tis one to P atients on D ialys is
246
Do Not Give Fludrocortisone to Patients on
Dialysis
At t imes of st ress or af t er major t raumat ic injury, t he hypot halamic-pit uit ary-
adrenal axis provides an import ant mechanism t o support t he body t hrough
diurnal rhyt hms. The hypot halamus releases cort icot ropin-releasing f act or, w hich
st imulat es t he ant erior pit uit ary gland t o produce adrenocort icot ropic hormone
(ACTH). The ACTH t hen st imulat es t he cort ex of t he adrenal gland t o release
glucocort icoids and mineralocort icoids. G lucocort icoids primarily aff ect
carbohydrat e and prot ein met abolism. These chemicals include cort isol,
cort isone, and hydrocort isone. Mineralocort icoids primarily aff ect sodium and
w at er met abolism.
Mineralocort icoids lead t o t he ret ent ion of sodium and secret ion of pot assium. I n
addit ion, calcium, phosphorous, and bicarbonat e are also reabsorbed. These
act ions are primarily mediat ed by t he reninangiot ensin-aldost erone syst em.
Angiot ensin 2 produced by t his syst em leads t o aldost erone product ion, w hich is
a nat urally occurring mineralocort icoid. Fludrocort isone is t he pharmacological
f orm of mineralocort icoids and is used t o provide mineralocort icoid eff ect s in
condit ions like syncope, ort host at ic hypot ension, and Addison disease.
What Not to Do
Pat ient s w it h end-st age renal disease and a low glomerular f ilt rat ion rat e of t en
require dialysis t o appropriat ely remove urea, f luid, and elect rolyt es f rom t heir
body in t he absence of f unct ioning kidneys. The administ rat ion of f ludrocort isone
t o t his group of pat ient s may be relat ively cont raindicat ed f or a number of
reasons, including t he exacerbat ion of hypert ension and t he ext rarenal
absorpt ion of many of t he elect rolyt es t hat dialysis is at t empt ing t o remove.
These agent s lead t o sodium reabsorpt ion and t heref ore f luid ret ent ion. The
sodium reabsorpt ion occurs f rom sit es t hat are remot e f rom t he kidney, including
t he sw eat glands and gast roint est inal t ract . I n addit ion, calcium and phosphorous
ret ent ion can occur w it h t his agent . Pat ient s on dialysis are of t en t aking calcium
and phosphorous binders t o improve t he excret ion of t hese element s. The
administ rat ion of mineralocort icoids may w orsen t he ret ent ion of t hese
elect rolyt es and lead t o an increase in t he medicat ions necessary t o cont rol t he
underlying renal f ailure.
Suggested Readings
Furuya R, Kumagai H, Sakao T, et al. Pot assium low ering eff ect of
mineralocort icoid t herapy in pat ient s undergoing hemodialysis. Nephron.
2002; 92: 576–581.
Vlassopoulos D, Sonikian M, Dardiot i V, et al. I nsulin and mineralocort icoids

inf luence on ext rarenal pot assium met abolism in chronic hemodialysis
pat ient s. Ren Fail. 2001; 23: 833–842.
> Table of C ontents > R enal > 247 - Avoid the S ubc lavian Vein for C entr al Ac c es s of any Type in a
D ialys is P atient or P os s ible D ialys is P atient
247
Avoid the Subclavian Vein for Central Access of
any Type in a Dialysis Patient or Possible
Dialysis Patient
There are almost 300, 000 hemodialysis pat ient s in t he Unit ed St at es t oday, and
t he number increases by about 4% t o 5% annually. The median age of t he
dialysis pat ient in t he Unit ed St at es has risen t o 65 years. The aging populat ion
and t he relat ive scarcit y of renal t ransplant s (about 10, 000 done annually in t he
Unit ed St at es) means t hat dialysis w ill be required f or longer periods of t ime and
in many pat ient s w ill be lif elong. I n parallel, t he crit ical nat ure of t he vascular
access in hemodialysis pat ient s is amplif ied by longer dependence on t his
access f or dialysis. Provision of vascular access is t he great est problem in
dialysis t oday, w it h t he solut ion involving t he pat ient , nephrologist , surgeon, and
int ervent ional radiologist .
I n t he opt imal sit uat ion, t he pat ient present s early enough t hat t he surgeon can
place a permanent access in advance of dialysis. How ever, more commonly,
pat ient s present w it h an acut e need f or dialysis t hat requires t emporary dialysis
access via percut aneous cat het ers. As t he majorit y of t hese pat ient s w ill go on
t o require chronic dialysis, it is vit al t hat t he t emporary-access cat het er not
compromise t he anat omy required f or permanent -access procedures. The
pref erred sit es f or t emporary-access cat het er placement are t he int ernal jugular
veins and t he f emoral veins; t he subclavian veins should be avoided.
What Not to Do
Temporary percut aneous dialysis access cat het ers are placed via cent ral veins.
They are very large in diamet er compared w it h ot her venous cat het ers and can
be associat ed w it h cannulat ed vein and an increased risk of t hrombosis or
st enosis of t he vein, w hich can occur acut ely or can present much lat er. The
venous damage can be at t he sit e of ent rance int o t he vein or more cent rally. I n
st udies examining damage f rom short -t erm use (2 t o 4 w eeks), dialysis cat het ers
placed by t he subclavian rout e result ed in venous st enosis or t hrombosis in 50%
t o 70% of veins, in cont rast t o a 0% t o 10% incidence in cat het ers placed by t he
int ernal jugular rout e. Alt hough t he risk of injury t o t he vein is less t han w it h
dialysis cat het ers, in dialysis pat ient s t he subclavian vein should be avoided f or
placement of smaller-bore cent ral
venous cat het ers such as t riple-lumen cat het ers, TPN (t ot al parent eral nut rit ion)
lines, or cat het ers t hat permit pulmonary art ery cat het ers t o be t hreaded t hrough
t hem.
Permanent accesses are art erial-t o-venous connect ions, eit her as a f ist ula
(direct art ery-t o-vein anast omosis) or via a bridge prost het ic graf t bet w een
art ery and vein. The best , longest -last ing permanent accesses are placed in t he
ext remit ies, and upper ext remit ies are pref erred over low er ext remit ies f or
reasons of inf ect ion, edema, and pat ient comf ort . Permanent accesses creat e
high f low t hrough t he vessels of t he ext remit y, and t he success of t he access is
highly dependent on adequat e venous out f low f rom t he access t o t he right
at rium. Compromise of venous ret urn result s in access f ailure or edema of t he
access ext remit y t o a degree t hat can be prof ound and limb t hreat ening. Thus,
subclavian vein st enosis or t hrombosis (even if clinically silent ) result s in loss of
all pot ent ial access sit es in t he ipsilat eral upper ext remit y. Placement of a
permanent access in such an ext remit y w ill result in immediat e clinical sympt oms
of venous insuff iciency (edema, cellulit is) and usually requires undoing t he
art eriovenous connect ion.
Pat ient s w ho require a permanent dialysis access af t er prior subclavian venous
punct ure should be screened f or pat ency of t he cent ral veins. Screening w it h
Doppler ult rasound is noninvasive and is sensit ive f or veins t hat are easily
visualized (e. g. , jugular or axillary veins), but it is not sensit ive f or det ect ion of
venous st enosis or t hrombosis of int rat horacic cent ral veins such as subclavian
and brachiocephalic (innominat e) veins. I f t here is any reason t o suspect
problems w it h more cent ral veins, a more sensit ive st udy such as magnet ic
resonance imaging (MRI ) or venography w ould be indicat ed.
Finally, pat ient s w it h cent ral venous occlusion may have spont aneous
recanalizat ion af t er 3 t o 6 mont hs, alt hough t he vein w ill never be complet ely
normal. Pat ient s w it h sympt omat ic st enosis of cent ral veins can be t reat ed w it h
angioplast y and st ent placement . The result s overall are not as good as f or
art erial disease but can be helpf ul in pat ient s w it h f ew or disappearing dialysis
access alt ernat ives.
Suggested Readings
Bander SJ, Schw ab SJ. Cent ral venous angioaccess f or hemodialysis and it s
complicat ions. Semin Dial. 1992; 5: 121–128.
Ciniochow ski G E, Worley E, Rut herf ord WE, et al. Superiorit y of t he int ernal
jugular over t he subclavian access f or t emporary hemodialysis. Nephron.
1990; 54: 154–161.
Criado E, Marst on WA, Jaques PF, et al. Primary venous out f low obst ruct ion
in pat ient s w it h upper ext remit y art eriovenous dialysis access. Ann Vase Surg.
1994; 8: 530–535.
Nat ional Kidney Foundat ion. DO Q I clinical pract ice guidelines f or vascular
access and anemia of chronic renal f ailure. Am J Kidney Disease. l997; 30
(suppl 3): 7–240.
Schillinger F, Schillinger D, Mont agnac R, et al. Post cat het erisat ion vein
st enosis in haemodialysis: comparat ive angiographic st udy of 50 subclavian
and 50 int ernal jugular accesses. Nephrol Dial Transplant . l991; 6: 722–724.
> Table of C ontents > R enal > 248 - Us e C aution w hen Us ing Milr inone in R enal Failur e
248
Use Caution when Using Milrinone in Renal
Failure
Ying Wei Lum MD
Milrinone is a select ive inhibit or of t ype I I I cyclic adenosine monophosphat e
(cAMP) phosphodiest erase isozyme in cardiac and vascular smoot h muscle. I t s
inhibit ory act ion on phosphodiest erase result s in increased cAMP levels, w hich in
t urn increases cont ract ilit y in cardiac muscle and st imulat es vasodilat at ion in
blood vessels. This causes an increase in cardiac out put and decrease in
pulmonary w edge pressure. These hemodynamic changes are obt ained w it hout
excessive changes in heart rat e or increase in myocardial oxygen consumpt ion.
Indication
The use of milrinone has been best st udied in pat ient s w it h congest ive heart
f ailure. I t appears t o be very eff icacious in nonhypot ensive pat ient s w it h acut e
nonischemic cardiomyopat hy, despit e t reat ment w it h diuret ics. These pat ient s
benef it f rom an enhancement in cont ract ilit y and af t er load reduct ion. Durat ion of
t herapy should last f or 48 t o 72 hours. There are no st udies t o dat e t hat support
it s use f or a longer period. Long-t erm oral t herapy w it h milrinone has been
associat ed w it h increased mort alit y.
Dose
The recommended dose in pat ient s w it h normal renal f unct ion is 50 µg/ kg bolus
f ollow ed by a cont inuous inf usion at 0. 375 t o 0. 75 µg/ kg/ min. Since milrinone is
excret ed mainly t hrough t he kidneys, it s dose in pat ient s w it h renal impairment
should be adjust ed accordingly (by surf ace area):
Creat inine clearance (CrCl) 50 mL/ min/ 1. 73 m2 : Administ er 0. 43 µg/ kg/ min.
CrCl 40 mL/ min/ 1. 73 m2 : Administ er 0. 38 µg/ kg/ min.
An immediat e improvement in hemodynamic f unct ion is seen w it hin 5 t o 15

minut es af t er init iat ion of t herapy. The mean half -lif e of milrinone is
approximat ely 2. 4 hours and pat ient s reach a st eady-st at e plasma milrinone
concent rat ion (of 200 ng/ mL) w it hin 6 t o 12 hours of a cont inuous maint enance
inf usion of 0. 50 mcg/ kg/ min. The impact of t he half -lif e of milrinone is clinically
import ant as providers must recognize t hat t he eff ect s of t he drug cannot be
t urned off rapidly like many of t he ot her commonly used inot ropic agent s such as
epinephrine.
Caution
Milrinone decreases at riovent ricular nodal conduct ion t ime, allow ing a pot ent ial
f or increased vent ricular response rat es (up t o 3. 8%) f or pat ient s w it h
supravent ricular arrhyt hmias. Vent ricular arrhyt hmias (vent ricular ect opy,
sust ained and nonsust ained vent ricular t achycardia) have also been report ed in
up t o 12% of pat ient s. Lif e-t hreat ening vent ricular arrhyt hmia appears relat ed t o
t he presence of ot her underlying f act ors such as a pre-exist ing arrhyt hmia and
met abolic abnormalit ies. Ext reme caut ion must be used in pat ient s w it h renal
compromise. Fat al vent ricular arrhyt hmias developed in six of nine pat ient s in a
recent st udy of t he pharmacokinet ics of milrinone in pat ient s on cont inuous
venovenous hemof ilt rat ion (CVVH). These pat ient s w ere oliguric (<400 mL/ d) and
had a serum creat inine of more t han 2. 0mg/ dL. All pat ient s received
0. 25µg/ kg/ min of milrinone and developed a mean st eady-st at e concent rat ion of
845 ng/ mL, f our t imes higher t han t hat of pat ient s w it h normal renal f unct ion. The
high prot ein-binding aff init y of milrinone and decreased urinary excret ion could
have cont ribut ed t o t his high concent rat ion, t hereby leading t o t he increased
incidence of vent ricular arrhyt hmias.
Suggested Readings
The Milrinone Source. ht t p: / / milrinone. com/
Taniguchi T, Shibat a K, Sait o S, et al. Pharmacokinet ics of milrinone in

pat ient s w it h congest ive heart f ailure during cont inuous venovenous
hemof ilt rat ion. I nt ens Care Med. 2000; 26(8): 1089–1093.
> Table of C ontents > R enal > 249 - D ec r eas e the D os e of Ganc ic lovir in R enal Ins uffic ienc y
249
Decrease the Dose of Ganciclovir in Renal
Insufficiency
Angela D. Shoher MD
G anciclovir is an ant iviral agent similar t o acyclovir. I t is act ive against all herpes
viruses but is especially eff ect ive f or Cyt omegalovirus (CMV). G anciclovir is
of t en used in t he int ensive care unit (I CU) t o prevent CMV disease in t ransplant
pat ient s and is also eff ect ive f or chronic suppression of CMV ret init is in
immunocompromised pat ient s. I t is an acyclic guanine nucleoside analog t hat
act s by inhibit ing viral DNA synt hesis. O ne f orm used in I CU pat ient s is
valganciclovir, w hich is t he L-valyl est er prodrug of ganciclovir. Valganciclovir is
w ell absorbed, is hydrolyzed t o ganciclovir rapidly, and produces levels
comparable t o int ravenous (I V) ganciclovir. The plasma half -lif e is about 2 t o 4
hours in pat ient s w it h normal renal f unct ion. More t han 90% of ganciclovir is
eliminat ed unchanged by renal excret ion. The half -lif e may be signif icant ly
increased in pat ient s w it h severe renal f ailure and t hus requires dosing based on
renal f unct ion.
What to Do
For t he t reat ment of CMV ret init is in a pat ient w it h normal renal f unct ion
(creat inine clearance >70), t he recommended dosage f or induct ion is 5 mg/ kg
int ravenously every 12 hours f or 2 t o 3 w eeks. Maint enance dosing af t er t he
induct ion period is 5 mg/ kg int ravenously once per day 7 days a w eek or 6 mg/ kg
once per day 5 days a w eek. O ral ganciclovir is not predict ably absorbed, so f or
pat ient s t aking oral valganciclovir, t he recommended dose is 900 mg by mout h
t w ice a day. Reinduct ion is recommended f or pat ient s w ho experience
progression of CMV ret init is w hile on maint enance dosing. To prevent CMV
disease in t ransplant pat ient s w it h normal renal f unct ion, t he dosage is also 5
mg/ kg every 12 hours f or 2 t o 3 w eeks, f ollow ed by daily dosing f or int ravenous
maint enance. The oral maint enance regimen is also 1, 000 mg t hree t imes a day
w it h f ood. O ral valganciclovir maint enance is 900 mg per day. For pat ient s in
renal f ailure, see Tabl e 249. 1 f or dosing guidelines.
The side eff ect s of ganciclovir include myelosuppression, w hich can cause
neut ropenia (w orse w hen not dose reduced f or renal f ailure), anemia, and
t hrombocyt openia. Neut ropenia occurs in 15% t o 40% and t hrombocyt openia in
5% t o 20% of pat ient s. The myelosuppressive eff ect s can be f at al but usually
reverse w it hin 1 w eek of cessat ion. O ral valganciclovir may also cause
gast roint est inal sympt oms such as
nausea and diarrhea as w ell as headache. Cent ral nervous syst em eff ect s of
ganciclovir occur in 5% t o 15% of pat ient s and can range f rom behavioral
changes t o coma. Approximat ely one-t hird of pat ient s must st op using eit her
ganciclovir or valganciclovir secondary t o t hese side eff ect s.
TABLE 249-1 GANCICLOVIR DOSING IN RENAL FAI
DOSING
CREAT ININE INDUCT ION MAINT ENANCE
INT ERVAL
CLEARANCE (MG/KG) (MG/KG)
(HOURS)
50–69 2.5 12 2.5
25–49 2.5 24 1.25
10–24 1.25 24 0.625
3 times a
week
<10 1.25 0.625
after
dialysis
Suggested Readings
Kat zung BG , ed. Basic & Clinical Pharmacology. 9t h ed. New York: McG raw -
Hill; 2004: 823–844.
Wickersham RM, G remillion S, eds. Drug Fact s & Comparisons Pocket

Version. 10t h ed. St . Louis: Wolt ers Kluw er Healt h, I nc. ; 2005: 963–1010.
> Table of C ontents > R enal > 250 - R em em ber that Tr im ethopr im - S ulfam ethoxaz ole ( B ac tr im )
C r ys tals c an P r ec ipitate in the K idney and C aus e R enal D am age and Failur e
250
Remember that Trimethoprim-Sulfamethoxazole
(Bactrim) Crystals can Precipitate in the Kidney
and Cause Renal Damage and Failure
Praveen Kalra MD
Bact rim is an ant ibiot ic drug f ormed by combining t he sulf onamide
sulf amet hoxazole (5 part s) w it h t rimet hoprim (1 part ). This combinat ion (SMX-
TMP) has in vivo act ivit y against bot h G ram-posit ive and G ram-negat ive aerobic
organisms. I t has minimal act ivit y against anaerobic organisms. Sulf onamides
are st ruct ural analogs of para-aminobenzoic acid (PABA) and compet it ively
inhibit t he bact erial enzyme dihydropt eroat e synt het ase t hat is responsible f or
incorporat ion of PABA int o dihydrof olic acid, result ing in a decrease in t he
amount of met abolically act ive t et rahydrof olic acid.
Trimet hoprim is a bact eriost at ic lipophilic w eak base st ruct urally relat ed t o
pyrimet hamine. I t binds and reversibly inhibit s t he enzyme dihydrof olat e
reduct ase and, like sulf onamide, select ively blocks t he conversion of dihydrof olic
acid t o t et rahydrof olic acid, t hus leading t o deplet ion of f olat e, w hich result s in
int erf erence of bact erial nucleic acid product ion. The individual eff ect s of each
component drug are magnif ied in combinat ion and t oget her may provide
bact ericidal eff ect s.
SMX-TMP is rapidly absorbed af t er oral administ rat ion, w it h t he bioavailabilit y
est imat ed at 90% t o 100%. Sulf amet hoxazole is primarily met abolized in t he liver
by acet ylat ion t o inact ive met abolit es t hat ret ain t he t oxicit y of t he parent
compound. Trimet hoprim is also met abolized in t he liver, w it h 10% t o 20%
met abolized t o inact ive met abolit es by o-demet hylat ion ring N-oxidat ion and alp-
hahydroxylat ion. Sulf amet hoxazole is excret ed renally by glomerular f ilt rat ion
w it h some t ubular secret ion. Excret ion is increased in alkaline urine.
Trimet hoprim is 40% t o 60% excret ed w it hin 24 hours primarily by t ubular
secret ion. Excret ion is increased in acid urine and decreased in alkaline urine.
Watch Out For

Alt hough Bact rim is generally an eff ect ive ant ibiot ic, it s administ rat ion must be
accompanied by know ledge of it s t oxicit y. Hemat ological side eff ect s include
agranulocyt osis, aplast ic anemia, t hrombocyt openia, leukopenia, hemolyt ic
anemia, and met hemoglobinemia. Largely
because of t he sulf onamide moiet y, Bact rim has been associat ed w it h severe
dermat ological sequelae including eryt hema mult if orme, St evens-Johnson
syndrome, and t oxic epidermal necrolysis. Again, largely because of t he
sulf onamide component , use of Bact rim has been associat ed w it h hypoglycemia;
it is st ruct urally similar t o t he sulf onylureas, t hus mimicking t he eff ect of oral
hypoglycemic agent s. I n addit ion, art hralgias and myalgias have been report ed
w it h t he use of Bact rim.
Alt hough rare, t he renal eff ect s of SMX-TMP can be st riking. Wit h only
t herapeut ic doses, t he development of cryst alluria, int erst it ial nephrit is, and
renal f ailure has been report ed, part icularly in pat ient s w it h pre-exist ing renal
impairment . Sulf amet hoxazole is relat ively insoluble in acid urine and w ill cause
cryst al f ormat ion w hen t he pH is <5. 5. Most commonly, cryst als are needle
shaped or roset t es. The presence of sulf onamide cryst als can be conf irmed at
t he bedside by t he lignin t est . This is perf ormed by placing one drop of urine on
a paper and adding one drop of 10% hydrochloric acid. The appearance of
yellow ish-orange color is considered a posit ive t est .
The best st rat egy t o prevent t he f ormat ion of cryst als and t he ensuing renal
damage is by maint aining f luid int ake above 3 L/ d and by monit oring t he urine f or
cryst als if t here is an unexplained decrease in renal f unct ion. I f cryst als are
seen, t he drug should be discont inued if possible w it h considerat ion given t o
alkalinizing t he urine. Management of est ablished renal insuff iciency includes
volume management , dialysis support if necessary, adjust ment of drug doses,
and avoidance of f urt her exposure t o nephrot oxins.
Suggested Readings
Perazella MA. Cryst al induced renal f ailure. Am J Med. 1999; 106(4): 459–465.
Vree T, Mart ea M, Hekst er YA, et al. Det eriorat ion of t he kidney f unct ion by
high doses of cot rimaxozole in man. Pharm World Sci. 1987; 9(2): 117–124.
> Table of C ontents > R enal > 251 - B e Aw ar e that Lipid- B as ed Am photer ic in P r oduc ts ar e As s oc iated
w ith Les s R enal Toxic ity than R egular Am photer ic in but c an S till C aus e R enal Injur y
251
Be Aware that Lipid-Based Amphotericin
Products are Associated with Less Renal
Toxicity than Regular Amphotericin but can Still
Cause Renal Injury
John J. Lewin III PharmD
Amphot ericin B is a polyene ant if ungal agent t hat act s by binding direct ly t o
ergost erol on t he f ungal cell membrane, result ing in disrupt ion of membrane
int egrit y and cell deat h. Amphot ericin B has act ivit y against a very w ide range of
clinically import ant yeast s and molds. Resist ance t o amphot ericin B is rare and
has been report ed in only a f ew species (Candi da l usi tani ae, Candi da
gui l l i ermondi i , Aspergi l l us terreus, and Fusari um species). Amphot ericin B has
been t he gold st andard f or t he t reat ment of invasive f ungal diseases f or many
years. Despit e t he more recent advent of new er ant if ungal agent s w it h broad
spect rums of act ivit y (e. g. , caspof ungin and voriconazole), amphot ericin B
remains t he st andard t reat ment f or many severe inf ect ions.
Watch Out For

Amphot ericin B is of t en associat ed w it h numerous side eff ect s including
nephrot oxicit y and an inf usion-relat ed phenomenon marked by hypot ension,
f ever, chills, and rigors. I n an eff ort t o alleviat e t hese side eff ect s, t hree
diff erent lipid-based product s have been developed: liposomal amphot ericin B (L-
AMB), amphot ericin B lipid complex (ABLC), and amphot ericin B colloidal
dispersion (ABCD). I t is import ant t o not e t hat t hese product s should not be
considered int erchangeable, as each has diff erent pharmacokinet ic and
physiochemical charact erist ics (see Table 251. 1).
Amphot ericin B–relat ed nephrot oxicit y occurs presumably as a result of it s
vasoconst rict ive eff ect s, as w ell as direct eff ect s on t he epit helial cell
membranes and t ubular dysf unct ion. O t her renal t oxicit ies associat ed w it h
amphot ericin B include pot assium- and magnesium-w ast ing syndromes as w ell as
renal t ubular acidosis. As a result , a reversible decline in glomerular f ilt rat ion
rat e (G FR) and a rise in serum creat inine up t o as high as 2. 5 mg/ dL is not
uncommon during a course of t herapy. How ever, t he result ant reduct ion in renal
blood f low and G FR can also precipit at e necrosis and acut e renal f ailure (ARF)
in some pat ient s. Pat ient s at great est risk f or developing ARF
are t he crit ically ill and pat ient s receiving concomit ant nephrot oxins (e. g. ,
cyclosporine, aminoglycosides). Numerous st udies consist ent ly suggest t hat
lipid-based amphot ericin product s are associat ed w it h at least 50% less renal
t oxicit y as compared w it h convent ional amphot ericin B. How ever, it is st ill
import ant t o not e t hat nephrot oxicit y and renal f ailure can st ill develop w it h t he
lipid-based product s; t heref ore, close monit oring of urine out put , elect rolyt es,
blood urea nit rogen, and serum creat inine is w arrant ed in all pat ient s.
TABLE 251-1 SELECTED CHARACTERISTICS OF AM PHO
LIPOSOMAL AMPHOT E
AMPHOT ERICIN
AMPHOT ERICIN B LIPID
B
B COMPLEX
Trade
Fungizone AmBisome Abelcet
name
Usual
0.7–1 mg/kg/d 3–5 mg/kg/d 5 mg/kg/d
dose
Vd (L/kg) 2–4 0.1–0.4 131
Plasma
half-life 24–48 8.7–11.2 19.7–23.5
(hours)
Clearance
40.6 9.4 17.8
(mL/hr/kg)
Cost per 1 mg/kg/d 5 mg/kg/d 5 mg/kg/d
day a $17.64 $1,318.80 $840.00
Vd , volume of distribution.
a Cost calculated based on a 70-kg patient using average wh
I nt ravascular volume expansion w it h isot onic cryst alloid solut ions is indicat ed in
all pat ient s receiving amphot ericin B product s, as it may ameliorat e t he decline
in G FR. A 500-mL bolus of 0. 9% sodium chloride solut ion prior t o each dose is
recommended, provided t he clinical sit uat ion permit s. I n many cases, aggressive
elect rolyt e replacement and correct ion of acid-base disorders may be
w arrant ed.
Suggested Readings
Cost a S, Nucci M. Can w e decrease amphot ericin nephrot oxicit y? Curr O pin
Crit Care. 2001; 7(6): 379–383.
Deray G . Amphot ericin B nephrot oxicit y. J Ant imicrob Chemot her. 2002; 49
(suppl 1): 341.
Herbrecht R, Nat arajan-Arne S, Nivioix Y, et al. The lipid f ormulat ions of

amphot ericin B. Expert O pin Pharmacot her. 2003; 4(8): 1277–1287.
> Table of C ontents > R enal > 252 - Have a High Level of S us pic ion for D r ug- Induc ed Ac ute Inter s titial
Nephr itis
252
Have a High Level of Suspicion for Drug-Induced
Acute Interstitial Nephritis
Adam R. Berliner MD
Derek M. Fine MD
Drugs cause t he vast majorit y of acut e int erst it ial nephrit is (AI N), part icularly in
hospit alized pat ient s. The seminal charact erizat ion of drug-induced AI N w as a
1978 series of 13 pat ient s t reat ed w it h met hicillin, in w hich AI N w as
accompanied by f ever, rash (usually maculopapular), and peripheral eosinophilia
(t he classic t riad of AI N), w hich occurred in 58%, 100%, and 100% of pat ient s,
respect ively. How ever, met hicillin use is no longer common, and t he reliance on
t his group of signs t o diagnose AI N in t he cont ext of more modern, diverse drugs
may lead t o missed diagnoses. I ndeed, t he complet e t riad is seen in only 5% of
nonmet hicillin AI N. A st udy of 150 AI N cases illust rat es how cert ain signs and
sympt oms may be poorly sensit ive f or diagnosis of AI N (Fig. 252. 1).
O f not e, int erst it ial nephrit is f rom nonst eroidal ant i-inf lammat ory drugs (NSAI Ds)
may somet imes present w it h bland urinary sediment but heavy prot einuria, even
in t he nephrit ic range.
Dozens of drugs are know n t o cause AI N and t heoret ically any drug can be
blamed.
Drugs commonly associat ed w it h AI N include penicillin derivat ives,

cephalosporins, aminoglycosides, sulf a-cont aining drugs (including
f urosemide and t orsemide), rif ampin, allopurinol, mesalamine, NSAI Ds, and
prot on pump inhibit ors.
The t ime f rom f irst drug dose t o evidence of AI N may be only a f ew days
(part icularly upon re-exposure) or as long as 2 t o 3 w eeks.
I n some cases, one dose of a medicat ion may be suff icient t o cause AI N.
Prior t olerance of a given drug does not rule it out as a cause of AI N in t he
present .
Renal biopsy is t he gold st andard f or diagnosis of AI N and show s inf lammat ory
int erst it ial inf ilt rat es, w it h or w it hout int erst it ial eosinophils. I t is import ant t o
not e t hat even biopsy can give unexpect ed result s. I n one series of 32 pat ient s
w it h acut e renal f ailure, only 44% t hought t o have AI N on clinical grounds
act ually had it on biopsy. Conversely, acut e int erst it ial nephrit is is f requent ly
f ound on biopsy w hen prebiopsy suspicion is low. I n addit ion, t he urine may be
examined w it h eit her Hansel st ain or Wright st ain f or eosinophiluria (considered
posit ive w hen >1% of urinary leukocyt es show posit ive st aining). I n one series of
200 pat ient s w it h urinary eosinophils, t he sensit ivit y, specif icit y, and posit ive
predict ive value f or AI N (w it h renal biopsy t he gold st andard) w ere only 40%,
72%, and 38%, respect ively. Eosinophiluria can also be seen in prost at it is,
rapidly progressive glomerular nephrit is, bladder cancer, and renal at heroembolic
disease. Renal ult rasound is very nonspecif ic and usually show s normal-sized
kidneys, occasionally enlarged, w it h increased echogenicit y.
FIG URE 252. 1. Approximat ed f requency w it h w hich clinical manif est at ions
occur during t he course of met hicillin-induced AI N (A), AI N induced by drugs
ot her t han met hicillin (B), or AI N induced by NSAI Ds and associat ed w it h a
nephrit ic syndrome (C). Prot einuria w as considered posit ive w hen it w as at
least 0. 5 g/ L or posit ive by dipst ick. Dat a are derived f rom case report s and
analysis of cases. Reused w it h permission f rom Rossert J. Drug-induced
acut e int erst it ial nephrit is. Kidney I nt . 2001; 60: 804–817.
What to Do
O t her t han discont inuing t he suspect ed drug, t reat ment is limit ed. Support ive
care of renal f ailure w it h dialysis if needed is t he mainst ay of t herapy. Earlier
cessat ion of an off ending drug holds bet t er hope f or ret urn t o baseline renal
f unct ion, since t ubular at rophy and int erst it ial f ibrosis increase w it h t ime. I n one
series of biopsy-proven AI N, resolut ion of acut e renal f ailure w it hin 2 w eeks w as
associat ed w it h a low er residual serumcreat inine (1. 4 mg/ dL) t han longer-
resolving acut e renal f ailure (3. 4 mg/ dL).
Cort icost eroids can be considered af t er discussion w it h a nephrologist , t hough

pref erably w it h biopsy-proven AI N. Cort icost eroid-t reat ed pat ient s in t he 1978
met hicillin st udy t ended t o ret urn t o a low er serum creat inine, and in great er
numbers, t han unt reat ed pat ient s. This, how ever, w as a small ret rospect ive
st udy, and only 8 of t he 14 pat ient s had biopsy-proven AI N. A larger
ret rospect ive st udy of cort icost eroids f or AI N f ound no benef it .
Suggested Readings
G alpin J, et al. Acut e int erst it ial nephrit is due t o met hicillin. Am J Med.
1978; 65: 756–765.
Michel DM, Kelly CJ. Acut e int erst it ial nephrit is. J Am Soc Nephrol. 1998; 9(3):
506–515.
Rossert J. Drug-induced acut e int erst it ial nephrit is. Kidney I nt . 2001; 60: 804–
817.
> Table of C ontents > R enal > 253 - C ons ider R habdom yolys is in the P atient w ho D evelops O ligur ic
R enal Failur e after a P r olonged S ur ger y w her e Mus c le C om pr es s ion m ay have O c c ur r ed
253
Consider Rhabdomyolysis in the Patient who
Develops Oliguric Renal Failure after a
Prolonged Surgery where Muscle Compression
may have Occurred
William R. Burns III MD
O liguria, charact erized by an hourly urine out put below 0. 5 cc per kilogram of
body w eight or a daily urine out put less t han 400 cc, is a common clinical
problem. Early recognit ion and successf ul diagnosis are required t o prompt ly
init iat e appropriat e t herapy and t o limit t he risk of renal f ailure. While t here are
a number of pot ent ial causes f or t his condit ion, one import ant et iology t o
consider (especially in post operat ive pat ient s and t rauma vict ims) is
rhabdomyolysis, a condit ion t hat of t en progresses t o myoglobinuric acut e renal
f ailure.
Rhabdomyolysis is def ined by an excessive breakdow n of st riat ed muscle. When
t his occurs, cellular mat erials (including a number of muscle-specif ic enzymes)
are released int o t he bloodst ream. Elevat ed blood levels of creat ine kinase, as
w ell as lact at e dehydrogenase or aldolase, are of t en used t o conf irm t he
diagnosis. Myoglobin, an oxygen-carrying hemeprot ein f ound in muscle cells,
also ent ers t he bloodst ream during rhabdomyolysis. Follow ing excessive muscle
injury, t he blood is unable t o bind and clear t he increased myoglobin load, w hich
is t hen excret ed by t he kidneys. When urinary myoglobin becomes elevat ed,
pat ient s develop dark, t ea-colored urine charact erist ic of myoglobinuria.
Most clinicians are able t o recognize t he common risk of rhabdomyolysis in
t rauma scenarios w it h ext remit y ischemia, reperf usion injury, compart ment
syndrome, crush injury, subst ant ial blunt t rauma, or elect rical injury.
I dent if icat ion in post operat ive pat ient s, on t he ot her hand, is more diff icult .
Posit ioning during prolonged surgical procedures has been w ell described in
associat ion w it h rhabdomyolysis and myoglobinuric renal f ailure. Whet her relat ed
t o nonsupine posit ions (lit hot omy or lat eral decubit us) or subopt imal padding
(hard t ables or back boards), muscle compression of any et iology impairs
perf usion and ischemic injury is likely t o occur. Morbidly obese are also at
increased rat e of developing signif icant muscle injury because of excessive
pressure and of t en ext ended operat ive t imes. Likew ise, muscle injury f rom
malignant hypert hermia and prolonged paralysis (of t en in conjunct ion w it h st eroid
use) can result in
rhabdomyolysis. Theref ore, specif ic at -risk populat ions should be rout inely
screened t o ensure prompt diagnosis. The ast ut e clinician w ill also recognize
t hat t he incit ing event f or rhabdomyolysis can be as minimal as compressions
delivered by sequent ial compression devices.
Pat ient s w it h myoglobinuria are at increased risk of acut e renal f ailure. A rout ine
urinalysis is an eff ect ive screening t est , revealing occult blood but no
eryt hrocyt es. This is because of t he det ect ion of myoglobin's heme molecules in
t he absence of hemat uria. I dent if icat ion of urinary myoglobin is also used t o
conf irm t he diagnosis. The associat ion of myoglobinuria and acut e renal f ailure is
most at t ribut ed t o an obst ruct ive process, but ot her mechanisms, some
independent of myoglobin, play a role as w ell. I n an acidic environment (pH <5),
myoglobin w ill precipit at e and f orm cast s t hat impair f ilt rat ion and damage t he
nephron. React ive oxygen species, w hich are also generat ed in rhabdomyolysis,
can init iat e direct t oxic eff ect s on t he renal t ubules. I n addit ion, a number of
vasoact ive agent s released during rhabdomyolysis cause renal art eriolar
const rict ion and subsequent ischemic injury. Regardless of t he exact mechanism
of injury, early diagnosis and prompt init iat ion of t herapy are crit ical t o limit ing
renal damage.
While st rat egies t o prevent , limit , and recognize muscle injury have improved,
successf ul t reat ment of rhabdomyolysis remains crit ical in prevent ing
myoglobinuric renal f ailure. The hallmarks of t herapy are reversal of shock,
minimizat ion of ongoing muscle damage, and aggressive volume resuscit at ion.
Af t er an init ial cryst alloid f luid bolus, inf usion rat es are usually t it rat ed t o
maint ain a robust hourly urine out put (of t en as high as 200 cm3 ). Dosing of
medicat ion of t en needs t o be reduced and nephrot oxic agent s should be
discont inued w henever possible. Addit ional t herapies include t he use of mannit ol
bicarbonat e, and diuret ics alt hough t hese remain cont roversial. Mannit ol is f elt t o
be of benef it because of a number of physiologic eff ect s including volume
expansion, renal vasodilat at ion, and hydroxyl radical scavenging. Alt hough
cont roversial Bicarbonat e administ rat ion can be used t o alkalinize t he urine,
w hich may limit myoglobin precipit at ion and t he result ant t ubular obst ruct ion w it h
myoglobin cast s. Therapies such as plasmapheresis and hemof ilt rat ion are very
eff ect ive at clearing circulat ing heme pigment s but do not seem t o slow t he
progression t o renal f ailure.
Selected Readings
Anema JG , Morey AF, McAninch JW, et al. Complicat ions relat ed t o t he high
lit hot omy posit ion during uret hral reconst ruct ion. J Urol. 2000; 164: 360–363.
Khurana RN, Baudendist el TE, Morgan EF, et al. Post operat ive
rhabdomyolysis f ollow ing laparoscopic gast ric bypass in t he morbidly obese.
Arch Surg. 2004; 139: 73–76.
Slat er MS, Mullins RJ. Rhabdomyolysis and myoglobinuric renal f ailure in

t rauma and surgical pat ient s: a review. J Am Coll Surg. 1998; 186: 693–716.
> Table of C ontents > R enal > 254 - Aim for 2 Milliliter s P er K ilogr am P er Hour of Ur ine O utput in
R habdom yolys is
254
Aim for 2 Milliliters Per Kilogram Per Hour of
Urine Output in Rhabdomyolysis
Awori J. Hayanga MD
Elliott R. Haut MD
Rhabdomyolysis is a syndrome charact erized by muscle necrosis and t he release
of int racellular muscle const it uent s int o t he circulat ion. The severit y of illness
ranges f rom asympt omat ic elevat ions of serum muscle enzymes t o lif e-
t hreat ening cases associat ed w it h severe elect rolyt e imbalances, acut e renal
f ailure, disseminat ed int ravascular coagulat ion, and deat h.
The classic present at ion of rhabdomyolysis includes myalgias, pigment uria due
t o myoglobinuria, and elevat ed serum muscle enzymes. The most commonly
measured enzyme is serum creat inine kinase (CK), w hich is t ypically great er
t han 10, 000 I U/ L. I t should be not ed t hat serum CK levels may remain elevat ed
in t he absence of myoglobinuria since myoglobin is cleared f rom t he serum more
rapidly t han CK. Since serum and/ or urine myoglobin levels of t en t ake at least
hours (if not days) t o obt ain result s, t hese should not be relied upon t o make t he
diagnosis. Rhabdomyolysis can be reliably diagnosed w it h t he combinat ion of t he
urine dipst ick t hat is posit ive f or heme (because of urine myoglobin) and urine
microscopy show ing an absence of red blood cells. O t her abnormal elect rolyt e
f indings include hyperkalemia, hyperphosphat emia, hypocalcemia, and met abolic
acidosis.
Rhabdomyolysis has many varied et iologies, w hich are diff icult t o cat egorize.
Direct mechanical injury result ing in rhabdomyolysis can be caused by t rauma,
elect rocut ions, prolonged immobilizat ion, ischemic limb injury, and crush injuries.
O t her cases can be caused by heat st roke and exert ional rhabdomyolysis
f ollow ing vigorous exercise (e. g. , st rong-man t riat hlons). I n addit ion,
rhabdomyolysis can be caused by drugs and t oxins, w hich can exert eit her direct
myot oxicit y (e. g. , st at ins) or cause indirect muscle damage (e. g. , alcohol or
cocaine). I nf ect ions, inf lammat ory disorders, endocrine, and met abolic et iologies
are also included in t he long list of diff erent ial diagnoses. G enet ic causes must
be considered if no ot her cause is readily apparent .
The most import ant goal of t reat ment in rhabdomyolysis is preservat ion of renal
f unct ion. Plasma volume expansion w it h int ravenous isot onic saline should be
given as soon as possible, even w hile t rying t o est ablish t he cause of t he
rhabdomyolysis. As an example, saline
inf usion may be st art ed bef ore reperf usion of t he t rapped body part in t he case
of severe crush injury. The t ime t o adequat e f luid volume rest orat ion direct ly
inf luences t he rat e of renal f ailure. Massive amount s of f luids (w ell over 10 L)
are of t en required t o compensat e f or t he amount of f luid sequest ered by necrot ic
muscle. Urine out put is t he most import ant early marker of adequat e hydrat ion
and many experienced physicians aim f or 2 mL/ kg/ h.
Cont roversy st ill exist s about t he use of mannit ol and alkalinizat ion of t he urine
as addit ional possible t herapies t o help prevent renal f ailure. While some
physicians have very st rong belief s regarding t he ut ilit y of t hese maneuvers,
most agree t hat neit her approach should be used in pat ient s w it h oliguria. There
is no clear evidence t hat alkalinizat ion is benef icial and t here is a risk t hat
alkalinizat ion may w orsen hypocalcemia. Likew ise, t here is no good evidence of
t he benef it of mannit ol use. Saline diuresis seems t o be t he primary t herapeut ic
act ion. Monit oring w it h serial measurement s of serum pot assium, calcium,
phosphat e, and creat inine is recommended. I t should be not ed t hat hypocalcemia
should not be correct ed unless t he pat ient is sympt omat ic t o avoid w orsening t he
common rebound hypercalcemia seen during t he recovery phase.
Acut e renal f ailure secondary t o rhabdomyolysis is managed expect ant ly and
renal replacement t herapy begun t o cont rol hyperkalemia and/ or volume
overload. Rhabdomyolysis-induced renal f ailure behaves somew hat diff erent ly
t han renal f ailure f rom ot her causes. Serum creat inine rises t o a higher level
more quickly, yet t he pat ient s have a bet t er prognosis f or recovery of renal
f unct ion.
Suggested Readings
Melli G , Chaudhry V, Cornblat h DR. Rhabdomyolysis: an evaluat ion of 475
hospit alized pat ient s. Medicine (Balt imore). 2005; 84(6): 377–385.
Sauret JM, Marinides G , Wang G K. Rhabdomyolysis. Am Fam Phys.

2002; 65(5): 907–912.
> Table of C ontents > R enal > 255 - D o not Attem pt to C onver t O ligur ic to Nonoligur ic R enal Failur e
w ith D iur etic s
255
Do not Attempt to Convert Oliguric to
Nonoliguric Renal Failure with Diuretics
Brandon R. Bruns MD
Heidi L. Frankel MD
O liguria, recent ly def ined as urine out put of less t han 0. 3 mL/ kg/ h f or a 24-hour
period, is a common problem in t he int ensive care unit (I CU). O liguria is an
import ant harbinger of acut e renal f ailure. Nearly 70% of I CU pat ient s w ho
develop acut e renal f ailure are oliguric. Classically, renal f ailure has been
ascribed t o t hree causes: prerenal, renal, and post renal.
Prerenal
Absolut e decrease in int ravascular volume
Hemorrhage
Fluid sequest rat ion (e. g. , pancreat it is, af t er an explorat ory laparot omy,
burns, diarrhea, and vomit ing)
Renal perf usion impairment
Thrombosis of renal vasculat ure
Emboli t o t he renal art ery
Abdominal compart ment syndrome
Dissect ion of a renal art ery
Relat ive decrease in int ravascular volume
Vasodilat at ion associat ed w it h sepsis
Severe right -sided heart f ailure
Cirrhosis
Nephrosis
Renal
Acut e t ubular necrosis
Myoglobin
Aminoglycosides
I nt ravenous cont rast
Acut e int erst it ial nephrit is, w hich can be caused by a w ide variet y of drugs
(e. g. , ant ibiot ics)
Postrenal
Prost at ic hypert rophy
Uret eral obst ruct ion
Foley cat het er malf unct ion
O t her t han t he avoidance of unnecessary t oxins or at t ent ion t o t he appropriat e

volume resuscit at ion of t he I CU pat ient , t here are f ew modalit ies t hat can
prevent incipient acut e renal f ailure. I n general, pat ient s w it h nonoliguric renal
f ailure are easier t o manage t han pat ient s w it h oliguric renal f ailure. How ever,
t he use of loop diuret ics in an at t empt t o convert f rom oliguric t o nonoliguric
renal f ailure is associat ed w it h an increased mort alit y and a decreased rat e of
recovery of renal f unct ion. Similarly, agent s such as dopamine and f enoldopam
have been unsuccessf ul.
A new renal f ailure grading syst em—RI FLE—uses glomerular f ilt rat ion rat e
(G FR) crit eria or urine out put t o predict w het her acut e renal f ailure w ill result in
chronic def icit s. The cat egories include risk, i njury, f ailure, and t he
subcat egories of l oss and end-st age renal disease. I n general, t he longer it
t akes f or renal f unct ion t o ret urn, t he less t he chance it ever w ill. Af t er 6 mont hs
of acut e renal f ailure, t here is lit t le hope f or recovery. Thus, w hereas avoidance
of nephrot oxins (and int ravenous cont rast ) is crit ical in t hose w it h incipient or
acut e renal f ailure of short st anding, t here is lit t le need t o w it hhold t hese
t herapies in t hose w it h chronic renal f ailure; t hus t he old saying “You can't kill a
nephron t w ice. ”
Treat ment f or acut e renal f ailure involves resolut ion of t he underlying cause, f or
example, rest oring int ravascular volume in a t rauma pat ient or removal of an
off ending nephrot oxic agent . O t her opt ions include mechanical means of renal
replacement —dialysis. Recovery f rom t he eff ect s of renal f ailure depends on t he
severit y of t he underlying damage.
Suggested Readings
Bell M, Liljest am E, G ranat h F, et al. O pt imal f ollow -up t ime af t er cont inuous
renal replacement t herapy in acut e renal f ailure pat ient s st rat if ied by t he
RI FLE crit eria. Nephrol Dial Transplant . 2005; 20: 354–360.
Civet t a JM, Taylor RW, Kirby RR. Crit ical Care. 3rd ed. Philadelphia:
Lippincot t – Raven Publishers; 1997: 2081–2091.
Fink MP, Abraham E, Vincent JL, eds. Text book of Crit ical Care. 5t h ed.
> Table of C ontents > R enal > 256 - C ons ider N- Ac etylc ys teine or S odium B ic ar bonate P r ophylaxis
Along w ith Adequate Hydr ation to C om bat C ontr as t- Induc ed Nephr opathy
256
Consider N-Acetylcysteine or Sodium
Bicarbonate Prophylaxis Along with Adequate
Hydration to Combat Contrast-Induced
Nephropathy
Cont rast -induced nephropat hy is t he t hird most common cause of in-hospit al
acut e renal f ailure and has been report ed t o occur in bet w een 6% and 15% of
pat ient s receiving int ravenous cont rast dye. I odinat ed cont rast is used f or
art eriography (including cardiac cat het erizat ion) and venography and is
administ ered int ravenously (I V) f or cont rast -enhanced comput ed t omography
(CT) scans. All iodinat ed cont rast agent s used f or int ravascular radiography have
pot ent ial nephrot oxicit y. The mechanism of renal injury is unknow n but may
include oxidat ive injury, a prerenal component f rom t he diuret ic eff ect of t he
cont rast agent , and ot her et iologies. Nephrot oxicit y depends on pat ient f act ors
and increases w it h t he f ollow ing: renal insuff iciency; diabet es or myeloma as t he
et iology of chronic renal insuff iciency; dehydrat ion; and increasing age.
Nephrot oxicit y is also a f unct ion of t he specif ic agent , w it h an increased
incidence w it h t he older, ionic, higher-osmolarit y agent s and a decreased
incidence w it h new er, more expensive, nonionic, low er-osmolarit y agent s.
What to Do
To reduce t he risk of cont rast -induced nephrot oxicit y, pat ient s should be
adequat ely hydrat ed bef ore exposure t o cont rast . Varying regimens have been
used. O ne common regimen consist s of 0. 45% saline pre-exposure at 1 mL/ kg/ h
f or 12 hours f or inpat ient s or 2 mL/ kg/ h f or at least 4 hours f or out pat ient s and
post exposure at 75 mL/ h f or 12 hours. I n addit ion t o hydrat ion, considerat ion
should be given t o administ ering N-acet ylcyst eine (NAC), especially in pat ient s
w it h renal insuff iciency. O ne recent met a-analysis show ed t hat NAC (w hich is
relat ively inexpensive and has f ew side eff ect s if t aken by mout h) reduced t he
risk of cont rast -induced nephrot oxicit y in pat ient s w it h pre-exist ing chronic renal
insuff iciency. How ever, in t his analysis, single doses varied w idely, w it h oral
doses of 400 t o 1, 500 mg and I V doses of 50 t o 150 mg/ kg given at varying
int ervals and durat ions. Some clinicians say t hat if oral dosing is used,
apparent ly more t han one pre-exposure dose is needed. O ne common regimen
consist s of pre-exposure NAC 600 mg orally f or t w o doses at least 4 hours apart
and post exposure
t w o doses 12 hours apart . For oral administ rat ion, NAC is best administ ered in a
st rongly f lavored drink such as a carbonat ed cola or Fresca t o part ially mask t he
disagreeable odor and t ast e. I n emergencies, I V administ rat ion of 150 mg/ kg in
500 mL of 0. 9% saline over 30 minut es pre-exposure and 50 mg/ kg in 500 mL of
0. 9% saline over 4 hours post exposure has been used, alt hough I V use is more
common f or t he t reat ment of acet aminophen overdose. I nt ravenous NAC has a
higher risk of anaphylact oid react ions t han does oral administ rat ion and t he
physician must consider t he risk-t o-benef it rat io of using I V NAC f or renal
prophylaxis.
Perhaps a bet t er choice t han using I V NAC f or cont rast prophylaxis is
administ ering a sodium bicarbonat e inf usion of 3 amps (t ot al 150 milliequivalent s
sodium bicarbonat e w hich is usually 150 cc) in 850 cc of D5W (dext rose 5%
w at er) f or a t ot al net volume of 1 L. A recent randomized, cont rolled t rial has
show n t he eff ect iveness of reducing cont rast -induced nephropat hy w it h t he use
of sodium bicarbonat e in pat ient s w it h bot h normal renal f unct ion and renal
insuff iciency pre–cont rast administ rat ion. I n t his st udy pat ient s received a bolus
of 3 mg/ kg/ h f or 1 hour bef ore iopamidol cont rast f ollow ed by an inf usion of 1
mL/ kg/ h f or 6 hours af t er t he procedure. Cont rast -induced nephropat hy occurred
in 1. 7% of pat ient s in t he bicarbonat e group, as compared w it h 13. 6% in t he
cont rol group of sodium chloride inf usion. Use of sodium bicarbonat e may have
an advant age over NAC in t he urgent set t ing because t he pret reat ment t ime is
only 1 hour.
O ne f inal not e is t hat NAC or sodium bicarbonat e is not needed w it h magnet ic
resonance imaging cont rast agent s. These cont rast agent s are not iodine based,
but rat her use a met al w it h magnet ic propert ies (e. g. , iron or t he rare eart h
gadolinium) and have diff erent t oxicit ies.
Suggested Readings
Alonso A, Lau J, Jaber BL, et al. Prevent ion of radiocont rast nephropat hy
w it h Nacet ylcyst eine in pat ient s w it h chronic kidney disease: a met a-analysis
of randomized, cont rolled t rials. Am J Kidney Dis. 2004; 43: 1–9.
Fishbane S, Durham JH, Mat zo K, et al. N-acet ylcyst eine in t he prevent ion of
radiocont rast -induced nephropat hy. J Am Soc Nephrol. 2004; 15: 251–260.
MacNeill BC, Harding SA, Bazari H, et al. Prophylaxis of cont rast -induced
nephropat hy in pat ient s undergoing coronary angiography. Cat het er
Cardiovasc I nt erv. 2003; 60: 458–461.
Mert en G J, Burgess WP, G ray LV, et al. Prevent ion of cont rast -induced
nephropat hy w it h sodium bicarbonat e: a randomized cont rolled t rial. JAMA.
2004; 291: 2328– 2334.
> Table of C ontents > B lood > 257 - R em em ber that Tr ans fus ion- R elated Ac ute Lung Injur y is Not D os e
D ependent
257
Remember that Transfusion-Related Acute Lung
Injury is Not Dose Dependent
Transf usion-relat ed acut e lung injury (TRALI ) is a syndrome of sudden-onset
noncardiogenic pulmonary edema occurring during or a f ew hours af t er
t ransf usion of a blood product . TRALI is believed t o occur in approximat ely 1 in
every 5, 000 t ransf usions and it usually occurs af t er administ rat ion of product s
cont aining large amount s of plasma, alt hough it has been report ed t o occur af t er
administ rat ion of as lit t le as 50 mL of w hole blood or any plasma-cont aining
blood product s including packed red blood cells and int ravenous
immunoglobulins. Leukoagglut inat ion and pooling of granulocyt es in t he
recipient 's lungs may occur, w it h t he release of t he cont ent s of leukocyt e
granules t hat result s in injury t o cellular membranes, endot helial surf aces, and
pot ent ially t o lung parenchyma. I n most cases, leukoagglut inat ion result s in mild
dyspnea, hypoxia, hypovolemia, hypot ension, f ever, chills, and pulmonary
inf ilt rat es w it hin about 6 hours of t ransf usion and spont aneously resolves.
O ccasionally, more severe lung injury occurs as a result of t his phenomenon and
acut e respirat ory dist ress syndrome (ARDS) result s. TRALI has a report ed
mort alit y rat e of 5% t o 8% (t he t hird most common cause of t ransf usion-relat ed
mort alit y). Host f act ors such as inf ect ion, lung disease, and recent surgery may
cont ribut e t o t he incidence and severit y of TRALI . Laborat ory f indings include
hemoconcent rat ion, hypoalbuminemia, and neut ropenia or neut rophilia. Typically,
chest radiograph reveals bilat eral inf ilt rat es.
What to Do
Treat ment is largely support ive. The t ransf usion should be st opped if t he
react ion is recognized in t ime. The pat ient should be given supplement al oxygen
and vent ilat ory support as necessary (required in approximat ely t w o-t hirds of
cases), ideally using t he same low -t idal-volume lung-prot ect ive st rat egies t hat
are employed in ARDS. The pulmonary edema is noncardiogenic, and hence
diuret ics are not w arrant ed. G lucocort icoids have been used, but t here are no
dat a t o support t his pract ice. Prevent ion is t he most import ant measure,
including avoiding unnecessary t ransf usions and increased use of red cells
cont aining less plasma. Plasma derived f rom mult iparous w omen has been
ident if ied
as a signif icant source of ant ileukocyt e ant ibodies, and it has been proposed
t hat t hese “f emme f at ales” be excluded f rom t he donor pool.
Suggested Readings
Looney MR, G ropper MA, Mat t hay MA. Transf usion-relat ed acut e lung injury:
a review. Chest 2004; 126: 249–258.
Siliman CC, Boshkov LK, Mehidzadehkashi Z, et al. Transf usion-relat ed acut e

lung injury: epidemiology and a prospect ive analysis of et iologic f act ors.
Blood 2003; 101: 452–464.
Toy P, Popovsky MA, Abraham E, et al. Transf usion relat ed acut e lung injury:
def init ion and review. Crit Care Med 2005; 33(4): 721–726.
> Table of C ontents > B lood > 258 - K now the S igns of a Tr ans fus ion R eac tion
258
Know the Signs of a Transfusion Reaction
Eugenie S. Heitmiller MD
Signs and Symptoms

An acut e hemolyt ic t ransf usion react ion occurs w hen immunologic incompat ibilit y
bet w een t he donor and t he recipient result s in lysis of red blood cells. Most
hemolyt ic t ransf usion react ions are due t o t he t ransf usion of ABO -incompat ible
packed red blood cells secondary t o clerical or syst em errors. The severit y of
t he react ion is relat ive t o t he amount of incompat ible blood received, t he t ype of
incompat ibilit y, and t he lengt h of t ime bef ore t reat ment is init iat ed. These include
chills, anxiet y, dyspnea, rash, nausea, f ever or rise in t emperat ure of 1° C or
more f rom baseline, hypert ension or hypot ension, headache, and chest and f lank
pain. I n a deeply sedat ed or anest het ized pat ient , hemoglobinuria, hypot ension,
and coagulopat hy may be t he only signs. A hemolyt ic t ransf usion react ion is
usually manif est ed during t he t ransf usion and can occur af t er receiving as lit t le
as 10 mL of incompat ible blood.
An increased t emperat ure during a t ransf usion may be t he f irst sign of a
hemolyt ic t ransf usion react ion, or it may be a sign of bact erial cont aminat ion of
t he blood product . Fever w it h hypot ension is charact erist ic of bact erial
cont aminat ion. A G ram st ain of t he blood product is helpf ul t o conf irm t he
diagnosis. Bact erial cont aminat ion may occur f rom cont aminat ion at t he
phlebot omy sit e during blood collect ion, f rom an unrecognized inf ect ion in t he
donor, or f rom improper st orage. The risk of inf ect ion has been report ed as 1 in
2, 000 t o 10, 000 unit s f or plat elet s and f rom 1 in 250, 000 t o 1, 500, 000 f or
packed red blood cells
What to Do
I f a t ransf usion react ion is suspect ed, t he t ransf usion should be st opped
immediat ely. The blood t ubing should be disconnect ed and normal saline solut ion
hung w it h new int ravenous (I V) t ubing t o produce a urine out put of 1 t o 2
mL/ kg/ h t o reduce t he risk of acut e renal f ailure. Alt hough cont roversial, some
clinicians use diuret ics t o increase urine out put . New er biologic agent s are being
developed t hat t arget complement int ermediat es or proinf lammat ory cyt okines
and may be eff ect ive agent s in t he t reat ment of severe hemolyt ic t ransf usion
react ions w hen available.
I f a pat ient develops urt icaria during a t ransf usion w it h no ot her signs or
sympt oms, it is not necessary t o st op t he t ransf usion. Administ rat ion of an
ant ihist amine (i. e. , Benadryl) may help t o decrease t he urt icaria, w hich is usually
due t o t ransf used allergens t hat int eract w it h t he pat ient 's mast cells, result ing in
degranulat ion of t he mast cells. I n addit ion, pat ient s w ho receive repeat ed
t ransf usions are more likely t o experience f ebrile nonhemolyt ic t ransf usion
react ions, w hich can be t reat ed w it h acet aminophen.
I t is t he responsibilit y of t he clinician t o know his or her inst it ut ion's prot ocol f or
t est ing blood and urine samples in suspect ed cases of hemolyt ic t ransf usion
react ion.
Suggested Readings
American Societ y of Anest hesiologist s. Pract ice G uidelines f or Perioperat ive
Blood Transf usion and Adjuvant Therapies.
ht t p: / / w w w. asahq. org/ publicat ionsAndServices/ pract iceparam. ht m#blood
Blajchman BA, Beckers EA, Dickmeiss E, et al. Bact erial det ect ion of
plat elet s: current problems and possible resolut ions. Transf us Med Rev
2005; 19(4): 259–272.
Davenport RD. Pat hophysiology of hemolyt ic t ransf usion react ions. Semin
Hemat ol 2005; 42(3): 165–168.
G oodnough L. Risks of blood t ransf usion. Anest hesiol Clin Nort h Am

2005; 23: 241– 252.
Yazdanakhsh K. Review : complement recept or 1 t herapeut ics f or prevent ion

of immune hemolysis. I mmunohemat ology 2005; 21(3): 109–118.
> Table of C ontents > B lood > 259 - Have a High Thr es hold in Tr ans fus ing P latelets , E s pec ially in
Nonbleeding P atients w ho ar e not P r eoper ative
259
Have a High Threshold in Transfusing Platelets,
Especially in Nonbleeding Patients who are not
Preoperative
Michael J. Haut MD
Plat elet t ransf usion has made a signif icant cont ribut ion t o t he care of cert ain
pat ient populat ions. These groups include t hose undergoing high-dose
chemot herapy (e. g. , f or acut e leukemia or st em cell t ransplant ) and pat ient s
sust aining major t rauma. Plat elet s f or t ransf usion are obt ained eit her by
cent rif ugat ion of unit s of w hole blood (random donor plat elet s) or by pheresis. I n
general, unit s available in t he Unit ed St at es are leukoreduced. Leukoreduct ion
reduces t he incidence of plat elet t ransf usion react ions and may reduce t he
incidence of alloimmunizat ion.
What to Do
Considerable eff ort has gone int o det ermining w hat t he t rigger should be f or t he
t ransf usion of plat elet s and how many plat elet s should be t ransf used.
Hist orically, t he plat elet t ransf usion t rigger f or pat ient s w it h chemot herapy-
induced t hrombocyt openia w as 20 × 109 / µL. How ever, several recent large
st udies have suggest ed t hat 10 × 109 / µL is accept able. A large ret rospect ive
st udy has show n t hat t he most signif icant predict or of bleeding in
t hrombocyt openic pat ient s is not t he plat elet count but a hist ory of bleeding in
t he previous 5 days. These dat a suggest t hat at t ent ion should be f ocused on
providing aggressive plat elet t herapy f or act ive bleeding rat her t han on
prophylact ically t ransf using plat elet s. Such an approach may not be appropriat e
in t rauma pat ient s, in w hom hypot hermia, acidosis, and ot her f act ors may aff ect
t he f unct ion of t he pat ient 's ow n plat elet s and t he plat elet s t hat are t ransf used.
Specif ic sit uat ions requiring plat elet t ransf usions in t hrombocyt openic pat ient s
include t he presence of bleeding and t he need t o perf orm invasive procedures
such as lumbar punct ure, bronchoscopy, and surgery.
I n t he asympt omat ic medical pat ient , t ransf usion should be given w hen t he
plat elet count is 10 × 109 plat elet s/ µL or w hen t he pat ient is bleeding and t he
plat elet count is 20, 000 × 109 / µL or less. I n many inst it ut ions, eit her one
pheresis single-donor pack (equal t o 6 unit s of plat elet s) or 6 unit s of random-
donor plat elet s is used. More recent ly, many inst it ut ions have opt ed t o give 4
unit s of random-donor plat elet s, w hich is eff ect ive in many sit uat ions.
Watch Out For

A major problem af t er plat elet t ransf usion is t he development of alloant ibodies.
This result s in subopt imal rise, or no rise at all, in plat elet count af t er plat elet
t ransf usion. This can happen af t er eit her random-donor or single-donor plat elet s,
and of t en occurs af t er only a f ew plat elet t ransf usions. To achieve adequat e
hemost asis af t er a pat ient has developed aut oant ibodies, HLA-mat ched plat elet s
should be t ransf used. When t his does not raise t he plat elet count , plat elet s can
be cross-mat ched. I n pat ient s f or w hom HLA-mat ched plat elet s do not raise t he
plat elet count , hemost asis can somet imes be achieved w it h epsilon-aminocaproic
acid.
Suggested Readings
Hoff man R, Benz E, Shat t il SI , et al. Hemat ology: Basic Principles and
Pract ice, 4t h ed. New York: Churchill Livingst one, 2005: 2433–2440.
Slicht er SJ. Relat ionship bet w een plat elet count and bleeding risk in
t hrombocyt openic pat ient s. Transf us Med Rev 2004; 18(3): 153–167.
Spahn DR, Rossaint R. Coagulopat hy and blood component t ransf usion in

t rauma. Br J Anaest h 2005; 95(2): 130–139.
St rauss RG . Pret ransf usion t rigger plat elet count s and dose f or prophylact ic
plat elet t ransf usions. Curr O pin Hemat ol 2005; 12(6): 499–502.
> Table of C ontents > B lood > 260 - D o not Adm inis ter P latelets in Im m une Thr om boc ytopenic P ur pur a
260
Do not Administer Platelets in Immune
Thrombocytopenic Purpura
Laith Altaweel MD
I mmune t hrombocyt openic purpura (I TP) is an aut oimmune disorder
charact erized by low plat elet count and mucocut aneous bleeding. The est imat ed
incidence is 100 per 1 million persons per year. I TP can be a primary disorder or
secondary t o an underlying disorder and can have an acut e (6 mont hs or less) or
chronic present at ion. I TP in children and adult s is very diff erent . I n children, t he
incidence is equal bet w een boys and girls, w ho are usually in good healt h unt il
t hey present w it h pet echiae or purpura a f ew days or w eeks af t er an acut e
inf ect ious illness. I n more t han 70% of t he cases, t he illness resolves w it hin 6
mont hs. I n adult s, t he illness is more chronic and insidious, and is more common
in w omen.
The pat hophysiology is relat ed t o aut oant ibodies against plat elet s, w hich t hen
undergo accelerat ed clearance in t he spleen. The diagnosis of I TP remains one
of exclusion. Disorders t hat I TP occurs secondary t o include syst emic lupus
eryt hemat osus, ant iphospholipid syndrome, immunodef iciency st at es such as
common variable hypogammaglobulinemia and immunoglobulin A def iciency,
lympho-prolif erat ive disorders (chronic lymphocyt ic leukemia, large granular
lymphocyt ic leukemia, and lymphoma), inf ect ion w it h human immunodef iciency
virus and hepat it is C virus, and drugs such as heparin and quinidine.
Watch Out For

Marked splenomegaly is not common in I TP, and should prompt a search f or an
alt ernat ive diagnosis. The prof ile of t he complet e blood count should be normal
except f or a low plat elet count and possibly low hemoglobin in t he set t ing of
hemorrhage. The blood smear should be evaluat ed t o rule out
pseudot hrombocyt openia, inherit ed giant plat elet s, and ot her hemat ologic
disorders. Large immat ure plat elet s are of t en seen in I TP. For pat ient s older
t han 60 years of age, a bone marrow aspirat ion is recommended.
Pat ient s w it h plat elet count s of great er t han 50, 000/ mL3 are usually diagnosed
incident ally. For count s bet w een 10, 000/ mL3 and 30, 000/ mL3 , pet echiae or
ecchymoses develop spont aneously, and plat elet count s less t han 10, 000/ mL3
can result in spont aneous int ernal hemorrhage.
What to Do
The init ial t reat ment of I TP consist s of cort icost eroid t reat ment , and a response
is seen in most pat ient s w it hin t hree w eeks. For very low plat elet count s and
pat ient s ref ract ory t o cort icost eroids, int ravenous immune globulin can be used.
For urgent t reat ment , as might occur w it h int ernal bleeding or a need f or
emergency surgery, int ravenous cort icost eroids and int ravenous immunoglobulin
should be given. I n addit ion, ant if ibrinolyt ic t herapy w it h aminocaproic acid or
act ivat ed f act or VI I can also be administ ered.
Chronic t reat ment is only necessary f or pat ient s w it h plat elet count s less t han
30, 000 plat elet s/ mm3 . Aside f rom cont inued medical t herapy w it h
cort icost eroids, danazol, dapsone, and int ravenous immunoglobulin, a surgical
opt ion may be necessary. Splenect omy, w het her laparoscopic or open, is t he
surgical t reat ment f or chronic I TP. Approximat ely t w o t hirds of pat ient s have a
response af t er surgery. Prior t o surgery, pat ient s need t o be immunized f or t he
encapsulat ed organisms including pneumococcus, Haemophi l us i nf l uenzae t ype
b, and meningococcus. For t hose w ho do not respond t o splenect omy, opt imal
medical t herapy w it h combinat ion t herapy as not ed bef ore is at t empt ed w it h a
goal of maint aining plat elet count great er t han 30, 000/ mm What To Do. I f t he
plat elet count does not respond t o surgical or st andard medical t herapy,
t reat ment w it h immunosuppressive agent s, such as azat hioprine or
cyclophosphamide, may need t o be init iat ed. Finally, t here are a number of
invest igat ional t herapies available f or cases t hat are ref ract ory t o st andard
t herapeut ic approaches.
Suggested Readings
Cines D, Blanchet t e V. I mmune t hrombocyt openic purpura. N Engl J Med
2002; 346: 995â 1 008.
G uidelines f or t he invest igat ion and management of idiopat hic

t hrombocyt openic purpura in adult s, children and in pregnancy. Br J Haemat ol
2003; 120: 574.
> Table of C ontents > B lood > 261 - D o Not Adm inis ter P latelets in Type 2 Hepar in- Induc ed
Thr om boc ytopenia
261
Do Not Administer Platelets in Type 2 Heparin-
Induced Thrombocytopenia
Laith Altaweel MD
Heparin is an import ant ant icoagulant in bot h medical and surgical pat ient
populat ions. O ne of t he most serious complicat ions of it s use is t he development
of heparin-induced t hrombocyt openia (HI T). There are t w o t ypes of HI T: t ype 1
and t ype 2. Type 1 is a nonimmune-mediat edd reduct ion in t he plat elet count t hat
occurs a f ew days af t er st art ing heparin, causes a mild t hrombocyt openia, and
normalizes w it h cont inued heparin use. Type 2 HI T is considerably more
dangerous t han t ype 1, is immune mediat ed, and of t en result s in t hrombosis.
Watch Out For

Type 2 HI T result s in a reduct ion of plat elet s by at least 50%. HI T is seen in
approximat ely 3% t o 5% of pat ient s exposed t o unf ract ionat ed heparin and 1%
of t hose exposed t o low -molecular-w eight heparin. I n heparin-naï ve pat ient s, t he
condit ion usually present s 5 days af t er t he init iat ion of heparin t herapy. I n
pat ient s exposed t o heparin w it hin t he prior 100 days, HI T may develop sooner,
of t en w it hin 24 hours of heparin init iat ion. Heparin of bovine origin is more likely
t han porcine heparin t o cause HI T t ype 2. The development of HI T does not
depend on t he quant it y or dose of exposure. HI T t ype 2 has been demonst rat ed
in pat ient s exposed t o minut e amount s of heparin as may be f ound in heparin
f lushes and heparin-coat ed cat het ers. Discont inuat ion of heparin does not
eliminat e t he risk f or complicat ions. HI T-associat ed t hrombosis may occur up t o
4 t o 6 w eeks af t er t he discont inuat ion of heparin.
The plat elet count is import ant in considering a diagnosis of HI T t ype 2. HI T
should be considered w henever t he plat elet count drops by more t han 50% at 5
t o 10 days af t er t he onset of heparin t reat ment in naï ve pat ient s. I t is import ant
t o not e t hat t he plat elet count may st ill be normal in pat ient s w hose baseline
plat elet count is high. Prof ound t hrombocyt openia is unusual and of t en w arrant s
a search f or an alt ernat ive diagnosis. Pat ient s w ho have received heparin in t he
recent past have a rapid onset of HI T t ype 2, usually w it hin 24 hours of heparin
exposure.
The laborat ory conf irmat ion is made by solid-phase assays or plat elet act ivat ion
assays. Solid-phase assays, such as t he solid-phase enzyme-linked
immunosorbent assay (ELI SA), t end t o be very
nonspecif ic. Plat elet act ivat ion assays, such as serot onin release assays
(SRAs), are more specif ic but are also t echnically demanding. Anot her met hod of
t est ing plat elet act ivat ion assay is t he plat elet aggregat ion t est , w hich is less
demanding t echnically but also less sensit ive t han SRA and solid-phase ELI SA.
The result s f rom laborat ory dat a must be considered w it hin t he cont ext of t he
pret est clinical probabilit y.
The most common clinical present at ion includes t hrombosis, usually in t he deep
venous syst em of t he low er limbs. Venous t hrombosis in t he arms can occur
part icularly at sit es of indw elling venous cat het ers. O t her sit es can develop
t hrombosis including t he cavernous sinus, adrenal cort ex, and skin lesions at
sit es of cut aneous heparin inject ions. O t her present at ions include disseminat ed
int ravascular coagulat ion, myocardial inf arct ion, mesent eric inf arct ion, renal
art ery t hrombosis, st roke, and t ransient global amnesia. Theref ore, it is crit ical
t hat t he clinician have a high index of suspicion f or t his condit ion and empirically
discont inue heparin or use alt ernat ive f orms of ant i-coagulat ion if HI T is present .
What to Do
The t reat ment of HI T t ype 2 consist s of discont inuing heparin w hen t he plat elet
count drops or w hen t hrombosis develops w hile t he pat ient is on heparin. Af t er
discont inuing t he heparin, t he pat ient w ho needs cont inued ant i-coagulat ion can
be st art ed on a t hrombin inhibit or, such as argat roban or lepirudin. I t is import ant
t o remember t hat low -molecular-w eight heparin can cross-react w it h HI T
ant ibodies and should not be used. O nce t he pat ient is on a t hrombin inhibit or,
w arf arin can be st art ed if cont inued ant i-coagulat ion is indicat ed f or t he pat ient 's
care. I t must be realized t hat st art ing w arf arin in t he acut e set t ing may result in
acut e skin necrosis and addit ional t hrombosis. Because t he risk of t hrombosis
persist s f or several w eeks af t er discont inuat ion of heparin, pat ient s should be
ant i-coagulat ed f or 30 t o 60 days af t er t he diagnosis is made. Plat elet
t ransf usion should be avoided because t his increases t he severit y of t hrombosis.
Suggested Readings
Davoren A, Ast er R. Heparin-induced t hrombocyt openia and t hrombosis. Am J
Hemat ol 2006; 81: 36–44.
Warkent in T, Kelt on J. Temporal aspect s of heparin-induced

t hrombocyt openia. N Engl J Med 2006; 344(17): 1286–1292.
> Table of C ontents > B lood > 262 - R em em ber that R etic uloc yte C ount is not Ac c ur ate After B lood
Tr ans fus ion
262
Remember that Reticulocyte Count is not
Accurate After Blood Transfusion
Anthony Slonim MD, DrPH
The presence of bot h an adequat e number of eryt hrocyt es and circulat ing blood
volume is essent ial f or t he delivery of oxygen and subst rat es t o t he cells in t he
body. An inadequat e number of eryt hrocyt es result s in anemia and compromises
oxygen delivery by reducing t he oxygen cont ent of t he blood. O xygen cont ent is
comprised of t w o subcomponent s. The cont ribut ion of oxygen bound t o
hemoglobin represent s a considerably larger component of t he oxygen delivery
t han t he dissolved component , w hich highlight s t he relat ive import ance of
hemoglobin t o overall oxygen delivery. The numerical equat ion f or t his
relat ionship is
O 2 = (hbg × % sat × 1. 32) + (PaO2 × 0. 0032)
Watch Out For

There are a number of t ypes and causes of anemia t hat aff ect int ensive care unit
(I CU) pat ient s, and t he init ial w orkup f or t his condit ion is import ant in correct ing
t he underlying dysf unct ion. For example, t he I CU pat ient may present w it h acut e
blood loss f rom a bleeding duodenal ulcer or divert iculosis. I n t his case, t he
administ rat ion of packed red blood cell t ransf usions and t he t reat ment of t he
underlying cause are of import ance. I f , f or example, t he pat ient had a
malignancy such as lymphoma, t he anemia may result f rom an overcrow ded
marrow and insuff icient eryt hrocyt e product ion. Underst anding t he cont ext of t he
anemia w ill not only allow f or appropriat e t herapy and t he correct ion of any
underlying oxygen def icit s, but w ill also provide an opport unit y t o appropriat ely
t reat t he underlying condit ion.
The init ial classif icat ion of anemia is improved w it h t he use of t he ret iculocyt e
count . The ret iculocyt e count represent s t he number of immat ure eryt hrocyt es
t hat are present in t he peripheral blood pool. They can be ident if ied by t he use
of a specif ic st ain on t he peripheral smear and t heir number provides an
est imat e of bone marrow f unct ioning. I t is import ant t o remember t hat normal
eryt hrocyt es have an average lif espan of approximat ely 120 days. As t hese cells
are cleared f rom t he circulat ion by t he spleen, t hey are replaced by ret iculocyt es
f rom t he bone marrow. The normal ret iculocyt e count is 1% t o
2%, w hich represent s t he normal t urnover of eryt hrocyt es f rom t he circulat ion.
The ret iculocyt e count increases severalf old w it hin t he f irst 2 w eeks af t er t he
onset of anemia. This assumes t hat t here is an adequat e st ore of iron and
eryt hropoiet in t o st imulat e product ion of eryt hrocyt es in t he bone marrow w hen
t he hemoglobin drops below 10g/ dL. I t is import ant t o not e t hat t he ret iculocyt e
count needs t o be correct ed f or t he number of circulat ing eryt hrocyt es. This is
done w it h t he use of t he f ollow ing f ormula:
Correct ed ret iculocyt e count = Act ual ret iculocyt e count × (Act ual hemoglobin or
[ hemat ocrit ] )/ (Normal hemoglobin [ or hemat ocrit ] )
This f ormula also highlight s one of t he major problems w it h checking t he
ret iculocyt e count af t er a t ransf usion. The ret iculocyt e count might be expect ed
t o decrease af t er a blood t ransf usion because t he bone marrow no longer needs
t o provide t he same level of peripheral eryt hrocyt es. The correct ed ret iculocyt e
count as displayed in t he equat ion w ill be alt ered because t he act ual hemoglobin
w ill be art if icially elevat ed, t hereby providing an inappropriat e rat io of act ual t o
normal hemoglobin t o allow f or t he correct ed ret iculocyt e count . O nce in t he
periphery, t he ret iculocyt e usually requires an addit ional 1 t o 2 days t o reach f ull
mat urat ion and f unct ion. Theref ore, an addit ional correct ion know n as t he
ret iculocyt e product ion index can be perf ormed t o account f or t his addit ional
mat urat ion.
Suggested Readings
Kasper DL, Braunw ald E, Fauci AS, et al. , eds. Harrison's Principles of
I nt ernal Medicine, 16t h ed. New York: McG raw -Hill, 2005.
> Table of C ontents > B lood > 263 - C ons ider Leukoc yte- D epleted B lood in P atients w ho ar e
Im m unos uppr es s ed
263
Consider Leukocyte-Depleted Blood in Patients
who are Immunosuppressed
Mehmet O zcan MD
Praveen Kalra MD
Leukocyt es in allogenic blood product s have been show n t o cause several
t ransf usion-relat ed adverse react ions and complicat ions. Leukocyt e deplet ion of
cellular blood product s has been show n t o reduce risks such as alloimmunizat ion
against donor leukocyt e ant igens, t ransmission of cyt omegalovirus (CMV), and
f ebrile nonhemolyt ic t ransf usion react ion. I mmunosuppressed pat ient s and
pot ent ial recipient s of organ t ransplant s are especially vulnerable t o t hose
adverse eff ect s and deserve considerat ion of administ rat ion of leukocyt e-
deplet ed blood product s w hen indicat ed.
Watch Out For

Alloimmunizat ion against donor leukocyt e ant igens may pot ent ially cause
reject ion of solid-organ t ransplant s via cross-immunizat ion. This is mediat ed by
t he HLA-ant igen syst em. Mat ching of HLA ant igens is most import ant f or kidney
and heart t ransplant s and least import ant f or liver t ransplant s. Theref ore, using
leukocyt e-deplet ed blood product s t o prevent HLA alloimmunizat ion is most
logical f or pot ent ial recipient s of kidney and heart t ransplant s. I n liver
t ransplant at ion, leukocyt es in t he int raoperat ively administ ered blood have been
suggest ed as a cont ribut or of acut e cellular reject ion via an inf lammat ory
cascade, alt hough t here are no randomized, cont rolled t rials t o support t he
benef it of leukocyt e deplet ion f or t his indicat ion. O verall, t here is reasonable
evidence f or t he benef it of using leukocyt e-deplet ed blood product s t o prevent
graf t reject ion in f ut ure recipient s of kidney t ransplant s, w hereas t he dat a are
insuff icient regarding heart t ransplant s. Leukocyt e-deplet ed blood product s are
not indicat ed f or liver t ransplant candidat es, at least f or t he purpose of
prevent ing HLA alloimmunizat ion.
CMV inf ect ion is a major cause of morbidit y and mort alit y in t he
immunosuppressed pat ient s. Allocat ing CMV-negat ive blood product s f or t hose
pat ient s is challenging because up t o 80% of blood donors are seroposit ive.
Because CMV is t ransmit t ed by leukocyt es, leukocyt e-deplet ed blood has been
suggest ed as an alt ernat ive t o CMV-seronegat ive blood. Even w it h CMV-
seronegat ive blood, t here is up t o a 4% chance of seroconversion of t he
recipient . Alt hough
leukocyt e-deplet ed blood is not absolut ely CMV saf e, it s risk is comparable t o
t hat of seronegat ive blood. There is not a clear consensus on w het her serologic
t est ing f or CMV should be abandoned if universal leukocyt e reduct ion is
implement ed. I n summary, bot h serologic t est ing and leukocyt e deplet ion are
eff ect ive but somew hat imperf ect w ays t o reduce t he risk of CMV t ransmission
t o t he immunosuppressed.
The t erm “leukocyt e-deplet ed” or “leukoreduced” ref ers t o a unit of blood
product cont aining less t han 5 × 106 leukocyt es. There are approximat ely 109
leukocyt es in a unit of w hole blood. Packed red blood cell (PRBC) unit s cont ain
roughly half t hat number (5 × 108 ). “Washed” and “f rozen-deglycerolized” PRBCs
t ypically have 107 and 5 × 106 leukocyt es per unit , respect ively. As f or plat elet s,
apheresis single-donor unit s cont ain 5 × 108 leukocyt es, w hereas random donor
concent rat es cont ain 5 × 107 leukocyt es per unit . Fresh-f rozen plasma and
cryoprecipit at e do not cont ain any cellular element s including leukocyt es.
Leukocyt e deplet ion is most commonly achieved by f ilt rat ion. Leukocyt e f ilt ers
(t hird-generat ion f ilt ers w it h a pore size <40 µm) achieve a 1, 000-f old decrease
in leukocyt e count s of cellular blood product s (i. e. , <10 leukocyt es per µL of
PRBC). Leukocyt e f ilt rat ion can be perf ormed pre- or post -st orage. How ever,
prest orage f ilt rat ion is t he pref erred met hod because it prevent s t he release of
met abolit es f rom leukocyt es during blood product st orage, w hich may also be
responsible f or alloimmunizat ion. Anot her disadvant age of post st orage f ilt rat ion
is a risk of hypot ension at t he t ime of t ransf usion, probably due t o t he act ivat ion
of t he bradykininogen/ kininogen syst em during f ilt ering. This hypot ension f rom
post st orage f ilt ering can be more pronounced in pat ient s using angiot ensin-
convert ing-enzyme inhibit ors.
Perhaps t he biggest concern regarding t he use of universal leukocyt e deplet ion
is f inancial: it cost s about $100 t o leukof ilt er one unit of blood. I n t he Unit ed
St at es, universal leukocyt e deplet ion of all cellular blood product s w ould cost an
addit ional hundreds of millions of dollars annually (2004 est imat e).
Suggested Readings
Bordin JO , Heddle NM, Blajchman MA. Biologic eff ect s of leukocyt es present
in t ransf used cellular blood product s. Blood 1994; 84: 1703–1721.
Dzik WH. Leukoreduct ion of blood component s. Curr O pin Hemat ol

2002; 9: 521–526.
Shapiro MJ. To f ilt er blood or universal leukoreduct ion: w hat is t he answ er?
Crit Care 2004; 8 (Suppl 2): S27–S30.
> Table of C ontents > B lood > 264 - Adm inis ter O c tr eotide in Var ic eal B leeding W hile W aiting for
E ndos c opy
264
Administer Octreotide in Variceal Bleeding While
Waiting for Endoscopy
Madhavi Meka MD
Cirrhosis aff ect s 3 of 1, 000 adult s in Nort h America and is responsible f or more
t han 30, 000 deat hs annually. A major consequence of cirrhosis is variceal
bleeding, w hich cont ribut es signif icant ly t o t he morbidit y and mort alit y
associat ed w it h cirrhosis. O bst ruct ion t o t he port al venous blood f low as seen in
cirrhosis, Budd-Chiari syndrome, port al venous t hrombosis, port al f ibrosis, and
ot her inf ilt rat ive diseases causes port al hypert ension. Event ually, varices
develop t o decompress t he hypert ensive port al vein and ret urn blood t o t he
syst emic circulat ion. Varices usually develop w hen t he port al venous pressure
rises above 12 mm Hg (normal value is <5 mm Hg).
What to Do
Various modalit ies of t reat ment have been used over t he years f or t reat ing
varices. Previously, vasopressin and t erlipressin, w hich direct ly const rict t he
mesent eric art erioles and reduce t he port al venous f low, w ere used in medical
management , but t hey are f alling out of f avor somew hat due t o t heir ischemic
eff ect s on heart , brain, bow el, and limbs. Today, more commonly used drugs in
t he medical management of variceal bleeds include somat ost at in and oct reot ide.
These drugs inhibit t he release of vasodilat or hormones such as glucagon and
indirect ly cause splanchnic vasoconst rict ion. O ct reot ide is a long-act ing synt het ic
analogue of somat ost at in. An int ravenous inject ion of 50-µg bolus f ollow ed by
cont inuous inf usion of 50 µg/ h has been show n t o cause signif icant decrease in
t he port al pressure. The eff ect of a single dose of oct reot ide on reducing port al
pressure is eff ect ive but short -lived (t 1/2 is 3 h in pat ient s w it h liver disease).
Endoscopic t herapy is a mainst ay of invasive t reat ment f or bleeding varices.
There are t w o t herapeut ic opt ions available w it h endoscopy: sclerot herapy and
ligat ion. O ct reot ide in combinat ion w it h sclerot herapy has been show n t o be
superior t o sclerot herapy alone in t hat it decreases t he incidence of early
rebleeding. A similar eff ect is observed w hen oct reot ide is used in combinat ion
w it h variceal ligat ion. I n addit ion, oct reot ide reduces t he bleeding and aids in
bet t er visualizat ion of t he varices during endoscopy w hen used prior t o t he
procedure. Most hepat ologist s recommend cont inuing it f or at least 24 t o 48
hours af t er endoscopy t o reduce t he risk of rebleeding.
I t is import ant t o not e t hat oct reot ide has eff ect s on ot her hormones such as
insulin, vasoact ive int est inal pept ide, and grow t h hormone. Side eff ect s of
oct reot ide include nausea, diarrhea, vomit ing, const ipat ion, st omach upset , gas,
bloat ing, dizziness, headache, light -headedness, f at igue, f lushing, and dry
mout h. Pain and irrit at ion have been report ed at t he inject ion sit e.
Suggested Readings
D'Amico G , Criscuoli V. Met a-analysis of t rials f or variceal bleeding.
Hepat ology 2002; 36 (4 pt 1): 1023–1024.
D'Amico G , Polit , G , Morabit o A. O ct reot ide compared w it h placebo in a

t reat ment st rat egy f or early rebleeding in cirrhosis. A double blind,
randomized pragmat ic t rial. Hepat ology 1998; 28(5): 1206–1214.
> Table of C ontents > B lood > 265 - C ons ider B leeding Ar ound a C hes t Tube to be a S ign of B leeding
in the C hes t C avity Until P r oven O ther w is e
265
Consider Bleeding Around a Chest Tube to be a
Sign of Bleeding in the Chest Cavity Until
Proven Otherwise
David J. Caparrelli MD
Est ablishing adequat e drainage of t he mediast inumand pleural cavit ies at t he
t ime of surgery plays an ext remely import ant role in t he post operat ive care of
pat ient s at risk f or int rat horacic or mediast inal hemorrhage. Adequat e drainage
allow s early recognit ion of surgically correct able bleeding and t he prevent ion of
lif e-t hreat ening complicat ions such as cardiac t amponade and t ension
pneumot horax. Tradit ionally, such drainage has been achieved w it h large-bore
plast ic chest t ubes. Typically a combinat ion of st raight and right -angle t ubes
measuring bet w een 28 French (F) and 32F in diamet er are used af t er open
cardiac surgery (e. g. , coronary art ery bypass), w hereas st raight t ubes alone are
most of t en used af t er elect ive t horacic procedures (e. g. , lobect omy). How ever,
t hese rigid t ubes can be quit e painf ul, damage bypass graf t s, impair vent ilat ion,
and cause cardiac arrhyt hmias. For t hese reasons, many surgeons have recent ly
adopt ed t he use of smaller (19F or 24F), more f lexible, f lut ed Silast ic drains.
Clinical dat a have demonst rat ed t hat t hese drains cause less pain and are as
eff ect ive as t heir more rigid predecessors. Current ly, in cardiac surgery, many
advocat e t he use of one rigid chest t ube in t he mediast inum (32F) in conjunct ion
w it h f lexible Silast ic drains (19F) t o drain t he pleural spaces. Typically, t he
mediast inal t ube is removed on post operat ive day one, w hereas t he Silast ic
drains remain in place unt il drainage is deemed appropriat ely low (usually 100
cc/ 24 h) f or removal. For noncardiac t horacic procedures, w here re-expansion of
t he lung and t he management of air leaks are more of a priorit y t han risk of
hemorrhage, large-bore rigid chest t ubes st ill predominat e.
Watch Out For

Regardless of t he t ype of t ube chosen, early recognit ion of signif icant
int rat horacic hemorrhage is vit al in t he care of cardiac and t horacic surgical
pat ient s. I n t he early post operat ive period, f requent monit oring of bot h t he
quant it y and t he qualit y of t he chest t ube out put is essent ial. O f t en chest t ube
out put w ill be high (>100 cc/ h) over t he f irst f ew hours as eff ort s are made t o
w arm t he pat ient and correct residual coagulopat hy. O ngoing bleeding despit e
replet ion of coagulat ion f act ors (usually w it h f resh-f rozen plasma) and plat elet
t ransf usions is
suggest ive of surgically correct able hemorrhage. I f chest t ube out put is low
and/ or t here is signif icant clot f orming in t he t ubes, one must be concerned w it h
undrained ongoing hemorrhage in any pat ient w it h persist ent t ransf usion
requirement s. Moreover, signif icant bleeding around a chest t ube should be
considered a sign of hemorrhage in t he chest cavit y unt il proven ot herw ise; t he
chest t ube may w ell allow capillary drainage around t he t ube, even if t he lumen
of t he t ube is not draining, perhaps due t o a blockage.
When evaluat ing an int ensive care unit pat ient in a sit uat ion in w hich int rat horacic
or mediast inal hemorrhage is a concern, physical exam can play a role but
usually t akes a back seat t o more object ive dat a. Muff led heart sounds or
decreased breat h sounds over one hemit horax may indicat e f luid around t he
heart or lung, but t hese f indings are neit her sensit ive nor specif ic. Furt hermore,
evaluat ion of jugular venous dist ension can be hindered by cent ral venous access
in t he neck. Hypot ension and a decreasing cardiac out put in t he cont ext of
increasing cent ral venous pressure may suggest cardiac t amponade physiology;
how ever, t hese f indings are also consist ent w it h vent ricular f ailure, f or w hich t he
t reat ment is very diff erent . Theref ore, prompt accurat e diagnosis of t he primary
problem is ext remely import ant . All pat ient s being evaluat ed f or int rat horacic
hemorrhage should have a st at chest radiograph (CXR). An enlarged cardiac
silhouet t e or w idened mediast inum may be indicat ive pericardial eff usion, but
CXR provides no inf ormat ion on t his f luid t hat is physiologically signif icant .
O pacif icat ion of t he pleural space on CXR is a good indicat ion of pleural
eff usion; how ever, t he charact er of t he f luid cannot be det ermined by t his
modalit y. Evaluat ion of t he heart and mediast inum is best undert aken by
t ranst horacic echocardiography. Not only can t his modalit y eff ect ively diagnose
pericardial eff usion, it also provides dat a w it h regard t o t he physiologic
signif icance of t he eff usion (e. g. , right vent ricular collapse). Alt hough very
sensit ive and specif ic f or int rat horacic hemorrhage, comput ed t omography (CT)
scan is not a diagnost ic modalit y of t en used in t he acut e set t ing. How ever, in a
st able pat ient w ho is several days t o w eeks out f rom surgery, CT scan can be
very usef ul in ident if ying signif icant int rat horacic or mediast inal f luid collect ions.
Such inf ormat ion can help t o guide int ervent ions such as pericardial w indow,
t horacent esis, and video-assist ed t horacoscopic decort icat ion.
What to Do
O nce t he diagnosis is ent ert ained t hat t here is new or ongoing hemorrhage in t he
chest , t he surgical t eam should be not if ied immediat ely. Signif icant undrained
f luid in t he mediast inum af t er cardiac surgery can be evacuat ed at
t he bedside by pericardiocent esis or in t he operat ing room t hrough a subxiphoid
pericardial w indow or redo st ernot omy. Similarly, f luid in t he pleural spaces can
be drained by t horacent esis, t ube t horacost omy (chest t ube insert ion), or in t he
operat ing room via an open or t horacoscopic approach. Thoracent esis is usually
reserved f or t he st able pat ient w ho is several days t o w eeks out f rom surgery
w it h a large pleural eff usion (unresponsive t o diuret ics) and persist ent short ness
of breat h or f ever. I n t he early post operat ive period, t ube t horacost omy is most
appropriat e f or drainage of pleural f luid, and a large-bore (28F t o 32F) t ube is
recommended. Typically, smaller chest t ubes (e. g. , 20F) can be used w hen t he
indicat ion f or chest t ube placement is pneumot horax w it hout associat ed pleural
eff usion. When persist ent bleeding f rom a chest t ube occurs (part icularly af t er a
bedside insert ion) hemorrhage f rom t he chest w all or int ercost al vessel must be
considered. This invariably requires more def init ive t reat ment , such as repeat
t horacot omy or an int ervent ional radiology procedure.
Suggested Readings
Braunw ald E, ed. Heart Disease: A Text book of Cardiovascular Medicine.
Philadelphia: WB Saunders, 1997: 229–230, 1731–1732.
O bney JA, Barnes MJ, Lisagor PG , Cohen DJ. A met hod f or mediast inal
drainage af t er cardiac procedures using small Silast ic drains. Ann Thorac
Surg 2000; 70: 1109–1110.
Yang SC, Cameron DE, eds. Current Therapy in Thoracic and Cardiovascular
Surgery. Philadelphia: Mosby, 2004: 532–535.
> Table of C ontents > B lood > 266 - S tar t a P r oton P um p Inhibitor Infus ion for Gas tr ic and D uodenal
B leeding
266
Start a Proton Pump Inhibitor Infusion for
Gastric and Duodenal Bleeding
Hari Nathan MD
Upper gast roint est inal bleeding f rom t he st omach or duodenum may be t he
primary reason f or int ensive care unit admission or may be a subsequent
complicat ion in an already crit ically ill pat ient . Upper gast roint est inal bleeding
carries a mort alit y rat e of approximat ely 12%. Af t er cont rol of t he airw ay as
w arrant ed, init ial resuscit at ion, and placement of a nasogast ric t ube, at t ent ion
must be t urned t o ident if ying t he cause of t he bleeding. Pot ent ial et iologies
include vascular abnormalit ies (e. g. , art eriovenous malf ormat ions, Dieulaf oy's
lesions, or varices), t raumat ic injury (e. g. , Mallory-Weiss t ears), gast rit is or
duodenit is, and pept ic ulcers.
I nit ial localizat ion and cont rol of bleeding may be achieved by endoscopy or
angiography w it h embolizat ion. Uncont rollable hemorrhage may require surgical
int ervent ion. G ast ric acid pot ent iat es bleeding by causing ongoing t issue
damage, inhibit ing plat elet aggregat ion, and promot ing clot lysis. I nhibit ion of
gast ric acid secret ion may t heref ore be an import ant adjunct t o prevent f urt her
bleeding. The eff icacy of t his st rat egy has been demonst rat ed in upper
gast roint est inal bleeding f rom pept ic ulcers. I nt ravenous prot on pump inhibit or
(PPI ) t herapy has been show n t o reduce t he risk of rebleeding f rom pept ic ulcers
(odds rat io 0. 49) af t er init ial hemost asis (e. g. , by endoscopic t herapy).
I nt ravenous PPI s also reduce t he need f or surgery (odds rat io 0. 61). I t should
be not ed, how ever, t hat PPI t herapy has not demonst rat ed any consist ent
mort alit y benef it . I n addit ion, it should also be not ed t hat H2-recept or
ant agonist s have f ailed t o show any benef it in prevent ing rebleeding.
To brief ly review, prot on pump inhibit ors covalent ly bind and irreversibly inhibit
t he H+ , K+ -ATPase responsible f or pariet al cell acid product ion. Most st udies of
PPI s in upper gast roint est inal bleeding have evaluat ed omeprazole, w hich is
unavailable in int ravenous f orm in t he Unit ed St at es. I nt ravenous agent s in t he
class t hat are available in t he Unit ed St at es include pant oprazole, lansoprazole,
and esomeprazole. The suggest ed pant oprazole dosage regimen is an
int ravenous bolus of 80 mg f ollow ed by a cont inuous inf usion of 8 mg/ h f or 72 h.
The pat ient should t hen be placed on 40 mg int ravenously t w ice daily as
maint enance t herapy.
Suggested Readings
I rw in RS, Rippe JM, eds. I rw in and Rippe's I nt ensive Care Medicine, 5t h ed.
Philadelphia: Lippincot t Williams & Wilkins, 2003: 1089â 1 093.
Leont iadis G I , Sharma VK, How den CW. Prot on pump inhibit or t reat ment f or
acut e pept ic ulcer bleeding. Cochrane Dat abase Syst Rev 2006(1): CD002094.
> Table of C ontents > B lood > 267 - R em em ber that B leeding As s oc iated w ith D ir ec t Thr om bin
Inhibitor s is not C or r ec table w ith P r otam ine, Fr es h- Fr oz en P las m a, or P latelets
267
Remember that Bleeding Associated with Direct
Thrombin Inhibitors is not Correctable with
Protamine, Fresh-Frozen Plasma, or Platelets
Michael B. Streiff MD
Thrombocyt openia is a common occurrence among hospit alized pat ient s,
part icularly among pat ient s in t he int ensive care unit (I CU). When a pat ient is
diagnosed w it h, or suspect ed of having, heparin-induced t hrombocyt openia (HI T)
t ype 2, I CU caregivers should be f amiliar w it h a clear t reat ment plan. I n addit ion
t o discont inuat ion of all heparin product s, t herapy f or HI T may include alt ernat ive
int ravenous ant icoagulant s, specif ically argat roban or lepirudin. I t is import ant t o
underst and alt ernat ives f or t herapy and how t o manage t hese and t heir pot ent ial
complicat ions.
1. Therapy f or pat ient s w it h HI T consist s of t he f ollow ing:

a. Eliminat e al l exposure t o heparin
b. Place a sign above t he pat ient 's bed and on t he chart indicat ing heparin
allergy
c. St art a direct t hrombin inhibit or (DTI )
i. Argat roban dosing (half -lif e 45 min, w it h normal hepat ic f unct ion)
1. For I CU pat ient s or pat ient s w it h mild hepat ic insuff iciency (see
Table 267. 1), begin argat roban at 0. 5 µcg/ kg/ min, and check t he
f irst act ivat ed part ial t hrom-boplast in t ime (aPTT) in 4 h (t arget
aPTT rat io 1. 5 t o 2. 5)
2. Monit or aPTT every 4 h unt il t w o consecut ive aPTT values are in
t he t herapeut ic range, t hen check aPTT 12 h lat er and, if st ill
t herapeut ic, t hen at least daily
3. Dose adjust ment s (argat roban)
a. I f argat roban is subt herapeut ic, increase t he inf usion rat e by
20% and recheck t he aPTT in 4 h
b. I f suprat herapeut ic, decrease t he inf usion rat e by 50% and
recheck t he aPTT in 4 h
i. I f aPTT rat io is >3. 0, hold t he inf usion 30 min
ii. I f aPTT rat io is >4. 0, hold 1 h bef ore rest art ing at t he
low er rat e
c. Af t er any dose adjust ment , more f requent aPTT t est ing is
w arrant ed (e. g. , every 4 h unt il aPTT is t herapeut ic
f or t w o consecut ive t est s, t hen 12 h lat er, t hen at least daily)

d. Special not e: Argat roban also increases t he rat io of
prot hrombin t ime t o int ernat ional normalized rat io (PT/ I NR);
how ever, t he aPTT should be used f or dose adjust ment s
(prolongat ion of t he PT is an issue w it h w arf arin cot herapy;
see lat er comment )
TABLE 267-1 ARGATROBAN DOSING IN

HEPATIC INSUFFICIENCY
PERCENTAGE OF
TOTAL
AST /ALT STANDARD INFUSION
BILIRUBIN
RAT E
150–
1.8–3.6
600 25
mg/dL
IU/L
>3.6 >600
Avoid, use lepirudin
mg/dL IU/L
AST, aspartate aminotransferase; ALT, alanine

aminotransferase.
ii. Lepirudin dosing (half -lif e 80 min)

1. For pat ient s w it h normal renal f unct ion (>60 mL/ min), begin
lepirudin at 0. 1 mg/ kg/ h cont inuous inf usion and check t he f irst
aPTT in 6 h (t arget aPTT rat io 1. 5 t o 2. 5)
2. For pat ient s w it h renal insuff iciency ref er t o Table 267. 2 f or
dosing or use argat roban
3. Monit or aPTT every 6 h unt il t w o consecut ive values are in t he
t herapeut ic range, t hen check t he aPTT 12 h lat er and, if st ill
t herapeut ic, at least daily t hereaf t er
TABLE 267-2 LEPIRUDIN DOSING FOR

RENAL INSUFFICIENCY
CREAT ININE PERCENTAGE OF

CLEARANCE STANDARD INFUSION
(ML/MIN) RAT E
45–60 50
30–44 25
15–29 10
<15 Use argatroban
4. Dose adjust ment s (lepirudin)

a. I f lepirudin is subt herapeut ic, increase t he inf usion rat e by
20% and recheck t he aPTT in 6 h
b. I f suprat herapeut ic, decrease t he inf usion rat e by 50% and
recheck t he aPTT in 6 h
i. I f aPTT rat io is >3. 0, hold t he inf usion 1 h, t hen rest art at
t he low er rat e
ii. I f aPTT rat io is >4. 0, hold 2 h, t hen rest art at t he low er

rat e
c. Af t er any dose adjust ment , more f requent aPTT t est ing is
w arrant ed (e. g. , every 6 h unt il t w o consecut ive aPTT values
are t herapeut ic, t hen 12h alt er, t hen at least daily)
d. Special not e: Approximat ely 50% of pat ient s t reat ed w it h
lepirudin f or more t han 5 days develop ant ibodies t o lepirudin
t hat generally increase it s plasma half -lif e; t heref ore t he
need t o make dose reduct ions is common af t er 5 t o 7 days of
t herapy; adjust according t o t he aPTT
d. Warf arin dosing
i. Do not st art w arf arin unt il t he pat ient 's plat elet count has ret urned t o
t he normal range and t he pat ient is on a t herapeut ic dose of a direct
t hrombin inhibit or (DTI )
ii. St art w arf arin at a dose t hat is appropriat e f or t he pat ient 's clinical
sit uat ion (no more t han 5 mg; probably 2. 5 mg is more appropriat e
f or most I CU pat ient s w it h HI T)
iii. Argat roban w ill increase t he I NR disproport ionally in pat ient s on
w arf arin; at inf usion rat es of 2 µg/ kg/ min, it generally doubles t he
I NR; t heref ore, w hen t he pat ient 's I NR is double t he int ended t arget
range in t he set t ing of t herapeut ic argat roban t herapy (I NR 4 t o 6 f or
int ended range 2 t o 3), t he argat roban inf usion may be discont inued
w it h close f ollow up of t he aPTT and I NR (every 4 h t o make sure t he
pat ient is t herapeut ic on w arf arin [ I NR 2. 0 or more] ) w hen all t races
of argat roban eff ect are gone (e. g. , aPTT rat io is normal)
2. Management of DTI -associat ed bleeding
DTI s are not reversible w it h prot amine: G enerally, because argat roban and
lepirudin have short half -lives, discont inuat ion of t he inf usion can be used in
most sit uat ions t o reverse t he ant icoagulant eff ect ; if lif e-t hreat ening
bleeding occurs or an urgent surgical procedure is necessary, recombinant
human f act or VI I a (NovoSeven, Novo Nordisk, Princet on, NJ) has been used
w it h some success in normalizing hemost asis in pat ient s being t reat ed w it h a
DTI : The use of f resh-f rozen plasma and/ or plat elet concent rat es is less
likely t o be benef icial.
Suggested Readings
Warkent in TE. Heparin-induced t hrombocyt openia: pat hogenesis and
management . Br J Haemat ol 2003; 121: 535–555.
Warkent in TE, G reinacher A. Heparin-induced t hrombocyt openia and cardiac

surgery. Ann Thorac Surg 2003; 76: 2121–2131.
> Table of C ontents > B lood > 268 - Attem pt to D ec r eas e P hlebotom y
268
Attempt to Decrease Phlebotomy
Elliott Haut R. MD
Blood t ransf usion is commonly used t o t reat anemia in int ensive care unit s
(I CUs). By I CU day t hree, 95% of all I CU pat ient s have an abnormal hemoglobin
level. Af t er a 1-w eek st ay in t he I CU, pat ient s have an 85% chance of being
t ransf used. According t o t he American Red Cross, 14 million unit s of packed red
blood cells (PRBCs) are t ransf used every year in t he Unit ed St at es. Blood
component t herapy can be a lif esaving measure in many pat ient s and t hese
t ransf usions have a sound physiologic rat ionale: t o increase hemoglobin and
oxygen-carrying capacit y. How ever, blood t ransf usion poses signif icant risks
including viral t ransmission, hemolyt ic t ransf usion react ion, volume overload, and
t he uncommon but devast at ing clerical error leading t o an ABO -incompat ible
t ransf usion. Transf usion of PRBCs in t he int ensive care unit is associat ed w it h
increased nosocomial inf ect ions, diminished organ f unct ion scores, and I CU
mort alit y.
As st at ed above, hemoglobin levels decrease in I CU, even in non-bleeding
pat ient s. A large percent age of t his blood loss is associat ed w it h phlebot omy,
w hich account s f or 30% of all blood t ransf used in t he I CU. The number of blood
draw s and t he volume of blood draw n correlat e w it h w orse I CU organ
dysf unct ion scores. At t empt s t o decrease red blood cell loss f rom phlebot omy in
t he I CU can be w ide ranging.
Watch Out For

Few er numbers of samples can be draw n if pract it ioners are caref ul t o order
only w hat labs are t ruly necessary. There are many indicat ions f or specif ic lab
t est s in t he I CU. The number of lab t est s perf ormed obviously varies according
t o pat ient -specif ic f act ors (e. g. , diagnosis, severit y of illness, comorbidit ies).
How ever, ot her, pat ient -independent f act ors also inf luence t he number and t ype
of labs perf ormed. These include t he presence of art erial lines, admission t o a
t eaching versus nont eaching I CU, and w rit t en or unw rit t en “prot ocols” f or
ordering cert ain labs. Not every pat ient needs every t est every day, especially
chronically crit ically ill pat ient s.
I CU pract it ioners can reduce blood w ast e w hen draw ing diagnost ic samples f rom
indw elling lines by reinf using blood t hat is normally discarded. This can be
accomplished by using a syst em such as t he SAFESET Blood Sampling Syst em
(Hospira) or t he Venous
Art erial blood Management Prot ect ion (VAMP) syst em (Edw ards Lif esciences).
This t ype of closed syst em connect s t o an art erial or venous line and makes it
easy t o reinf use t he init ially draw n-off dilut ed sample, w hich w ould ot herw ise be
discarded. The use of smaller “pediat ric t ubes” and point -of -care t est ing can
lead t o less blood w ast e by using smaller blood volumes f or diagnost ic t est ing.
As w it h many issues in t he I CU, blood conservat ion is mult if act orial. O t her
aspect s t o reducing blood loss include using adjunct s such as pharmacologic
hemost asis (i. e. , aprot inin, f act or VI I a) and red blood cell salvage t echniques
such as an int raoperat ive cell saver. Research int o blood subst it ut es is ongoing
t o decrease t he need f or PRBC t ransf usions.
I CU physicians can also help pat ient s t o increase red cell mass. This can
somet imes be accomplished by t he administ rat ion of exogenous eryt hropoiet in.
Eryt hropoiet in may improve hemoglobin levels and decrease blood t ransf usions in
specif ic pat ient populat ions, such as t hose w it h renal f ailure and possibly f or
t hose t ransf erred t o long-t erm acut e care f acilit ies. I n pat ient s w ho have
concomit ant iron def iciency, enhanced nut rit ional support w it h supplement al iron
and vit amin C may be benef icial.
Anot her import ant w ay t o avoid exposure t o PRBCs is t o only order appropriat e
PRBC t ransf usions. The commonly quot ed goal of a hemoglobin level of 10 g/ dL
has been used f or many years w it h very lit t le dat a t o support t his number. Low er
t ransf usion t riggers are saf e and may act ually improve pat ient out comes. There
is a w idely held misconcept ion t hat st at es “give one, give t w o, ” meaning t hat all
pat ient s w ho receive a t ransf usion should get at least t w o unit s of blood. This
adage is ent irely unf ounded. Some pat ient s really only need one unit of PRBCs.
Suggested Readings
Barie PS, Hydo LJ. Learning t o not know : result s of a program f or ancillary
cost reduct ion in surgical crit ical care. J Trauma 1996; 41(4): 714–720.
Fow ler RA, Rizoli SB, Levin PD, et al. Blood conservat ion f or crit ically ill
pat ient s. Crit Care Clin 2004; 20(2): 313–324.
Nguyen BV, Bot a DP, Melot C, et al. Time course of hemoglobin

concent rat ions in nonbleeding int ensive care unit pat ient s. Crit Care Med
2003; 31(2): 406–410.
Vincent JL, Baron JF, Reinhart K, et al. Anemia and blood t ransf usion in
crit ically ill pat ient s. JAMA 2002; 288: 1499–1507.
> Table of C ontents > B lood > 269 - C ons ider the Us e of Fac tor viia to Tr eat Medic al B leeding in a
S ur gic al or Tr aum a P atient
269
Consider the Use of Factor viia to Treat Medical
Bleeding in a Surgical or Trauma Patient
Suneel Khetarpal MD
Andrew J. Kerwin MD
The primary init ial goal in t he care of t he t rauma pat ient has alw ays been cont rol
of hemorrhage. Under most circumst ances t his can be achieved by surgical
cont rol or more recent ly by ot her adjunct s such as angiography. How ever, in t he
massively injured pat ient a second f orm of bleeding commonly ref erred t o as
nonsurgical bleeding can pose a signif icant t hreat t o lif e. The et iology of t his
nonsurgical bleeding is mult if act orial; it result s f rom a signif icant decrease in
hemost at ic component s (plat elet s, f ibrinogen, and coagulat ion f act ors) as w ell
as unf avorable condit ions f or enzyme kinet ics such as acidosis, hypot hermia, and
hypocalcemia. The end result is t hat t he pat ient develops diff use bleeding t hat
cannot be cont rolled by surgical met hods.
Watch Out For

The primary t reat ment of nonsurgical bleeding is prevent ion. Early recognit ion of
predisposing f act ors such as hypot hermia, correct ion of acidosis by improved
resuscit at ion, and replacement of blood and blood product s can help prot ect
against t he development of coagulopat hy in t rauma pat ient s. How ever,
coagulopat hic bleeding may st ill be present in as much as 35% of severely
t raumat ized individuals and it s incidence increases w it h t he severit y of t rauma.
The t reat ment of such individuals has t radit ionally been support ive. Recent ly,
t here have been an increasing number of cases support ing t he use of
recombinant f act or VI I in t he t herapy of t his uncont rolled nonsurgical bleeding.
Recombinant f act or VI I a (rFVI I ) (NovoSeven; Novo Nordisk A/ S, Denmark) has
been approved by t he U. S. Food and Drug Administ rat ion in t he t reat ment of
bleeding episodes in hemophilic pat ient s w it h inhibit ors t o f act or VI I I or f act or
I X. The f irst report ed use of t his f act or in a t rauma pat ient w as in 1999. The
principle behind t he use of t his f act or in t he t rauma pat ient is t he cell-based
model of coagulat ion. This suggest s t hat rFVI I enhances hemost asis at t he sit e
of injury. Tissue f act or exposed at t he sit e of injury f orms a complex w it h f act or
VI I , w hich in t urn init iat es t he coagulat ion cascade by act ivat ing f act or X and
f act or I X. Fact or X complexes w it h it s cof act or, f act or V, w hich act ivat es
prot hrombin t o produce t hrombin. Whereas t his can
cont ract f ibrinogen t o f ibrin clot , it also accelerat es t he coagulat ion cascade by
act ivat ing f act or V, f act or VI I I , f act or XI , and addit ional plat elet s. This last
accelerat ion causes a large amount of t hrombin t o be f ormed, w hich changes
f ibrinogen t o insoluble f ibrin, result ing in hemost asis. This clot should be
localized and, at least t heoret ically, should avoid any syst emic
hypercoagulabilit y.
Since t he f irst report ed use of f act or VI I a f or t rauma in 1999, t here has been an
increasing number of successf ul cases. I n t hese cases, t he dose varied f rom 40
t o 140 µg/ kg. I n 2005, a mult icent er randomized, cont rolled t rial f rom Europe,
Sout h Af rica, and Canada show ed t hat use of t he drug w as (1) saf e w it h no
increase in t hrombogenic act ivit y and (2) helped t o reduce t he number of
t ransf usions in severely injured t rauma pat ient s. The st udy did not show a
mort alit y benef it . Furt hermore, t he st udy suggest ed a dose of bet w een 100 and
200 µg/ kg given int ravenously over 5 min, f ollow ed by repeat ed doses of 100
µg/ kg x 2 at hourly int ervals if t here is ongoing evidence of bleeding. The cost of
t he drug is signif icant , averaging approximat ely $1, 000/ mg (result ing in an
average cost per dose of $10, 000 t o $20, 000). Alt hough t his cost may init ially
seem prohibit ive, use of t he drug may be cost eff ect ive if it signif icant ly
decreases t he need f or blood and blood product s and massive t ransf usions can
be eliminat ed. Furt her w ork is necessary t o est ablish t he opt imal t iming, dose,
and indicat ion f or f act or VI I a. How ever, it is clear t hat t here must be ongoing
resuscit at ion t o correct hypot hermia and acidosis, and t here be signif icant
f ibrinogen f or t he drug t o w ork. Finally, f act or VI I a is not meant , nor w ill it w ork,
as a subst it ut e f or sound surgical management in t he t rauma pat ient .
Suggested Readings
Boff ard KD, Riou B, Warren B, et al. NovoSeven Trauma St udy G roup.
Recombinant f act or VI I a as adjunct ive t herapy f or bleeding cont rol in severely
injured t rauma pat ient s: t w o parallel randomized, placebo-cont rolled, double-
blind clinical t rials. J Trauma 2005; 59(1): 8–15; discussion 15–18.
Kenet G , Walden R, Eldad A, et al. Treat ment of t raumat ic bleeding w it h

recombinant f act or VI I a. Lancet 1999; 354(9193): 1879.
Mohr AM, Holcomb JB, Dut t on RP, et al. Recombinant act ivat ed f act or VI I a
and hemost asis in crit ical care: a f ocus on t rauma. Crit Care 2005; 9 (Suppl
5): 537–542.
> Table of C ontents > B lood > 270 - C ons ider Angiogr aphy as an Adjunc t in C ontr olling S olid- O r gan
B leeding After D am age C ontr ol S ur ger y
270
Consider Angiography as an Adjunct in
Controlling Solid-Organ Bleeding After Damage
Control Surgery
Pat ient s t ypically ent er t he int ensive care unit (I CU) w it h an open abdomen
f ollow ing t rauma “damage cont rol” surgery or t reat ment or prevent ion of
abdominal compart ment syndrome. Damage cont rol surgery is perf ormed w hen
excessive bleeding result s in t he let hal t riad of acidosis, hypot hermia, and
coagulopat hy. O bt aining cont rol of bleeding in t he operat ing room usually
involves packing and occasionally t he use of balloon occlusion cat het ers.
Surgery is t erminat ed bef ore t he complet ion of t he procedure in order t o ret urn
t he pat ient t o t he I CU. This I CU resuscit at ion phase of rew arming and reversal of
coagulopat hy is ref erred t o as t he second phase of damage cont rol surgery. O n
ret urn t o t he I CU, resuscit at ion w ill only be successf ul if bot h medical and
surgical bleeding has been adequat ely cont rolled.
I n general, inabilit y t o correct hypot hermia, acidosis, and coagulopat hy w it h
ongoing requirement f or t ransf usion indicat es t he presence of ongoing surgical
bleeding. Reexplorat ion and repacking may be at t empt ed, even in t he I CU.
Considerat ion of angiographic embolizat ion or placement of balloon occlusion
cat het ers is appropriat e in cert ain cases.
Embolizat ion represent s an accept able primary f orm of t herapy in many solid-
organ injuries in w hich t he goal is t o avoid laparot omy and/ or loss of an organ. I n
addit ion, f or high-grade liver injury, bleeding pelvic f ract ure, and ret roperit oneal
bleeding, embolizat ion reduces t he overall blood loss associat ed w it h a more
invasive operat ive approach t hat int errupt s t he compressive eff ect s of nat ural
t issue planes. O nce t he abdomen is open, in t he set t ing of phase I I damage
cont rol, t he role of angiography is more specif ic.
What to Do
Success of angioembolizat ion is more likely w hen coagulat ion can be correct ed
(int ernat ional normalized rat io is ≤2. 0). Because damage cont rol pat ient s are
physiologically unst able, highly select ive embolizat ion is of t en avoided in f avor of
opt ions t hat are more expedit ious. For example, embolizat ion of t he right hepat ic
art ery in a hypot ensive bleeding damage cont rol pat ient may be chosen as
opposed t o a segment al branch occlusion in a more st able pat ient . Hypogast ric
(int ernal iliac) art ery occlusion
may be pref erable t o select ive occlusion of a pudendal art ery branch. I n cert ain
cases clinical judgment may dict at e t hat it is w ort h sacraf icing a paired organ
such as t he kidney t o avoid reoperat ing on an unst able pat ient w ho w ould require
a nephrect omy t o cont rol hemorrhage. Embolizat ion under t hese circumst ances
may result in necrosis of viable t issue leading t o f ever or lat e inf ect ion, bile
leaks (liver), or renovascular hypert ension.
Alt hough angioembolizat ion has provided exponent ial improvement in t he opt ions
available f or cont rolling cont rol t raumat ic int ra-abdominal hemorrhage, clinicians
must be caref ul not t o ask t oo much of t heir radiology colleagues. Transport of
an unst able, hypot ensive, hypot hermic, acidemic pat ient t o t he int ervent ional
radiology suit e should only be done w hen t he t rauma surgeon is present and
direct ing t he overall approach t o care. The adage t hat “t he I CU is a philosophy,
not a physical locat ion” can be lif e-saving in t hese pat ient s w ho require ongoing
resuscit at ion.
O ne f inal comment is t hat alt hough bleeding may be f rom sources t hat can be
addressed by reoperat ion and direct surgical repair (i. e. , sut ure ligat ion),
bleeding of t en result s f rom diff use oozing secondary t o coagulopat hy or bleeding
f rom sit es not easily amenable t o surgical repair. Diff erent iat ion of t hese ent it ies
can be complex and requires a t horough underst anding of t he injuries
encount ered and how t hey w ere managed surgically. O pt ions f or medical
hemorrhage cont rol include administ rat ion of coagulat ion f act ors in t he f orm of
f resh-f rozen plasma, plat elet s, cryoprecipit at e, and f act or VI I in addit ion t o
correct ion of hypot hermia.
Suggested Readings
G ourlay D, Hoff er E, Rout t M, et al. Pelvic angiography f or recurrent
t raumat ic pelvic art erial hemorrhage. J Trauma 2005; 59(5): 1168–1174.
Mohr AM, Lavery RF, Barone A. Angiographic embolizat ion of liver injuries:
low mort alit y, high morbidit y. J Trauma 2003; 55(6): 1077–1082.
Rot ondo MF, Zonies DH. The damage cont rol sequence and underlying logic.
Surg Clin Nort h Am 1997; 77(4): 761–777.
> Table of C ontents > B lood > 271 - C ons ult S ur ger y E m er gently if a P atient W ith a B leeding P eptic
Ulc er R ebleeds after E ndos c opic C ontr ol
271
Consult Surgery Emergently if a Patient With a
Bleeding Peptic Ulcer Rebleeds after Endoscopic
Control
Lee Ann Lau MD
Heidi L. Frankel MD
Pept ic ulcer bleeding is a signif icant concern, w it h an incidence report ed at 100
t o 150 per 100, 000 populat ion per year. Approximat ely 5% of all int ensive care
unit (I CU) pat ient s develop upper gast roint est inal bleeding, usually due t o pept ic
ulcer disease or erosive gast rit is. Most pat ient s w ill not require endoscopy f or
diagnosis or t reat ment .
Signs and Symptoms

I n t he I CU pat ient , t he diagnosis of bleeding due t o pept ic ulcerat ion is made by
bloody/ coff ee-ground nasogast ric aspirat e, emesis, or melena. Upper endoscopy
is w arrant ed w hen t he bleeding is associat ed w it h decreasing hemat ocrit and/ or
hemodynamic inst abilit y. I f act ive bleeding or signs of recent bleeding are
ident if ied during endoscopy, hemost at ic t herapies can be perf ormed, including
epinephrine inject ion t o cause vasospasm, placement of clips t o direct ly occlude
t he vessel, or heat er probe applicat ion t o coagulat e t he vessel. Nonst eroidal
ant i-inf lammat ory drug/ aspirin t reat ment should be st opped in pat ient s diagnosed
w it h upper gast roint est inal bleeding. I n addit ion, acid reduct ion must be
inst it ut ed because low er gast ric pH may cause clot lysis. Evidence support s t he
use of int ravenous prot on pump inhibit or in t his set t ing. Despit e aggressive
endoscopic and medical management , rebleeding occurs in 15% t o 20% of
pat ient s; t heref ore surgical consult at ion is imperat ive in all pat ient s requiring
urgent upper endoscopy.
The clinical f act ors t hat increase rebleeding risk include: age great er t han 60
years, medical comorbidit ies (especially cardiopulmonary disease or liver
disease), and init ial present at ion consist ent w it h severe hemorrhage (f rank
hemat emesis or copious melena, low hemat ocrit or coagulopat hy, shock, and
need f or >5 unit s of packed red blood cells). The endoscopic f indings linked t o
rebleeding include art erial bleeding (“spurt ing”) f rom t he ulcer base, visible
vessel in t he ulcer bed, adherent clot in t he ulcer bed, and ulcers locat ed on t he
lesser curvat ure or t he post erior duodenum due t o proximit y t o major vessels
(lef t gast ric art ery and gast roduodenal art ery,
respect ively). Technically, bleeding f rom t hese locat ions is more challenging t o
cont rol endoscopically.
What to Do
Because endoscopy is a successf ul diagnost ic and t herapeut ic t ool, it is t he
f irst -line t reat ment f or upper gast roint est inal bleeding. How ever, surgical
int ervent ion must be considered w henever t here is endoscopic f ailure t o st op
hemorrhage. I f t he pat ient is physiologically unable t o t olerat e t he hemodynamic
st ress of a signif icant bleed, if endoscopic t herapy is not immediat ely available,
or if t he ulcer charact erist ics suggest addit ional endoscopic t reat ment w ill f ail
(especially large ulcers), t he pat ient should undergo immediat e surgery on
rebleeding rat her t han addit ional endoscopy. I nt ervent ional t herapy w it h
embolizat ion is also an opt ion. Most pat ient s, how ever, should undergo a second
endoscopic at t empt at hemost asis. Wit h endoscopic ret reat ment , approximat ely
one t hird of pat ient s w ill st ill event ually need operat ive int ervent ion.
Suggested Readings
Lau JY, Sung JJ, Lam Y H, et al. Endoscopic ret reat ment compared w it h
surgery in pat ient s w it h recurrent bleeding af t er init ial endoscopic cont rol of
bleeding ulcers. N Engl J Med 1999; 340(10): 751–756.
Lau JY, Sung JJ, Lee KK, et al. Eff ect of int ravenous omeprazole on recurrent
bleeding af t er endoscopic t reat ment of bleeding pept ic ulcers. N Engl J Med
2000; 343(5): 310–316.
Schoenberg MH. Surgical t herapy f or pept ic ulcer and nonvariceal bleeding.

Langenbecks Arch Surg 2001; 386(2): 98–103.
> Table of C ontents > B lood > 272 - C all for a S engs taken- B lakem or e or Minnes ota Tube W hen a
C ir r hotic P atient has an Upper Gas tr ointes tinal B leed
272
Call for a Sengstaken-Blakemore or Minnesota
Tube When a Cirrhotic Patient has an Upper
Gastrointestinal Bleed
Cirrhot ic liver disease is a condit ion t hat is caused by chronic scarring and
f ibrosis of t he hepat ic parenchyma. The causes of cirrhosis are quit e varied and
include alcohol abuse, met abolic diseases (α -1 ant it rypsin def iciency, Wilson
disease), biliary disorders (bile duct obst ruct ion, sarcoidosis, cyst ic f ibrosis),
t oxins (carbon t et rachloride, hypervit aminosis A), and inf ect ions (hepat it is A, B,
or C). Regardless of t he et iology, t he pat hophysiology and clinical f indings are
similar. The liver becomes enlarged on clinical examinat ion unt il lat e in t he
course. There may be jaundice, hyperbilirubinemia, t ransaminit is, or overt liver
f ailure. Wit h noninvasive imaging, t he liver appears nodular and may elicit
st igmat a of port al hypert ension including splenomegaly, esophageal varices, and
reversal of port al blood f low. A def init ive diagnosis is usually made by liver
biopsy.
Upper gast roint est inal bleeding is a common int ervening f inding in pat ient s w it h
cirrhosis and can occur f rom t he more t radit ional t ypes of lesions including
gast rit is, Mallory-Weiss t ears, and pept ic ulcer disease. An import ant and crit ical
complicat ion of port al hypert ension t hat is responsible f or nearly one t hird of all
deat hs in pat ient s w it h cirrhot ic liver disease is variceal bleeding. The bleeding
f rom esophageal varices can be abrupt and massive. Wit h t he occurrence of a
variceal bleed, t radit ional approaches t o hemodynamic st abilizat ion need t o be
implement ed including t he use of t w o large-bore int ravenous access devices,
isot onic saline inf usions and blood t o maint ain int ravascular volume, and
ident if icat ion of t he underlying et iology. For emergent variceal bleeding, most
pat ient s can be cared f or w it h endoscopic and pharmacologic int ervent ions,
alt hough each of t hese int ervent ions w ill f ail bet w een 15% and 20% of t he t ime.
When endoscopic and pharmacologic int ervent ions f ail, t he use of balloon
t amponade may be used as a t emporizing measure t o cont rol bleeding.
Balloon t amponade had it s origins in t he early 1930s. The placement of a t ube
w it h f our port s, an esophageal and gast ric port s f or aspirat ion of gast roint est inal
cont ent s, and an esophageal and gast ric balloons t hat w hen f illed w it h air and
lodged against t he bleeding sit e
can lead t o t amponade of a bleeding varix. These t ubes are available under a
number of names, including t he Sengst aken-Blakemore t ube and t he Minnesot a
t ubes, but t he concept ual approaches and applicat ions are similar regardless of
t he specif ic t ube used.
Bef ore placement of t his t ube, t he airw ay should be prot ect ed by endot racheal
int ubat ion. The t amponade t ube is t hen insert ed t hrough t he nares and int o t he
st omach in much t he same w ay as a nasogast ric t ube. A radiograph is obt ained
t o verif y posit ion below t he diaphragm. Then, t he gast ric balloon is inf lat ed w it h
30 t o 50 mL of air and a radiograph is repeat ed t o again verif y t hat t he balloon
is below t he diaphragm. The t ube is pulled back unt il t ension is f elt and t ract ion
is placed t o assure t hat t he balloon applies pressure t o t he esophagogast ric
junct ion w it h a count erw eight device. I f t here is an esophageal balloon present ,
it is inf lat ed and t he esophageal and gast ric port s are t hen placed t o suct ion t o
remove any debris, secret ions, or blood f rom below t he balloon and t o prevent
regurgit at ion. There is cont roversy regarding how long t he t amponade balloons
can remain inf lat ed w it hout undue risk of gast ric or esophageal perf orat ion. The
classic recommendat ion is 24 hours, alt hough many pract it ioners leave t he
balloon inf lat ed longer if need be w hile def init ive t reat ment is being arranged.
Ult imat ely, it must be remembered t hat balloon t amponade is a t emporizing
measure only and t he pat ient w ill need a decompressive procedure t o reduce t he
pressure in t he port al circulat ion by eit her a t ransjugular int rahepat ic
port osyst emic shunt procedure or syst emic shunt .
Suggested Readings
Sharara AI , Rockey DC. G ast roesophageal variceal hemorrhage. N Engl J
Med 2001; 345: 669–681.
> Table of C ontents > Im aging and Tes ts > 273 - Tr eat Los s of D oppler S ignals in a Fr ee Flap as a
S ur gic al E m er genc y
273
Treat Loss of Doppler Signals in a Free Flap as a
Surgical Emergency
Mazen Bedri MD
Surgical Technique
The use of aut ologous f ree t issue t ransplant at ion w as pioneered in t he 1950s
and has since been an import ant rung in t he “reconst ruct ive ladder” commonly
described in plast ic surgery. This reconst ruct ive ladder is based on t he concept
of using t he simplest t echnique possible t o provide adequat e closure or
coverage, w it h increasingly complex t echniques employed as needed. The f ree
f lap is generally considered t he highest and most complex rung of t his ladder.
Bot h f unct ional and aest het ic considerat ions inf luence a surgeon's decision t o
ut ilize a f ree f lap, of t en in t he set t ing of post oncologic reconst ruct ion of t he head
and neck or breast , af t er t raumat ic loss of sof t t issue, or in f acial reanimat ion
procedures. Depending on t he nat ure of t he def ect , t he f lap may include
innervat ed muscle, as w ell as myocut aneous, f asciocut aneous, or
osseocut aneous component s.
Preoperat ive evaluat ion of a pat ient f or a pot ent ial f ree f lap reconst ruct ion must
consider t he pat ient 's general clinical st at us and t he qualit y and condit ion of bot h
t he donor and recipient sit es. The pat ient 's nut rit ional st at us and age, as w ell as
associat ed comorbidit ies such as diabet es, peripheral vascular disease, and
cardiovascular disease, are import ant f act ors t o be w eighed w hen making t he
assessment . The eff ect of t obacco use on f lap viabilit y is cont roversial, alt hough
it is not an absolut e cont raindicat ion t o creat ing a f ree f lap. Pat ient s w it h
mult iple comorbidit ies w arrant preoperat ive medical risk st rat if icat ion.
Specif ic t o t he surgical sit es, f act ors t o consider are t he lengt h of t he vascular
pedicle, t he qualit y and caliber of recipient vessels, t he size mat ch of donor and
recipient vessels, t he volume and geomet ry of t he f lap t issue, and t he general
condit ion of t he recipient sit e (prior irradiat ion, vascular disease, t raumat ic
injury, and inf ect ion can aff ect f lap survival).
The most import ant det erminant in graf t survival is int raoperat ive t echnique. I n
addit ion t o prophylact ic ant ibiot ics, most microvascular surgeons administ er
eit her a one-t ime bolus of 5, 000 U of heparin prior t o graf t harvest or a low er-
dose bolus of heparin f ollow ed post operat ively by a cont inuous low -dose
inf usion. Topical lidocaine or
papaverine is used f or vasodilat ion. Surgical t echnique should emphasize t he

delicat e handling of vasculat ure t o prevent vasospasm and t hrombosis.
Excessive t ract ion and drying should be avoided. Vessels should be 1 t o 3 mm in
diamet er, and t he ends should be t rimmed of loose advent it ia. Alignment of t he
donor and recipient vessels is of ut most import ance and is f acilit at ed by ensuring
adequat e pedicle lengt h (2 t o 3 cm), appropriat e mat ching of vessel caliber, and
t he met iculous placement of int errupt ed sut ures symmet rically and
circumf erent ially. The anast omoses should be f ree of t ension but also of
redundancy t o prevent kinking and t w ist ing. Flap ischemia t ime should be
minimized, alt hough ischemia t imes short er t han 3 t o 4 hours should not
cont ribut e t o t he risk of f lap loss. O n inset t ing t he f lap, t he vascular pedicle
should be inspect ed t o ensure t he vascular pedicle is not compromised by
kinking, t w ist ing, or compression. Post operat ively, pat ient s are of t en maint ained
on heparin, dext ran, or aspirin, alt hough st udies do not conclusively show benef it
of any part icular regimen.
How Graft Viability Is M onitored

Many met hods of f lap monit oring have been described and used w it h varying
degrees of success, including laser Doppler, t ranscut aneous and int ravascular
measurement of oxygen t ension, and t emperat ure monit oring. How ever, diligent
clinical assessment w it h hourly exams in t he init ial post operat ive period remains
t he st andard in post operat ive care f or f ree f laps. Free f laps w it h a cut aneous
component should be visually inspect ed f or color, w hich should be similar t o t he
color of t he surrounding recipient bed. The f lap should be w arm t o t ouch, w it h
relat ively normal t urgor and a capillary ref ill of 1 t o 2 seconds. Art erial occlusion
is suggest ed by a pale color, cool t emperat ure, and sluggish or absent capillary
ref ill. Problems w it h venous out f low are suggest ed by congest ion, of t en w it h
darkening of color, w armt h and edema, and brisk capillary ref ill. Bedside Doppler
is helpf ul in t he assessment of blood f low, alt hough a signal may be f alsely
reassuring in t he set t ing of venous compromise, or if t he signal is f rom
surrounding t issue. A pinprick may also be usef ul in evaluat ion of t he f lap and
should produce t w o t o t hree drops of bright red blood. Brisk venous bleeding or
absence of bleeding suggest s venous or art erial compromise, respect ively.
A unique challenge in monit oring f lap viabilit y is posed by t he buried f lap, in
w hich clinical assessment is impossible and surf ace Doppler is unreliable. The
t emporary placement of implant able Doppler probes can f acilit at e post operat ive
care.
Loss Rate of Flaps, What to do if Loss of Signal Occurs,
Rate of Salvage w ith Re-exploration
The overall success rat e of f ree f laps has increased signif icant ly since t he
t echnique w as init ially int roduced (current ly 95% t o 99%). Some clinical series
have associat ed cert ain f act ors w it h higher f ailure rat es, including f ree f laps w it h
an osseous component , previously irradiat ed recipient sit es, or t he use of vein
graf t s. The use of medicinal leeches w it h proper ant ibiot ic prophylaxis has been
described w it h some success in t he set t ing of venous congest ion due t o
inadequat e out f low. How ever, t he most common cause of f lap f ailure is venous
t hrombosis, w it h early re-explorat ion rat es as high as 15% in some series.
Almost 80% of such f ailures occur w it hin t he crit ical period of t he f irst t w o
post operat ive days. Wit h t he t imely det ect ion and management of vascular
compromise—w it hin 6 t o 8 hours—salvage rat es in recent series have ranged
f rom 60% t o 75%.
The det ect ion of vascular compromise should prompt immediat e mobilizat ion of
t he operat ive t eam. Re-explorat ion may begin w it h int ravenous administ rat ion of
heparin, f ollow ed by an inspect ion of t he anast omosis sit e f or kinking or t w ist ing,
hemat oma, or any ot her ext rinsic compression. I nt ravascular t hrombosis
necessit at es t akedow n of t he anast omoses and embolect omy w it h a Fogart y
cat het er. Persist ent f ailure t o f reely irrigat e suggest s t hrombosis of t he
microvasculat ure and may respond t o high-dose inf usions of st rept okinase,
urokinase, or recombinant t issue plasminogen act ivat or w hile t he f lap is
complet ely disconnect ed t o avoid syst emic dosing. The f lap is t hen re-
anast omosed and inset , f ollow ed by maint enance on heparin or dext ran. Failure
t o re-est ablish circulat ion necessit at es removal of t he nonviable f lap.
Suggested Readings
Jones NF. I nt raoperat ive and post operat ive monit oring of microsurgical f ree
t issue t ransf ers. Clin Plast Surg 1992; 19(4): 783–797.
Tow nsend C, ed. Sabist on Text book of Surgery, 17t h ed. Philadelphia: WB
Saunders, 2004: 2186–2188.
> Table of C ontents > Im aging and Tes ts > 274 - R em em ber W hen R eview ing D oppler Ultr as ound
R es ults that the S uper fic ial Fem or al Vein is a C om ponent of the D eep Venous S ys tem
274
Remember When Reviewing Doppler Ultrasound
Results that the Superficial Femoral Vein is a
Component of the Deep Venous System
Patrick Schaner MD
Signs and Symptoms

Doppler ult rasound of t he veins is a w idely used t est t o search f or t he presence
of deep vein t hrombosis (phlebit is) t hat requires ant icoagulat ion and t hat may
result in pulmonary embolism. This t est is of t en ordered f or a pat ient w it h signs
or sympt oms of a suspect ed pulmonary embolism (pleurit ic chest pain,
t achypnea, t achycardia, hypocapnia, hypoxia, hypot ension) or deep vein
t hrombosis (leg sw elling, calf pain, posit ive Homan sign) or f or a f ever w orkup.
Commonly, a preliminary reading t hat ident if ies w hich veins may be harboring
clot s is placed in t he chart or report ed verbally by t he t echnician. The decision
t o ant icoagulat e a pat ient hinges on t hese result s. I n t his set t ing, it is import ant
t o know t he venous anat omy of t he leg and t he risk of pulmonary embolism f or
phlebit is of t he relevant veins.
The venous drainage of t he leg is comprised of t w o syst ems: t he deep veins and
t he superf icial veins. The t w o chief superf icial veins are t he great er and lesser
saphenous veins. O t her superf icial veins are t he superf icial epigast ric,
superf icial circumf lex iliac, and ext ernal pudendal veins and surf ace varicosit ies.
Component s of t he deep venous syst em include t he soleal sinusoids, ant erior
and post erior t ibial, common f emoral, prof unda f emoral, superf icial f emoral,
circumf lex f emoral, and iliac veins. The deep veins course in proximit y t o t he
major art eries (venae comit ant es). Clot s in t he deep syst em put a pat ient at
much higher risk of pulmonary embolism and usually require ant icoagulat ion (20%
of pat ient s w it h clot s in t he superf icial syst em also have coexist ing clot s in t he
deep syst em).
I t is import ant t o remember w hen review ing t he locat ion of blood clot s on t he
Doppler report t hat t he superf icial f emoral vein t hat accompanies t he superf icial
f emoral art ery is a major deep vein. This vein is “superf icial” only relat ive t o t he
prof unda (i. e. , deep) f emoris vein but is a major deep st ruct ure despit e t he
superf icial appellat ion. Clot s in t he more proximal veins (such as t he superf icial
f emoral) have a great er risk of lif e-t hreat ening embolism t han t hose more dist al,
and a f inding of a clot here should never be dismissed as inconsequent ial.
Suggested Readings
April EW. Anat omy. Media, PA: Harw al, 1984: 381–382.
Ernst CB, St anley JC, eds. Current Therapy in Vascular Surgery. St . Louis:
Mosby, 1995: 875–876.
Moore KL, Dailey AF, eds. Clinically O rient ed Anat omy, 4t h ed. Philadelphia:
Lippincot t Williams & Wilkins, 1999: 524–526.
> Table of C ontents > Im aging and Tes ts > 275 - P os t a S ign O uts ide of the D oor of P atients W ho Have
R ec eived a Tagged W hite B lood C ell S c an W ar ning of the P r egnanc y R is k
275
Post a Sign Outside of the Door of Patients Who
Have Received a Tagged White Blood Cell Scan
Warning of the Pregnancy Risk
Tonya N. Walker MD
A t agged w hit e blood cell (WBC) scan is a nuclear medicine st udy used t o
diagnose and localize areas of inf ect ion and/ or inf lammat ion. I t can be helpf ul
w hen an occult inf ect ion such as pyelonephrit is, abdominal abscess or
ost eomyelit is is suspect ed. A major component of t he immune syst em's def ense
mechanism is t he act ivat ion and upregulat ion of w hit e blood cells. These cells
can be f ound in large numbers in areas w here t here is act ive inf lammat ion and/ or
inf ect ion; t he abilit y t o “t ag” t hese cells allow s t he det ect ion of such f oci in t he
evaluat ion of pat ient s w it h a suspect ed source of inf ect ion.
To perf orm a t agged WBC scan, a sample of t he pat ient 's WBCs is mixed w it h
t he radioact ive mat erial indium oxine, isot ype 111 (111 I n). The t agged sample is
t hen reinf used int ravenously t o t he pat ient . I f t here is an area of act ive
inf lammat ion/ inf ect ion, a scanner w ill det ect localized areas of t hese t agged
w hit e cells and convert t hem int o an image view ed on a screen or recorded on
f ilm. The scan t akes approximat ely 1 t o 2 hours and requires pat ient t ransport t o
t he nuclear medicine suit e.
There are several special considerat ions w hen deciding w het her a t agged WBC
scan is t he appropriat e diagnost ic t ool. False-negat ive result s can occur due t o
ant ibiot ic use or chronic inf ect ion, and because WBCs are usually f ound in t he
spleen and liver, a t rue inf ect ion in t hese organs can be missed w it h t he scan.
False-posit ive result s can occur due t o bleeding and t he presence of t ubes,
drains, or cat het ers in t he body. Because t he pat ient has t o w ait 6 t o 24 hours t o
be scanned af t er t he reinf usion, t his st udy may not alw ays be t he best opt ion f or
t he crit ically ill pat ient .
The radiat ion dose of int erest f or pat ient s undergoing a nuclear medicine scan is
t he radiat ion absorbed dose (rads). Dose limit s are expressed in sievert s (Sv) of
phot ons or elect ron exposure, w here 1 millisievert (mSv) = 0. 1 rad. The average
annual radiat ion exposure f rom nat ural sources t o an individual in t he Unit ed
St at es is approximat ely 3 mSv. O ne chest x-ray exposes a pat ient t o an
addit ional 6 mSv (0. 6 rad), w hereas a t ransat lant ic f light exposes an individual
t o 0. 25 mSv (0. 025 rad). O ne t agged WBC scan provides an eff ect ive dose of
7. 2 mSv (0. 72 rad) t o t he pat ient undergoing t he st udy. This is equivalent t o
slight ly more t han a t ypical chest x-ray and about 30 t ransat lant ic f light s. Those
individuals, visit ors, or caregivers exposed t o t he pat ient w ould be exposed t o 10
t o 100 t imes less t han t his.
Watch Out For

O nce a pat ient ret urns t o t he care area f rom a t agged WBC st udy, t he radiat ion
emit t ed f rom t he pat ient present s t w o t ypes of risks: t he risk t o t he pat ient and
t he risk t o groups exposed t o t he pat ient . As indicat ed previously, t he risk t o t he
pat ient is minimal. Exposed groups include w ard st aff and visit ors. The medical,
nursing, and ancillary st aff caring f or a “radioact ive” pat ient can be, and in many
inst ances are, w omen of childbearing age. These w omen may also be pregnant
w hile on t he w ard w it h t he pat ient . I t is import ant t o inf orm visit ors and hospit al
st aff of t he slight ly increased radiat ion risk by post ing a sign out side t he
pat ient 's door. According t o t he U. S. Nuclear Regulat ory Commission, di rect
abdomi nal / pel vi c exposure at t he level of 2 rads (20 mSv) is t he amount of
radiat ion considered t o be t he low est exposure level f or possible damage t o
f et uses. This is many t imes higher t han t he likely radiat ion exposure a nurse or
physician w ould sust ain f rom caring f or a pat ient w ho had a t agged WBC scan.
How ever, because most pregnant w omen w ould likely w ant no ext ra radiat ion if it
could be avoided (it must be remembered t hat radiat ion eff ect s on a f et us are
cumulat ive), it seems prudent and reasonable t o post a w arning sign.
Tw o days af t er t he WBC scan is complet e, t he sign indicat ing increased
radiat ion exposure may be removed because virt ually all radioact ivit y has
dissipat ed.
Suggested Readings
Low e SA. Diagnost ic radiography in pregnancy: risks and realit ies. Aust r N Z
J O bst et G ynaecol 2004; 44: 191–196.
MedLine Plus Medical Encyclopedia. WBC (nuclear) scan.

ht t p: / / w w w. nlm. nih. gov/ medlineplus/ ency/ art icle/ 003834. ht m
Mount f ord PJ. Risk assessment of t he nuclear medicine pat ient . Br J Radiol
1997; 70: 671–684.
Unit ed St at es Nuclear Regulat ory Commission, O ff ice of Public Aff airs. Fact
Sheet . Biological Eff ect s of Radiat ion, December 2003.
> Table of C ontents > Im aging and Tes ts > 276 - O btain an E c hoc ar diogr am to R ule O ut B ac ter ial
E ndoc ar ditis in Gr am - P os itive B ac ter em ia
276
Obtain an Echocardiogram to Rule Out Bacterial
Endocarditis in Gram-Positive Bacteremia
Mike Faulkner MD
Bact erial endocardit is is an uncommon condit ion w it h signif icant morbidit y and
mort alit y. The classic clinical t riad of bact eremia, valvular pat hology, and
peripheral embolic phenomena unf ort unat ely only present s in a minorit y of
aff ect ed individuals. Cardiac imaging w it h echocardiography has enhanced t he
abilit y t o diagnose inf ect ive endocardit is, current ly comprising one of t he major
crit eria in t he Duke clinical crit eria f or diagnosis of t he disorder (Fi g. 276. 1).
What to Do
I n pat ient s w it h low clinical suspicion and low init ial risk (f ever, previous heart
murmur, no peripheral st igmat a), t ranst horacic echocardiography (TTE) is a
reasonable init ial screening t est . TTE is noninvasive and has an excellent
specif icit y f or veget at ions (98%). Unf ort unat ely, TTE has relat ively poor
sensit ivit y f or diagnosing inf ect ive endocardit is, det ect ing approximat ely 60% of
nat ive valve inf ect ions and 20% of prost het ic valve inf ect ions. I t poorly resolves
lesions less t han 2 mm or lesions of t he lef t heart . I n addit ion, a signif icant
number of pat ient s may suff er f rom condit ions such as chronic obst ruct ive
pulmonary disease or obesit y t hat make imaging t echnically diff icult . The limit ed
resolut ion of TTE also hinders diagnosing complicat ions of inf ect ive endocardit is
such as valvular leaf let perf orat ions, perivalvular abscesses, and int racardiac
f ist ulae.
Transesophageal echocardiography (TEE) may be indicat ed as an init ial imaging
t echnique in high-risk pat ient s (prost het ic valves, congenit al heart disease,
previous endocardit is, new murmur, heart f ailure, or ot her st igmat a of
endocardit is) or in pat ient s w it h moderat e t o high clinical suspicion of inf ect ive
endocardit is. TEE is a more invasive t echnique, but it allow s placement of t he
echo t ransducer closer t o t he valvular st ruct ures of t he heart . Sensit ivit y of TEE
has been report ed t o be great er t han 90% in det ect ing veget at ions. Det ailed
imaging of complicat ions f rom inf ect ive endocardit is such as perivalvular abscess
is also enhanced by t he t ransducer's proximit y t o cardiac st ruct ures.
Bact eremia w it h G ram-posit ive organisms such as Streptococcus vi ri dans and
Streptococcus bovi s are uncommonly seen in pat ient s w it hout inf ect ive
endocardit is. The presence of t hese pat hogens in t w o
separat e blood cult ures def ines a major crit erion in t he Duke clinical crit eria f or
t he diagnosis of inf ect ive endocardit is.
FIG URE 276. 1. Bact erial endocardit is w it h large veget at ion on t he mit ral
valve (arrow). (Reprint ed w it h permission f rom Aurigemma G P, Tighe DA.
Echocardiography in t he int ensive care unit . I n: I rw in RS, Rippe JM. I rw in &
Rippe's I nt ensive Care Medicine, 5t h ed. Philadelphia: Lippincot t Williams &
Wilkins, 2003: 91.)
Staphyl ococcus aureus inf ect ion has increased dramat ically over t he last 30
years and is now t he most common organism f ound in all t ypes of inf ect ive
endocardit is. Risk f act ors f or inf ect ion include int ravenous (I V) drug abuse,
indw elling cent ral venous cat het ers, and t he presence of condit ions such as
malignancy, renal f ailure, diabet es, alcoholism, and st eroid use. I nf ect ion w it h
t his organism t ends t o f ollow an aggressive course, of t en result ing in perivalvular
ext ension. More t han 50% of valve inf ect ions involving I V drug abuse involve t he
t ricuspid valve.
I nf ect ion w it h st rept ococcal organisms t ends t o f ollow a more subacut e
present at ion. Lef t -sided st ruct ures are aff ect ed most of t en, w it h mit ral valve
involvement more common t han aort ic valve inf ect ion. S. vi ri dans is responsible
f or 50% t o 60% of subacut e inf ect ions. Ent erococci are t he t hird-most -common
cause of inf ect ive endocardit is and have high rat es of ant imicrobial resist ance.
The nonent erococcal group D st rept ococci (especially S. bovi s) are associat ed
w it h gast roint est inal pat hologies such as colon cancer and polyps. I nf ect ion w it h
group B st rept ococci is most common in pregnant w omen or in elderly pat ient s
w it h chronic diseases. Embolic complicat ions are common and t reat ment of
t hese inf ect ions w it h ant ibiot ics alone is of t en unsuccessf ul.
Prost het ic valve inf ect ions make up 15% t o 20% of inf ect ive endocardit is. Early
(<2 mont hs af t er surgery) cases usually result f rom perioperat ive cont aminat ion
by organisms such as Staphyl ococcus epi dermi di s or S. aureus and of t en involve
mechanical valves. Lat e prost het ic valve inf ect ions can result f rom t ransient
bact eremia f rom dent al or gast roint est inal sources. These lat er inf ect ions are
of t en seen in bioprost het ic valves implant ed f or longer t han 1 year.
Suggested Readings
Bayer AS, Bolger AF, Taubert KA, et al. Diagnosis and management of
inf ect ive endocardit is and it s complicat ions. Circulat ion 1998; 98: 2936–2948.
Durack DT, Lukes AS, Bright DK. New crit eria f or diagnosis of inf ect ive
endocardit is: ut ilizat ion of specif ic echocardiographic f indings: Duke
Endocardit is Service. Am J Med 1994; 96: 200–209.
Mylonakis E, Calderw ood SB. I nf ect ive endocardit is in adult s. N Engl J Med
2001; 345(18): 1318–1330.
Sachdev M, Pet erson G E, Jollis JG . I maging t echniques f or diagnosis of

inf ect ive endocardit is. Cardiol Clin 2003; 21: 185–195.
> Table of C ontents > Im aging and Tes ts > 277 - D iagnos e Tam ponade B as ed on C linic al Findings and
not E c hoc ar diogr am
277
Diagnose Tamponade Based on Clinical Findings
and not Echocardiogram
Jay Weller MD
Signs and Symptoms

Cardiac t amponade occurs w hen increased int rapericardial pressure f or any
reason (e. g. , blood, f luid, air) impedes venous f illing of t he right vent ricle.
Sympt oms may include t achycardia, t achypnea, dyspnea, ort hopnea, and
diaphoresis. Beck's t riad of elevat ed cent ral venous pressure (CVP), syst emic
hypot ension w it h decreased pulse pressure, and dist ant / muff led heart sounds
describes t he classic clinical signs, but one or more f indings is f requent ly
absent . Elect rocardiography may reveal decreased amplit ude or elect rical
alt ernans. Ult imat ely, as int rapericardial pressure (I PP) dict at es int racardiac
chamber pressures, diast olic f illing pressures equilibrat e (i. e. , CVP = PADP =
PCWP = I PP, w here PADP is t he pulmonary art ery diast olic pressure, and PCWP
is t he pulmonary capillary w edge pressure). The hemodynamics of t amponade
physiology are uniquely suscept ible t o w ide respirat ory variat ions. I n part icular,
pulsus paradoxus is an exaggerat ion of t he normal physiologic decrease in
syst emic art erial blood pressure w it h spont aneous inspirat ion. I n t he post –
cardiac surgery pat ient , compression may be isolat ed t o a single chamber of t he
heart (e. g. , due t o loculat ed blood clot ), making t he classic signs and sympt oms
less usef ul; a high degree of clinical suspicion is required t o make t he diagnosis.
I ncreased availabilit y has led t o an increase in reliance on perioperat ive
echocardiography f or t he diagnosis of cardiac t amponade. Echo f indings may
include presence of a pericardial eff usion, right at rial collapse during syst ole,
right vent ricular collapse during diast ole, and inf erior vena caval plet hora.
Echocardiographic correlat es of pulsus paradoxus may be seen, including
respirat ory variat ion in bot h lef t and right vent ricular f illing as evidenced eit her
by Doppler inf low pat t erns or by variat ions in sept al w all mot ion. Right at rial
collapse last ing longer t han one t hird of syst ole has a 94% sensit ivit y and 100%
specif icit y f or t he diagnosis of t amponade. I n t he post –cardiac surgery pat ient ,
t ranst horacic imaging may be t echnically limit ed, necessit at ing a
t ransesophageal st udy (Fi gure 277. 1).
I mmediat e hemodynamic management goals f or t he pat ient w it h pericardial
t amponade are t o keep t he pat ient “f ull, f ast , and t ight . ” Because right
vent ricular f illing is impeded by increased int rapericardial
pressure, hypovolemia and vasodilat ion are poorly t olerat ed. The rest rict ive
nat ure of t amponade physiology prevent s compensat ory increases in st roke
volume, leaving cardiac out put heavily dependent on heart rat e. As discussed
previously, t he pat ient w it h pericardial t amponade is part icularly suscept ible t o
respirat ory variat ion in syst emic blood pressure. Conversion f rom spont aneous
vent ilat ion (w it h it s inspirat ory augment at ion of venous ret urn t o t he right
vent ricle) t o posit ive-pressure vent ilat ion (w it h increased int rat horacic pressure
inhibit ing venous ret urn) can lead t o cat ast rophic hemodynamic collapse.
FIG URE 277. 1. Transesophogeal echocardiogram show ing a large amount of

f luid in t he pericardial see causing right vent ricular collapse (arrow ).
Def init ive t reat ment of pericardial t amponade requires drainage of t he f luid f illing
t he pericardial sac. For t he pat ient in ext remis, percut aneous drainage should be
considered. I n t he case of penet rat ing t horacic t rauma, emergency depart ment
t horacot omy may be indicat ed. Echocardiography may be usef ul in guiding
needle placement . When t he eff usion is loculat ed or post erior, percut aneous
drainage may not be f easible. Because of it s sympat homimet ic propert ies,
ket amine is generally considered t o be t he induct ion agent of choice w hen
general anest hesia is required f or surgical drainage. For t he acut ely
sympt omat ic pat ient , it may be w ise f or t he surgeon t o prep and drape prior t o
induct ion of general anest hesia. Spont aneous vent ilat ion can
be maint ained t hrough t he induct ion phase t o sust ain hemodynamics unt il t he
eff usion is drained.
Suggested Readings
O t t o C, ed. Text book of Clinical Echocardiography, 2nd ed. Philadelphia: WB
Saunders, 2000: 218–222.
Thomas S, Kramer J, eds. Manual of Cardiac Anest hesia, 2nd ed. New York:
Churchill Livingst one, 1993: 129–136.
> Table of C ontents > Im aging and Tes ts > 278 - C ons ider P uls us Alter nans as a S ign of Im pending
Tam ponade
278
Consider Pulsus Alternans as a Sign of
Impending Tamponade
Melissa Camp MD
Cardiac t amponade occurs w hen f luid accumulat es in t he pericardial space and
leads t o impaired diast olic f illing due t o increased pericardial pressure. The
pericardium is relat ively inelast ic and in t he acut e set t ing an accumulat ion of 100
t o 200 cc of f luid can cause t amponade physiology. I n int ensive care unit
pat ient s t he most common scenarios involving t amponade include af t er
cardiot horacic surgery, af t er cent ral line placement w it h inadvert ent punct ure of
t he superior vena cava or right at rium, and af t er blunt or penet rat ing chest
t rauma. O f not e, f ollow ing cardiot horacic surgery, t amponade may also result
f rom a small clot compressing t he f illing of t he heart .
Signs and Symptoms

Early signs of t amponade include t achycardia and elevat ed cent ral venous
pressure, f ollow ed by hypot ension and decreased cardiac out put . Beck's t riad
describes t he classic f indings of jugular venous dist ent ion, muff led heart sounds,
and hypot ension. I f a pulmonary art ery cat het er is present , t he equalizat ion of
f illing pressures may be evident w it h right at rial pressure = right vent ricular end-
diast olic pressure = pulmonary art ery diast olic pressure = pulmonary art ery
w edge pressure = lef t vent ricular end diast olic pressure. I t should be not ed t hat
pressure equalizat ion may not be evident post cardiac surgery because t he
pericardium has invariably been opened. Pulsus paradoxicus, t he decrease in
syst olic blood pressure of great er t han 10 t o 15 mm Hg on inspirat ion, is also
observed in t amponade. Pulsus paradoxicus is a result of syst emic venous ret urn
during inspirat ion causing rapid right vent ricular f illing, w hich compresses and
impairs lef t vent ricular f illing.
Elect rical alt ernans is one of t he f indings on t he elect rocardiogram t hat is
suggest ive of cardiac t amponade. Wit h each beat , t he amplit ude of t he Q RS
complex varies f rom large t o small. Elect rical alt ernans is due t o a change in t he
Q RS axis f rom a “bobbing” movement of t he heart w it hin a large pericardial
eff usion (Fi g. 278. 1).
Emergent t reat ment of cardiac t amponade is pericardiocent esis using an 18-
gauge spinal needle. A needle w it h an elect rocardiogram lead at t ached w it h an
alligat or clip f or monit oring of needle posit ion
is insert ed 0. 5 cm t o t he lef t of t he xiphoid at an angle of 45 degrees and

direct ed post eriorly w hile being aimed t ow ard t he lef t shoulder and cont inuously
aspirat ed. The needle w ill produce a Q RS complex inversion once t he epicardium
of t he pericardial sac is cont act ed. The needle should be advanced unt il t he
pericardial space is ent ered and f luid or blood is aspirat ed. I f t he needle is
advanced t oo f ar, t ouching t he myocardium, ST-segment elevat ion w ill be
observed. I f ST elevat ion occurs, t he needle should be w it hdraw n back int o t he
pericardial space. O nce t he pericardial eff usion is drained and t amponade
relieved, diast olic f illing w ill be improved. As a result , cardiac out put and blood
pressure w ill increase.
FIG URE 278. 1. Pulsus alt ernans in t he carot id art ery. (Reprint ed w it h
permission f rom Const ant J. Bedside Cardiology, 5t h ed. Philadelphia:
Lippincot t Williams & Wilkins, 1999: 52.)
The best imaging st udy (alt hough not 100% sensit ive) t o conf irm diagnosis of a
suspect ed pericardial eff usion is an echocardiogram; it can also guide
pericardiocent esis.
Suggested Readings
Chen H, Sonnenday C, Lillemoe K, eds. Manual of Common Bedside Surgical
Procedures. Philadelphia: Lippincot t Williams & Wilkins, 2000: 90–94.
Lanken P, ed. The I nt ensive Care Unit Manual. Philadelphia: WB Saunders,
2001: 615–626.
Longo M, Jaff e C. I mages in clinical medicine—elect rical alt ernans. N Engl J

Med 1999; 341: 2060.
> Table of C ontents > Im aging and Tes ts > 279 - D o Not Us e a Negative Foc us ed As s es s m ent W ith
S onogr aphy for Tr aum a ( Fas t) E xam to R ule O ut B ow el Injur y or Injur y to the R etr oper itoneum or as
the O nly Tes t in P enetr ating Tr aum a
279
Do Not Use a Negative Focused Assessment
With Sonography for Trauma (Fast) Exam to
Rule Out Bowel Injury or Injury to the
Retroperitoneum or as the Only Test in
Penetrating Trauma
Brendan G . Carr MD
Patrick K. Kim MD
Focused assessment w it h sonography f or t rauma (FAST) is a noninvasive
ult rasound procedure t hat is a quick, reliable exam used t o det ect f ree
int raperit oneal f luid in t he injured pat ient . I nit ially used in Europe as a diagnost ic
modalit y, t he ease of use and abilit y t o evaluat e t he injured pat ient in t he
resuscit at ion area has led t o it s w idespread use. The FAST exam aids in t he
t riage of injured pat ient s direct ly t o t he operat ing room w it hout obt aining f urt her
imaging st udies. The t radit ional FAST exam consist s of t he f ollow ing f our view s
(one pericardial and t hree abdominal) (Fi g. 279. 1):
1. Subxiphoid view (t o evaluat e t he pericardium)

2. Right upper quadrant view (hepat orenal recess or Morrison's pouch)
3. Lef t upper quadrant view (splenorenal recess)
4. Suprapubic view
For decades, t he st andard f or evaluat ion of t he abdomen in unst able blunt

t rauma w as diagnost ic perit oneal lavage (DPL). This invasive modalit y is
sensit ive but nonspecif ic and reliance on DPL result s in high rat es of
nont herapeut ic laparot omies. Comput ed t omography (CT) has largely replaced
DPL in t he evaluat ion of t he st able blunt t rauma pat ient and is highly sensit ive
and specif ic. How ever, CT requires t ransport of t he pat ient t o t he radiology
suit e, w it h t he inevit able delay in obt aining and evaluat ing t he images.
I n cont rast , ult rasound is available at t he bedside, readily learnable by
nonradiologist s, and very quick t o perf orm (2 t o 4 minut es). I n t he blunt t rauma
populat ion, report ed sensit ivit y and specif icit y approximat e 83. 3% and 99. 7%,
respect ively. More import ant , in t he evaluat ion of hypot ensive t rauma pat ient s,
t he sensit ivit y and specif icit y of ult rasound approach 100% in experienced
hands.
FIG URE 279. 1. Transducer posit ions f or f ocused assessment w it h

sonography f or t rauma (FAST): (1) pericardial area, (2) right and (3) lef t
upper quadrant s, and (4) pelvis. (Reprint ed w it h permission f rom Rozycki
G S, Ballard RB, Feliciano DV, et al. Surgeon-perf ormed ult rasound f or t he
assessment of t runcal injuries: lessons learned f rom 1540 pat ient s. Ann Surg
1998; 228(4): 557–567. )
Watch Out For

The sensit ivit y of t he FAST exam depends on mult iple f act ors including t he
volume of hemoperit oneum, pat ient posit ioning, and ult rasonographer experience.
Alt hough t here is subst ant ial variabilit y, about 500 cc of int raperit oneal f luid is
necessary t o enable ult rasonographic visualizat ion (Fi g. 279. 2). I t must be not ed
t hat t he small amount of f ree f luid t ypically result ing f rom bow el perf orat ion is
not suff icient f or reliable ult rasound diagnosis. Ret roperit oneal injuries may not
communicat e w it h t he perit oneal cavit y and do not alw ays cause
hemoperit oneum. As a result , ult rasound remains unreliable in diagnosing
ret roperit oneal and hollow viscus injuries.
FIG URE 279. 2. (A) Sagit t al sect ion of liver, kidney, and diaphragm. Normal
f indings. (B) Sagit t al sect ion of right upper quadrant show ing blood bet w een
t he liver and kidney and bet w een t he liver and diaphragm. (Reprint ed w it h
permission f rom Rozycki G S, Ballard RB, Feliciano DV, et al. Surgeon-
perf ormed ult rasound f or t he assessment of t runcal injuries: lessons learned
f rom 1540 pat ient s. Ann Surg 1998; 228(4): 557–567. )
Alt hough t he FAST exam has hist orically been used only f or blunt t rauma
pat ient s, many cent ers have begun using ult rasound in penet rat ing t rauma
pat ient s. Many clinicians have f ound t hat t he FAST exam is helpf ul not only in
t riaging pat ient s direct ly t o t he operat ing room, but also in priorit izing body
cavit y explorat ion (i. e. , deciding on t horacot omy vs. laparot omy f irst ). The
overall sensit ivit y of FAST
in penet rat ing t rauma has been report ed t o be 46%, and as a result , FAST
should not be used as t he sole screening t ool in penet rat ing t rauma.
Alt hough it is not a sensit ive screening t est in overall penet rat ing t rauma, t he
FAST exam may be usef ul in penet rat ing t horacic t rauma. Ult rasound is a usef ul
screening t est in t he assessment of hemopericardium. Wit h a sensit ivit y of 100%
and a specif icit y of 97%, t he FAST exam has become a popular adjunct in
screening penet rat ing t horacic t rauma pat ient s. Pleural ult rasound may be as
sensit ive as chest x-ray in t he det ect ion of pneumot horax in t he t rauma
populat ion. Researchers have called f or ext ending t he FAST exam t o include
pleural view s t o diagnose bot h hemot horax and pneumot horax. An import ant
limit at ion t o t he use of ult rasound in t horacic t rauma is t hat cardiac injury cannot
be eff ect ively ruled out in t he cont ext of a f luid-f illed t horacic cavit y
(hemot horax) because large amount s of pleural f luid can obscure a small
pericardial collect ion and t raumat ic hemopericardium can decompress int o t he
pleural space. Despit e it s limit at ions, given t he ready availabilit y of ult rasound in
t rauma set t ings, ult rasound should be used early and liberally in t he evaluat ion of
penet rat ing t horacic w ounds.
Suggested Readings
Kirkpat rick AW, Sirois M, Laupland KB, et al. Hand-held t horacic sonography
f or det ect ing post -t raumat ic pneumot horaces: t he ext ended f ocused
assessment w it h sonography f or t rauma (EFAST). J Trauma 2004; 57(2): 288–
295.
Meyer D, Jessen ME, G rayburn PA. The use of echocardiography t o det ect
cardiac injury af t er penet rat ing t horacic t rauma: a prospect ive st udy. J
Trauma 1995; 39: 902–907.
Rozycki G S, Ballard RB, Feliciano DV, et al. Surgeon-perf ormed ult rasound
f or t he assessment of t runcal injuries: lessons learned f rom 1540 pat ient s.
Ann Surg 1998; 228(4): 557–567.
Rozycki G S, Feliciano DV, O chsner MG , et al. The role of ult rasound in

pat ient s w it h possible penet rat ing cardiac w ounds: a prospect ive mult icent er
st udy. J Trauma 1999; 46(4): 543–551.
Udobi KF, Rodriguez A, Chiu WC, et al. Role of ult rasonography in penet rat ing
abdominal t rauma: a prospect ive clinical st udy. J Trauma 2001; 50(3): 475–
479.
> Table of C ontents > Im aging and Tes ts > 280 - C ons ider Us ing C om puted Tom ogr aphy
R ec ons tr uc tions of the C hes t/Abdom en/P elvis to R ule O ut Thor ac ic and Lum bar Fr ac tur es and
D is loc ations
280
Consider Using Computed Tomography
Reconstructions of the Chest/Abdomen/Pelvis to
Rule Out Thoracic and Lumbar Fractures and
Dislocations
Suneel Khetarpal MD
Barbara Haas MD
Evaluat ion f or t he presence of t horacolumbar spine injury is f requent ly necessary
in t he mult iply injured pat ient based on clinical f indings, mechanism of injury,
inabilit y t o adequat ely assess f or injury due t o decreased level of consciousness,
or dist ract ing injuries. Tradit ionally, t he t horacolumbar spine w as evaluat ed using
ant eropost erior and lat eral plain f ilms, but t his modalit y is considered less t han
ideal f or several reasons. Because t he t horacolumbar spine cannot be
adequat ely evaluat ed w it h port able radiographs, obt aining such f ilms requires
t ransport at ion of t he crit ically injured pat ient aw ay f rom areas equipped f or
emergency resuscit at ion. Moreover, a signif icant t ime delay bet w een arrival in
t he t rauma bay and complet ion of t hese f ilms is document ed. Most import ant , t he
sensit ivit y of plain f ilms has been report ed t o be as low as 58%, w it h a high
proport ion of f ilms being described as t echnically inadequat e and requiring
f urt her comput ed t omography (CT) imaging.
What to Do
The mechanism of injury leading t o t he need f or t horacolumbar spine imaging
f requent ly also requires t hat pat ient s have CT imaging of t he chest and
abdomen, and an increasing volume of lit erat ure suggest s t hat imaging ordered
t o rule out visceral injury can be used t o expedit e screening f or t horacolumbar
spine injuries. CT scans of t he chest , abdomen, and pelvis w ere f ound t o be
more sensit ive and specif ic f or t he det ect ion of spinal injury t han plain f ilms,
even if no specif ic spine prot ocol w as included. I n one prospect ive st udy, CT
w as f ound t o have a sensit ivit y of 97% and a specif icit y of 99%. Alt hough neit her
t echnique missed unst able f ract ures, CT images w ere more likely t o det ect
st able f ract ures. O t her st udies report ed similar result s.
Published report s also suggest t hat t he use of CT decreases t he delay bet w een
arrival of t he pat ient in t he t rauma bay and t horacolumbar spine clearance, w hich
reduces demand on emergency room resources and permit s t he bet t er ut ilizat ion
of personnel w ho previously w ere engaged in pat ient t ransport (nurses,
resident s, and respirat ory
t echnicians). Moreover, because pat ient s at highest risk of t horacolumbar spine

injuries are being scanned f or ot her indicat ions, including t horacolumbar spine
reconst ruct ions does not increase t he ut ilizat ion of t he CT scan or expose t he
pat ient t o addit ional radiat ion, and t his new approach is very likely more cost
eff ect ive, part icularly in high-risk pat ient s.
> Table of C ontents > Im aging and Tes ts > 281 - D o not R ule O ut C er vic al S pine or S pinal C or d Injur y
on B ony Film s or C om puted Tom ogr aphy Alone
281
Do not Rule Out Cervical Spine or Spinal Cord
Injury on Bony Films or Computed Tomography
Alone
The incidence of cervical spine injury in blunt t rauma pat ient s varies according t o
t he crit eria used t o select pat ient s f or screening, t he manner in w hich pat ient s
are screened, and t he crit eria used t o def ine injury relat ive t o a given imaging
modalit y. The Nat ional Emergency X-radiography Ut ilizat ion St udy (NEXUS)
report ed an incidence of 2. 8% among pat ient s evaluat ed in emergency
depart ment s f ollow ing blunt t rauma. G ross man and colleagues report ed an
incidence of 4. 3% in more t han 100, 000 pat ient s admit t ed t o t rauma cent ers. I t
has been show n t hat t he incidence of cervical spine injury increases w it h injury
severit y, severe closed head injury, severe f acial f ract ures, associat ed spinal
column injury, and t he use of comput ed t omography (CT) scanning as a screening
t ool. The overall clinical relevance of rout ine use of CT as a screening t ool f or
cervical spine injury is not know n w it h respect t o adequat e t hree-view cervical
spine x-rays. Sensit ivit y f or det ect ion of bony injury is clearly bet t er. How ever,
t reat ment mandat ed by det ect ion of injury t hat may be clinically silent has
increased t he use of rigid cervical collars w it h an unknow n eff ect on out comes.
I t is import ant t o not e t hat exclusion of bony injury does not “clear” t he cervical
spine in every case. There is a possibilit y t hat pat ient s w it hout bony injury may
sust ain clinically signif icant ligament ous injury t hat could result in subluxat ion and
neurologic sequelae if unrecognized. The incidence of such injuries is diff icult t o
est imat e based on considerat ions out lined earlier, but is most likely less t han 1%
in most t rauma cent ers. Similarly t he incidence of spinal cord injury w it hout
radiographic abnormalit y (SCI WO RA) is diff icult t o est imat e and most likely
represent s operat ional t erminology prior t o t he w idespread use of magnet ic
resonance imaging (MRI ) t o image pat ient s w it h any degree of neurologic
dysf unct ion f ollow ing injury. Dysf unct ion includes but is not limit ed t o anat omic
syndromes (cent ral cord, ant erior cord), dysest hesias, t ransient sympt oms
(spinal “concussion”), and disc herniat ions. MRI has high sensit ivit y in det ect ing
anat omic lesions, part icularly in t he presence of acquired chronic spinal st enosis
or spondylolist hesis.
Watch Out For

Many opt ions exist f or clearance of t he cervical spine. For pat ient s w ho are
aw ake, alert , nonint oxicat ed, and can provide reliable physical examinat ion,
clearance t hrough clinical exam is reasonable. Det erminat ion of “reliable physical
exam” is a mat t er of clinical judgment inf ormed by t he clinical hist ory. High-
energy mechanism, associat ed low er spine f ract ure, advanced age, or
dist ract ing injury part icularly in proximit y (upper ribs, clavicle) t o t he cervical
spine may inf luence t he pract it ioner t o obt ain addit ional inf ormat ion.
Flexion-ext ension f ilms of t he cervical spine have a role in est ablishing st abilit y
provided adequat e range of mot ion (30 degrees in bot h direct ions) is present
and t he cervicot horacic junct ion can be visualized. G iven t hese considerat ions,
t his modalit y is most eff ect ive f or aw ake, alert pat ient s w ho can cooperat e w it h
inst ruct ions. Cervical MRI has evolved as a gold st andard despit e lack of
specif icit y f or injuries t hat are t ruly “unst able” (i. e. , comparison t o a st andard
t hat allow s correlat ion of sof t t issue f indings w it h subluxat ion). How ever,
sensit ivit y of MRI is good, and a normal MRI provides t he clinician w it h t he
assurance t hat ligament ous injury has been excluded t o t he best degree of
medical cert aint y.
Suggested Readings
G rossman MD, Reilly PM, G illet t T, et al. Nat ional survey of t he incidence of
cervical spine injury and approach t o cervical spine clearance in US t rauma
cent ers. J Trauma 1999; 47: 684–690.
G uidelines f or t he Management of Cervical Spine I njury. Neurosurgery

2002; 50(3): S1– S199.
Hoff man JR, Mow er WR, Wolf son AB, et al. The Nat ional Emergency X-
radiography Ut ilizat ion St udy (NEXUS) group. Validit y of a set of clinical
crit eria t o rule out injury t o t he cervical spine in pat ient s w it h blunt t rauma.
NEngl J Med 2000; 343: 94– 99.
I nsko EK, G racias VH, G upt a R, et al. Ut ilit y of f lexion and ext ension
radiographs of t he cervical spine in t he acut e evaluat ion of blunt t rauma. J
Trauma 2002; 53(3): 426–429.
> Table of C ontents > Im aging and Tes ts > 282 - Us e the P os ition of the Medias tinum as a C Lue to the
D iagnos is of A W hite- out on C hes t R adiogr aph
282
Use the Position of the Mediastinum as a CLue
to the Diagnosis of A White-out on Chest
Radiograph
Noah Lechtzin MD, MHS
Watch Out For

The chest radiograph (CXR) is of t en t he f irst radiologic st udy obt ained in t he
evaluat ion of dyspnea. I n t he int ensive care unit (I CU), port able chest
radiographs are f requent ly ordered “STAT” t o evaluat e pat ient s w it h new or
acut e onset of short ness of breat h. O ccasionally, t here is diff use opacif icat ion or
“w hit e-out ” of a hemit horax. Though t he complet e diff erent ial diagnosis f or
“w hit e-out ” is quit e large, t he complet e opacif icat ion of a hemit horax w it h no
visible lung markings is usually caused by one of t w o main processes:
at elect asis/ lung collapse or massive pleural eff usion, w it h t he main t ypes of
f luids in t he pleural space being serous f luid (hydrot horax), blood (hemot horax),
lipid (chylot horax), and pus (pyot horax or empyema). Less common causes of
w hit e-out include a consolidat ive process w it h f illing of t he air spaces, such as
bact erial pneumonia, sof t t issue t umor, or post -pneumonect omy w it h t he residual
space f illing w it h f luid and f ibrot ic mat erial.
The prompt recognit ion of t he correct diagnosis is import ant f or appropriat e and
t imely t herapeut ic int ervent ion. O f t en, t he most helpf ul clue t o t he diagnosis of
w hit e-out of a hemit horax is t he posit ion of t he mediast inum. Normally, t he
t rachea is midline; how ever, t he mediast inum is a mobile st ruct ure and responds
t o diff erences in pressure bet w een t he t w o sides of t he t horax. I n complet e lung
collapse, of t en caused by proximal obst ruct ion of t he bronchus and post -
pneumonect omy, t here is loss of volume on t he opacif ied side. The CXR reveals
complet e opacif icat ion of t he hemit horax w it h shif t of t he mediast inum t ow ard t he
collapsed side (t o t he side of t he w hit e-out ). O t her radiologic signs relat ed t o
t he reduct ion in int rapleural pressure include diaphragmat ic elevat ion, narrow ing
of t he int ercost al spaces of t he aff ect ed lung, and hyperinf lat ion of t he opposit e
lung.
I n t he case of a pleural eff usion, t he opacif icat ion of t he t horax is usually
homogeneous, and if t he pat ient is erect , t here may be a visible meniscus. I n
massive eff usions (usually >1, 000 mL), t he int rapleural pressure on t he aff ect ed
side of t he lung is increased and t he mediast inum shif t s t o t he cont ralat eral side.
I n t he case of consolidat ion, t he size of t he hemit horax is of t en unchanged and
t he mediast inum remains midline. I f t he pneumonic consolidat ion is massive and
t he lung volume increases or if t he consolidat ion is associat ed w it h a
concomit ant eff usion, t here may be mediast inal shif t t o t he cont ralat eral side.
How ever, in t he lat er st ages of t he consolidat ive process secret ions may cause
airw ay obst ruct ion w it h associat ed at elect asis and volume loss, result ing in
mediast inal shif t t o t he ipsilat eral side. An addit ional radiologic clue t o t he
diagnosis of a consolidat ive process is t he presence of air bronchograms.
Suggested Readings
Muller NL, Colman N, Pare PD, et al. , eds. Fraser and Pare's Diagnosis of
Diseases of t he Chest , 4t h ed. Philadelphia: WB Saunders, 1999: 525–534.
Reed JC. Chest Radiology, 5t h ed. Philadelphia: Mosby, 2003.

> Table of C ontents > Im aging and Tes ts > 283 - C ons ider Angiogr aphy for B lunt S olid- O r gan Injur y
283
Consider Angiography for Blunt Solid-Organ
Injury
Suneel Khetarpal MD
Barbara Haas MD
The abilit y t o manage t he majorit y of solid-organ injuries nonoperat ively has
become a cornerst one in t he opt imal approach t o blunt t rauma. Bot h diagnost ic
and t herapeut ic angiography is gaining increasing prominence as a t echnique by
w hich t his goal can be achieved.
The spleen is t he most f requent ly injured organ f ollow ing blunt abdominal t rauma,
w it h splenic injuries being present in 25% t o 30% of cases. Liver injury is less
f requent (15% t o 20% of pat ient s), and renal t rauma is present in approximat ely
10% of pat ient s. Alt hough t he circumst ances in w hich angiography can be used in
t he management of t hese injuries are const ant ly evolving, a number of report s
have demonst rat ed it s value as an adjunct t hat can increase overall organ
salvage and decrease t he need f or surgical int ervent ion.
All grades of splenic injury have been successf ully managed nonoperat ively in
hemodynamically st able pat ient s. How ever, increasing grade of injury, as w ell as
cont rast ext ravasat ion or blush on admission comput ed t omography (w hich MUST
be perf ormed w it h int ravenous cont rast ), has been associat ed w it h increasing
likelihood of nonoperat ive management f ailure. Angioembolizat ion has been
considered usef ul in avoiding surgery in t his subset of pat ient s, w ho are most
likely t o f ail observat ion in t he int ensive care unit . I n recent ly published report s,
bet w een 5% and 15% of pat ient s w it h splenic injuries w ere successf ully
managed w it h angioembolizat ion, w it h success rat es great er t han 90% report ed.
The indicat ions and t echnique f or angioembolizat ion in splenic t rauma have been
evolving rapidly and are t he subject of considerable debat e. Advocat es of
proximal splenic art ery embolizat ion t heorize t hat t his approach decreases
splenic blood pressure, w hich prevent s delayed hemorrhage and accelerat es
splenic healing w hile maint aining splenic perf usion t hrough collat eral blood f low.
O t hers use angioembolizat ion at t he level of more dist al art eries t o t he spleen,
w hich requires more t ime and manipulat ion but preserves blood f low t o t he
spleen.
Watch Out For
I t is import ant t o not e t hat approximat ely 8% t o 10% of pat ient s w ho init ially
appear t o have been successf ully managed nonoperat ively are at
risk f or delayed complicat ions requiring int ervent ion. Complicat ions include
persist ent hemorrhage, delayed splenic rupt ure due t o expanding hemat oma, and
splenic art ery pseudoaneurysm f ormat ion. Angioembolizat ion should be
considered in t he t reat ment of t hese complicat ions, given t hat pat ient s w ho f ail
nonoperat ive management and require delayed operat ive int ervent ion have
signif icant ly w orse out comes t han bot h pat ient s managed nonoperat ively and
pat ient s managed surgically at init ial present at ion.
The use of angiography and art erial embolizat ion f or hepat ic injury has also
undergone considerable evolut ion. I nit ially int roduced as a viable t herapeut ic
modalit y f or hemodynamically st able pat ient s, t he t echnique has been show n t o
be a successf ul alt ernat ive t o surgery in pat ient s w ho are hemodynamically
st able only w it h ongoing resuscit at ion. These are generally pat ient s w it h isolat ed
liver injuries or liver injuries associat ed w it h concomit ant int ra-abdominal t rauma
not requiring emergent laparot omy. Moreover, emphasis has been placed on t he
use of angioembolizat ion in pat ient s w it h high grades of injury and evidence of
ongoing art erial bleeding on comput ed t omography in t he f orm of cont rast blush.
Hist orically, t hese pat ient s are most likely t o f ail nonoperat ive management .
A separat e group of pat ient s w ho benef it f rom hepat ic art ery angioembolizat ion
are severely injured pat ient s f or w hom t he t echnique is a component of
post operat ive st abilizat ion and resuscit at ion. I n part icular, pat ient s w it h grade I V
and grade V liver injuries appear t o benef it f rom angioembolizat ion, alt hough t he
dat a support ing t his approach combine pat ient s w it h bot h blunt and penet rat ing
mechanisms of injury. Finally, angioembolizat ion has successf ully been used in
t he management of delayed complicat ions in pat ient s w it h hepat ic injuries.
Specif ically, up t o 5% of pat ient s managed nonoperat ively subsequent ly develop
delayed or recurrent hemorrhage, 1% t o 2% develop vascular abnormalit ies,
such as pseudoaneurysm, and 1% develop hemobilia. I n most cases of delayed
vascular complicat ions, angioembolizat ion is t he primary t reat ment of choice.
Angioembolizat ion has also been used w it h success in t he management of blunt
renal injuries, alt hough t here is less experience managing t his t ype of injury w it h
int ervent ional radiology. G iven t hat 85% of renal injuries are grade I I I or less,
t hese injuries are less likely t o require int ervent ion. Successf ul angioembolizat ion
of renal vascular injuries in pat ient s w ho are hemodynamically st able has been
report ed by a number of groups, w it h emphasis being placed on highly select ive
embolizat ion t o at t empt maximum t issue preservat ion and t hus preserve renal
f unct ion.
As w it h any int ervent ion, angiography and subsequent embolizat ion is not w it hout
risk. Angiography is associat ed w it h risk of hemat oma, vessel t hrombosis, and
vascular injury, such as vessel dissect ion or pseudoaneurysm f ormat ion. Vessel
embolizat ion can be associat ed w it h t issue inf arct ion and subsequent abscess
f ormat ion, part icularly w hen less select ive t echniques are used. Nevert heless,
t he ut ilit y of t his t echnique in t he successf ul nonoperat ive and operat ive
management of blunt solid-organ injuries has been demonst rat ed in a variet y of
pat ient populat ions.
Suggested Readings
Haan JM, Biff l W, Knudson MM, et al. West ern Trauma Associat ion Mult i-
I nst it ut ional Trials Commit t ee. Splenic embolizat ion revisit ed: a mult icent er
review. J Trauma 2004; 56(3): 542–547.
Harbrecht BG . I s anyt hing new in adult blunt splenic t rauma? Am J Surg 2005;
190(2): 273–278.
> Table of C ontents > Im aging and Tes ts > 284 - B e Aler t for C om par tm ent S yndr om es
284
Be Alert for Compartment Syndromes
Brett M. Cascio MD
Derek Papp MD
Compart ment syndromes are common f ollow ing t rauma and ot her condit ions in
crit ically ill pat ient s. I rreversible muscle and nerve injuries of t en occur in
unrecognized compart ment syndromes. Many healt h care providers are unf amiliar
w it h t he sympt oms and signs of an incipient compart ment syndrome and are
unable t o make an early diagnosis.
Pathophysiology
Compart ment syndromes are caused by a signif icant increase in t he t issue
pressure w it hin a closed space result ing in reduced circulat ion t o muscles and
nerves. As t he t issue pressure increases, t he perf usion gradient decreases and
muscle and nerve ischemia occurs. Working w it h dogs, Heckman f ound
irreversible muscle inf arct ion w hen t he t issue pressure w as w it hin 10 mm Hg of
t he diast olic pressure. Bernot not ed t hat muscle subject ed t o ischemia bef ore an
increase in compart ment pressure developed hypoxic met abolic changes w hen
t he perf usion pressure w as w it hin 40 mm Hg of t he mean art erial pressure.
Cellular anoxia is t he f inal common denominat or of all compart ment syndromes,
w it h myoglobin released f rom t he injured muscle cells. Pat ient s w ho have
compart ment syndromes may experience myoglobinuria, met abolic acidosis, and
hyperkalemia, w hich may result in renal f ailure, shock, hypot hermia, and cardiac
arrhyt hmias or f ailure.
Risk Factors
The most common cause of compart ment syndromes is a f ract ure. How ever,
virt ually any condit ion t hat causes marked muscle or compart ment sw elling may
cause a compart ment syndrome, including bleeding f rom coagulopat hies or
ant icoagulat ion; crush injuries; prolonged dependency (int ravenous drug abusers,
alcoholics, and st roke vict ims); t ight cast s or dressings; ost eot omies; burns;
reperf usion; prolonged operat ing room posit ion; and milit ary ant ishock t rousers.
Signs and Symptoms
Pain out of proport ion t o injury is t he hallmark f inding in pat ient s w it h a
compart ment syndrome. Pat ient s not e a severe, unrelent ing
discomf ort , w hich is of t en not relieved by narcot ics. Pat ient s may explain t heir
pain by st at ing t hat it is int olerable, or t hey may lay quiet ly w it h t heir eyes
closed in an at t empt t o t olerat e t he pain. Escalat ing doses of narcot ics are of t en
necessary t o cont rol t he pain. There is a signif icant risk t hat clinicians may
prescribe suff icient narcot ics t o mask t he compart ment syndrome, leading t o a
signif icant delay in diagnosis. Pat ient s may also describe neurologic sympt oms
such as parest hesias and a loss of mot or f unct ion.
Pain w it h passive st ret ch of t he muscles t raversing t he compart ment , t enseness,
pain on palpat ion of t he involved compart ment , and loss of mot or or sensory
f unct ion are t he import ant physical examinat ion f indings. Unf ort unat ely,
t enseness and pain on palpat ion are not specif ic f or compart ment syndrome and
are of t en present in pat ient s w it h f ract ures. Pat ient s w it h a compart ment
syndrome in t he leg exhibit severe discomf ort w hen t he t oes are f lexed or
ext ended; t hey of t en do not allow t he examiner t o repeat t he t est . Many pat ient s
w it h compart ment syndrome are not able t o act ively f lex and ext end t heir t oes.
I t is very import ant t o perf orm serial mot or and sensory examinat ions. Loss of
sensat ion is a common f inding as a compart ment syndrome evolves. I n t he leg,
loss of sensat ion in t he dist ribut ion of t he deep peroneal nerve (f irst w eb space
of t he f oot ) is a very common early f inding. Weakness of ankle and great t oe
dorsif lexion are ot her common early f indings. Pallor, pulselessness, and
paralysis are very lat e f indings and indicat e t hat t here is complet e muscle loss
and possible t ot al ischemia if t here are no pulses.
Pressure M easurement/Indications for Fasciotomy

The measurement of compart ment pressures is usef ul f or det ermining w het her a
pat ient w ho is experiencing severe pain or t he unconscious pat ient w ho has
signif icant sw elling has a compart ment syndrome. When an at -risk pat ient has
pain out of proport ion t o injury, a t ense compart ment , and pain on passive
st ret ch, one can make t he diagnosis of compart ment syndrome and t reat t he
pat ient w it h an emergency release of t he f ascial compart ment w it hout measuring
pressures. I n many pat ient s, it can be diff icult t o separat e t he severe pain of t he
init ial injury f rom t hat of a compart ment syndrome. I n t his scenario measurement
of t he pressures assist s t he clinician in det ermining w het her a compart ment
syndrome is present .
I nt racompart ment al pressures are easily measured w it h commercially available
devices, such as t he St ryker Monit or. The device has
a needle w it h a side port , and t he needle is advanced int o t he compart ment . A

small amount of saline is inject ed int o t he compart ment , and t issue pressure is
det ermined (normal pressure is 0 t o 7) and t hen compared t o t he diast olic
pressure. I f t he t issue pressure is w it hin 30 mm Hg of t he diast olic pressure,
t here is signif icant risk of muscle and nerve ischemia and f asciot omy should be
perf ormed. I f t he diff erence bet w een t he t issue pressure and t he compart ment
pressure is great er t han 30 mm Hg, a compart ment syndrome is not present and
t he pat ient can be observed.
Emergency Fasciotomy
Pat ient s w it h compart ment syndromes require emergency f asciot omy t o reduce
t he t issue pressure and increase t he perf usion pressure t o nerves and
muscle. The f asciot omies are ideally perf ormed w it hin 45 t o 90 minut es of
diagnosis. Fasciot omies are perf ormed t hrough st andard incisions t o allow
complet e release of t he f ascia. The f ascia is released f rom t he originat ion of t he
muscle t o t he musculot endinous junct ion. The skin and subcut aneous t issues are
lef t open and eit her closed by delayed primary closure or skin graf t s at a lat er
dat e.
O ccasionally, a compart ment syndrome goes unrecognized f or great er t han 24
hours. I n t his scenario, pat ient s have complet e and irreversible loss of muscle
and nerve f unct ion. There is no pot ent ial f or recovery of f unct ion. When a
compart ment syndrome has been present f or more t han 24 hours and t here is
complet e mot or and sensory loss, t here is no indicat ion f or f asciot omy.
Fasciot omy may be delet erious in t hat t here may be an increased release of
myoglobin and ot her t oxic met abolit es. Even more problemat ic is t he diff icult y
w it h at t aining w ound healing af t er f asciot omy in an ext remit y w it h an
unrecognized compart ment syndrome. The risk of amput at ion is very high.
Special Considerations in the Intensive Care Unit

Pat ient s are especially at risk f or missed compart ment syndromes in t he set t ing
of an int ensive care unit because t hey of t en are unable t o communicat e t hat t heir
limb hurt s. The cardinal f inding of pain out of proport ion t o injury may not be
apparent . Caref ul physical examinat ion of all ext remit ies in t he unconscious
pat ient should be perf ormed on a daily basis. Pat ient s w ho might be especially
at risk include t he f ollow ing:
1. I n pat ient s w ho have had long operat ions, t he lat eral decubit us or lit hot omy
posit ion may predispose t he pat ient t o a compart ment syndrome in a
nont raumat ized leg.
2. Pat ient s w ho have reperf usion of an ischemic limb (i. e. , vascular surgery)
are very much at risk f or t he development of compart ment syndrome of t he
leg.
3. O bese pat ient s w ho are not able t o t urn t hemselves are at risk f or
glut eal/ but t ock compart ment syndromes.
4. Pat ient s w ho have sust ained massive t rauma and have been int ubat ed at t he
scene of t he accident or in t he emergency room of t en have missed ext remit y
lesions because t he clinician cannot obt ain a hist ory of painf ul areas.
Compart ment syndrome may occur in t he upper and low er ext remit ies in t his
group of pat ient s.
5. Pat ient s w ho have had a t hrombot ic event or a massive int ravenous
inf ilt rat ion are at risk secondary t o sw elling.
Suggested Readings
Bernot , M, G upt a, R, Dobrasz, J, et al. The eff ect of ant ecedent ischemia on
t he t olerance of skelet al muscle t o increased int erst it ial pressure. J O rt hop
Trauma 1996; 10: 555– 559.
Cascio BM, Buchow ski JM, Frassica FJ. Well-limb compart ment syndrome
af t er prolonged lat eral decubit us posit ioning: a report of t w o cases. J Bone
Joint Surg 2004; 86A(9): 2038–2040.
Cascio BM, Wilckens JH, Ain MC, et al. Document at ion of acut e compart ment
syndrome at an academic healt h-care cent er. J Bone Joint Surg
2005; 87A(2): 346–350.
Heckmann MM, Whit esides TE Jr, G rew e SR, et al. Hist ologic det erminat ion
of t he ischemic t hreshold of muscle in t he canine compart ment syndrome
model. J O rt hop Trauma 1993; 7: 199–210.
> Table of C ontents > Im aging and Tes ts > 285 - B e Aler t for Ac alc ulous C holec ys titis
285
Be Alert for Acalculous Cholecystitis
Meredith S. T inti MD
Patrick K. Kim MD
Acut e acalculous cholecyst it is is an inf lammat ory process of t he gallbladder in
t he absence of calculi. I n cont rast t o acut e cholecyst it is, in w hich anat omic
obst ruct ion by gallst ones is t he primary process, obst ruct ion is only variably
present and is not t he causat ive process in acalculous cholecyst it is. The
pat hophysiology of acut e acalculous cholecyst it is is not clearly def ined but is
believed t o be produced by a combinat ion of syst emic mediat ors of inf lammat ion,
biliary st asis, and ischemia. Alt hough 60% t o 90% of all post operat ive or post -
t rauma cases of cholecyst it is are acalculous, t he overall incidence of acalculous
cholecyst it is in t he int ensive care unit is only 0. 2%.
Signs and Symptoms

Diagnosis of acalculous cholecyst it is is of t en diff icult and f requent ly delayed. I n
part , t his is due t o it s preponderance in t he diff icult -t o-examine crit ically ill
pat ient s w it h signif icant comorbidit ies. Early considerat ion of t he disease and a
high degree of suspicion are required t o ensure prompt diagnosis. The delay in
diagnosis and t he ill pat ient populat ion lead t o a high rat e of complicat ion, such
as gangrene or perf orat ion (40% t o 60%). The clinical f indings are nonspecif ic
and include f ever, right upper quadrant abdominal pain, leukocyt osis, and
elevat ion of liver enzymes (LFTs) and bilirubin. Abdominal pain is t he most
consist ent f inding, but it is not alw ays localized t o t he right upper quadrant ;
leukocyt osis and elevat ed LFTs are present in approximat ely 75% of cases.
Def init ive diagnosis can be made by many diff erent t est s. Abdominal
ult rasonography is usually t he f irst t est perf ormed in suspect ed cases of
acalculous cholecyst it is because it can be done at t he bedside of crit ically ill
pat ient s and has a sensit ivit y and a specif icit y of 70%. Hepat obiliary scint igraphy
(hepat obiliary iminodiacet ic acid [ HI DA] scan) requires t ransport of t he pat ient t o
t he nuclear medicine suit e but has a sensit ivit y of 80% t o 90% and a specif icit y
of 90% t o 100%. Comput ed t omography (CT) also requires t ransport at ion of t he
pat ient but is usef ul f or evaluat ing ot her sources of abdominal signs and
sympt oms. CT has a specif icit y of 90% t o 95% if cert ain diagnost ic crit eria are
met .
Surgical cholecyst ect omy (laparoscopic or open) is t he t reat ment of choice
because t he incidence of gangrene, perf orat ion, and empyema is signif icant ly
higher t han f or calculous cholecyst it is. How ever, many pat ient s w it h acalculous
cholecyst it is have an ext remely high operat ive mort alit y risk. Pat ient s can
somet imes be managed w it h percut aneous cholecyst ost omy. This procedure has
a 95% t o 100% t echnical success rat e, but a clinical response of only 70% t o
85%. I f t here is no improvement af t er 72 hours or if t here are signs of gangrene
or emphysema at t he t ime of diagnosis, surgical cholecyst ect omy is required.
Follow ing successf ul t reat ment of acalculous cholecyst it is w it h a percut aneous
cholecyst ost omy t ube, t he t ube may be removed if t he pat ient is no longer
crit ically ill, a t ract has f ormed over a 3- t o 6-w eek period, t he cyst ic and
common bile duct s are demonst rat ed t o be open, and t here is no ascit es
present . There is no consensus in t he lit erat ure as t o w het her cholecyst ect omy
is mandat ory f ollow ing cholecyst ost omy; many surgeons opt t o observe
sympt om-f ree pat ient s.
Suggested Readings
Cameron J, ed. Current Surgical Therapy, 8t h ed. Philadelphia: Elsevier
Mosby, 2004: 385–392.
Mullholl and MW, Lillemoe KD, Dohert y G M, et al. , eds. G reenf ield's Surgery:
Scient if ic Principles and Pract ice, 4t h ed. Philadelphia: Lippincot t Williams &
Wilkins, 2005: 990–991.
Lane JD, Lomis N. Cholecyst it is, acalculous.

w w w. emedicine. com/ RADI O / t opic157. ht m
> Table of C ontents > Im aging and Tes ts > 286 - Avoid Giving Intr avenous C ontr as t D ye m or e O ften
Than E ver y 48 Hour s if P os s ible
286
Avoid Giving Intravenous Contrast Dye more
Often Than Every 48 Hours if Possible
Amisha Barochia MD
Radiocont rast -induced acut e renal f ailure is t he t hird-leading cause of acut e
renal f ailure in hospit alized pat ient s in t he Unit ed St at es. More t han 10 million
radiocont rast st udies are perf ormed annually in t he Unit ed St at es. O f
import ance, cont rast -induced nephropat hy, alt hough largely prevent able, is
associat ed w it h 34% mort alit y. Radiocont rast -induced nephropat hy is due t o
acut e t ubular necrosis t hat result s f rom reduced renal perf usion secondary t o
vasoconst rict ion, t ubular obst ruct ion, and direct t ubular t oxicit y f rom t he cont rast
medium, w hich is t hought t o be mediat ed by oxygen f ree radicals. The def init ion
of radiocont rast -induced nephropat hy is variable, w it h some def ining it as a
great er t han 25% increase f rom baseline serum creat inine and ot hers def ining it
as an absolut e increase of great er t han 0. 5 mg/ dL of creat inine f rom baseline
w it hin 48 hours of exposure t o cont rast . Because absolut e serum creat inine
values are an imprecise met hod f or measuring renal f unct ion, t he calculat ion of
est imat ed glomerular f ilt rat ion rat e is recommended bef ore and af t er
administ rat ion of cont rast dye t o evaluat e f or acut e renal f ailure.
The f ollow ing risk f act ors predispose pat ient s t o nephropat hy secondary t o
radiocont rast agent s: pre-exist ing chronic renal insuff iciency, diabet es mellit us,
reduced eff ect ive circulat ing volume as in pat ient s w it h int ravascular deplet ion or
heart f ailure, and t he use of large volumes of cont rast dye. Most pat ient s w ho
develop radiocont rast -induced nephropat hy recover t heir renal f unct ion w it h no
last ing sequelae. The serum creat inine value usually peaks at 3 t o 5 days and
ret urns t o baseline w it hin 10 days. How ever, a f ew pat ient s have a residual
requirement f or renal replacement t herapy. These pat ient s have a report ed 2-
year survival rat e of 18%. I n addit ion, t he risk-adjust ed odds rat io f or deat h
af t er t he development of radiocont rast -induced nephropat hy is nearly 5 (p < . 01).
O f t he various prevent ive st rat egies available, several st udies have show n t hat
hydrat ion bef ore exposure t o radiocont rast mat erial reduces t he risk of
nephropat hy. I nt ravenous hydrat ion appears t o be bet t er t han oral hydrat ion, and
normal saline, if t olerat ed, seems t o be superior t o 0. 45% saline. All pat ient s
should receive hydrat ion bef ore
any st udy t hat requires t he administ rat ion of int ravenous radiocont rast . O ne
recommendat ion is t o use at least 1 cc/ kg/ hr f or 12 hours bef ore and af t er t he
st udy. O ne relat ively small t rial show ed a signif icant benef it w it h use of sodium
bicarbonat e inst ead of saline 1 hour bef ore and f or 6 hours af t er t he
administ rat ion of radiocont rast , but t he result s have not been validat ed in a
larger t rial. I n addit ion, medicat ions such as met f ormin, aminoglycosides,
angiot ensin-convert ing enzyme inhibit ors, f urosemide, or nonst eroidal ant i-
inf lammat ory drugs t hat are likely t o have an added det riment al eff ect on t he
kidneys should be st opped w ell bef ore t he st udy is undert aken. Some
radiocont rast mat erials, such as nonosmolar and nonionic agent s, appear t o be
less nephrot oxic. These alt ernat ives should be considered in high-risk pat ient s.
Many st udies have invest igat ed t he role of N-acet ylcyst eine or sodium
bicarbonat e inf usion in prevent ing radiocont rast -induced nephropat hy. Anot her
approach t hat needs t o be validat ed is t he use of high-dose ascorbic acid (a
t ot al of 7 g) administ ered in t hree doses bef ore and af t er cont rast t o reduce t he
risk of radiocont rast -induced nephropat hy. Therapies t hat are not recommended
f or prevent ing radiocont rast -induced nephropat hy include f enoldopam, calcium-
channel blockers, dopamine, at rial nat riuret ic pept ide, mannit ol, f urosemide, and
t heophylline.
Suggested Readings
Lin J, Bonvent re JV. Prevent ion of radiocont rast nephropat hy. Curr O pin
Nephrol Hypert ens 2005; 14: 105–110.
Weisbord SD, Palevsky PM. Radiocont rast induced acut e renal f ailure. J
I nt ens Care Med 2005; 20: 63–75.
> Table of C ontents > P r egnanc y > 287 - P lac e P r egnant P atients w ith R ight S ide E levated 15 D egr ees
287
Place Pregnant Patients with Right Side Elevated
15 Degrees
G len T inkoff MD
A w oman in t he t hird t rimest er of pregnancy is predisposed t o hypot ension w hile
in t he supine or sit t ing posit ion due t o t he hemodynamic and anat omic changes of
pregnancy. The large ut erus of lat e pregnancy can compress t he inf erior vena
cava (I VC) such t hat venous ret urn is signif icant ly reduced. This decreased
preload can lead t o decreased cardiac f illing and hence decreased cardiac
out put and hypot ension. This can be especially delet erious in t he usual set t ing of
increased cardiac demand in pregnancy.
O f normovolemic pregnant pat ient s, only 8% t o 10% display supine hypot ension
due t o adequat e physiologic compensat ion; how ever, w hen f aced w it h blood or
ot her f luid losses such as in t rauma or crit ical illness, supine hypot ension of lat e
pregnancy is more prevalent . I n t hese inst ances, simple reposit ioning can be lif e
saving. Est ablishing lef t ut erine displacement by elevat ing t he pat ient 's right side
great er t han 15 degrees allow s t he ut erus t o be displaced off t he inf erior vena
cava. Af t er a t raumat ic injury, bef ore t his maneuver is perf ormed assessment of
t he st abilit y of t he pat ient 's spinal cord must be undert aken, and if uncert ainly
exist s, t he pat ient should not be moved w it hout using f ormal spinal precaut ions.
Suggested Readings
Kinsella SM, Lohman G . Supine hypot ensive syndrome. O bst et G ynecol
1994; 83: 774.
Ueland K, Met caf e J. Circulat ory changes in pregnancy. Clin O bst et G ynecol
1975; 18: 41.
> Table of C ontents > P r egnanc y > 288 - Avoid the Us e of D r ugs Har m ful to the Fetus if at all P os s ible
288
Avoid the Use of Drugs Harmful to the Fetus if at
all Possible
Lee Ann Lau MD
Heidi Frankel MD
When t reat ing pregnant pat ient s, it is t he physician's responsibilit y t o t reat t he
mot her as t he primary pat ient . Fet al considerat ions are secondary t o ensuring
mat ernal w ell-being because mat ernal demise w ill generally result in subsequent
f et al deat h. Wit h t his in mind, how ever, w hen opt ions f or t reat ment exist ,
considering f et al eff ect s is prudent .
Many common crit ical care drugs are saf e in pregnancy, w hereas some should
be used w it h caut ion and ot hers avoided ent irely. I t is also essent ial t o
remember t hat f ew drugs are ext ensively t est ed in pregnant humans, so caut ion
should be exercised w it h administ rat ion of any pharmaceut icals in t he pregnant
w oman. Consult ing t he Physi ci ans' Desk Ref erence f or t he current saf et y rat ings
in pregnancy w ill ensure adherence t o t he most current recommendat ions.
What to Do
For analgesia, acet aminophen is generally considered saf e w hen used in doses
saf e f or t he mot her. Narcot ics do not have report ed t erat ogenicit y; how ever,
t hey can cause f et al respirat ory depression if administ ered near t he t ime of
delivery. Aspirin should be avoided due t o risks of int raut erine grow t h ret ardat ion
and prolonged labor and bleeding. Alt hough t here is some cont roversy,
nonst eroidal ant i-inf lammat ory drugs are generally considered t o be saf e in t he
lat er part of pregnancy. Most inhalat ional agent s and neuromuscular blockers are
saf e. Local anest het ics are able t o cross t he placent a but do not have f et al
eff ect s at common doses.
Specif ic ant ibiot ics need t o be assessed on an individual basis (e. g. , quinolones
are t hought t o cause t endon def ect s). Aminoglycosides are know n t o cause f et al
ot ot oxicit y, sulf onamides lead t o neonat al kernict erus, t et racyclines inhibit f et al
bone grow t h, and f luconazole is t erat ogenic. Penicillins, cephalosporins,
vancomycin, and clindamycin are considered saf e.
Hydralazine, alpha-met hyldopa, and nit roglycerine are saf e f or t he t reat ment of
hypert ension. Sodium nit roprusside can cause high cyanide levels in t he f et us.
Angiot ensin-convert ing enzyme inhibit ors cause f et al renal f ailure injury or deat h
and should be avoided during t he second and t hird t rimest ers.
For gast roint est inal prophylaxis, sucralf at e, hist amine-2 blockers, and
pant oprazole are not associat ed w it h know n f et al problems. Heparin and low -
molecular-w eight heparin are saf e, but Coumadin causes f et al w arf arin syndrome
(f acial dysmorphism, heart def ect s, grow t h def ect s). Use of ant iconvulsant s or
sedat ives is associat ed w it h some risk. Benzodiazepines, barbit urat es, and
propof ol cause f et al respirat ory depression. I n addit ion, benzodiazepines can
cause “f loppy baby” syndrome and w it hdraw al syndromes. They are also
quest ionably associat ed w it h clef t palat e w hen used early. Haloperidol,
phenyt oin, and valproic acid are all t erat ogenic.
For hemodynamic inst abilit y, dobut amine is generally considered saf e but is
associat ed w it h decreased placent al f low in some animal t rials. Ephedrine
preserves ut erine blood f low. Phenylephrine is eff ect ive and saf e f or
count eract ing hypot ension due t o spinal or epidural anest hesia.
Treat ment of nausea and vomit ing in t he pregnant pat ient deserves special
ment ion. Vit amin B6 and ginger have been proven t o be saf e and eff ect ive;
vit amin B6 may reduce t he risk of congenit al heart def ect s. Ant ihist amines (H1
blockers) are considered saf e and generally eff ect ive. Dopamine-recept or
ant agonist s (met oclopramide, droperidol, phenot hiazines) have not been st udied
ext ensively, but limit ed dat a suggest saf et y and eff icacy. O ndanset ron has not
show n t oxicit y in animal st udies but saf et y and eff icacy st udies have not been
done f or human pregnancy.
Suggested Readings
Davis M. Nausea and vomit ing of pregnancy: an evidence-based review. J
Perinat Neonat Nurs 2004; 18: 312–328.
Hall JB, Schmidt G A, Wood LDH, ed. Principles of Crit ical Care, 3rd ed. New
York: McG raw -Hill, 2005: 1593–1614.
I rw in R, Rippe J, eds. Manual of I nt ensive Care Medicine, 5t h ed.

Philadelphia: Lippincot t Williams & Wilkins, 2003: 1751–1758.
> Table of C ontents > P r egnanc y > 289 - S tr ongly C ons ider the Us e of an E lec tr onic Fetal Monitor in
C ar ing for a P atient at Gr eater than 24 W eeks ' Ges tation in the Intens ive C ar e Unit
289
Strongly Consider the Use of an Electronic Fetal
Monitor in Caring for a Patient at Greater than
24 Weeks' Gestation in the Intensive Care Unit
I t is not uncommon f or a pregnant pat ient t o be aff ect ed by crit ical illness.
Examples of nonobst et ric maladies t hat lead t o int ensive care unit admission
among pregnant w omen are hemat ologic issues (venous t hromboembolism and
pulmonary embolus, pre-pregnancy hypercoagulable st at es), t rauma (including
mot or vehicle accident s, f alls, assault s, burns, et c. ), ast hma, valvular and
congenit al heart disease, and acut e abdominal condit ions (appendicit is, rupt ured
viscus associat ed w it h pept ic ulcer disease, and inf lammat ory bow el disease).
Watch Out For

Af t er approximat ely 24 w eeks of gest at ion, most aut horit ies consider t he f et us
viable, so f or many pregnant pat ient s in t he int ensive care unit (I CU) care must
also be direct ed t o t he f et us. Alt hough mat ernal lif e should never be jeopardized
f or t he care of t his second pat ient , t he crit ical care physician should be able t o
recognize signs of f et al dist ress and collaborat e w it h obst et ric and neonat ology
colleagues t o det ermine opt imal t herapy f or bot h pat ient s. Fet al dist ress can be
caused by a host of f act ors; most not ably, hypoxemia and hypovolemia are
dangerous f or t he f et us. Fet al t achycardia, lat e f et al heart rat e decelerat ions,
and loss of heart rat e variabilit y are signs of signif icant f et al dist ress and call
f or immediat e act ion. These are best ident if ied w it h t he use of t he cont inuous
elect ronic f et al monit or.
Alt hough t here is a lack of high-qualit y evidence, it is probably rat ional t o use
elect ronic f et al monit oring f or any pat ient s w it h est imat ed gest at ional age of 24
w eeks or great er. The elect ronic f et al monit or consist s of t w o probes placed on
t he mot her's abdomen. O ne probe uses ult rasound (Doppler) t o t ransmit and
record t he f et al heart rat e. The second sensor is a pressure-sensit ive device
t hat records f requency and durat ion of ut erine cont ract ions. Fet al monit oring
may demonst rat e f et al dist ress, but changes in f et al heart rat e are nonspecif ic
and are of t en a lat e sign of inadequat e f et al oxygen delivery. I t is most accurat e
w hen perf ormed at 32 w eeks of gest at ion or lat er. The most specif ic indicat or of
f et al dist ress is probably f et al pH
monit oring, w hich is perf ormed by scalp blood sampling. How ever, t o access t he
f et al scalp, cervical dilat ion must have begun and membranes must be rupt ured,
so t his in not pract ical f or many of t he gravid pat ient s in t he I CU.
What to Do
I f f et al dist ress or premat ure labor occurs, t he f irst call should be t o t he
obst et ric service. Their expert ise in assessment of gest at ional age, elect ronic
f et al monit oring, maint enance of ut erine blood f low, medicat ion use, and t he
decision t o perf orm emergency cesarean sect ion are invaluable t o t he mot her
and child. O xygen delivery t o t he f et us should be opt imized by providing
supplement al oxygen and rest oring adequat e circulat ing volume and cardiac
f unct ion t o t he mot her. Prevent ion of compression of t he inf erior vena cava by
t he gravid ut erus is accomplished w it h placement of t he mot her in t he lef t lat eral
decubit us posit ion. I f premat ure labor has commenced, t ocolysis may be
indicat ed t o halt labor or delay delivery. Mat ernal st eroid administ rat ion has
been demonst rat ed t o promot e f et al lung mat urit y. How ever, delivery of t he f et us
may be t he best opt ion f or t he mot her and t he f et us.
Suggested Readings
Hall JB, Schmidt G A, Wood LDH, eds. Principles of Crit ical Care, 3rd ed. New
York: McG raw -Hill, 2005: 1593â 1 614.
> Table of C ontents > P r egnanc y > 290 - K now the Nor m al P hys iologic C hanges and As s oc iated
Labor ator y Values that O c c ur in P r egnanc y
290
Know the Normal Physiologic Changes and
Associated Laboratory Values that Occur in
Pregnancy
G len T inkoff MD
The physiologic adapt at ions of pregnancy occur in response t o f et al development
and event ual delivery. I t is incumbent on t he int ensive care physician t o have an
underst anding of t he normal variant s t hat occur in pregnancy t o correct ly
int erpret diagnost ic t est s and eff ect ively t reat clinical condit ions. Some changes
t hat occur in pregnancy include t he f ollow ing.
Endocrine
By f ar t he most signif icant physiologic changes of pregnancy are endocrinologic.
Signif icant increases in product ion and serum levels of grow t h hormone,
prolact in, t hyroxin, parat hyroid hormone, calcit onin, and progest erone occur in
support of f et al development and part urit ion.
I n addit ion, pregnancy is associat ed w it h posit ive nit rogen balance and
post prandial hyperglycemia and hyperinsulinemia. This lat t er response is
consist ent w it h a pregnancy-induced st at e of peripheral resist ance t o insulin, t he
purpose of w hich is t o ensure sust ained supply of glucose t o t he f et us.
Furt hermore, t he concent rat ions of plasma lipids, lipoprot eins, and
lipopolysaccharides increase appreciably during pregnancy and lead t o cent ral
f at deposit ion t o supply f et al demands.
Cardiovascular
Blood volume at or near t erm is approximat ely 40% above t hat in nonpregnancy.
This hypervolemia of pregnancy meet s t he demands of pregnancy and prot ect s
t he mot her against impaired venous ret urn and t he blood loss of child birt h.
Along w it h t he increase in blood volume, lef t vent ricular size also increases w it h
increased cont ract ilit y. Cardiac out put increases as early as t he f if t h w eek of
pregnancy w it h rest ing pulse rat e increasing by approximat ely 10 beat s per
minut e. Peripheral vascular resist ance decreases. These cardiovascular changes
are similar t o changes occurring w it h moderat e exercise.
Hematologic
Alt hough blood volume expansion occurs and result s in increases in bot h plasma
volume and eryt hrocyt e number, more plasma is made t han red cells. Thus,
hemoglobin concent rat ion and hemat ocrit decrease (at t erm hemoglobin is
approximat ely 12. 5 g/ dL). Furt hermore, w it h t he increase in red cell mass, iron
requirement s are increased in pregnancy such t hat iron supplement at ion is
necessary.
Despit e t he f act t hat pregnancy is associat ed w it h immunologic suppression t o
accommodat e t he semi-allogeneic f et us, t he leukocyt e count becomes elevat ed
during lat e pregnancy, averaging 14, 000 t o 16, 000. I n addit ion, t he coagulat ion
cascade is also act ivat ed during pregnancy, w it h increases in most clot t ing
f act ors and f ibrinogen.
Electrolyte
Pregnancy is marked by hypermet abolism and salt and w at er ret ent ion. These
eff ect s are mediat ed predominat ely t hrough t he act ions of t hyroxin and
vasopressin. Wat er ret ent ion induced by increased vasopressin secret ion is such
t hat t he average w oman accrues an ext ra 6. 5L during normal pregnancy. Sodium
and pot assium are ret ained due t o enhanced renal t ubular resorpt ion; how ever,
concent rat ions of t hese elect rolyt es are slight ly decreased as a result of t he
expanded plasma volume. I n addit ion, serum magnesium levels decline during
pregnancy such t hat pregnancy causes a st at e of cont inuous ext racellular
magnesium deplet ion.
Hepatic
O t her laborat ory abnormalit ies t hat occur during pregnancy include an increase
in t ot al alkaline phosphat ase, w it h serum levels almost doubling during normal
pregnancy. Serum t ransaminases and bilirubin levels are usually slight ly low er in
pregnancy as compared w it h nonpregnant normal values. The concent rat ion of
serum albumin decreases during pregnancy, and t heref ore t ot al calcium also
decreases.
Suggested Readings
Cunningham FG , Loveno K, Bloom S, et al. , eds. Williams O bst et rics, 22nd
ed. New York: McG raw -Hill, 2005: 121–149.
Piet zman AB, Rhodes M, Schw ab CW, et al. , eds. The Trauma Manual, 2nd
ed. Philadelphia: Lippincot t Williams & Wilkins, 2002: 461–468.
> Table of C ontents > B ur ns > 291 - D o not Adm inis ter P r ophylac tic Antim ic r obials to B ur n P atients
291
Do not Administer Prophylactic Antimicrobials
to Burn Patients
G ary T. Marshall MD
Despit e t he great st rides made over t he last decades in improving survival af t er
t hermal injury, inf ect ious complicat ions remain an import ant cause of morbidit y
and mort alit y. Nearly 50% of all burn deat hs are relat ed t o inf ect ion. Several
f act ors cont ribut e t o t he high incidence and severit y of inf ect ion in burn pat ient s.
First , a prof ound immune suppression is induced t hat is proport ional t o t he
ext ent of burn injury. I n addit ion, t he environment at t he sit e of t he w ound is
f avorable f or t he mult iplicat ion of inf ect ing organisms, w hich colonize t he w ound
w it hin 72 hours of injury. Despit e t he t empt at ion t o administ er syst emic
ant ibiot ics, t here is no rol e f or prophyl acti c systemi c anti mi crobi al s. This has no
i mpact on survi val and l eads to the emergence of resi stant organi sms. Frequent
exposure t o ant ibiot ics also est ablishes an environment f avorable f or yeast and
f ungi.
Wit hin 3 days of injury, G ram-posit ive organisms colonize t he w ound. G ram-
negat ive organisms t hen colonize t he eschar, and t hey become dominant by t he
end of t he f irst w eek. The t ypical f lora of burn w ounds has changed over t he
years. Bef ore t he advent of penicillin, β -hemolyt ic st rept ococci w ere t he most
f requent cause of burn w ound and lif e-t hreat ening syst emic inf ect ions. Penicillin
largely eliminat ed mort alit y due t o st rept ococci, and Staphyl ococcus aureus
became t he more common early colonizer. As broad-spect rum ant ibiot ics
eff ect ive against Staphyl ococcus w ere developed, gram-negat ive organisms,
especially Pseudomonas aerugi nosa, became t he predominant organisms. These
organisms have great er invasive pot ent ial due t o t heir product ion of t oxins and
prot eolyt ic enzymes. The use of eff ect ive t opical agent s has reduced t he
incidence of deat h secondary t o burn w ound sepsis, but f ungi have emerged as
t he most common cause of invasive burn w ound inf ect ions.
The t hree most commonly used t opical agent s used t o f ight w ound inf ect ions are
silver sulf adiazine (SSD), maf enide acet at e, and silver nit rat e. SSD is t he most
f requent ly used. I t has broad-spect rum act ivit y as w ell as ant if ungal propert ies.
SSD is limit ed by it s inabilit y t o penet rat e t he burn eschar. Maf enide acet at e has
good eschar penet rat ion and can suppress subeschar prolif erat ion of bact eria.
Maf enide acet at e is a carbonic anhydrase inhibit or t hat int erf eres w it h t he renal
buff ering. A hyperchloremic met abolic acidosis develops due t o t he consumpt ion
of bicarbonat e and chloride ret ent ion. A compensat ory respirat ory alkalosis is
generat ed by hypervent ilat ion but is rarely of clinical signif icance. Silver nit rat e
is eff ect ive if applied bef ore bact erial penet rat ion int o t he w ound. Side eff ect s of
silver nit rat e include hyponat remia and met hemoglobinemia.
Watch Out For

Despit e t hese maneuvers, invasive inf ect ions st ill occur. Risk f act ors include
severe injury (>30% t ot al body surf ace area burn), syst emic diseases such as
diabet es w it h poor glycemic cont rol, and delayed excision and graf t ing of t he
burn. The key t o successf ul management of inf ect ion in t he burn pat ient is early
diagnosis. This is best accomplished by daily scheduled monit oring of t he burn t o
ident if y inf ect ion at it s early st ages w hen surgical and ant ibiot ic t herapy w ill be
most eff ect ive. Burn w ound inf ect ions are classif ied as cellulit is, invasive
inf ect ions, or burn w ound impet igo. Cellulit is is diagnosed clinically based on
charact erist ic eryt hema, edema, and hyperest hesia of unburned skin at t he
margin of t he w ound. Cellulit is unrelat ed t o inf ect ion may also occur, usually
peaking on t he second or t hird day af t er t he burn and t hen receding; t his is
rarely accompanied by syst emic response. Expanding cellulit is should be t reat ed
w it h t he applicat ion of maf enide acet at e and syst emic penicillin if a β -hemolyt ic
st rept ococcus is involved or a broad-spect rum β -lact am if specif ic cult ure and
sensit ivit y are not available. A local sign of invasive inf ect ion is t he appearance
of dark brow n, black, or violaceous discolorat ion of t he w ound. O t her clinical
signs are t he conversion of a part ial-t hickness injury t o f ull t hickness and
necrosis of previously viable t issue in t he w ound bed. Pseudomonas may cause
changes in t he unburned skin at t he w ound margin. I nf ect ions caused by f ungi
display unexpect ed rapid eschar separat ion and rapid cent rif ugal spread of
subcut aneous edema w it h cent ral necrosis. Vesicular lesions appearing in healing
or healed burns on t he f ace are charact erist ic of herpes simplex virus t ype 1.
Signs And Symptoms

O nce a burn w ound inf ect ion is suspect ed, diagnosis is conf irmed by burn w ound
biopsy. Surf ace cult ures only ident if y t he colonizing organisms present and are
t heref ore inadequat e. Q uant it at ive bact erial count s of 105 or great er are
suggest ive of invasion but only correlat e w it h invasive inf ect ion on hist ologic
exam in f ew er t han 50% of paired samples. Hist ologic examinat ion of a burn
w ound biopsy is t he most reliable conf irmat ory t est f or invasive burn
w ound inf ect ion. Wit h a scalpel, a lens-shaped sample of t he eschar and
unburned underlying/ adjacent t issue are obt ained f rom t he area of t he w ound
show ing t he most pronounced changes. O n hist ologic exam, t he presence of
microorganisms in unburned t issue is diagnost ic. O t her f indings t hat indicat e burn
w ound inf ect ion are t he presence of hemorrhage, small-vessel t hrombosis, and
necrosis in unburned t issue. Af t er diagnosis, general support ive t reat ment is
used, and specif ic syst emic ant ibiot ic t herapy is init iat ed based on t he
inst it ut ion's usual f lora w it h subsequent modif icat ion as sensit ivit ies ret urn.
Maf enide acet at e should be applied t o t he w ound t w ice daily. Subeschar clysis
of a broad-spect rum penicillin may also be employed. Surgical excision should be
perf ormed as soon as possible. Topical clot rimazole should be applied w hen
invasive f ungi are t he culprit microorganisms. Syst emic t reat ment w it h
amphot ericin B or ot her, new er ant if ungal agent s is also used w hen t here is
evidence of syst emic f ungemia.
Burn impet igo occurs af t er burn healing or graf t ing and is charact erized by
mult if ocal small superf icial abscesses. S. aureus is t he responsible microbe.
Treat ment includes unroof ing of abscesses and f requent cleansing f ollow ed by
applicat ion of t opical ant ibiot ics. O ccasionally, syst emic sympt oms may be seen
w hen t he t oxic shock syndrome t oxin is produced. I n t his case int ravenous
vancomycin is used.
I n addit ion t o t he previously described clinical sit uat ions, appropriat e uses of
ant ibiot ics in t he burn pat ient include t arget ed t reat ment of pulmonary, urinary,
and cat het er-relat ed inf ect ions. I n addit ion, perioperat ive prophylact ic ant ibiot ics
should be administ ered just bef ore w ound excision. These are given based on
t he f indings of t ransient bact eremia in 21% of pat ient s f ollow ing w ound
manipulat ion, alt hough some st udies have show n t hat bact eremia is rare w hen
t here is less t han 40% t ot al body surf ace area burn.
Suggested Readings
Edw ards-Jones V, G reenw ood JE. What 's new in burn microbiology? Burns
2003; 29: 15–24.
Pruit t BA Jr, McManus AT, Kim SH. Burn w ound inf ect ion: current st at us.
World J Surg 1998; 22: 135–145.
> Table of C ontents > B ur ns > 292 - R em em ber that not All Fever in the B ur n P atient is D ue to
Infec tion
292
Remember that not All Fever in the Burn Patient
is Due to Infection
Virt ually all burn pat ient s have elevat ed core body t emperat ures and even a
leukocyt osis. Thus, f ever in burn pat ient s is not a reliable indicat or of inf ect ion.
O ne st udy in children f ound t hat f ever had no predict ive value f or t he presence of
inf ect ion and physical examinat ion w as a more reliable source of inf ormat ion
about w ound inf ect ion and sepsis. Furt hermore, rout ine blood cult ures have been
show n t o be of lit t le value in w orking up a f ever in burn pat ient s.
Watch Out For

Mild hypert hermia in t he f irst 24 hours f ollow ing injury is almost alw ays t he result
of pyrogen release. The t hree endogenous pyrogens t hat mediat e f ever are
int erleukin-1 (I L-1), t umor necrosis f act or-α (TNF-α , cachet in), and int erf eron-γ.
I n addit ion t o inducing f ever, t hese pyrogens also modulat e a large number of
host def ense responses. Af t er 72 hours, burn pat ient s develop syst emic
inf lammat ory response syndrome (SI RS), charact erized by t achycardia, relat ive
hypot ension, and hypert hermia, w hich are t he classic signs of sepsis t hat can
occur w it hout an inf ect ious source. I n f act , body t emperat ure may be as high as
39°C (102. 2°F) and t he leukocyt e count may be as high as 20, 000 cells/ mL
during sat isf act ory recovery. Thus, elevat ion of body t emperat ure above normal,
leukocyt osis, and ot her signs of inf lammat ion are common among burn pat ient s
and should be expect ed.
The hypermet abolic response elicit ed by a burn injury is marked by increased
energy expendit ure and muscle prot ein cat abolism. Fever f urt her increases
energy expendit ure and loss in burn pat ient s. Thus, t here is a t heoret ical
met abolic benef it in at t enuat ing f ever in such pat ient s. Mild elevat ion of body
t emperat ure is usually w ell t olerat ed and does not require specif ic t reat ment .
How ever, body t emperat ure higher t han 39C may be t reat ed w it h common
ant ipyret ics (e. g. , acet aminophen, aspirin, indomet hacin, or ibuprof en). Burn
pat ient s also pref er w armer ambient t emperat ure (26°C t o 33°C) t han do normal
individuals because t he f ever experienced by t he burn pat ient is an upw ard
adjust ment of t he t hermoregulat ory cent er in t he brain. This t emperat ure range
promot es pat ient comf ort and reduces t he physiologic demands of cold st ress.
Treat ing burn pat ient s by surf ace
cooling (ice packs, alcohol rubs, cooling blanket s) is not uncommon but should
be avoided because surf ace cooling creat es a cold st ress on an already crit ically
ill pat ient .
Nonet heless, inf ect ion is a relat ively common serious complicat ion in pat ient s
w it h major t hermal injuries. I t is est imat ed t hat up t o 75% of deat hs f ollow ing
burn injury are relat ed t o inf ect ion. I mmunosuppression is a universal f eat ure of
major t hermal injury and burn pat ient s are more suscept ible t o microbial
colonizat ion and inf ect ion. Special at t ent ion t o t echnique during bedside
procedures and t he use of universal precaut ions are a must . I n one st udy, up t o
100% of burn pat ient s developed an inf ect ion f rom one or more sources during
t heir hospit al st ay. O ne might assume t hat prophylact ic ant ibiot ics, part icularly
against gram-posit ive organisms, are indicat ed; how ever, t his pract ice has been
show n t o lead t o t he development of mult idrug-resist ant , gram-negat ive bact erial
inf ect ions and f ungal inf ect ions. I n f act , st udies have now verif ied t hat
prophylact ic ant ibiot ics are not only unnecessary, but are also cont raindicat ed in
pat ient s w it h t hermal injuries. Except ions t o t his st at ement are t et anus
prophylaxis and perioperat ive ant ibiot ics in pat ient s w ho undergo excision and
graf t ing. O t herw ise, daily w ound care and t he applicat ion of a t opical, broad-
spect rum ant imicrobial agent are t he mainst ays f or t he init ial prevent ion of
w ound inf ect ion. Ult imat ely, excision and graf t ing is t he most eff icacious means
of reducing t he incidence of burn w ound sepsis.
Suggested Readings
Bessey PQ . Met abolic response t o crit ical illness. I n: ACS Surgery: Principles
and Pract ice. Chicago: American College of Surgeons, 2006.
w w w. acssurgery. com
G ibran NS, Heimbach DM. Management of t he pat ient w it h t hermal injuries.

I n: ACS Surgery: Principles and Pract ice. Chicago: American College of
Surgeons, 2006. w w w. acssurgery. com
Parish RA, Novack AH, Heimbach, et al. Fever as a predict or of inf ect ion in
burned children. J Trauma 1987; 27(1): 69–71.
> Table of C ontents > B ur ns > 293 - Never Under es tim ate the S ever ity of an E lec tr ic al B ur n
293
Never Underestimate the Severity of an
Electrical Burn
Travis L. Perry MD
Elect rical injuries cont inue t o be clinically and surgically challenging f or surgeons
and crit ical care physicians w orldw ide. G ross underest imat ion of t he init ial injury
has repeat edly proven t o increase morbidit y and det riment al t o t he overall
out come.
To brief ly review, elect ricit y is t he f low of elect rons t hrough a conduct or via t he
f orce of volt age. Volt age is cat egorized int o low (<1, 000 volt s) and high (≥1, 000
volt s). Alt ernat ing current (AC) and direct current (DC) are t he t w o f orms of
elect rical energy. Alt ernat ing current produces cyclic back and f ort h movement of
elect rons and is t he common current in most households. Direct current is t he
f low of energy in one direct ion.
Severit y of elect rical injury is a f unct ion of t hree f act ors: (1) current source or
volt age, (2) durat ion of cont act , and (3) pat hw ay of current f low. The clinical
int erplay of t hese f act ors can best be underst ood by O hm's law (I = V/ R), w hich
st at es t hat current (I) is direct ly proport ional t o volt age (V) and inversely
proport ional t o t he resist ance (R) of t he conduct or. Theref ore, current w ill f ollow
t he pat h of least resist ance. Hist ologically, t issues w it h high f luid and elect rolyt e
cont ent s (i. e. , nerves, vessels, muscle, and mucosal membranes) have low er
resist ance. Theref ore, t he pat hw ay of elect rical current has a higher aff init y f or
t hese t issues and w ill pref erent ially damage t hese sit es. Tissues such as f at ,
bone, t endon, and dry skin have low er f luid and elect rolyt e cont ent and t heref ore
have higher resist ance. These t issues do generat e an int ense amount of t hermal
injury, how ever, due t o t heir poor abilit y t o conduct elect rical current . Because of
t his, t he init ial assessment of t he severit y of an elect rical injury t ypically
underest imat es t he ext ent of sof t t issue injury.
I t is import ant t o not e t hat elect rical burns w it h alt ernat ing current , especially
low -volt age current , have t he propensit y t o produce muscle t et any due t o it s
cyclic f low of energy. This, in t urn, prolongs t he durat ion of cont act , w hich
increases t he pot ent ial of injury developing, especially f rom current f low t hrough
t issues w it h high resist ance. This impeded f low of current can generat e a
t remendous amount of t hermal injury t hat is of t en hidden in t he f orm of
coagulat ed necrosis of underlying subcut aneous f at , bone, and t endon.
High-volt age exposure t ypically produces a single violent muscle cont ract ion
leading t o a short er durat ion of cont act due t o eject ion aw ay f rom t he primary
source of current . This eject ion increases t he risk of f ract ures, dislocat ions, loss
of consciousness, and closed-head injuries. High-volt age cont act also generat es
severe skin burns due t o arcing and f lashing of t he elect rical current . As alw ays,
t he pat h of current f low det ermines organ syst em involvement and overall injury
severit y. A f low of current parallel t o t he body's vert ical axis increases t he
pot ent ial of mult isyst em injury. A horizont al f low of current is associat ed w it h
less organ syst em involvement (e. g. , t he t ransf er f rom hand t o hand).
Light ning is a f orm of ext remely high-volt age direct current , and it s st rikes are
f ort unat ely rare occurrences. Light ning can st rike direct ly or indirect ly via ground
or inanimat e object t ransf er. I t also has t he abilit y t o f low along t he body's
surf ace w it hout direct ly ent ering t he body. This f lashing phenomenon produces a
pat hognomonic f ern-like pat t ern (Fig. 293. 1) over t he body t hat resolves in about
24 hours.
Elect rical injuries should be view ed and assessed f rom a mult isyst em
perspect ive. I nit ial evaluat ion should proceed as in any major t rauma, w it h
primary and secondary surveys. Special at t ent ion should be given t o high-volt age
injury complicat ions (e. g. , ext remit y or digit amput at ion, muscle necrosis,
compart ment syndrome, and sepsis). Alt hough mult iple organ syst ems may be
involved, it is damage t o t he major syst ems t hat may require int ensive care unit
evaluat ion and t reat ment .
Cardiovascular
The cardiovascular syst em may be aff ect ed via cellular necrosis or, more
commonly, dysrhyt hmias. Myocardial injury could involve direct necrosis of
myocyt es, pacing nodes, and/ or coronary vessels. Fat al dysrhyt hmias include
vent ricular f ibrillat ion and asyst ole result ing f rom cell membrane inst abilit y,
w hereas a myriad of less dangerous dysrhyt hmias can be precipit at ed. Large
blood vessels are suscept ible t o rupt ure or aneurysmal f ormat ion. Smaller
vessels are more vulnerable t o coagulat ion necrosis, w hich may lead t o a
compart ment syndrome in t he peripheral circulat ion. The cardiovascular syst em
should be monit ored via cont inuous elect rocardiogram (ECG ) if any dysrhyt hmias
are det ect ed on an init ial 12-lead ECG . Cardiac enzymes should be checked and
f ollow ed as clinically indicat ed. Physical examinat ion and measurement of
compart ment pressures as needed w ill det ect pot ent ial peripheral vascular
complicat ions.
FIG URE 293. 1. Light ning injury.
Skin
Cut aneous elect rical injury varies depending on volt age, cont act durat ion, and
resist ance of t he involved skin. Normal adult skin has a high resist ance and
t heref ore w ill generat e underlying coagulat ion necrosis if prolong cont act
ensues. How ever, young children w ho nat urally have higher w at er cont ent in t heir
skin and adult s exposed t o w at er suff er more superf icial cut aneous injury due t o
t he low ered resist ance in t heir skin. These cut aneous burns should be
approached as any burns w ould be. Caref ul resuscit at ion should be given t o
t hose burns t hat may involve subcut aneous f at , t endon, or bone necrosis. Do not
under-resuscit at e. How ever, t he init ial assessment of t he ext ent and severit y of
an elect rical burn is quit e diff icult and usually underest imat es t he amount of sof t
t issue involved. Thus, mul ti pl e exci si ons to establ i sh a heal thy wound bed f or
graf ti ng are the rul e rather than the excepti on when exci si ng an el ectri cal burn.
Perioral burns f rom an elect rical cord are t he most common cut aneous elect rical
burns in children. These should primarily be monit ored f or labial art ery bleeding.
Special at t ent ion t o pot ent ial f eeding diff icult ies is of paramount import ance in
inf ant s and w arrant s hospit al admission unt il a regular f eeding pat t ern is re-
est ablished.
Pulmonary
Respirat ory arrest is relat ed t o direct injury of t he brain st em respirat ory cent er
or diaphragmat ic muscle t et any impeding respirat ory eff ort . Typically, respirat ory
arrest f rom elect rical injuries manif est s at t he scene. This is monit ored f or and
t reat ed sympt omat ically via mechanical vent ilat ion.
Other
Renal dysf unct ion can occur direct ly or indirect ly via exposure t o hypoxemia,
myoglobin, and/ or creat inine phosphokinase. Serial urine myoglobin and
creat inine phosphokinase should be measured as clinically indicat ed. Adequat e
resuscit at ion and hydrat ion should be provided. Renal f ailure should be t reat ed
as indicat ed. Volume resuscit at ion should be based on a urinary out put of 30 cc
per hour in an adult w it hout myoglobinuria. A minimal urine out put of 100 mL/ h is
t he goal in a pat ient w it h myoglobinuria. Cent ral and peripheral nervous syst em
injuries should be evaluat ed and recognized during t he init ial t rauma assessment .
O pht halmologic examinat ion should be perf ormed on all pat ient s suff ering high-
volt age injuries t o assess f or premorbid cat aract s because cat aract development
is a delayed complicat ion of high-volt age injuries.
Selected Readings
Arnold B. Elect rical injuries: A 20-year review. J Burn Care Rehabil
2004; 25(6): 479–484.
Herndon D, Muehlberger T, eds. Tot al Burn Care, 2nd ed. Philadelphia: WB

Saunders, 2002: 455–469.
Koumbourlis, AC. Elect rical injuries: scient if ic review s. Crit Care Med 2002;
30(11): 424–430.
> Table of C ontents > B ur ns > 294 - D o not Us e Hyper bar ic O xygen Ther apy in B ur ns
294
Do not Use Hyperbaric Oxygen Therapy in Burns
Travis L. Perry MD
Hyperbaric oxygen t herapy is current ly used in various clinical t reat ment
regimens. These include decompression sickness, carbon monoxide poisoning,
cyanide poisoning, gas embolus, gas gangrene, resist ant anaerobic inf ect ions,
and t hreat ened split -t hickness skin graf t s. The mechanism of act ion purport edly
involves increasing t issue oxygenat ion, w hich increases collagen and f ibroblast
f ormat ion and suppresses Cl ostri di a t oxin product ion. I t also enhances t he killing
abilit y of t he leukocyt e and capillary prolif erat ion. This mechanism has primarily
been est ablished in animal models and limit ed human t rials. At best , hyperbaric
oxygen in t he management of sof t t issue inf ect ions can be promot ed only as part
of a coordinat ed medical and surgical approach.
I t has been post ulat ed t hat hyperbaric oxygen in burn w ounds pot ent ially exhibit s
some benef it by st imulat ing vasoconst rict ion and count eract ing hypoxia. This
t heory is based on t he possibilit y t hat hyperbaric oxygen may decrease acut e
edema, f luid requirement s, and inf ect ion rat es and promot e re-epit hializat ion.
There have been several animal models t hat demonst rat ed varied result s.
How ever, t he use and eff icacy of hyperbaric oxygen in burn w ound t herapy is not
est ablished. Recent st udies have not been able t o demonst rat e a st at ist ically
signif icant diff erence in t he lengt h of hospit al st ay, t he number of operat ions
required, and morbidit y and/ or mort alit y w hen comparing hyperbaric oxygen
supplement at ion versus st andard burn t herapy. Thus, hyperbaric oxygen has no
place in t he acut e management of t hermal injuries.
Carbon monoxide (CO ) poisoning should be suspect ed in virt ually every f ire
vict im. I t remains one of t he most f requent causes of deat h in smoke inhalat ion
injuries. Hyperbaric oxygen has demonst rat ed value in t he t reat ment of isolat ed
CO poisoning. I t decreases t he high aff init y t hat CO has f or hemoglobin.
Hyperbaric oxygen decreases t he half -lif e of CO f rom about 4 hours t o about 25
minut es, w hich leads t o quicker rest orat ion of oxygenat ion. Thus, hyperbaric
oxygen t herapy may have a place in t he care f or burn pat ient s w it h CO
poisoning. How ever, hyperbaric oxygen t herapy in burn pat ient s w it h CO
poisoning is of unknow n t herapeut ic value or eff icacy and
should only be used w hen (1) t he CO bound t o hemoglobin is great er t han 25%,
(2) a neurologic def icit exist s, (3) no f ormal burn resuscit at ion is required
(t ypically, t ot al body surf ace area burned is <10% t o 15%), (4) pulmonary
f unct ion is st able w it h an int act airw ay, and (5) int erf acilit y t ransf er does not
compromise burn care.
Selected Readings
Brannen LA, Law EA. Randomized prospect ive t rial of hyperbaric oxygen in a
ref erral burn cent er populat ion. Am Surg 1997; 63: 205–208.
Herndon D, Traber D, eds. Tot al Burn Care, 2nd ed. Philadelphia: WB

Saunders, 2002: 221–231.
Wiseman D, G rossman RA. Hyperbaric oxygen in t he t reat ment of burns. Crit

Care Clin 1985; 1(1): 129–143.
> Table of C ontents > B ur ns > 295 - D o not Us e P ar enter al Nutr ition, if at all P os s ible, in B ur n P atients
295
Do not Use Parenteral Nutrition, if at all
Possible, in Burn Patients
Jeremy W. Pyle MD
Years of w ork and hundreds of st udies have been perf ormed t o elucidat e t he
best and most benef icial means of supplying nut rit ion t o burn pat ient s. Focus in
t he recent past has shif t ed f rom t he overall import ance of nut rit ion t o a highly
specialized and improved collect ion of recommendat ions. These include not only
w hat and how much, but also w here and w hen, w hy and how.
Burn w ound pat hophysiology and how it aff ect s t he body as a w hole are f airly
w ell underst ood, and t he degree t o w hich t hermal injury alt ers nut rit ion
requirement s is direct ly relat ed t o t he t ot al body surf ace area (TBSA) of t he
burn. Burn w ound derangement s include grossly elevat ed local met abolic needs
and increased syst emic requirement s. The bat t le t o provide adequat e nut rit ion in
t he f ace of t his alt erat ion has t hree f ront s. First , a burn w ound requires ramped-
up repair eff ort s and immune support . Second, t he body's response t o t he insult
includes signif icant syst emic derangement s in neuroendocrine and cyt okine-
mediat ed processes. Finally, t he gast roint est inal t ract and it s response t o t hese
changes are unique and import ant . Designing a nut rit ion regimen must address
all t hree of t hese areas, paying special at t ent ion t o t he carbohydrat e and prot ein
requirement s needed t o avoid muscle and mucosal cat abolism and avoiding a
heavy reliance on lipids as a source of calories.
What not to Do
Post burn hypermet abolism has been show n t o respond t o a low -f at , high-prot ein
ent eral diet w it h f ew er w ound complicat ions and a low er overall morbidit y.
Conversely, parenteral nutri ti on has consi stentl y been l i nked to i ncreased
mortal i ty i n the burn pati ent. Burn w ounds almost exclusively ut ilize glucose.
This, coupled w it h t he increased overall glucose requirement s t hat come w it h
t hermal injury, means t hat glucose supply must be eit her increased in t he diet or
come f rom t he cat abolism of ot her t issues. For t his reason, carbohydrat es
should serve as t he major nonprot ein source of nut rit ion. Doing so ensures not
only t hat t he w ound w ill be bet t er supplied w it h energy, but also t hat t he rest of
t he body operat es w it hout quit e so severe a demand f or
gluconeogenic subst rat es, and so prot ein st ores can be preserved. I f a pat ient is
poorly f ed, t he end result of t his demand in such a pat ient has an energy def icit
t hat has been show n t o develop very early post burn. Hence, early and
aggressive f eedings are recommended. They have been show n t o posit ively
aff ect prot ein requirement s, energy def iciency, and insulin secret ion. Because
hyperglycemia and prot ein malnut rit ion are w ell know n f or t heir det riment al
eff ect s on w ound healing, it seems reasonable t o assume t hat increased insulin
and a decreased need f or endogenous gluconeogenic subst rat es provide a
benef it . I deal energy balance is met w it h 5 mg/ kg/ min of glucose.
The syst emic eff ect s of a burn w ound, as mediat ed by cyt okine release and
neuroendocrine f lux, are considerably more complex t han t he increased
met abolism of a local w ound. The alt ered int eract ions of glucagon and insulin
aff ect blood glucose levels and prot ein met abolism w hile producing t he
hyperglycemia of t en seen in burns. Combining glucose and prot ein in ent eral
f eeding provides a synergist ic response t hat provides bet t er nit rogen balance
t han eit her one alone. The eff ect s of glucose are seen in decreased cat abolism,
and t he eff ect s of prot ein provide f or t he maint enance of amino acid st ores. The
import ance of preburn prot ein levels has been delineat ed in st udies looking at
elderly pat ient s w ho are prot ein def icient bef ore burns, w ho have more
complicat ions post burn t han mat ched, w ell-nourished pat ient s. Current nut rit ion
recommendat ions include 7 g/ kg/ day of glucose and about 2 g/ kg/ day of prot ein.
O nce t he basic component s of nut rit ion are det ermined, at t ent ion is t urned t o t he
rout e and t iming of administ rat ion. Research indicat es t hat provision of ent eral
nut rit ion very early in t he st ay of a severely burned pat ient provides a benef it
w it hout sacrif icing saf et y. Perhaps no organ benef it s more f rom early ent eral
nut rit ion t han t he gut . Cell number, t issue mass, and t he immunologic propert ies
of t he gast roint est inal t ract are all improved by early ent eral f eeding, eit her
int ragast ric or post pyloric. Manif est at ions of t his benef it come in t he f orm of
f ew er gast roint est inal bleeds and a decline in t he incidence of t he classic Curling
ulcer. I n f act , int ragast ric ent eral f eeding provides t he equivalent of
pharmacologic acid suppression. I n t he process of doing so, ent eral f eeding also
increases splanchnic perf usion, t hus providing bet t er oxygen balance and
decreased bact erial t ranslocat ion. Preeminent among t he reasons f or ent eral
rat her t han parent eral nut rit ion is t he prot ect ion t hat local maint enance of
mucosal int egrit y provides against t he perils of syst emic inf ect ion, poor w ound
healing, and gut at rophy.
What to Do
I mmediat e inst it ut ion of t ot al ent eral nut rit ion should be maint ained f or as long as
a pat ient is in t he hospit al and has nut rit ional requirement s t hat cannot be met by
t he pat ient t aking an oral diet . This includes const ant f eeding, up t o and
t hroughout int raoperat ive periods, because bot h int raoperat ive and perioperat ive
t ot al ent eral nut rit ion have been proven t o be saf e if post pyloric. Eff ort s t o
ensure supranormal nut rit ion should cont inue w ell past t he immediat e post burn
period. That children w ho sust ain large burn injuries can have persist ent grow t h
delay f or up t o 2 years support s t he f act t hat energy needs are elevat ed f or
mont hs t o years af t er a burn appears t o be adequat ely healing. Pat ient s w it h
large t ot al body surf ace area burns can have a rest ing met abolic rat e of 120% of
normal even 9 mont hs af t er t he injury.
Similarly, periodic evaluat ion of t he requirement s f or and overall eff ect s of t ot al
ent eral nut rit ion is recommended. I ndirect calorimet ry provides a noninvasive
means of evaluat ing energy expendit ure and a usef ul means of garnering caloric
requirement s. Urine urea nit rogen (UUN), a represent at ive measurement of
prot ein cat abolism, can be used t o det ermine t he prot ein-sparing eff ect iveness
of a part icular regimen, and f eeding goals should include a consist ent ly posit ive
nit rogen balance. Use of t he visceral prot eins such as prealbumin and t ransf errin
is most helpf ul w hen view ed as a part of a t rend and not as a point in t ime.
Suggested Readings
Magnot t i LJ, Deit ch EA. Burns, bact erial t ranslocat ion, gut barrier f unct ion,
and f ailure. J Burn Care Rehabil 2005; 26(5): 383–391.
Pereira CT, Herndon DN. The pharmacologic modulat ion of t he hypermet abolic
response t o burns. Adv Surg 2005; 39: 245–261.
Raff T, Hart mann B, G ermann G . Early int ragast ric f eeding of seriously
burned and long-t erm vent ilat ed pat ient s: a review of 55 pat ient s. Burns
1997; 23: 19.
Ret t mer RL, Williamson JC, Labbe RF, Heimbach DM. Laborat ory monit oring
of nut rit ional st at us in burn pat ient s. Clin Chem 1992; 38: 334.
> Table of C ontents > B ur ns > 296 - S tr ongly C ons ider E s c har otom y in C ir c um fer ential B ur ns
296
Strongly Consider Escharotomy in
Circumferential Burns
Richard Wong She MBChB
Eschar ref ers t o relat ively inelast ic burned skin. An escharot omy is t he
procedure of “opening t he eschar, ” eit her t o rest ore or improve t he perf usion t o
an ext remit y or t o allow or improve vent ilat ion w hen t he t orso is involved.
I nf lammat ory mediat ors released in response t o t he burn injury, in combinat ion
w it h f luid resuscit at ion, result in generalized edema and t hird spacing. Unburned
skin is elast ic and can accommodat e t his increased sof t t issue volume, w hereas
burned skin, part icularly w hen t he burn is a deep part ial-t hickness or f ull-
t hickness burn, is relat ively inelast ic and cannot . Thus, w it h t ime, ongoing t issue
edema and sw elling beneat h t he inelast ic burn eschar result s in decreased t issue
perf usion and t issue compliance. Recognit ion of t his evolving clinical sit uat ion
requires a high index of suspicion and ongoing clinical assessment . The clinical
manif est at ions of const rict ing eschar t hat mandat e release depend on t he sit e of
involvement .
Extremities
I n t he ext remit ies, t he signs and sympt oms are similar t o t hose of compart ment
syndrome (w hich mandat es f asciot omy) and are described by t he w ell-know n
“p's”: pain out of proport ion, pain on passive f lexion/ ext ension, perishingly cold,
palpably sw ollen, pallor, parest hesia, and pulselessness. Act ual loss of a pulse
is a lat e sign, w hereas a diminished signal on Doppler exam is t he hallmark.
Treat ment involves t he release of t he const rict ing, or rat her limit ing, envelope of
t issue (t he eschar).
The index of suspicion f or t he need f or escharot omy should be raised in t he
pat ient w ho requires large f luid volumes f or resuscit at ion because of t he size of
his or her burn and in t he pat ient w it h a circumf erent ial or near-circumf erent ial
deep part ial-t hickness or f ull-t hickness burn. I n t his set t ing, clinical monit oring
consist s of assessment of capillary ref ill and pulses (eit her t hrough palpat ion or
Doppler ult rasound). Any change in t he clinical examinat ion suggest ive of
circulat ory compromise should be invest igat ed. Reversible causes of poor
perf usion, namely hypot hermia, hypovolemia, hypot ension, and
ext ernal causes of const rict ion (e. g. , dressings) should be addressed. I f poor
perf usion remains, t hen escharot omy should be perf ormed in t he aff ect ed limb.
Escharot omies are made in t he medial and lat eral, midaxial lines w it h t he limbs
in t he anat omic posit ion, beginning a f ew millimet ers proximal and ending a f ew
millimet ers dist al t o t he ext ent of t he burn. There is a t endency f or t he upper
limbs t o pronat e and f lex, w hich should be correct ed prior t o marking t he lines
f or t he escharot omy. Care should be t aken around t he medial epicondyle of t he
humerus and t he head of t he f ibula f or t he ulnar and common peroneal nerves,
respect ively. O ccasionally, release of t he int ermet acarpal (lumbrical) spaces
dorsally is required f or a hand burn. Digit al escharot omies are of limit ed use due
t o t he absence of ischemic-sensit ive muscle w it hin t he f inger and pot ent ial
damage t o t he neurovascular bundles t hat run along t he midaxial line of t he
digit s.
Torso
Circumf erent ial chest w all burns decrease compliance, w it h result ant diff icult ies
in vent ilat ion. Children, w ho are predominant ly diaphragmat ic breat hers, are at
more risk of compromised vent ilat ion even if just t he ant erior aspect of t he chest
and abdomen are involved. I ncreased peak vent ilat ory pressures and decreased
compliance not responding t o sedat ion or paralysis in t hese set t ings are
indicat ions f or chest escharot omies. Escharot omies of t he chest are made t o
creat e an independent “breast plat e, ” much like a suit of armor. I ncisions are
made along t he ant erior axillary lines t o t he cost al margins. Transverse incisions
connect ing t hese t hat run parallel t o t he cost al margins can be added, if needed.
Pat ient s requiring escharot omies t ypically have major burns and are int ubat ed in
t he int ensive care unit , so it is usually easy t o produce sat isf act ory analgesia
and sedat ion. Nonint ubat ed pat ient s may require a general anest het ic, even
t hough t he eschar is said t o be insensat e. The procedure can be done at t he
bedside as a clean (as opposed t o st erile) procedure. A scalpel or
elect rocaut ery is used t o cut t hrough t he eschar. The dept h of t he incision is
t hrough t he ent iret y of t he eschar t o expose t he underlying f at but not t hrough
t he underlying f ascia. Hemost asis is import ant because t here can be signif icant
blood loss f rom t hese incisions. Complet eness of t he escharot omy is conf irmed
by running a f inger along t he w oundand addressing any residual bands of t issue.
The eff ect iveness of t he escharot omy is det ermined by t he improvement in
physiologic endpoint s, namely perf usion in t he limbs and vent ilat ion in
t he t orso. I f t here is no improvement in limb perf usion, considerat ion should be

made f or f ormal f asciot omies.
Af t er escharot omy, dressings should prevent desiccat ion of t he underlying
subcut aneous t issue and not be so t ight as t o duplicat e t he original problem.
Because t hese w ounds are t ypically w it hin t he burn, t hey are most easily
dressed w it h a t opical ant imicrobial and t he usual gauze. Ult imat ely, t he
escharot omy w ound w ill be excised along w it h t he burn and resurf aced.
Selected Readings
Het t iarat chy S, Papini R. ABC of burns. I nit ial management of a major burn:
I I . Assessment and resuscit at ion. BMJ 2004; 329: 101–103.
Kao CC, G arner WL. Acut e burns. Plast Reconst r Surg 2000; 104: 2482–2500.
> Table of C ontents > B ur ns > 297 - B e Aw ar e of the P itfalls in the Managem ent of S tevens - Johns on
S yndr om e and Toxic E pider m al Nec r olys is
297
Be Aware of the Pitfalls in the Management of
Stevens-Johnson Syndrome and Toxic
Epidermal Necrolysis
Vijay A. Singh MD
Stephen M. Milner MD, FACS
There is cont roversy about t he classif icat ion of St evens-Johnson syndrome and
t oxic epidermal necrolysis. These rare but pot ent ially debilit at ing dermat ologic
abnormalit ies have a mort alit y rat e as high as 40%. The debat e has cent ered on
w het her t hese condit ions are relat ed or separat e ent it ies. Current consensus is
t hat t hey are variant s of t he same disease w it h diff ering severit ies, w it h t he
ext ent of epidermal det achment being t he key component t o it s classif icat ion. A
pract ical clinical classif icat ion considers less t han 10% of t he t ot al body surf ace
area involved t o indicat e St evens-Johnson syndrome. I f t he epidermal
det achment involves more t han 30% of t he t ot al body surf ace area, t hen t oxic
epidermal necrolysis is t he presumed diagnosis. An area bet w een 10% and 30%
is less def init ive and is considered t o represent St evens-Johnson syndrome and
t oxic epidermal necrolysis overlap.
Watch Out For

A mult it ude of medicat ions have been associat ed w it h t he development of t hese
condit ions, but most commonly Dilant in, vancomycin, and sulf a drugs have been
implicat ed. Suspicion should be raised in pat ient s w ho have f ever and
const it ut ional sympt oms f or 2 t o 3 days preceding a rash t hat appears f ollow ing
t he administ rat ion of medicat ion. O t her prodromal signs and sympt oms may
include conjunct ivit is, pharyngit is, and prurit us. The acut e phase of t he disease
is charact erized by persist ent f ever, mucosit is, and generalized epidermal
sloughing. The skin lesions begin as a painf ul or burning erupt ion, usually st art ing
symmet rically on t he f ace and t horax and spreading t o t he ent ire body.
Denudat ion and erosions of t he mucous membranes usually are corresponding.
Mucosit is can occur on conjunct ival, buccal, t racheal, bronchial, pharyngeal,
esophageal, nasal, anal, vaginal, and perineal mucosa. The skin involvement can
be ext ensive. The eryt hemat ous skin lesions of t oxic epidermal necrolysis exhibit
t he Nikolsky sign (slipping of t he superf icial skin layer induced by slight rubbing
pressure).
Diagnosis is by def init ion hist ologic. Theref ore, skin biopsy at t he earliest
possible st age is import ant in est ablishing an accurat e diagnosis
and direct ing specif ic t herapeut ic modalit ies. Consequent ly, early consult at ion
w it h a dermat ologist is recommended. O nce diagnosis is made, t herapy is most ly
support ive t oget her w it h cessat ion of any suspect ed medicat ions used.
G enerally, t reat ment is most eff icient ly provided in an int ensive care unit or a
burn care cent er because t hese pat ient s are or have t he pot ent ial t o be
ext remely ill and require mechanical vent ilat ion, hemodynamic support , and
specialized w ound care. There is no def init e t reat ment f or t he ailment . G iven it s
aut oimmune nat ure, immunoglobulin and/ or st eroid t herapy has periodically been
used w it h varying success. The out comes of clinical cont rolled st udies have yet
t o be det ermined.
Suggested Reading
Let ko E. St evens-Johnson syndrome and t oxic epidermal necrolysis: a review
of t he lit erat ure. Ann Allergy Ast hma I mmunol 2005; 94: 419â 4 36.
> Table of C ontents > B ur ns > 298 - Adopt a P hilos ophy of E ar ly E xc is ion and Gr afting of B ur n W ounds
298
Adopt a Philosophy of Early Excision and
Grafting of Burn Wounds
Jeremy W. Pyle MD
No measure has had a great er eff ect on t he out come f or burn pat ient s t han t he
shif t aw ay f rom conservat ive management and t ow ard early and aggressive
surgical management . I n t he 1970s, t he st andard of care f or burns w as t o
perf orm dressing changes w it h daily debridement s, w it h bact erial collagenase
aiding in t he t hinning of t he eschar and t he development of granulat ion t issue. I n
due t ime, split -t hickness skin graf t s (STSG ) w ould be applied, and a t ake of
about 50% w as expect ed. A gradual evolut ion has produced a new st andard of
care in t erms of bot h result ant graf t t ake and individual f unct ion. A 95% graf t
t ake is now t he norm. Current pract ice has evolved around one principle: early
excision of w ounds and immediat e coverage provides ext raordinarily bet t er
result s. No ot her int ervent ion has reduced burn mort alit y as much.
Benef it s of early excision and graf t ing include decreased lengt h of st ay, reduced
cost s, less t ime aw ay f rom w ork or school, f ew er debridement s, f ew er inf ect ious
w ound complicat ions, and reduced mort alit y. Furt hermore, t here is evidence t o
suggest t hat early excision and graf t ing leads t o less int ense complicat ions w hen
t hey do occur. Aest het ic and f unct ional benef it s are relat ed t o a less int ense
inf lammat ory response in t he w ound, w hich is reduced w it h early removal of
eschar and provision of coverage.
What to Do
Timing of early excision and graf t ing is not exact . Many w ounds benef it f rom 3 t o
7 days of conservat ive management w hile t he t issue in t he zone of st asis
declares it self as eit her viable or not . Wit h t he except ion of w ounds over a joint ,
if t he t issue appears able t o heal in less t han 2 t o 3 w eeks, surgical
management is likely unnecessary. Those w ounds t hat do not require excision
have a good vascular supply providing a pink or w hit e and shiny base w it h rapid
capillary ref ill. These are charact erist ics of superf icial part ial-t hickness burns,
w hich do w ell w it h local w ound care and t ime. When indicat ed, as in obvious f ull-
t hickness burns, excision and graf t ing should be perf ormed as soon as t he
pat ient is st abilized. I deally, t his is w it hin 5 days and cert ainly w it hin 1 w eek.
Dat a
suggest t hat morbidit y, lengt h of st ay, local inf ect ion, and sepsis are more
common if w ounds needing excision are t reat ed conservat ively beyond 1 w eek.
Tangent ial and f ascial excisions are t he t w o w ell-accept ed met hods of burn
w ound excision. Alt hough each has a place in current st andards, t angent ial
excisions const it ut e t he majorit y of procedures. Tangent ial, also called
sequent ial, excision represent s a st rat egy t hat conserves as much of t he nat ive
t issue as possible. Risks involved w it h t his approach hinge on t he int ense blood
loss t hat can occur. The procedure uses a hand dermat ome t o remove t hin layers
of eschar and unhealt hy t issue. I f removal of one layer provides an adequat ely
vascular surf ace, t hen t he excision is done. I f t he w ound bed appears dry or
gray or if t hrombosed vessels are st ill readily apparent , excision must cont inue.
Excision ends w it h t he removal of all dead t issue. Hemost asis is accomplished
w it h epinephrine (1: 10, 000)-soaked pads covering f resh excision sit es, f ollow ed
by caref ul elect rocaut ery t o act ively bleeding vessels. Fascial excision
represent s t he second major met hod of burn w ound excision. Dissect ing dow n t o
a f ascial plane w it h elect rocaut ery provides a consist ent and viable w ound bed
f or graf t ing. Fascial excision can signif icant ly decrease operat ive blood loss
w hile removing some of t he guessw ork in evaluat ing eff ect ive excision.
Disadvant ages t o t his t echnique include longer operat ive t ime, great er damage
t o superf icial neurovascular st ruct ures and lymph passages, and inf erior
cosmesis t hat can result f rom obvious t issue def icit s.
What To Do Next
Bot h excision t echniques are immediat ely f ollow ed by reconst ruct ive eff ort s,
w henever possible. Evaluat ion of t he w ound bed and t he expect ed overall course
of t he pat ient help t o dict at e w here and w hen def init ive coverage is at t empt ed.
When f easible, STSG t o t he ent ire w ound is ideal. When t he problem of limit ed
donor sit es arises, skin subst it ut es are commonly used. Most common among
t hese are t he dermal subst it ut es I nt egra, AlloDerm, and Dermagraf t . Their
primary purpose is t o serve as a dermal replacement allow ing t hinner graf t s t o
survive over t hem; secondarily t hey provide a t emporary means of part ially
closing a w ound. Using a t hin STSG (<0. 010 in. ) allow s donor sit es t o be more
f requent ly reharvest ed. AlloDerm is cryopreserved allogeneic dermis t hat has
been st ripped of it s dermal element s, t hereby allow ing nat ive f ibroblast
inf ilt rat ion and t he event ual development of a st able dermis. I nt egra consist s of
a deep, porous collagen-chondroit in 6-sulf at e f ibrillar layer covered w it h a
removable layer of silicone. The f ibrillar layer provides scaff olding f or ingress
of aut ogenous f ibroblast s and subsequent creat ion of a “neo-dermis. ” The t hin
silicone layer is a t emporary epidermis and is ext irpat ed af t er vascularizat ion
w hen ult imat e coverage is applied. I nt egra is t he most st udied of t he dermal
subst it ut es. Dermagraf t is a cult ure of human neonat al f ibroblast s on a layer of
Biobrane. Tradit ional w isdom has been t hat Alloderm can be immediat ely covered
w it h graf t , w hereas I nt egra and Dermagraf t should be allow ed t o vascularize f or
about 14 days bef ore placing a graf t .
O nce applied, t here are mult iple means of securing graf t s. For decades,
st aples/ sut ures w it h a bolst er dressing has been t he dressing of choice because
it helps t o mechanically secure t he graf t and inhibit seroma or hemat oma
f ormat ion, t hereby eliminat ing t hree of t he most common causes f or graf t f ailure.
The advent of negat ive-pressure dressings (e. g. , VACTM by KCI USA, I nc. , San
Ant onio, TX) has provided a means of limit ing shear f orces, f luid collect ions,
bridging, and bact erial cont aminat ion w hile increasing granulat ion t issue
f ormat ion, vascularit y, oxygen t ension, and t hus more rapid re-epit helializat ion of
w ounds t han w it h bolst er dressings. Negat ive-pressure dressings are applied at
t he t ime of excision and graf t ing and removed 4 t o 5 days lat er. A single layer of
a nonadherent dressing is applied bet w een t he graf t and t he negat ive-pressure
dressing t o keep t he t w o easily separable.
Much experiment al w ork has been done on t he use of cult ured aut ograf t s f or
coverage of ext ensively burned individuals. The U. S. Food and Drug
Administ rat ion has approved one cult ured epit helial aut ograf t . Epicel requires a
f ull-t hickness biopsy of a pat ient 's skin and 2 t o 3 w eeks t o grow. Use is
signif icant ly limit ed by it s t remendous cost and legit imat e quest ions of graf t
durabilit y.
Suggested Readings
At iyeh BS, Hayek SN, G unn SW. New t echnologies f or burn w ound closure
and healing. Review of t he lit erat ure. Burns 2005; 31(8): 944–956.
Barret JP, Herndon DN. Eff ect s of burn w ound excision on bact erial
colonizat ion and invasion. Plast Reconst r Surg 2003; 111(2): 744–750;
discussion 751–752.
Thourani VH, I ngram EL, Feliciano DV. Fact ors aff ect ing success of split -
t hickness skin graf t s in t he modern burn unit . J Trauma 2003; 54(3): 562–568.
w w w. w oundvac. com
> Table of C ontents > B ur ns > 299 - B e Aggr es s ive in Mobiliz ing B ur n W ounds
299
Be Aggressive in Mobilizing Burn Wounds
Dana Nakamura O T
Burn pat ient s in t he int ensive care unit should be seen on admission by t he
physical and occupat ional t herapy t eam. The key to successf ul rehabi l i tati on i s
earl y i nterventi on. Scarring is virt ually inevit able w it h major burns and t herapy is
int egral t o limit ing it s eff ect on pat ient out comes. I nit ial assessment s include
det erminat ion of posit ioning and splint ing needs, range of mot ion (RO M) and
st rengt h, prior level of f unct ional mobilit y and act ivit y independence, cognit ion,
and abilit y t o learn/ part icipat e in t he t herapy program. Early int ervent ions include
posit ioning and splint ing, edema management , exercise, f unct ional
mobilit y/ act ivit y ret raining, and pat ient / f amily educat ion. Early int ervent ion and
ant icipat ion of pot ent ial complicat ions are crit ical t o opt imizing f unct ional
out come. Al l members of t he burn t eam need t o be able t o ant icipat e pot ent ial
complicat ions t o be proact ive and t ake a prevent at ive approach t o t reat ing burn
pat ient s. St aff educat ion w it h t raining in t he roles of all t eam members and daily
collaborat ion are import ant f or smoot h carryover of t reat ment plans.
Communicat ion of t herapy plans is vit ally import ant so t hat all st aff involved in
t he pat ient 's care underst and t he t echniques and procedures used and t he
rat ionale f or such. Pat ient and f amily educat ion is also import ant t o f acilit at e
underst anding of t he t herapy process and build rapport f or a t herapeut ic
relat ionship. I t is imperat ive t o communicat e t hat “t herapy” needs t o cont inue 24-
7 because t he healing and scarring process is a cont inuous and relent less one.
Communicat ion can occur by a number of w ays such as one-on-one t eaching, in-
services, w rit t en inf ormat ion in t he medical record, on-line document at ion, and
phot os/ diagrammed inst ruct ions on bedside boards.
Positioning and Splinting

Posit ioning goals are t o minimize edema, prevent t issue dest ruct ion, and
maint ain sof t t issue in an elongat ed st at e t o f acilit at e f unct ional recovery.
Likew ise, splint ing prevent s loss of mot ion and def ormit y, promot es f unct ional
independence, prot ect s anat omic st ruct ures, preserves skin graf t int egrit y, and
rest ores f unct ion. The met hods used must prevent cont ract ure and def ormit y but
not compromise t he
pat ient 's mobilit y and f unct ion. I ndividual programs are developed based on t he
pat ient 's medical st at us, range of mot ion, locat ion of burn, burn dept h, and level
of cooperat ion. I t should alw ays be st ressed t hat t he “posit ion of
comf ort =posit ion of cont ract ure” w hen det ermining appropriat e posit ioning and
splint ing programs. Frequent posit ioning checks and splint f it assessment s are
required t o prevent misalignment and pressure sore complicat ions.
Exercise
Frequent exercises and movement are crit ical. Treat ment w it h t he t herapist once
or t w ice a day f or 30 t o 60 minut es is not suff icient . Pat ient s w ho are aw ake and
able t o part icipat e are inst ruct ed in self -range of mot ion exercises t o st ret ch t he
aff ect ed body part and skin. The benef it s of exercise include preservat ion of
range of mot ion, f unct ional st rengt h, endurance, and coordinat ion; act ivat ion of
t he muscle pump act ion f or edema mobilizat ion; prevent ion of venous
t hromboembolism; endorphin release; and psychological w ell being including
f eelings of having some sense of cont rol in a new and of t en chaot ic medical
set t ing.
Functional M obility
Ambulat ion is crit ical t o preserve skin f lexibilit y, muscle st rengt h, and
t endon/ ligament lengt h. Many t imes, physicians aut omat ically w rit e admit orders
f or “bed rest . ” Nursing and t herapy st aff members need t o assess t he individual
sit uat ion and quest ion t he bed rest order because being conf ined t o bed just
cont ribut es t o complicat ions such as het erot opic ossif icat ion, venous
t hromboembolism, muscle w ast ing, and pressure sores. Vent ilat or-dependent
pat ient s are also mobilized in bed or t o “cardiac” chairs and at some burn
cent ers are ambulat ed w it h respirat ory t herapy, nursing, and
occupat ional/ physical t herapy collaborat ing. Cont inued mobilizat ion in some f orm
is crit ical during vent ilat or support because it prevent s/ reduces respirat ory
complicat ions.
Functional Activity
All self -care act ivit ies are considered exercise, especially if t he hands or upper
ext remit ies are involved. Play act ivit ies f or bot h pediat ric and adult pat ient s are
also import ant , w het her it be shoot ing a basket ball, playing cards or a board
game, bouncing a ball, or assembling a puzzle/ building blocks. Many burn
cent ers have a child lif e specialist
or t herapeut ic recreat ion specialist w ho can assist w it h such act ivit ies. Pet
t herapy programs are benef icial f or social visit s w it h a t herapy dog, hands-on
t herapy w it h w alking t he dog on a leash f or balance and gait t raining, and
brushing/ grooming or playing f et ch f or arm range of mot ion.
Suggested Readings
Campbell SK, ed. Physical Therapy f or Children. Philadelphia: WB. Saunders,
1994: 763–786.
Robinson LR, ed. Trauma Rehabilit at ion. Philadelphia: Lippincot t Williams &
Wilkins, 2006: 181–204.
> Table of C ontents > B ur ns > 300 - Have a Low Thr es hold for Intubating a P atient w ith an Inhalation
B ur n Injur y
300
Have a Low Threshold for Intubating a Patient
with an Inhalation Burn Injury
Richard Wong She MBCHB
A burn pat ient w it h an inhalat ion injury has t w ice t he mort alit y rat e of a similar
burn pat ient w it hout an inhalat ion injury. I nit ial present at ion may be essent ially
asympt omat ic because t ime (and resuscit at ion f luid) is required t o generat e t he
edema and alveolar damage, w hich w ill manif est as progressive airw ay
obst ruct ion and dist urbances in pulmonary f unct ion. The mainst ay of t reat ment
f or inhalat ion injuries involves securing t he airw ay pri or t o obst ruct ion and
maint aining vent ilat ion w hile minimizing lung damage. I nhalat ion injury can result
f rom a combinat ion of t hree mechanisms. First , direct t hermal injury t o t he upper
airw ay can be caused by superheat ed gases or aspirat ion of hot liquids. Hot
liquids result in rapid edema f ormat ion and require emergent int ubat ion if
suspect ed. By cont rast , superheat ed gases result in relat ively gradual edema
f ormat ion. Second, product s of combust ion can dissolve in t he mucus of t he
low er airw ay, result ing in a chemical pneumonit is and alveolar damage, w hich
develops over hours t o days. Finally, syst emic poisoning can result f rom t he
absorpt ion of carbon monoxide (CO ). This poisoning should be suspect ed in all
inhalat ion injuries. A carboxyhemoglobin level should be measured on init ial
evaluat ion of t he pat ient . Humidif ied 100% oxygen is t he t reat ment of choice
because it decreases t he half -lif e of carboxyhemoglobin f rom 250 minut es on
room air respirat ion t o around 40 minut es. There is a limit ed role f or hyperbaric
oxygen. Supplement al oxygenat ion should cont inue f or up t o 48 h af t er
normalizat ion because t here can be delayed release of CO f rom t he int racellular
cyt ochrome syst em.
Signs and Symptoms

Diagnosis of an inhalat ion injury requires a high index of suspicion based on a
combinat ion of t he hist ory and physical f indings. A hist ory of being t rapped in a
conf ined space w it h a f ire is highly suggest ive. By cont rast , a f lash f ire, even
w hen associat ed w it h singed nasal hairs, is t ypically not associat ed w it h an
inhalat ion injury. Sympt oms and signs of an inhalat ion injury include a f acial burn,
soot around t he nose and mout h, int raoral soot or burns, singed f acial and nasal
hair, carbonaceous sput um, a hoarse voice, inspirat ory st ridor or
increased w ork of breat hing, and CO poisoning, w hich may present w it h signs as
subt le as an alt ered level of consciousness or a decreased rat io of mean
alveolar oxygen pressure t o f ract ion of oxygen in t he inspired air. Alt hough
conf irmat ion of an inhalat ion injury can be achieved w it h nasopharyngoscopy,
laryngoscopy, or bronchoscopy, t hese invasive diagnost ic t est s do not inf luence
t he out come of an inhalat ion injury.
Prophylact ic int ubat ion of a suspect ed inhalat ion injury is pref erable t o w ait ing
f or event ual airw ay compromise f rom edema. Emergent int ubat ion in t hese
circumst ances can be challenging. I n addit ion, if t he pat ient is t o be t ransport ed
t o a regional burn cent er, it is bet t er t hat t he airw ay be secured bef ore
t ransport . Acont rolled ext ubat ion on arrival is bet t er t han an emergent int ubat ion
en rout e. During int ubat ion, t he presence of mucosal involvement and
carbonaceous deposit s should be not ed.
Wit h inhalat ion injuries, det eriorat ion in pulmonary f unct ion can t ake several days
t o occur. The t reat ment of an inhalat ion injury is support ive. The goal is t o
maximize oxygenat ion/ vent ilat ion w hile minimizing alveolar damage, t hus allow ing
t he lungs t o heal t hemselves suff icient ly f or ext ubat ion. Prophylact ic st eroids and
ant ibiot ics have no role. Bronchospasm should be t reat ed w it h bronchodilat ors.
Small airw ay obst ruct ion can occur secondary t o debris and slough liberat ed
f rom t he airw ays. Aggressive pulmonary t oilet is essent ial. Nebulized heparin and
N-acet yl cyst ine may be usef ul, especially in children. Subsequent pulmonary
inf ect ion can occur, eit her as pneumonia or as a purulent t racheobronchit is.
Bronchoalveolar lavage may help t o quant if y and ident if y organisms t o be
specif ically t reat ed. Vent ilat or st rat egies f or maint aining oxygenat ion w hile
minimizing airw ay pressures include lung-prot ect ive vent ilat ion, permissive
hypercapnia, high-f requency percussive vent ilat ion, nit ric oxide, and even
ext racorporeal support . Prolonging survival permit s spont aneous lung recovery
and f inally ext ubat ion.
Selected Readings
Ansermino M, Hemsley C. ABC of burns; int ensive care management and
cont rol of inf ect ion. BMJ 2004; 329: 220–223.
Sheridan RL. Airw ay management and respirat ory care of t he burn pat ient . I nt
Anest hesiol Clin 2000; 38(3): 129–145.
> Table of C ontents > B ur ns > 301 - D o not Under - R es us c itate a B ur n P atient
301
Do not Under-Resuscitate a Burn Patient
Thermal injuries result in t he release of cyt okines and ot her inf lammat ory
mediat ors at t he sit e of injury. These mediat ors have a syst emic eff ect once t he
burn reaches about 30% of t he t ot al body surf ace area (TBSA). Among t he
cardiovascular eff ect s, t here is a loss of microvascular int egrit y, causing
int ravascular f luid t o leak int o t he int erst it ial space. Peripheral and splanchnic
vasoconst rict ion also occurs, and myocardial cont ract ilit y has been show n t o
decrease, possibly due t o a release of t umor necrosis f act or-α . The end result is
burn shock: syst emic hypot ension and end-organ hypoperf usion. The met abolic
rat e may also increase up t o t hree t imes it s basal rat e, f urt hering t he demand
f or t issue perf usion. I n addit ion, a concomit ant inhalat ion injury w ill t ypically
increase resuscit at ion f luid requirement s.
I t has been show n t hat maint aining adequat e int ravascular volume is import ant t o
perf use vit al organs and may also limit t he dept h/ ext ent of a burn. Use of
cryst alloid alone, specif ically lact at ed Ringer's solut ion (LR), is a saf e and
eff ect ive w ay t o resuscit at e burn pat ient s during t he f irst 24 h af t er injury. Using
t he Parkland-Baxt er f ormula, 4 mL LR × w eight (kg) × %TBSA burn, one can
est imat e t he t ot al volume of lact at ed Ringer's solut ion needed in t he f irst 24 h
af t er t hermal injury. Because t he increase in capillary permeabilit y is great est
over t he f irst 8 h af t er injury, half of t he volume of lact at ed Ringer's solut ion as
calculat ed by t he Parkland-Baxt er f ormula is given over t he f irst 8 hours f rom the
ti me of i njury and t he remainder is given over t he subsequent 16 hours. An
accurat e assessment of burn dept h and %TBSA is import ant f or t he Parkland-
Baxt er f ormula t o be eff ect ive. The “rule of nines” is commonly used as a quick
bedside assessment t o est imat e TBSA: each arm is about 9% of t he TBSA, each
leg is about 18%, t he ant erior t runk is about 18%, t he post erior t runk is about
18%, and t he head is about 9%, w it h t he perineum encompassing t he remaining
approximat ely 1%. Children have great er evaporat ive losses because t hey have
vast ly diff erent body proport ions and a great er percent age of t ot al body w at er.
Thus, age-appropriat e diagrams (e. g. , Lund-Brow der or Berkow ) are more
accurat e in det ermining t he TBSA burned in children.
Exampl e. A 70-kg adult w it h a 40% burn requires 11, 200 mL (4 × 70 × 40) over
24 h f rom t he t ime of injury, w it h 5, 600 mL being given in t he f irst 8 h f rom t he
t ime of injury and 5, 600 mL being given in t he subsequent 16 h.
I t should be st ressed t hat any resuscit at ion f ormula only provides an est imat e
f or required f luids. O bject ive paramet ers such as urine out put (0. 5 cc/ kg f or
adult s and 1. 0 cc/ kg f or children) or blood pressure should be used t o adjust
f luid needs. Mean art erial pressure (MAP) and urine out put are considered t o be
t he most reliable measures of adequat e t issue perf usion f or burn resuscit at ion. A
MAP maint ained above 60 mm Hg t ypically ensures adequat e cerebral perf usion.
Hypot ension or decreased urine out put should be t reat ed by judiciously
increasing t he resuscit at ion f luid rat e and not by bolusing f luids because boluses
w ill only increase t he volume of t he capillary leak, result ing in w orse edema.
There is no consist ent evidence t o support t he use of pulmonary art erial cat het er
measurement s f or rout ine resuscit at ion. I n f act , t heir use may lead t o over-
resuscit at ion and subsequent complicat ions. Nonet heless, invasive hemodynamic
monit oring may be necessary in pat ient s w it h signif icant premorbid
cardiovascular, pulmonary, or renal disease.
What to Do
Delay in burn resuscit at ion has unequivocally been show n t o increase mort alit y
rat es. I nt ravenous access is vit al and may be obt ained by placing t w o large-bore
peripheral int ravenous lines (16 t o 18 gauge) or, less commonly, a cent ral
venous line. Venous cut -dow ns are t o be avoided, given t he high incidence of
inf ect ion w it h t hem. Lact at ed Ringer's solut ion is t he most common cryst alloid
solut ion used during resuscit at ion in t he Unit ed St at es. Hypert onic saline has
been advocat ed by some cent ers, but has not been show n t o decrease f luid
requirement s. Normal sal i ne shoul d never be used to resusci tate burn pati ents
because t he volumes required w ill precipit at e a severe hyperchloremic met abolic
acidosis. Albumin is not advised in t he early post burn resuscit at ion period
because it simply passes out int o t he t issues, causing great er t issue edema.
G lucose should be administ ered in children t ypically w eighing less t han 20 kg t o
prevent hypoglycemia because t hey have relat ively diminished glycogen st ores.
This may be accomplished eit her int ravenously in t he f orm of w eight -based
dext rose-cont aining maint enance f luids (in addit ion t o resuscit at ion lact at ed
Ringer's volumes) or ent erally in t he f orm of t ot al ent eral nut rit ion.
Use of diuret ics and inot ropes should be avoided because t hey w ill not st op t he
capillary leak and may lead t o ischemia of t he w ound, possibly result ing in
conversion of a part ial-t hickness t o a f ull-t hickness w ound or ext ension of t he
TBSA of t he w ound. Pat ient s w it h myoglobinuria require increased volumes of
f luid or mannit ol t o maint ain urine out put of 100 cc/ h or great er. As t he capillary
leak resolves at about 24 hours af t er injury, t he amount of f luid needed t o
maint ain an appropriat e MAP and urine out put decreases. Colloids (albumin) may
be used t o f acilit at e resuscit at ion in a pat ient w it h persist ent low urine out put
and/ or hypot ension af t er t he f irst 24 hours. Wit h colloids, t he f ormula used is 0. 3
t o 0. 5 cc/ kg/ %TBSA burned over 24 hours; 5% albumin is t ypically used. I f t he
pat ient cont inues t o have inadequat e urine out put and MAP or his or her required
resuscit at ion volumes great ly exceed t he est imat ed volumes, plasma-pheresis
may be used because t here is a t heoret ical advant age t o removing inf lammat ory
mediat ors t hat cause vasodilat ion and capillary leak.
I t is w ell know n t hat inadequat e resuscit at ion result s in decreased t issue
perf usion w it h subsequent mult iorgan f ailure and increased mort alit y. How ever,
over-resuscit at ion also exposes pat ient s t o complicat ions. These include
compart ment syndromes involving t he abdomen or ext remit ies, pulmonary edema,
and pleural eff usions. Abdominal compart ment syndrome decreases lung
compliance and also impedes lung expansion, t hus causing elevat ed airw ay
pressures and hypovent ilat ion. The classic present at ion includes high peak
airw ay pressures, decreased venous ret urn, oliguria, and int ra-abdominal
pressures exceeding 25 mm Hg. Abdominal compart ment syndrome is largely
avoidable via a judicious resuscit at ion w it h close monit oring. I f abdominal
compart ment syndrome occurs, decompressive laparot omy is t he only proven
eff ect ive t reat ment in adult s, w hereas children may respond t o paracent esis.
What To Do Next
Af t er resuscit at ion has been complet ed, ongoing f luid administ rat ion should be
cont inued t o replace insensible losses t hat are associat ed w it h t he burn it self
and t he evolving hyperdynamic st at e. Maint enance f luid requirement s in t he
burned pat ient are classically 1. 5 t imes t hose of nonburned individuals.
Administ rat ion of f luids may be cont inued int ravenously; how ever, t he pref erable
rout e is ent eral. These should be in addit ion t o and combined w it h t ot al ent eral
nut rit ion support . Daily w eighings can help t o measure and assess ongoing
insensible f luid loss or f luid ret ent ion.
Suggested Readings
Bessey PQ . Met abolic response t o crit ical illness. I n: American College of
Surgeons: Principles and Pract ice. Chicago: American College of Surgeons,
2006: w w w. acssurgery. com
G ibran NS, Heimbach DM. Management of t he pat ient w it h t hermal injuries.

I n: American College of Surgeons: Principles and Pract ice. Chicago: American
College of Surgeons, 2006: w w w. acssurgery. com
> Table of C ontents > B ur ns > 302 - Avoid the P itfalls of Vas c ular Ac c es s in B ur n P atients
302
Avoid the Pitfalls of Vascular Access in Burn
Patients
Myron S. Powell MD
A burn pat ient w it h less t han 50% t ot al body surf ace area (TBSA) involved can
usually begin resuscit at ion via t w o large-bore peripheral lines. Wit h burns of
great er t han 50% TBSA or in pat ient s w it h severe premorbid diseases, cent ral
venous access is t ypically needed. Vascular access should ideally be est ablished
early in t he course of injury bef ore t he f ormat ion of signif icant edema. I f
resuscit at ion is inadequat e, burn shock may lead t o mult iorgan
dysf unct ion/ f ailure syndrome, w hich almost invariably result s in a f at al out come.
What to Do
When obt aining vascular access on a burn pat ient , t here are numerous opt ions
w it h t heir individual benef it s, risks, and complicat ions. Placement t hrough
unburned t issue, all t hings being equal, is pref erable t o placement t hrough
burned t issue; t hat is, an int ravenous line should be placed t hrough t he right
ant ecubit al space in t he set t ing of a lef t ant ecubit al burn. Peripheral lines in t he
upper ext remit ies are opt imal; how ever, cut -dow ns are cont raindicat ed, given t he
inordinat e inf ect ion rat e associat ed w it h t hem. Any line may be placed t hrough
burned t issue, if necessary, but lines t hrough eschar should be removed w it hin
72 hours, given t he high colonizat ion and inf ect ion rat es beyond t hat t ime period.
Cent ral venous cat het ers are benef icial f or t hose requiring prolonged t reat ment s
w it h large volumes of f luids (i. e. , large burns or inhalat ion injury), invasive
hemodynamic monit oring, t ot al parent eral nut rit ion, or cert ain medicat ions.
Theoret ically, peripherally insert ed cent ral cat het ers avoid some of t he
disadvant ages of classical cent ral lines, but t heir eff icacy and place in t he
management of t he severely burned pat ient have yet t o be clinically def ined.
The risks associat ed w it h placement of a vascular access device somew hat
depend on t he insert ion sit e; how ever, t hey all include bleeding, inf ect ion, and
injury t o adjacent st ruct ures. These risks can be minimized by correct ing
coagulopat hies, caref ully maint aining st erile t echnique and caring f or t he
insert ion sit es, and adhering t o sound surgical principles. Nonet heless, an
inf ect ion may st ill develop. Cat het er-relat ed bloodst ream inf ect ion (CRBSI ) risk
in burn pat ient s is t heoret ically less w hen obt aining cent ral venous access above
t he diaphragm
(i. e. , via t he int ernal jugular or subclavian veins). Furt hermore, cat het er-induced
deep venous t hrombosis appears t o be more common w it h f emoral lines t han
w it h ot her sit es. About 1% of pat ient s develop a pneumot horax f rom cent ral line
placement w hen using a vein in t he chest or t he neck. Changing cent ral venous
cat het ers on a set schedule does not reduce inf ect ion rat es nor does it predict
t he rat e of bact erial cont aminat ion of t he cat het er. I n f act , it only exposes
pat ient s t o increased complicat ion rat es. Cent ral lines should only be changed
w hen inf ect ion is clinically suspect ed.
I n sit uat ions involving an inhalat ion injury or burns of great er t han 50% TBSA,
invasive hemodynamic monit oring is t ypically required, necessit at ing bot h venous
and art erial vascular access. For serial art erial blood gas sampling, invasive
blood pressure monit oring, and rout ine lab draw s, art erial access is ideal. The
radial art ery is t he pref erred sit e f or several reasons. I t has an easily det ect able
pulse t hat lies relat ively superf icial t o t he overlying skin. I schemic complicat ions
are low w it h cannulat ion of t he radial art ery secondary t o collat eral f low via t he
ulnar art ery and t he superf icial/ deep palmar arches; how ever, t hese collat erals
are absent in 5% t o 7% of t he populat ion. There are ot her sit es f or obt aining
art erial access, but t hey have specif ic disadvant ages compared w it h t he radial
art ery. The f emoral art ery may be cannulat ed; how ever, access in t he groin
carries a t heoret ical increased risk of inf ect ion. The brachial art ery is anot her
opt ion but of t en provides inaccurat e readings, risk of median nerve injury, and
risk of ischemia due t o lack of collat erals. The axillary art ery has a signif icant
risk of injury t o t he brachial plexus due t o it s proximit y.
Suggested Readings
Ramos G E, Bolgiani AN, Pat ino O , et al. Cat het er inf ect ion risk relat ed t o t he
dist ance bet w een insert ion sit e and burned area. J Burn Care Rehab
2002; 23(4): 266–271.
Sheridan R, Nackel A, Lydon M, et al. I nf radiaphragmat ic cent ral venous

pressures ref lect supradiaphragmat ic pressures in st able burn pat ient s. J
Trauma I njury I nf ect Crit Care 1999; 47(2): 300–302.
St ill JM, Law E, Thiruvaiyaru D, et al. Cent ral line–relat ed sepsis in acut e
burn pat ient s. Am Surg 1998; 64(2): 165–170.
> Table of C ontents > B ur ns > 303 - K now How to E s tim ate B ur n S iz e and D epth
303
Know How to Estimate Burn Size and Depth
John Zannis MD
Est imat ing t he severit y of a burn is import ant in t reat ing t hese complicat ed yet
common injuries. Burns const it ut e t he t hird-leading cause of t rauma deat hs in t he
Unit ed St at es and more t han 2. 4 million Americans are t reat ed f or burns each
year. Scalding is t he most common cause of t hermal injury in bot h adult s and
children. House f ires, alt hough account ing f or only 5% of adult s t reat ed, are
responsible f or great er t han 40% of t he adult deat hs.
Caref ul init ial evaluat ion and w ound management is crit ical t o t he t reat ment of
burns. Aside f rom pat ient resuscit at ion, proper management begins w it h
est imat ing t he severit y of a burn, including dept h and size. The abilit y t o do t his
requires a basic underst anding of skin anat omy and physiology.
The skin is composed of epidermis, dermis, and subcut aneous f at . The epidermal
layer provides a vapor and bact erial barrier. The dermal layer provides f lexibilit y
and st rengt h. I t also cont ains t he dermal appendages t hat produce oils, t erminal
capillaries f or t hermoregulat ion, and nerve endings. These f unct ions are
compromised af t er t he skin suff ers a burn injury. Assessing w hich of t hese layers
are aff ect ed by a part icular burn is of t en diff icult t o accomplish in t he acut e
set t ing. Early burn dept h est imat es are most accurat e in very deep or very
superf icial w ounds.
Superficial Burns (First Degree)

A superf icial burn is limit ed t o t he epidermis. I t is charact erized by pain, heat ,
and reddening of t he burned surf ace but does not show blist ering or charring of
t issue. Superf icial burns heal in approximat ely 5 days by epidermal regenerat ion
and do not scar. Typical superf icial burns include sunburn and hot w at er scalds.
Partial-Thickness Burns (Second Degree)

The epidermis and dermis are involved in part ial-t hickness burns. Blist ering
occurs at t he epidermal-dermal junct ion. The skin appears pink or cherry red,
moist , and w eepy t o mot t led w hit e. These burns are t ypically very painf ul and
exhibit int ense sw elling. Healing of
superf icial, part ial-t hickness burns occurs w it hout scarring or cont ract ure in 10
t o 21 days.
Full-Thickness Burns (Third Degree)

Full-t hickness burns ext end deeply int o t he dermis and dest roy sensory nerve
endings. The t issue damage ext ends below hair f ollicles and sw eat glands t o
subcut aneous (f at ) t issue. Wit h t his degree of burn, t he skin becomes charred
and leat hery. I nit ially, t he skin can be bright red, w axy w hit e, t an, or brow n;
t here are no blist ers, and massive sw elling is common. Full-t hickness burns are
usually not painf ul because t he injury has dest royed nerve endings. Hair pulls out
easily. Skin graf t ing and/ or ot her coverage opt ions are required f or t reat ment of
a f ull-t hickness burn.
Fourth-Degree Burns
This is a t erm somet imes used t o describe very deep burns t hat involve
underlying st ruct ures such as muscle, bone, t endon, and/ or ligament . Surf ace
appearance and ot her charact erist ics are similar t o t hose of t hird-degree burns.
I n addit ion t o burn dept h, t he severit y of a burn is det ermined by t he size or
body surf ace area aff ect ed. An accurat e assessment of burn size can be made
early in t he pat ient 's evaluat ion and is an import ant f act or in det ermining
resuscit at ive f luid administ rat ion. There are several met hods t hat can be used t o
est imat e t he percent age of body surf ace area burned. The most w idely used
met hod f or rapid assessment is t he “rule of nines, ” w hich st at es t hat t he f ront
t orso, back, and each low er ext remit y represent s 18% t ot al body surf ace area
(TBSA). Each upper limb and t he head represent 9% TBSA. The perineum equals
1% TBSA. Small burns or scat t ered w ounds can also be rapidly measured using
t he pat ient 's palmar surf ace, w hich represent s 1% TBSA. Alt hough t he rule of
nines is helpf ul, it is less accurat e t han ot her met hods of est imat ing burn ext ent .
This is part icularly t rue w it h children, w hose body proport ions change w it h
grow t h. The inf ant head represent s 18% and t he legs 14% of TBSA (in cont rast
t o 9% and 18%, respect ively, in t he adult ). Theref ore, burn size in children is
best est imat ed using an age-specif ic chart such as t he Lund and Brow der Chart .
Along w it h burn dept h and size, it is import ant t o not e any circumf erent ial
component s t o burns of t he ext remit ies, neck, and t orso. Such burns may lead t o
compromise of peripheral circulat ion and a
compart ment syndrome in an ext remit y or diff icult y w it h vent ilat ion in t he neck or
t orso.
I t is very diff icult t o predict mort alit y in burned pat ient s. Age older t han 60
years, f ull-t hickness burns over 40% of TBSA, and inhalat ion injury are
signif icant prognost ic f act ors. A f ormula has been proposed st at ing t hat
percent age mort alit y = age + percent age TBSA burned. How ever, more recent
st udies have show n t hat t here is no reliable met hod f or predict ing survival on
admission.
Suggested Readings
Heimbach D, Engrav L, G rube B, et al. Burn dept h: a review. World J Surg
1992; 16: 10–15.
Mulholland M, Lillemoe KD, Dohert y G M, et al. G reenf ield's Surgery:

Scient if ic Principles and Pract ice, 4t h ed. Philadelphia: Lippincot t Williams &
Wilkins, 2005.
Saff le JR, Davis B, Williams P, et al. Recent out comes in t he t reat ment of
burn injury in t he Unit ed St at es: a report f rom t he American Burn Associat ion
Pat ient Regist ry. J Burn Care Rehabil 1995; 16(3 Pt 1): 219–232; discussion
288–289.
> Table of C ontents > Mis c ellanc eous > 304 - D o not Talk to Fam ilies about O r gan D onation
304
Do not Talk to Families about Organ Donation
Matthew J. Weiss MD
Solid-organ t ransplant at ion is t he t herapy of choice f or pat ient s w it h end-organ
f ailure. Despit e numerous advances in t he f ield, t he number of individuals
aw ait ing t ransplant at ion cont inues t o exceed t he supply of organs. O ne approach
t o increasing t he supply of organs is t o maximize donat ion rat es of eligible
pat ient s. Family consent rat es f or donat ion are current ly only 30% t o 40%.
Alt hough f amilies decline donat ion f or many reasons, it is indisput able t hat t here
are myt hs believed t o be t rue by f amilies t hat do not aid in increasing t he
donat ion rat e: pat ient s w ill be premat urely declared dead t o “get bet t er organs, ”
t he organ allocat ion syst em is corrupt , and rich people “get t he good organs. ”
Part icularly damaging is t he percept ion t hat t he physician caring f or a loved one
may have allegiance t o a f ut ure organ recipient . This can compromise t he t rust
necessary f or increasing donat ion rat es. Theref ore, it is import ant t hat
physicians caring f or t he pot ent ial donor do not discuss donat ion w it h t he f amily.
During t he last hours of lif e, t he physician caring f or t he pot ent ial donor must
cont inue t o provide aggressive care. The donor f amily must underst and and
believe t hat everyt hing has been done f or t heir loved one.
How ever, all f amilies of pot ent ial donors must be approached f or consent . For
heart -beat ing donors, discussions should be init iat ed in a privat e set t ing as early
as possible af t er declarat ion of brain deat h. A member of t he hospit al t eam
should be present t o answ er medical quest ions af t er and a t rained organ
procurement organizat ion (O PO ) represent at ive should approach t he f amily.
Many st at es have enact ed legislat ion t hat requires healt h prof essionals t o not if y
t heir regional O PO of pot ent ial candidat es and very likely t he O PO
represent at ive w ill be f amiliar w it h t he possible donor several hours bef ore deat h
is declared. O PO represent at ives are t rained t o eff ect ively communicat e w it h
donor f amilies, and st udies clearly demonst rat e t heir success in increasing
procurement rat es over t hose of physicians on t he care t eam. I n addit ion, t hese
represent at ives exercise innovat ive approaches t o minorit y f amilies, w ho are
hist orically less likely t o agree t o donat ion.
Suggested Readings
Razek T, O lt hoff K, Reilly PM. I ssues in pot ent ial organ donor management .
Surg Clin Nort h Am 2000; 80: 1021–1032.
Siminoff LA, G ordon N, Hew let t J, et al. Fact ors inf luencing f amilies' consent
f or donat ion of solid organs f or t ransplant at ion. JAMA 2001; 286(4): 71–77.
> Table of C ontents > Mis c ellanc eous > 305 - Aler t the Tr ans plant Team E m er gently if Ther e is an
Ac ute D ec r eas e in Ur ine O utput after a K idney Tr ans plant
305
Alert the Transplant Team Emergently if There is
an Acute Decrease in Urine Output after a
Kidney Transplant
Matthew J. Weiss MD
The post operat ive management of a kidney t ransplant recipient requires
met iculous monit oring of urine out put and elect rolyt e levels. Renal allograf t s w ill
not necessarily make urine as soon as t hey are reperf used. Depending on t he
part icular cent er, t he incidence of delayed graf t f unct ion ranges f rom 5% t o 15%
in cadaveric and f rom 0% t o 5% in live donor kidney t ransplant at ion. The ast ut e
clinician w ill query t he pat ient as t o how much urine w as produced pret rans-plant
so as not t o be lulled int o a f alse assurance t hat t he graf t is f unct ioning because
“t here is urine. ”
Alt hough no physician can predict t he post operat ive urine out put of a kidney
recipient w it h 100% cert aint y, t here are several f act ors t hat consist ent ly
cont ribut e t o delayed graf t f unct ion. Cadaveric donors, increased age of donors,
et hnicit y of donors, diabet ic donors, prolonged w arm and cold ischemia, and
ischemia/ reperf usion injury are all risk f act ors f or delayed graf t f ailure result ing
f rom acut e t ubular necrosis. I t must be remembered t hat delayed graf t f unct ion
begins at t he t ime of organ reperf usion and t he urine out put begins and remains
low and does not change abrupt ly.
Acut e changes in urine out put in a f unct ioning graf t can be a signal t hat a
cat ast rophic insult has occurred t o t he graf t and must be report ed t o t he
t ransplant t eam immediat ely. Several causes of acut e decreases in urine out put
are surgical complicat ions and are correct able if prompt at t ent ion is
appropriat ed. A list of t echnical causes of low urine out put includes renal
art ery/ vein t hrombosis or kinking, uret eral obst ruct ion, and compression of t he
graf t by a f luid collect ion (hemat oma, lymphocele, seroma, or urinary leak).
Hypovolemia can also cause oliguria as in ot her post operat ive pat ient s. Vascular,
urologic and lymphat ic complicat ions require prompt diagnosis and t reat ment t o
salvage t he graf t . The w orkup of decreased urine out put of t en depends on
inst it ut ion and t ransplant t eam, but usually consist s of ult rasound, color f low
Doppler, or radionuclide imaging of t he graf t . An ult rasound is invaluable in t his
sit uat ion because it can be perf ormed quickly at t he bedside and can evaluat e
f or bot h vascular compromise and f luid collect ions. How ever, t he call t o t he
t ransplant t eam must not be delayed w hile w ait ing f or ult rasound result s.
Suggested Readings
Halloran PF, Hunsicker LG . Delayed graf t f unct ion: st at e of t he art , November
10–11, 2000. Am J Transplant 2001; 1: 115–120.
Humar A, Leone JP, Mat as AJ. Kidney t ransplant at ion: a brief review. Front
Biosci 1997; 2: 41–47.
> Table of C ontents > Mis c ellanc eous > 306 - R em em ber that C ar diac O utput is not the S am e Thing as
E jec tion Fr ac tion
306
Remember that Cardiac Output is not the Same
Thing as Ejection Fraction
Frank Rosemeier MD
Cardiac out put equals t he volume of blood pumped by t he heart per minut e,
w hereas st roke volume (SV) is t he amount pumped on a single beat . Cardiac
out put can be det ermined using indicat or dilut ion met hods (Fick, t hermodilut ion),
Doppler velocit y dat a, vent ricular impedance, and radionucleot ide met hods.
Eject ion f ract ion (EF) is measured in percent : EF (%) = (SV/ EDV) × 100%;
w hereas SV is calculat ed as SV = EDV - ESV, w it h EDV t he end-diast olic volume
and ESV t he end-syst olic volume. Eject ion f ract ion is t ypically est imat ed using
qualit at ive t w o-dimensional echocardiography. How ever, t he accuracy depends
on t he observer experience. I nadequat e def init ion of t he endocardial border and
regional w all mot ion abnormalit ies w it h asymmet ric vent ricular cont ract ion may
lead t o inaccuracies, w hich may be overcome by various yet t ime-consuming
vent ricular volume f ormulas.
Watch Out For

Alt hough a sat isf act ory eject ion f ract ion w ould suggest a similar adequat e
cardiac out put , t his relat ionship does not alw ays hold t rue. I n mit ral
regurgit at ion, eject ion f ract ion can be excellent , yet part of t he lef t vent ricular
end-diast olic volume is eject ed back int o t he lef t at rium via t he incompet ent
mit ral valve during vent ricular syst ole. I n t his scenario, eject ion f ract ion can be
normal, yet cardiac out put is diminished. I n cont rast , cardiac out put is
supranormal in aort ic incompet ence, but part of t he f orw ard st roke volume
regurgit at es back int o t he lef t vent ricle during vent ricular diast ole. As t he
disease progresses, heart f ailure w ill develop w it h a reduct ion of eject ion
f ract ion pseudonormalizing cardiac out put , event ually reaching a point at w hich
bot h hemodynamic paramet ers are severely depressed. Regurgit ant volume and
regurgit ant f ract ion are used in grading t he severit y of t hese valvular lesions,
adding more valuable inf ormat ion t o t he assessment of cardiac f unct ion.
Temperat ure t hermodilut ion met hods f or cardiac out put det erminat ion remain
accurat e because t hese are t ypically measured in t he pulmonary art ery using a
Sw an-G anz cat het er and right -sided cardiac out put must equal lef t -sided cardiac
out put .
Suggested Readings
Fink MP, Abrahams E, Vincent J, et al. , eds. Text book of Crit ical Care, 5t h
ed. Philadelphia: Elsevier Saunders, 2005: 735–740.
Savage RM, Aronson S, eds. Comprehensive Text book of I nt raoperat ive

Transesophageal Echocardiography. Philadelphia: Lippincot t Williams &
Wilkins, 2005: 129–146.
> Table of C ontents > Mis c ellanc eous > 307 - D o Not “ R oc k the P elvis ” in a Fr ac tur e
307
Do Not “Rock the Pelvis” in a Fracture
J. Christopher DiG iacomo MD
Pelvic f ract ures f rom blunt f orce t rauma are broadly cat egorized as eit her
“st able” or “unst able” based on t he st at us of t he pelvic ring. The pelvis is
composed of t hree bones (t he iliopubic bones and t he sacrum), w hich are held
t oget her by st rong ligament s: t he symphysis pubis ant eriorly and t he ant erior
and post erior iliosacral ligament post erolat erally. A pelvic f ract ure is generally
considered unst able if t he ring is broken in t w o or more places.
The unst able pelvic f ract ure is somet imes ident if ied during t he secondary survey,
w hen t he pelvis is not ed t o “move” during eit her ant eropost erior or lat eral
compression on physical examinat ion. More commonly, t oday it is det ect ed on
plain radiograph or abdominal comput ed t omography (CT). How ever an unst able
pelvic f ract ure is ident if ied, once t he diagnosis has been made, all f urt her
manual manipulat ion (i. e. , “rocking t he pelvis”) must cease. Alt hough it is
t empt ing t o use an unst able pelvic exam as a prime t eaching point , t he unst able
pelvis should not be manipulat ed. Each t ime t he damaged joint is moved, t he clot
may be disrupt ed, result ing in addit ional bleeding f rom t he t orn veins and bony
surf aces. Consumpt ion coagulopat hy may occur as coagulat ion f act ors are used
f or new clot f ormat ion.
Unst able pelvic f ract ures include disrupt ion of t he ring ant eriorly at t he
symphysis pubis or t hrough a pubic bone and are broadly cat egorized int o t hree
t ypes: ant erior-post erior (open-book) f ract ure, lat eral compression f ract ure, and
vert ical shear (Malgaigne) f ract ure.
Anterior-Posterior (Open-Book) Fracture

This t ype of injury t ypically occurs af t er an ant erior-post erior blunt f orce injury t o
t he pelvis, in w hich t he ant erior ring f ract ure is accompanied by disrupt ion of t he
ant erior iliosacral ligament s (Fig. 307. 1). Wit h t he st abilit y of t he ant erior ring
lost , t he pelvis opens ant eriorly and lat erally and t he iliac bones hinge on t he
post erior pelvic ligament s, much as a book opens around it s spine. This disrupt s
t he iliosacral joint and t he veins t hat lie ant erior t o t he iliosacral joint . This
result s in a large amount of venous and bony bleeding int o t he
ret roperit oneal space. Approximat ely 10% of pat ient s have associat ed art erial
bleeding. I nit ial st abilizat ion should be in t he f orm of ext ernal binding of t he
pelvis w it h a bed sheet t ied t ight ly around t he pelvis or a pelvic binder device.
FIG URE 307. 1. Ant erior-post erior (open-book) f ract ure.
Lateral Compression Fracture

This t ype of injury t ypically occurs af t er lat eral blunt f orce injury t o t he pelvis,
such as a in a pedest rian st ruck by a car (Fig. 307. 2). The
ant erior ring f ract ure is accompanied by disrupt ion of t he post erior iliosacral
ligament s. The iliac bones hinge on t he ant erior iliosacral ligament s. Venous and
bony bleeding f rom t he disrupt ed iliosacral joint t ends t o be ext rapelvic. I n t his
t ype of f ract ure, at t empt ed st abilizat ion w it h a sheet or ext ernal binder w ill only
promot e f urt her inw ard collapse of t he iliac bones and cause increased bleeding.
Ext ernal f ixat ion of t he pelvis or open reduct ion and int ernal f ixat ion (O RI F) is
pref erable.
FIG URE 307. 2. Lat eral compression f ract ure.
Vertical Shear (M algaigne) Fracture

This t ype of injury t ypically occurs w hen a large vert ical load is placed on one
leg, such as in a f all f rom height (Fig. 307. 3). The iliac bones are complet ely
separat ed f rom t he remainder of t he pelvis by disrupt ion of t he ant erior ring and
bot h t he ant erior and post erior iliosacral ligament s. Venous and bony bleeding is
not cont ained w it hin t he pelvis, and art erial bleeding must be ruled out . I nit ial
st abilizat ion requires t ract ion on t he injured side leg t o reduce t he vert ical
displacement , f ollow ed by ext ernal binding of t he pelvis. This f ract ure t ype
carries t he highest mort alit y rat e due t o exsanguinat ion.
The init ial management of pelvic f ract ures f ollow s t he paradigm of t he ABCs of
t rauma resuscit at ion. Af t er t he airw ay (A) and breat hing (B) are considered,
circulat ion and cont rol of hemorrhage (C) are
imperat ive. Adequat e volume resuscit at ion is essent ial, and blood should be
administ ered early t o t he hemodynamically unst able or met ast able pat ient . The
pelvic x-ray w ill provide essent ial det ails about t he bony injury t hat are crucial
early in t he resuscit at ion phase. Trauma and ort hopedic surgical consult at ions
should be obt ained f or immediat e evaluat ion. Abdominal and pelvic CT scans w it h
int ravenous cont rast are perf ormed t o assess f or ot her injuries and f urt her
def ine t he pelvic f ract ure. I t is import ant t hat t he pelvic port ion be review ed by
t he physician in at t endance of t he pat ient immediat ely, looking f or art erial
ext ravasat ion of int ravenous cont rast . The f inding of art erial ext ravasat ion
w arrant s immediat e art eriography w it h t herapeut ic angioembolizat ion. O perat ive
at t empt s t o cont rol such bleeding are usually unsuccessf ul. Bleeding f rom ot her
solid-organ injuries (e. g. , liver, spleen, kidney) can also be addressed at t he
t ime of angiography.
FIG URE 307. 3. Vert ical shear (Malgaigne) f ract ure.
I f laparot omy is necessary, an ext ernal f ixat or should be applied prior t o t he

abdominal procedure. The bracing arms should be direct ed t ow ard t he f eet
inst ead of being placed in t he usual posit ion t ow ard t he head (overlying t he
low er abdomen). O rt hopedic surgeons w ill somet imes object because t his
int erf eres w it h t he pat ient sit t ing up. How ever, t he pat ient w ill not be sit t ing up in
t he near f ut ure, and abdominal exposure is more urgent . Def init ive f ixat ion can
be accomplished by revising t he bracing arms or perf orming int ernal f ixat ion
w hen t he pat ient is st able.
Suggested Readings
DiG iacomo JC, Bonadies JA, Cole FJ, et al. Pract ice management guidelines
f or hemorrhage in pelvic f ract ure. The East ern Associat ion f or t he Surgery of
Trauma. ht t p: / / w w w. east . org/ t pg.
DiG iacomo JC, McG onigal MD, Haskal ZJ, et al. Art erial bleeding diagnosed
by CT in hemodynamically st able vict ims of blunt t rauma. J Trauma
1996; 40(2): 249–252.
Moore EE, Feliciano DV, Mat t ox KL. Trauma. New York: McG raw -Hill,
2003: 779– 807.
> Table of C ontents > Mis c ellanc eous > 308 - C ons ider Tr eatm ent for Heter otopic O s s ific ation After
Tr aum a
308
Consider Treatment for Heterotopic Ossification
After Trauma
Constantine A. Demetracopoulos BS
Introduction
Het erot opic ossif icat ion may occur f ollow ing head or spinal cord t rauma, burns,
or major surgery on t he musculoskelet al syst em. A het erot rophic bone may
cause severe physical impairment by surrounding a joint such as t he hip or t he
elbow and f ixing t he joint in a nonf unct ional posit ion (i. e. , cont ract ure).
Het erot opic ossif icat ion is also very common f ollow ing open reduct ion and
int ernal f ixat ion of acet abular f ract ures.
Pathophysiology
Het erot opic ossif icat ion is t he process of bone f ormat ion in ext raskelet al sof t
t issue. Robert s in 1968 w as f irst t o not e t he associat ion bet w een head t rauma
and periart icular ect opic bone f ormat ion, and ot hers have since demonst rat ed a
posit ive correlat ion bet w een t he ext ent of head t rauma and f unct ional severit y of
het erot opic ossif icat ion. The pat hogenesis of het erot opic ossif icat ion f ollow ing
head injury is uncert ain. How ever, recent t heories suggest t hat head t rauma
causes an increase in inf lammat ory mediat ors and bone morphogenic prot eins,
w hich induce sof t t issue ost eoprogenit or st em cells t o diff erent iat e int o
ost eoblast s. I ncreased serum levels of prost aglandin E2 af t er head t rauma are
also believed t o play a role in t he prolif erat ion of diff erent iat ed ost eoblast s. The
ost eoblast s secret e ost eoid peripherally about t he area of t rauma w it hin 7 t o 10
days. Primit ive cart ilage appears af t er 14 days, and t rabecular bone f orms
w it hin 2 t o 5 w eeks. Mat ure lamellar bone in t he periphery surrounding immat ure
and undiff erent iat ed cent ral t issues is present by t he sixt h w eek. Plain
radiographs and comput erized t omography scans show a very charact erist ic
zoning pat t ern of mat ure bone at t he periphery and a lucency in t he cent er. The
lucent area represent s immat ure musculoskelet al t issues.
Clinical Features
The hip is t he most commonly involved joint , f ollow ed by t he elbow, shoulder, and
knee. The incidence of het erot opic ossif icat ion of t he
hip f ollow ing head t rauma varies f rom 11% t o 76%, w it h ankylosis of t he joint
occurring in 11% t o 20% of t hose pat ient s. Clinical signs t hat should cause one
t o suspect het erot opic ossif icat ion include pain and decreased range of mot ion
at t he joint , as w ell as inf lammat ory markers such as f ever, eryt hema, sw elling,
and w armt h at t he joint . Making t he diagnosis is also dependent on ident if ying
pat ient s w ho are more likely t han ot hers t o develop het erot opic ossif icat ion. Risk
f act ors include male gender, underlying syst emic disorders such as ankylosing
spondylit is and diff use idiopat hic skelet al hyperost osis, previous het erot opic
bone f ormat ion, and limb spast icit y.
The diagnosis of het erot opic ossif icat ion is t ypically made based on clinical
suspicion and radiographic f indings (Fig. 308. 1), w hich appear w it hin t he f irst 3
t o 6 w eeks af t er injury. Brooker devised t he classif icat ion syst em t hat is most
f requent ly used t o radiographically grade t he amount of het erot opic ossif icat ion
at t he hip. Class I includes islands of bone w it hin t he sof t t issue t hat is locat ed
about t he hip. Class I I consist s of bone spurs f rom t he pelvis or proximal f emur
leaving at least 1 cm bet w een opposing bone surf aces. Class I I I allow s f or less
t han 1 cm bet w een opposing bone surf aces, and class I V is apparent ankylosis
of t he hip in t he ant erior-post erior view. Triple-phase radionucleot ide bone scan
is t he pref erred met hod f or early det ect ion. I t w ill show increased blood f low and
increased concent rat ion of t he t racer w it hin t he sof t t issue of t he hip bef ore any
f indings are apparent on radiographs. Comput ed t omography is most commonly
used t o def ine t he areas of involvement . Alt hough serum alkaline phosphat ase
does not have diagnost ic or prognost ic value, levels correlat e w it h t he degree of
ongoing ossif icat ion, and t hus it used as a marker f or act ive disease.
Treatment
Nonst eroidal ant i-inf lammat ory drugs (NSAI DS), ext ernal beam radiat ion, and
disphosphonat es have been used w it h varying success f or prophylaxis against
het erot opic ossif icat ion. I n a randomized, double-blind, prospect ive st udy, 400
mg of ibuprof en t hree t imes daily administ ered w it hin 48 hours of t he injury and
given f or 8 days w as f ound t o be eff ect ive. I ndomet hacin has also been used f or
prophylaxis and is dosed at 75 t o 100 mg/ day f or 7 t o 14 days. How ever, pat ient
compliance w it h NSAI Ds is of concern because of t he incidence of
gast roint est inal dist ress. I onizing radiat ion has been show n t o inhibit
ost eogenesis. A recent met a-analysis comparing single-dose radiat ion w it h
NSAI Ds f ound radiot herapy t o be more eff icacious.
FIG URE 308. 1. Charact erist ic radiographic appearance of het erot rophic
ossif icat ion of t he knee. (Reprint ed w it h permission f rom Harris JH Jr. Knee.
I n: Harris JH Jr. , Harris WH, eds. The Radiology of Emergency Medicine, 4t h
ed. Philadelphia: Lippincot t Williams & Wilkins, 2000: 836.)
A one-t ime dose of 700 t o 800 cG y is recommended w it hin 24 hours of t rauma or

72 hours post operat ively. Alt hough t he risk of inducing malignant disease w it h
radiot herapy exist s, it is uncommon at such low doses. Finally, disphosphonat es
are pyrophosphat e analogues t hat delay mineralizat ion of ost eoid by inhibit ing
t he grow t h of hydroxyapat it e cryst als. They have not been show n t o be eff ect ive
and carry t he risk of ost eomalacia w it h long-t erm use.
Def init ive t reat ment of het erot opic ossif icat ion once it has developed and
become sympt omat ic is surgical excision f ollow ed by adjuvant t herapy t o
decrease t he likelihood of recurrence. Excision occurs w it h t he int ent ion t o
improve f unct ion because pat ient s can become severely limit ed in t heir act ivit ies
of daily living. How ever, surgery must w ait unt il t he lesion has mat ured. Act ive,
ongoing ossif icat ion is an absolut e cont raindicat ion. Thus, excision must aw ait
t he radiographic appearance of w ell-def ined cort ex and normal serum levels of
alkaline phosphat ase. This t ypically occurs around 18 mont hs af t er
t he injury. O t her indicat ors f or successf ul excision are good cognit ive recovery
f ollow ing head t rauma and good mot or cont rol of t he ext remit y. Post operat ive
radiat ion t herapy is recommended t o prevent recurrence.
Suggested Readings
G arland DE, Blum CE, Wat ers RL. Periart icular het erot rophic ossif icat ion in
head injured adult s. I ncidence and locat ion. J Bone Joint Surg Am
1980; 62(7): 1143– 1146.
Morrey BF, ed. Joint Replacement Art hroplast y. New York: Churchill
Livingst one, 1991: 867–876.
Saraf is KA, Karat zas G D, Yot is CL. Ankylosed hips caused by het erot rophic
ossif icat ion af t er t raumat ic brain injury: a diff icult problem. J Trauma
1999; 46(1): 104–109.
> Table of C ontents > Mis c ellanc eous > 309 - Look For Mis s ed E xtr em ity Fr ac tur es in P atients w ith a
D iagnos ed E xtr em ity Fr ac tur e
309
Look For Missed Extremity Fractures in Patients
with a Diagnosed Extremity Fracture
William S. Hoff MD
Missed injuries, commonly ref erred t o as “t he t rauma surgeon's nemesis, ” are an
expect ed occurrence in t he management of mult iply injured pat ient s. The
def init ion of a missed injury is inst it ut ion specif ic. G enerally speaking, how ever,
a missed injury is an injury ident if ied at some def ined t ime af t er t he init ial
assessment . Missed injuries are not f requent ly lif e-t hreat ening. How ever,
depending on t he exact circumst ances, a missed injury may result in long-t erm
disabilit y. Moreover, missed injuries may complicat e t he relat ionship bet w een t he
healt h care provider and his or her pat ient .
The init ial assessment of a t rauma pat ient consist s of a primary and a secondary
survey. The purpose of t he primary survey is t o simult aneously ident if y and
init iat e t reat ment of immediat ely lif e-t hreat ening condit ions. The secondary
survey consist s of a syst emat ic, head-t o-t oe physical examinat ion. A def init ive
care plan is est ablished based on injuries ident if ied in t he init ial assessment .
Despit e a caref ully perf ormed init ial assessment , t he incidence of missed injury
is approximat ely 10%. The majorit y of missed injuries are musculoskelet al
injuries, a large proport ion of w hich are ext remit y injuries. The f ollow ing
condit ions are associat ed w it h missed injuries in t rauma pat ient s:
Alt ered sensorium secondary t o et hanol/ drug int oxicat ion or t raumat ic brain
injury
Clinical inst abilit y w it h more urgent t reat ment priorit ies
Unappreciat ed physical f indings on init ial assessment
Failure t o obt ain necessary radiographic st udies
I nadequat ely perf ormed or misint erpret ed radiographic st udies
From a perf ormance improvement perspect ive, t he f irst t w o condit ions are
expect ed in severely injured pat ient s in w hom a reliable and complet e clinical
examinat ion may not be possible. The remaining f act ors represent errors in
judgment or management .
What to Do
The most eff ect ive w ay t o reduce missed injuries is t o perf orm a repeat
syst emat ic physical examinat ion by an experienced physician. This “t ert iary
survey”
should ideally be perf ormed w it hin t he f irst 24 hours af t er admission or w hen t he

pat ient has been st abilized. During t his examinat ion, at t ent ion should be paid t o
any subt le areas of cont usion, abrasion, sw elling, or def ormit y. I n a conscious
pat ient , areas of t enderness should also be recorded. I n t he pat ient w ho is more
mobile, pain w it h movement or ambulat ion is not ed. Appropriat e radiographic
st udies should be obt ained based on t he result s of t he t ert iary survey. Anot her
pot ent ial source of missed injuries may result w hen t he t rauma surgeon relies
ent irely on his or her int erpret at ion of radiographic st udies obt ained during t he
init ial assessment . A recent st udy demonst rat ed a 9. 7% incidence of new
diagnoses based on mandat ory review of admission x-rays w it hin 24 hours by a
radiologist .
Because of t hese considerat ions, t he experienced and skilled int ensive care
physician w ill perf orm his or her ow n t horough physical examinat ion and
independent ly review t he images and report s of all radiology invest igat ion in all
t rauma pat ient s admit t ed t o t he int ensive care unit .
Suggested Readings
Biff l WL, Harringt on DT, Cioff i WG . I mplement at ion of a t ert iary t rauma survey
decreases missed injuries. J Trauma 2003; 54: 38–44.
Born CT, Ross SE, I annocone WM, et al. Delayed ident if icat ion of skelet al
injury in mult isyst em t rauma: t he “missed” f ract ure. J Trauma 1989; 29: 1643–
1646.
Enderson BL, Reat h DB, Meadors J, et al. The t ert iary t rauma survey: a
prospect ive st udy of missed injury. J Trauma 1990; 30: 666–670.
Enderson BL, Maull KI . Missed injuries—t he t rauma surgeon's nemesis. Surg

Clin Nort h Am 1991; 71: 399–418.
Hoff WS, Sicout ris CP, Lee SY, et al. Formalized radiology rounds: t he f inal
component of t he t ert iary survey. J Trauma 2004; 56: 291–295.
> Table of C ontents > Mis c ellanc eous > 310 - Have a W or king K now ledge of Intens ive C ar e Unit
S c or ing S ys tem s
310
Have a Working Knowledge of Intensive Care
Unit Scoring Systems
Ron Pauldine MD
Severit y-of -illness scoring syst ems are used in t he int ensive care unit (I CU)
set t ing as t ools t o st rat if y risk f or t herapeut ic t rials or as predict ive inst rument s
t o evaluat e t he perf ormance of a specif ic I CU over t ime or in comparison t o
anot her unit . I t must be not ed at t he out set t hat t hey have limit ed ut ilit y w it h
regard t o individual pat ient out comes.
I nt ensive care unit scoring syst ems f all int o t w o broad cat egories. Risk
predict ion (also called “prognost ic”) models are designed t o assess prognosis at
t he t ime of admission t o t he I CU or w it hin t he f irst 24 hours. O rgan dysf unct ion
scores are used t o quant if y t he burden of organ dysf unct ion and can be
perf ormed repeat edly t hroughout t he I CU st ay. Several generalizat ions can be
made w hen comparing and cont rast ing t he applicat ion of prognost ic and organ
dysf unct ion scoring syst ems. Prognost ic scoring syst ems are concerned w it h
predict ing mort alit y, w hereas organ dysf unct ion scores describe morbidit y.
Prognost ic scores are obt ained at t ime of admission or w it hin t he f irst 24 hours
of t he I CU st ay and use a relat ively complex set of calculat ions. O rgan
dysf unct ion measures are usually easier t o use and can be obt ained repeat edly.
Prognost ic scores reveal no inf ormat ion on individual organ dysf unct ion, w hereas
organ dysf unct ion scores are primarily concerned w it h individual organ f unct ion.
Risk predict ion (prognost ic) models are const ruct ed f rom large dat abases
obt ained f rom pat ient s admit t ed t o hospit als t hroughout t he Unit ed St at es and
Europe. Logist ic regression met hods are applied t o est imat e t he probabilit y of
mort alit y based on dat a obt ained at or near t he t ime of I CU admission. The
accuracy of t he syst em is inf luenced by how t he crit eria w ere derived f or a
specif ic syst em and how t hey are applied t o a given pat ient . Pat ient f act ors,
issues w it h dat a collect ion, and innovat ions in t reat ment over t ime all have an
impact . The major out comes measured by t he most w idely used syst ems are
mort alit y and, in some cases, lengt h of st ay. The models assume t hat mort alit y
is aff ect ed by physiologic derangement s t hat occur early in t he course of illness.
The Acut e Physiology and Chronic Healt h Evaluat ion (APACHE) w as t he f irst risk
predict ion model described. APACHE I included 34 physiologic variables derived
f rom t he medical
lit erat ure and expert opinion. I t suff ered f rom complexit y and diff icult y in
applicat ion. APACHE I I w as designed t o be more user-f riendly and included 12
physiologic variables. The score w as validat ed in 5, 815 I CU admissions at 13
diff erent cent ers. Limit at ions include t he age of t he dat abase (out comes are
based on t reat ment f rom 1979 t o 1982), lack of applicabilit y t o individual
pat ient s, and select ion bias. APACHE I I I is an updat ed syst em t hat addresses
some of t hese issues but is more diff icult t o use and ut ilizes propriet ary
sof t w are. The Simplif ied Acut e Physiology Score (SAPS) w as int roduced in 1984
and has been updat ed as SAPS I I . SAPS I I uses dat a f rom 1991 t o 1992
collect ed in I CUs in t he Unit ed St at es and Europe. Sevent een variables
represent ing physiology, t ype of admission, and underlying disease comprise t he
model.
O rgan dysf unct ion scoring syst ems allow det erminat ion of organ dysf unct ion at
t ime of admission and at regular int ervals t hroughout t he I CU st ay. They allow
assessment of t he change in organ f unct ion, and t he accuracy of mort alit y
predict ions may be improved w it h repeat measurement s. The benef it of t his t ype
of scoring syst em lies in it s abilit y t o capt ure change. This t ype of assessment
may be usef ul in comparing a baseline level of f unct ion at t ime of ent ry int o a
clinical t rial or allow mort alit y predict ion bet w een groups of pat ient s. Scores
t ypically f ocus on six organ syst ems: cardiovascular, respirat ory, hemat ologic,
cent ral nervous syst em, renal, and hepat ic. Tw o f requent ly employed organ
scoring syst ems include t he Sequent ial O rgan Failure Assessment Score (SO FA)
and t he Mult iple O rgan Dysf unct ion Score (MO DS). They diff er mainly in t he
crit eria used t o assess t he cardiovascular syst em. SO FA uses t he w orst daily
value over a represent at ive value f or t he included variables and uses urine out put
t o assess renal f unct ion. O verall, high SO FA scores and signif icant increases in
SO FA score correlat e w it h mort alit y.
Suggested Readings
Herridge MS. Prognost icat ion and int ensive care unit out come: t he evolving
role of scoring syst ems. Clin Chest Med 2003; 24(4): 751–762.
Rosenberg AL. Recent innovat ions in int ensive care unit risk-predict ion
models. Curr O pin Crit Care 2002; 8(4): 321–330.
Sociét é Franaise d'Anest hésie et de Réanimat ion. Scoring syst ems f or I CU

and surgical pat ient s. w w w. sf ar. org/ s/ art icle. php3?id art icle=60, 2006.
> Table of C ontents > Mis c ellanc eous > 311 - Have a W or king K now ledge of the E m er genc y Medic al
Tr eatm ent and Ac tive Labor Ac t as it Applies to the Intens ive C ar e Unit
311
Have a Working Knowledge of the Emergency
Medical Treatment and Active Labor Act as it
Applies to the Intensive Care Unit
Christian Merlo MD, MPH
The Emergency Medical Treat ment and Act ive Labor Act (EMTALA) w as passed
by t he U. S. Congress in 1986 t o prevent hospit als f rom reject ing, ref using, or
t ransf erring pat ient s because t hey are unable t o pay or because t hey have public
healt h insurance. The main purpose of EMTALA is t o ensure nondiscriminat ory
pat ient access t o emergency medical care and t o prevent t he t ransf er of
uninsured pat ient s f rom privat e t o public hospit als w it hout considerat ion of
medical condit ion or clinical st abilit y.
Alt hough t he act has hist orically been associat ed w it h care in t he emergency
depart ment , it is import ant t o realize t hat EMTALA imposes t hree specif ic legal
dut ies on t he ent ire hospit al, including t he int ensive care unit (I CU). First ,
hospit als must perf orm a screening examinat ion on any person w ho comes t o t he
hospit al t o evaluat e w het her he or she has a medical emergency. Second, if an
emergency medical condit ion exist s, hospit al st aff must st abilize t he pat ient t o
t he best of t heir capabilit ies and t ransf er t he pat ient t o anot her hospit al if
specialized care is needed and not available at t heir ow n inst it ut ion. Finally,
hospit als w it h specialized capabilit ies are required t o accept pat ient t ransf ers if
t hey have t he capacit y t o care f or t hem.
I nt ensive care unit management can be provided at most hospit als, but cert ain
circumst ances may necessit at e specialt y care (e. g. , burns, t rauma, neonat al
I CU) t hat requires t ransf er of t he pat ient t o anot her inst it ut ion. I n t his case, t he
ref erring physician is obliged t o f irst st abilize t he pat ient . I f t he medical benef it s
of t ransf er out w eigh t he medical risks, t hen t he physician should eff ect ively
communicat e result s and t reat ment and t ransf er t he pat ient t o an accept ing
f acilit y w it h t he capabilit y t o t reat t he pat ient 's condit ion. Alt hough not required
by EMTALA, it is of t en a good idea t o f ollow up w it h t he accept ing physician
af t er t he t ransf er has occurred t o ensure proper cont inuit y of care.
The accept ing physician in t he I CU of a t ert iary care cent er also has obligat ions
under EMTALA rules and must have a good w orking know ledge of his or her
specif ic I CU w it h regard t o capacit y (beds
and st aff ing) and capabilit y. A pat ient should not be accept ed if t he I CU is f ull. A
t ransf er f or high-f requency oscillat ory vent ilat ion, ext racorporeal membrane
oxygenat ion, or cont inuous renal replacement t herapy should not be accept ed if
an I CU does not have t hese specif ic capabilit ies immediat ely available. This may
sound t rivial, but mist akes like t hese happen and compromise pat ient care. I f
beds and st aff ing are available and t he I CU can provide f or a st abilized pat ient
requiring specialized care, t hen not only it is appropriat e, but it is also required
by law t o accept t he pat ient . Again, it is f it t ing t o provide relevant f ollow -up
inf ormat ion regarding t he t ransf er or t he pat ient 's condit ion t o t he ref erring
physician. Many hospit al syst ems have a physician access line t o help f acilit at e
t his process.
Suggested Readings
EMTALA. com. w w w. emt ala. com
U. S. Depart ment of Healt h and Human Services, Cent er f or Medicare and

Medicaid Services. EMTALA. w w w. cms. hhs. gov/ EMTALA/
Williamson T, Crippen D. Do you have t o accept t his pat ient ? What int ensivist s
and hospit alist s need t o know about EMTALA. Cost Q ual 2001; 8–11.
Zibulew sky J. The Emergency Medical Treat ment and Act ive Labor Act
(EMTALA): w hat it is and w hat it means f or physicians. Proc Baylor Univ Med
Cent er 2001; 4: 339–346.
> Table of C ontents > Mis c ellanc eous > 312 - K now W hat the B as ic S tatis tic al Ter m s Mean
312
Know What the Basic Statistical Terms Mean
Peter F. Cronholm MD
Joseph B. Straton MD
How Clinicians M ake Decisions

To make inf ormed, evidence-based decisions, providers need t o have an
underst anding of several st at ist ical concept s. The decision-making process
begins by assessing inf ormat ion by means of a hist ory and physical examinat ion
f ramed w it hin an underst anding of t he relat ive probabilit ies of disease st at es.
The next st ep is t o det ermine t he likelihood t hat t he pat ient has t he disease in
quest ion. I f w e have enough inf ormat ion at t hat t ime, no f urt her t est ing is
needed, and w e may move on t o t reat ment . I f not , more inf ormat ion is needed,
and f urt her t est ing must be done t o bet t er det ermine w het her t he disease in
quest ion is in f act t he underlying et iology. For each st ep of t he decision-making
process, w e need t o underst and t est ing and disease charact erist ics.
Clinical decision making is based on an underst anding of t he incidence and
prevalence of diseases considered in diff erent ial diagnoses f or t he t ypes of
pat ient s considered. The prevalence of a disease t ells us how many people at a
given point or period of t ime have t he disease in quest ion. Prevalence combines
people w ho already have t he disease and t hose t hat w ill acquire t he disease
during t hat period of t ime. The incidence of a disease t ells us how many new
cases of a disease develop or are likely t o develop over a period of t ime.
I ncidence is a measure of risk, w hereas prevalence is more of a measure of t he
burden of disease f or a given populat ion. The process of developing a
diff erent ial diagnosis is a ranking of et iologies based on our underst anding of t he
incidence and prevalence of diseases f or a given set of hist orical and physical
dat a.
Decisions are made based on t he likelihood of a disease f or a given clinical
sit uat ion. The likelihood of a disease ranges f rom nil (0%) t o absolut e (100%).
There exist s a range of likelihoods t hat vary depending on t he balance of cost s
and benef it s of provider decisions t hat det ermine w hat our next st eps should be
in t erms of making choices t o t reat , not t o t reat , or t o conduct diagnost ic
t est ing. Diagnost ic t est ing should be considered w hen t here is a diff erence
bet w een t he likelihood t hat a pat ient has t he disease and t he t hreshold at w hich
a provider chooses t o move f orw ard w it h t reat ment opt ions. Tw o t hresholds must
be considered t hat det ermine t he range over w hich diagnost ic t est ing should be
considered. The f irst is t he “t est ing t hreshold, ” w hich is t he likelihood below
w hich a provider w ould consider t he disease suff icient ly rare a cause t hat he or
she w ould not empirically t reat f or t he disorder nor t est f or t he presence of t he
condit ion. To make t reat ment decisions, providers need t o have some point at
w hich t hey w ill decide t o t reat a pat ient , given t he w eight of t he evidence w it h no
f urt her t est ing. This point is ref erred t o as t he “t reat ment t hreshold” (Fig.
312. 1).
FIG URE 312. 1. The t hreshold approach t o medical decision making.
Shaping a diff erent ial diagnosis is a process of cat egorizing t he probabilit ies of
various diseases associat ed w it h t he pat ient hist ory and exam. The probabilit y
t hat a pat ient has t he disease bef ore any diagnost ic t est ing is done is know n as
t he “prior probabilit y” of disease. Diagnost ic t est ing increases or decreases t he
likelihood t hat a person has t he disease in quest ion. The likelihood t hat person
has t he disease af t er diagnost ic t est ing is know n as t he “post erior probabilit y. ”
Diagnost ic t est ing occurs along a cont inuum of likelihoods w hen t he prior
probabilit y is low er t han t he t reat ment t hreshold but higher t han t he t est ing
t hreshold (Fig. 312. 1). I f t he prior probabilit y is above t he t reat ment t hreshold,
one t reat s t he pat ient w it hout f urt her t est ing. I f t he prior probabilit y is below t he
t reat ment t hreshold, f urt her t est ing is required, t he result s of w hich eit her move
one aw ay f rom t he t reat ment t hreshold if t hey are negat ive or closer or over t he
t reat ment t hreshold if t hey are posit ive. A series of t est s may be necessary t o
move t he post erior probabilit y t hat a pat ient has t he disease t o t he point of
t reat ment . As an illust rat ion, if one operat es on a person only if one is 70% sure
t hat he or she has t he disease, but af t er t he hist ory and physical one is only
40% sure, t hen one needs t o do more t est ing t o move t he likelihood above 70%.
There is no need t o perf orm t est s w hen one is already over t he t reat ment
t hreshold (alt hough t hese are of t en done f or “academic” reasons) or if t he t est
w ill not provide enough evidence t o move one over t he t reat ment t hreshold if no
ot her t est ing is linked or available. For example, if a f ebrile pat ient has t ender
ant erior cervical ant erior cervical adenopat hy w it h an exudat ive pharynx and t he
absence of cough by hist ory, he or she should be t reat ed f or bact erial

pharyngit is w it hout f urt her t est ing.
How Clinicians Judge a Test

Sensit ivit y and specif icit y are charact erist ics of diagnost ic t est s. They inf orm t he
provider how t he t est w ill behave among people w it h or w it hout disease. The
sensit ivit y of a t est is a measure of how of t en t he t est w ill be posit ive w hen
t est ing people w ho have t he disease. I n cont rast , t he specif icit y of a t est
measures how of t en t he t est w ill be negat ive w hen t est ing pat ient s w ho do not
have t he disease. Highly sensit ive t est s are used f or screening because a highly
sensit ive t est , w hen negat ive, rules t he pat ient out f or t he disease
(SeNsit ive: O UT, or SNO UT). A highly specif ic t est , w hen posit ive, rules t he
pat ient in f or t he disease (SPecif ic: I N, or SPI N).
As clinicians, w hat w e really w ant t o know is, if t he t est comes back posit ive or
negat ive, w hat does t his mean f or t he pat ient : do t hey have t he disease or not ?
To answ er t his quest ion, one needs t o know t he sensit ivit y and specif icit y of t he
t est s used, but one also needs t o have a sense of t he prevalence of t he disease
among people like t he person being t est ed so t hat measures know n as posit ive
and negat ive predict ive values can be calculat ed. The posit ive predict ive value of
a t est is t he likelihood t hat a person has t he disease if t he t est is posit ive. The
negat ive predict ive value is t he likelihood t hat a pat ient does not have t he
disease if t he t est is negat ive. I f a t est comes back posit ive f or a person w ho is
very likely t o have t he disease (high prevalence), t he result is likely t o be a t rue
posit ive. How ever, if t he person is very unlikely t o have t he disease, t he result is
much more likely t o be a f alse posit ive. Similar st at ement s can be made f or
negat ive result s f or high and low prevalence condit ions.
How Clinicians know Whether their Treatments Will

Work
Measures of associat ion are means of est imat ing t he st rengt h of t he relat ionship
bet w een observed out comes and f act ors t hat may produce t he out come.
Common measures of associat ion are t he odds rat io, relat ive risk, absolut e risk
reduct ion, and number needed t o t reat . O dds rat ios are used in st udies in w hich
t he incidence (or t rue risk) of disease cannot be accurat ely assessed (case
series, case-cont rol st udies, or some ret rospect ive st udy designs). O dds rat ios
are det ermined by calculat ing t he rat io of t he odds of exposure among t he cases
compared w it h t he odds of exposure among t he cont rols. For example, if t he
rat io of t he odds of exposure t o art if icial t anning lamps among cases w it h

melanoma and cont rols is 3: 2, t he result ing odds rat io is 1. 5. Relat ive risk
specif ies t he risk of developing t he disease in t he exposed group relat ive t o
t hose w ho are not exposed and can be report ed f rom cert ain prospect ive st udy
designs and clinical t rials. Relat ive risk is t he best measure of t he associat ion
bet w een exposure and disease. How ever, odds rat ios can provide robust
est imat es of associat ion and can approximat e relat ive risk f or out comes t hat are
rare.
The absolut e risk reduct ion (ARR) is t he diff erence in disease rat es bet w een t he
exposed and unexposed groups. I t is import ant f or clinicians t o underst and t he
diff erence bet w een absolut e and relat ive risk reduct ions. For uncommon
diseases w it h high relat ive risks or common diseases w it h low or moderat e
relat ive risk reduct ions, t he absolut e risk reduct ion of an int ervent ion may be
quit e low. An int ervent ion may have a 10-f old relat ive risk reduct ion, w it h rat es
of disease near 0. 1 f or t hose w it h t he int ervent ion and 1. 0 f or pat ient s w it hout it
[ RR=(1. 0%)/ (0. 1%)=10] . How ever, t he same st udy f indings could be described
as producing an absolut e risk reduct ion of less t han 1% (ARR = 1. 0% - 0. 1% =
0. 99%) f or t hose pat ient s exposed t o t he int ervent ion.
A clinical use of absolut e risk reduct ion is t o t ake it s reciprocal, w hich is know n
as t he number needed t o t reat (NNT). The NNT f or an int ervent ion is t he number
of people w ho w ould need t o be exposed t o t he int ervent ion t o produce t he
desired out come f or one person. For example, if t he ARR is 8% w hen giving
clopidogrel t o pat ient s get t ing st ent s placed in t he set t ing of sympt omat ic
coronary art ery disease w it h t he int ent ion of reducing myocardial inf arct ion (MI ),
t he NNT w ould be 12 (NNT = 1/ ARR = 1/ 0. 08 = 12). I t f ollow s t hat w it h a NNT of
12, one w ould need t o give clopidogrel t o 12 pat ient s t o prevent 1 MI . Alt hough 1
MI w as prevent ed, 11 of t he 12 pat ient s received no benef it f rom t he
int ervent ion; how ever, all 12 pat ient s needed t o be t reat ed because one could
not predict w hich one w ould benef it in advance. There is no single number t hat
represent s a “good” NNT. The clinical impact of NNT is based on t he likelihood of
t he disease, t he cost of t he int ervent ion (medicat ions, procedures, harm
associat ed w it h int ervent ion), and t he cost of not doing t he int ervent ion (rat es
and values assigned t o pat ient morbidit y and mort alit y associat ed w it h t he
disease).
When int erpret ing st udy result s it is import ant t o underst and w hat is ref erred t o
as pow er as w ell as t ypes of errors t hat may be encount ered. The pow er of a
st udy, usually present ed as a percent age, is a measure of t he probabilit y t hat
t he null hypot hesis is reject ed w hen it
is f alse. That is, one w ill not f ind an associat ion bet w een t he dependent and
independent variables if t here is no t rue relat ionship bet w een t hem. I f t he
measures of associat ion (odds rat ios, relat ive risk, et c. ) of a st udy reach
signif icance, t hen t he pow er of t he st udy is irrelevant and t he f indings st and as
signif icant , assuming t he st udy met hods are valid. How ever, result s may not
reach signif icance eit her because t here is no t rue relat ionship bet w een t he
dependent and independent variables or t he st udy did not have enough pow er
(usually an issue of sample size) t o demonst rat e t he relat ionship (Fig. 312. 2).
FIG URE 312. 2. A. Measuring t he operat ing charact erist ics of a diagnost ic
t est . B. O f moderat e-prevalence diseases. C. O f low -prevalence diseases.
D. O f high-prevalence diseases. FN, f alse negat ive; FP, f alse posit ive; NPV,
negat ive predict ive value; PPV, posit ive predict ive value; TN, t rue negat ive;
TP, t rue posit ive.
Type I errors represent t he chance t hat t he null hypot hesis is reject ed w hen it is
act ually t rue, or t hat one f inds a result t hat is signif icant by chance alone and
t here is no t rue underlying relat ionship bet w een t he dependent and independent
variables. This is t he rat e of f alse alarms or f alse posit ives. Type I errors are
t he equivalent t o t he “signif icance” level report ed in st udies (e. g. , p values <. 05).
Type I I errors are t he chance t hat one does not reject t he null hypot hesis w hen
it is f alse. Type I I errors are t he complement of pow er (t ype I I error rat e = 1 -

pow er). Type I I errors are t he chance t hat one w ill miss an eff ect w hen it is
really t here. I n ot her w ords, it is t he rat e of f ailed alarms or f alse negat ives.
Suggested Readings
Cent re f or Healt h Evidence. Users’ G uides t o Evidence-Based Pract ice.
ht t p: / / w w w. cche. net / usersguides/ main. asp
Hennekens CH, Buring JE, Mayrent SL. Epidemiology in medicine. Bost on:
Lit t le, Brow n, 1987.
Rot hman KJ, G reenland S. Modern epidemiology, 2nd ed. Philadelphia:

Lippincot t -Raven, 1998.
Sw inscow TDV, Campbell MJ. St at ist ics at square one, 10t h ed. London: BMJ
Books, 2002.
> Table of C ontents > Mis c ellanc eous > 313 - C ons ider E m boli W hen Ther e is a C hange in Mental
S tatus After an Invas ive P r oc edur e
313
Consider Emboli When There is a Change in
Mental Status After an Invasive Procedure
Nirav G . Shah MD
Embolism is an uncommon complicat ion of alt ered ment at ion af t er invasive
procedures, but it must be placed on t he list of diff erent ial diagnoses. The f our
most common emboli seen in t he crit ical care pat ient populat ion are air, f at ,
cholest erol, and blood emboli (clot s).
Air embolism occurs w hen air ent ers t he syst emic vascular circulat ion during
placement or removal of a cent ral venous cat het er or at ot her t imes w hen a
cent ral venous cat het er is in use. Air can also be int roduced int o t he syst emic
circulat ion during cardiac surgery procedures, neurosurgical procedures, and
endoscopic procedures. Air t hat ent ers t he venous circulat ion can t ravel t o t he
right side of t he heart and cause cardiopulmonary compromise. The clinical
present at ion of a venous air embolism is highly variable and depends on t he
amount of air ent ry, t he speed of ent ry, and t he pat ient 's size and premorbid
condit ion. Sympt oms may range f rom mild chest discomf ort t o alt ered ment at ion
t o complet e cardiovascular collapse. Alt ernat ively, a right -t o-lef t shunt can result
in an embolus t hat ent ers t he art erial circulat ion and causes cerebral ischemia. I f
a cerebral air embolus is suspect ed, t hen t he head should be immediat ely
low ered, a f ract ion of inspired oxygen of 1. 0 should be delivered, and adequat e
vent ilat ion maint ained. O ne t herapy f or air embolus is hyperbaric oxygen t herapy,
w hich if desired, should be inst it ut ed w it hin 5 h of neurologic or cardiac
sympt oms.
Fat emboli are most commonly t he result of t rauma or long-bone f ract ure.
How ever, t hey can also present f ollow ing ort hopedic procedures such as spine
surgery and knee and hip replacement s. The classic f indings include pet echiae,
dyspnea, and alt ered ment al st at us. The sympt oms usually occur 1 t o 2 days
af t er t he precipit at ing event and are t he result of a diff use vasculit is secondary
t o f ree f at t y acids. Alt hough t his ent it y has a classic appearance, t he diagnosis
is one of exclusion. The t reat ment of f at emboli is support ive and includes
oxygen t herapy and mechanical vent ilat ion if needed.
Anot her uncommon cause of alt ered ment al st at us, especially f ollow ing cardiac
cat het erizat ion or art eriography, is a cholest erol embolus. St roke may result
f rom t he embolizat ion of at herosclerot ic mat erial
t hat is dist urbed by t he cat het er during t he procedure. The syndrome may
present w it h live do ret icularis, severe limb pain, renal f ailure, or f ocal
neurologic def icit s. The neurologic def icit s usually present acut ely and are
f requent ly reversible. I f skin changes are present , t he diagnosis can be
conf irmed w it h skin biopsy demonst rat ing cholest erol cryst als. I n addit ion,
emboli t o t he kidney may cause renal dysf unct ion and emboli t o t he viscera may
be a cause of ischemic bow el.
The most f requent embolic mat erial af t er invasive procedures is blood. Clot s t hat
subsequent ly embolize occur secondary t o int imal disrupt ion and venous st asis
f rom immobilit y. The t ime of onset can be variable, as can t he present ing
sympt oms. O rt hopedic surgery and neurosurgery are associat ed w it h t he highest
risks of venous emboli secondary t o t he release of t issue f act or, w hich is a
pow erf ul t rigger of blood clot t ing. Alt hough deep vein t hrombosis and pulmonary
emboli are common in t he post surgical populat ion, cerebrovascular accident s due
t o venous emboli require communicat ion bet w een t he lef t and right heart via a
sept al def ect or pat ent f oramen ovale (t he classic paradoxical emboli). The
clinical manif est at ions of cerebral emboli depend on t he specif ic vessel occluded
and t he dist ribut ion of blood f rom t hat vessel. For example, occlusion of t he
middle cerebral art ery result s in cont ralat eral hemiparesis and sensory loss w it h
more severe sympt oms in t he arms and f ace. Treat ment of a t hrombot ic st roke
includes t hrombolyt ic t herapy, blood pressure regulat ion, support ive care, and
aggressive physical t herapy. Mult iple st udies have demonst rat ed t he
eff ect iveness of t hrombolyt ic t herapy if inst it ut ed w it hin t he f irst 3 hours of
sympt om onset and af t er a noncont rast comput ed t omography scan excludes an
int racranial hemorrhage.
Suggested Readings
Blacker DJ. I n-hospit al st roke. Lancet Neurol 2003; 2(12): 741–746.
Cramer SC. Pat ent f oramen ovale and it s relat ionship t o st roke. Cardiol Clin
2005; 23(1): 7–11.
> Table of C ontents > Mis c ellanc eous > 314 - K now the Noninfec tious C aus es of Fever in the Intens ive
C ar e Unit
314
Know the Noninfectious Causes of Fever in the
Intensive Care Unit
Laith Altaweel MD
Fever in t he int ensive care unit (I CU) pat ient is a common problem t hat result s in
t he perf ormance of many diagnost ics t est s. This increases bot h t he cost s of
medical care and t he exposure of t he pat ient t o uncomf ort able procedures. The
ast ut e clinician w ill recognize t hat alt hough a common cause of f ever in t he I CU
is inf ect ion, many cases of f ever are caused by noninf ect ious et iologies. Proper
management demands implement ing a simult aneous algorit hm f or t he w orkup of
noninf ect ious causes.
Fever is t hought t o be a prot ect ive mechanism against inf ect ion. Many animal
species are know n t o develop f ever in response t o a microbiologic organism. I n
humans, f ever is t hought t o enhance several paramet ers of immune f unct ion and
pot ent ially enhance survival. I n addit ion, hypert hermia also increases cardiac
out put , oxygen consumpt ion, carbon dioxide product ion, and energy expendit ure,
w hich may be harmf ul t o pat ient s w it h low cardiopulmonary reserve or cere-
brovascular injury. Mat ernal f ever may also be a cause of f et al malf ormat ions or
spont aneous abort ions.
Normal body t emperat ure is 37. 0°C w it h circadian variat ion of bet w een 0. 5°C
and 1. 0°C. Several met hods are used t o measure body t emperat ure. Taking t he
pulmonary art ery mixed venous t emperat ure is t he most accurat e met hod f or
measuring core body t emperat ure. I nf rared ear t hermomet ry is nearly equivalent
t o t aking pulmonary art ery and brain t emperat ures. Rect al t emperat ures
obt ained w it h a mercury t hermomet er or elect ronic probe are of t en a f ew t ent hs
of a degree higher t han core body t emperat ure. O ral t emperat ures can be
inf luenced by drinking and eat ing or w armed air in vent ilat or circuit s. Axillary
measurement s are unreliable. The Societ y of Crit ical Care Medicine def ines a
f ever as a t emperat ure great er t han 38. 3°C. Most inf ect ious causes of f ever
f ollow a diurnal pat t ern.
Watch Out For

There are many noninf ect ious causes of f ever. Wit h t he except ion of drug f ever
and t ransf usion react ion, noninf ect ious causes of f ever usually do not lead t o a
f ever great er t han 39. 8°C. Most causes of noninf ect ious f ever (Table 314. 1) w ill
be suggest ed based on a good hist ory and
physical exam. A recent operat ion, chest pain w it h elect rocardiogram (ECG )
changes, and a quadriparet ic pat ient w it h asymmet ric low er ext remit y sw elling or
a large sacral decubit us ulcer are all commonly encount ered scenarios causing
f ever in t he I CU pat ient . Rout ine laborat ory t est s, such as a complet e blood
count t o look f or evidence of bleeding, leukocyt osis, and eosinophilia; a liver
panel and an amylase and lipase t est ; ECG w it h cardiac enzymes; and a check
of lact ic acid level can help t o quickly rule in or rule out many of t he most
common noninf ect ious causes of f ever. I maging st udies such as a chest
radiograph; comput ed t omography of t he head, chest , abdomen, and pelvis; right
upper quadrant ult rasound; and low er ext remit y venous duplex may also be
necessary t o complet e t he w orkup.
TABLE 314-1 NONINFECTIOUS CAUSES OF FEVER IN

THE INTENSIVE CARE UNIT
Alcohol/drug withdrawal
Acalculous cholecystitis
Postoperative fever (48 hours postoperative)
Decubitus ulcer
Drug fever
Cerebrovascular accident/subarachnoid hemorrhage
Myocardial infarction
Pancreatitis
Ischemic bowel
Cirrhosis
Gastrointestinal bleeding
Hematoma
Aspiration pneumonitis
Acute respiratory distress syndrome (both acute and

fibroproliferative phase)
Deep vein thrombosis/pulmonary embolism
Phlebitis/thrombophlebitis
Adrenal insufficiency
Intravenous contrast reactions
Neoplastic fevers
Drug w it hdraw al f ever should alw ays be considered in a pat ient w it h a hist ory of
subst ance abuse. I n many cases t his abuse hist ory may not be know n t o t he
clinician, so clinical suspicion should be high. Drugs are a commonly considered
et iology of f ever; in realit y very f ew cases are cit ed in t he lit erat ure. Drugs
commonly associat ed w it h drug f ever are t he β -lact am ant ibiot ics, procainamide,
and diphenylhydant oin.
Suggested Readings
Fink M. Text book of Crit ical Care, 5t h ed. Philadelphia: Saunders/ Elsevier,
2005: 1186.
Marik P. Fever in t he I CU. Chest 2000; 117: 855–869.
> Table of C ontents > Mis c ellanc eous > 315 - C ar diover t Uns table Tac hyc ar dias ( B oth Nar r ow and
W ide C om plex)
315
Cardiovert Unstable Tachycardias (Both Narrow
and Wide Complex)
Laith Altaweel MD
The assessment of a pat ient w it h a t achycardia requires a syst emat ic approach.
First , t he physician must det ermine w het her t he pat ient is experiencing evidence
of hemodynamic compromise as a result of t he t achycardia, w hich is generally
not seen unt il t he heart rat e is great er t han 150 beat s per minut e. Unst able
t achycardias can be manif est ed by chest pain, short ness of breat h, decreased
urine out put , ment al st at us changes, or hypot ension. O nce a pat ient is not ed t o
be hemodynamically unst able, t he clinician t hen needs t o assess t he t ype of
rhyt hm. This is import ant because t he rhyt hm dict at es w hat f urt her management
is needed. G enerally, t he rhyt hm is classif ied int o t w o broad cat egories: narrow
complex t achycardia or w ide complex t achycardia.
What to Do
Narrow complex t achycardias are rhyt hms w it h Q RS durat ion of less t han 120
ms. There are t hree cat egories of narrow complex t achycardias: junct ional
t achycardia, paroxysmal supravent ricular t achycardia, and at rial t achycardia. I n
most cases of unst able narrow complex t achycardias (w it h t he except ion of
junct ional t achycardia) immediat e synchronized cardioversion w it h 50 t o 100
joules is w arrant ed. I n t he conscious pat ient , premedicat ion w it h a sedat ive or
analgesic, such as diazepam, midazolam, et omidat e, propof ol, f ent anyl, or
morphine, should be at t empt ed. Junct ional t achycardia, w hich is rare and most
f requent ly a sign of digit alis or t heophylline t oxicit y, should not be cardiovert ed
because it represent s an escape rhyt hm. An accelerat ed junct ional rhyt hm is
rarely f ast er t han 120 beat s per minut e and should be t reat ed w it h a bet a-
blocker or, if t he underlying eject ion f ract ion is not know n, w it h amiodarone
w it hdraw al and possibly t reat ment of t he underlying cause (e. g. , digoxin).
Wide complex t achycardias (Q RS >120 ms) include narrow complex t achycardias
w it h aberrant conduct ion, ant idromic at riovent ricular nodal reent ry t achycardia in
pat ient s w it h Wolf -Parkinson-Whit e syndrome, and vent ricular t achycardia
(monomorphic or polymorphic). I mmediat e synchronized cardioversion w it h up t o
360 joules (or 200 joules if biphasic def ibrillat ion is used) should be at t empt ed
f or all unst able w ide complex t achycardia, w it h t he except ion of
polymorphic vent ricular t achycardia, w hich should be def ibrillat ed w it h 360

joules.
Suggested Readings
American Heart Associat ion in Collaborat ion w it h t he I nt ernat ional Liaison
Commit t ee on Resuscit at ion. G uidelines 2000 f or cardiopulmonary
resuscit at ion and emergency cardiovascular care. Part 6: Advanced
cardiovascular lif e support : 7d: The t achycardia algorit hms. Circulat ion
2000; 102: I 158.
> Table of C ontents > Mis c ellanc eous > 316 - B e C onc er ned About C hes t P ain E ven if it is Found to be
Nonc ar diac in Natur e
316
Be Concerned About Chest Pain Even if it is
Found to be Noncardiac in Nature
Laith Altaweel MD
Chest pain in t he int ensive care unit (I CU) is a common and pot ent ially serious
complaint . The diff erent ial diagnosis of chest pain is broad, and t he physician
must not be limit ed t o cardiac et iologies, alt hough myocardial inf arct ion and
angina must alw ays be considered. The init ial approach t o chest pain requires a
rapid evaluat ion, hist ory, physical, elect rocardiogram, chest radiograph, and t he
considerat ion of addit ional laborat ory and radiologic t est s.
The init ial approach should be t o ensure t hat t he pat ient has hemodynamic and
respirat ory st abilit y. This usually result s f rom an assessment of t he pat ient 's
vit al signs and clinical condit ion. The pat ient w ho is bradycardic and hypot ensive
requires more urgent diagnosis t han t he pat ient in pain w ho is aw ake and
conversant . I f t he condit ion is st able, a concise hist ory regarding t he nat ure of
t he pain should be obt ained. A mnemonic t hat may be helpf ul in asking t he
necessary quest ions is O LDCAAR (Table 316. 1). Classic sympt oms of
myocardial inf arct ion or ischemia include chest pain t hat may be charact erized
as sharp, dull, pressure, t earing, or crushing or a f eeling of doom. Pat ient s may
complain of radiat ion t o t he chin, lef t arm, or back. Associat ed sympt oms may
include nausea, vomit ing, diaphoresis, and palpit at ions. The ast ut e clinician
should consider at ypical sympt oms such as “gas” or heart burn t o be cardiac in
et iology unt il proven ot herw ise.
A f ocused physical exam should be perf ormed looking f irst f or cardiovascular
problems such as a diff erence in t he pulses bet w een t he limbs, pulses
paradoxus, pulse volume and rat e, new murmurs, rubs, or gallops. I t is import ant
t o not e t hat t he exam may be normal despit e a cardiac et iology. Addit ional
physical signs may provide insight t o ot her et iologies. For example, rhonchi or
rales, absent breat h sounds, or hyper-resonance may point t ow ard a pulmonary
et iology, w hereas abdominal t enderness, masses, absent or abnormal abdominal
sounds, guarding. and rebound may provide insight int o an abdominal component .
TABLE 316-1 OLDCAAR M NEM ONIC
O: Onset of pain: W hen? W hat was the patient doing at

the time?
L: Location of pain? Pinpoint or diffuse?
D: Duration of pain, days or seconds? Comes and goes

or persists?
C: Character of pain: dull, sharp, lancinating, ripping,

tearing, etc.
A: Associated symptoms such as nausea, vomiting,

diaphoresis, palpitations
A: Alleviating/aggravating factors such as position,

belching, deep breathing
R: Radiating to back, arm, abdomen, neck
Watch Out For

A chest radiograph should be obt ained t o look f or a pneumot horax, w idened
mediast inum, eff usion, new inf ilt rat es, f ree subdiaphragmat ic air, rib f ract ures,
and malposit ioned endot racheal, nasogast ric, or chest t ubes. An
elect rocardiogram should alw ays be obt ained. Changes suggest ive of cardiac
ischemia or inf arct ion can of t en be seen as changes in t he ST segment , T w ave
morphology, or t he presence of Q w aves, w hich can indicat e cardiac ischemia or
inf arct . Cardiac serum markers such as t roponin and creat inine phosphokinase
MB can also ident if y myocardial injury. I f t he clinical hist ory and ECG suggest
cardiac ischemia, t he t reat ment f or acut e cardiac syndromes should be init iat ed.
Pulmonary embolism (PE) is a const ant concern f or I CU pat ient s, many of w hom
have at least one risk f act or, w hich include immobilizat ion, burns, a
hypercoagulable st at e, and heart f ailure. PE can present w it h pleurit ic chest
pain, t achypnea, and dyspnea. Large PEs can result in cardiovascular collapse
w it h hypot ension f rom obst ruct ive cardiogenic shock. Echocardiogram may
reveal a dilat ed right vent ricle w it h reduced f unct ion and a sept al shif t . The chest
radiograph w ill likely be normal and a comput ed t omography (CT) angiogram or
vent ilat ion/ perf usion scan w ill be necessary t o conf irm t he clinical suspicion. The
t reat ment f or a PE in t he set t ing of hemodynamic st abilit y is ant icoagulat ion.
The sympt oms of aort ic dissect ion of t en overlap w it h t hose of myocardial
ischemic pain. The sudden onset of severe, sharp chest pain t hat may or may
not radiat e t o t he back is a t ypical sympt om. A chest x-ray is usually not helpf ul
but may reveal a w idened mediast inum, t he separat ion of int imal calcif icat ion
f rom t he aort ic knob, t he deviat ion of t he t rachea, or t he blurring of t he aort ic
margins. Comparison w it h a recent radiograph is helpf ul. Cont rast -enhanced CT
is usually t he best conf irmat ory t est . The init ial management should f ocus on
blood pressure cont rol and be f ollow ed by surgical consult at ion.
A pneumot horax can occur in I CU pat ient s secondary t o iat rogenic causes such
as cent ral venous cat het er placement or vent ilat or-associat ed barot rauma.
Pulmonary diseases such as chronic obst ruct ive pulmonary disease, ast hma, and
acut e respirat ory dist ress syndrome are risk f act ors. I f a pat ient w it h a
pneumot horax develops hypot ension, jugular venous dist ension, absence of
breat h sounds, hyper-resonance t o percussion, and t racheal deviat ion, t hen t he
development of a t ension pneumot horax is likely. A t ension pneumot horax is of
special concern f or pat ient s on mechanical vent ilat ion receiving posit ive
pressure. I mmediat e needle decompression and chest t ube placement may be
lif e saving in t his condit ion.
Esophageal rupt ure can also cause chest pain and is a lif e-t hreat ening condit ion
t hat can lead t o let hal mediast init is. The hist ory may suggest ingest ion of a
caust ic subst ance, f orcef ul vomit ing, or iat rogenic t rauma (e. g. , nasogast ric t ube
placement , esophageal dilat ion). Physical examinat ion may reveal subcut aneous
emphysema or mediast inal crackling on auscult at ion, know n as Hamman's
crunch. Chest radiograph may show pneumot horax, pneumomediast inum or
pneumoperit oneum, pleural eff usion, or subcut aneous emphysema. A w at er-
soluble cont rast st udy or esophagoscopy conf irms t he diagnosis.
I nt ravenous cont rast -enhanced spiral comput ed t omography can be helpf ul in
evaluat ing f or many of t hese condit ions w it hin t he diff erent ial diagnosis of chest
pain including pulmonary embolism, aort ic dissect ion, pericardial eff usion, or
ant erior pneumot horax. I f t he clinical scenario suggest s one of t hese diagnoses
but ot her, more f undament al t est ing is not def init ive, cont rast -enhanced spiral CT
may be indicat ed. I n addit ion, an echocardiogram may be usef ul f or evaluat ing
t he pat ient f or regional w all mot ion abnormalit ies, w hich may occur w it h coronary
ischemia, lef t and right vent ricular f unct ion, pulmonary hypert ension, valvular
disease, and pericardial eff usion or t amponade.
Suggested Readings
Fink MP, Abraham E, Vincent JL, et al. , eds. Text book of Crit ical Care, 5t h
ed. Philadelphia: Elsevier Saunders, 2005: 120–123.
Lee T. Chest discomf ort and palpit at ions. I n: Kasper D, ed. Harrison's O nline,
16t h edit ion. New York: McG raw -Hill, 2005.
> Table of C ontents > Mis c ellanc eous > 317 - K now the D iffer enc e B etw een m m Hg and c m H2O
317
Know the Difference Between mmHg and cm H2 O
Pressure is def ined as t he f orce per unit area. I t can be applied eit her t o a solid
or a liquid int erf ace at a point perpendicular t o t he surf ace and is represent ed by
t he height of t he column t hat exert s a pressure at it s base. Depending on t he
f luid used, t he height of a column of f luid w ill diff er because it s densit y and
t heref ore t he pressure it exert s w ill diff er.
Mercury (Hg) is a very dense f luid (13. 5951 g/ cm3 ) and is of t en used t o calibrat e
pressure gradient s and diff erent ials. Normal at mospheric pressure can support
760mmHg. O ne millimet er of Hg t heref ore represent s approximat ely 1/ 760 of an
at mosphere and is a usef ul measure f or comparing pressures. This measure is
of t en ref erred t o as a Torr, af t er t he I t alian physicist Torricelli. The unit mmHg is
very usef ul in measurement s f or biologic syst ems. I t is t he convent ional unit used
t o measure blood pressure.
Wat er has been used as a measure of pressure f or numerous physiologic
paramet ers and remains in use. Wat er is less dense t han mercury, and t he
conversion f rom measurement s using mercury t o t hose using w at er f ollow s t he
relat ionship 1 mmHg = 1. 36 cm H2 O . This unit is st ill used f or measurement s of
cent ral venous pressure, int racranial pressure, pressures in mechanical
vent ilat ion, and pulmonary capillary w edge pressures. I t is import ant f or t he
int ensive care unit clinician t o recognize w hich f luid scales are being used t o
correct ly int erpret t he measurement s and provide t herapeut ic int ervent ions
based on t he numeric values. Conf usion about t he scales used can lead t o
inappropriat e act ions and compromise t he pat ient 's care.
Suggested Readings
Cent imet re of w at er. Wikipedia. ht t p: / / en. w ikipedia. org/ w iki/ CmH2 O
Pressure. Wikipedia. ht t p: / / en. w ikipedia. org/ w iki/ Pressure

Torr. Wikipedia. ht t p: / / en. w ikipedia. org/ w iki/ MmHg
> Table of C ontents > Mis c ellanc eous > 318 - B e Aler t for Lupus Flar es
318
Be Alert for Lupus Flares
Syst emic lupus eryt hemat osus (SLE) is a mult isyst em aut oimmune disease
caused by aut oant ibodies and t he deposit ion of immune complexes. The disease
aff ect s primarily w omen and has a higher prevalence among Af rican Americans. A
series of clinical f indings is associat ed w it h SLE. These include a charact erist ic
malar rash, phot osensit ivit y, art hrit is, serosit is, and a discoid rash. I n addit ion,
involvement of t he renal, neurologic, and hemat ologic syst ems present w it h
syst em-specif ic sympt oms. The presence of aut oant ibodies (e. g. , ant i-Sm, ant i–
double-st randed DNA) can also assist in making t he diagnosis. When a pat ient
exhibit s f our or more of t hese crit eria t he diagnosis can be made w it h a
sensit ivit y and specif icit y of 75% and 95%, respect ively.
SLE can present w it h an acut e or subacut e onset . Most pat ient s undergo some
element s of acut e exacerbat ion and remission during t he nat ural course of t he
disease. This is import ant because t he goals of SLE t reat ment include mainly t he
cont rol of sympt oms and prevent ion of t he w orsening of organ f ailure. Treat ment
usually involves a number of syst emically administ ered medicat ions. For minor
disease, art hrit is and pain can be t reat ed w it h nonst eroidal ant iinf lammat ory
medicat ions or CO X-2 inhibit ors. Ant imalarial agent s (chloroquine, quinacrine)
can also benef it pat ient s w it h generalized sympt oms. For severe syst emic
disease, t he use of st eroids is t he mainst ay of t reat ment . These drugs can be
administ ered in high doses, w it h at t empt s t o reduce t he dose at f requent
int ervals once sympt oms are under cont rol. Cyt ot oxic agent s are used f or
aggressive disease including renal manif est at ions.
I t is import ant t hat int ensive care unit (I CU) clinicians underst and w het her t he
present at ion of a w orsening clinical condit ion in an SLE pat ient in t he I CU result s
f rom t he adverse eff ect s of t he medicat ions used t o t reat t he disease, an acut e
lupus f lare, or a separat e and unrelat ed condit ion. For inst ance, in severe cases
of SLE, t he medicat ions used f or immunosuppression increase t he possibilit y f or
severe sepsis and sept ic shock. I t is import ant t o remember t hat t he mechanism
of act ion of t hese drugs is aimed at t he cellular component of t he immune
syst em and complement . Theref ore, pat ient s present ing w it h sepsis are likely t o
be inf ect ed w it h organisms t hat
are usually prevent ed by t he appropriat e f unct ioning of t hese syst ems. Broad-
spect rum ant imicrobials should be used t o t reat sepsis in t hese pat ient s, but
considerat ion of ot her organisms (viruses, f ungi) should be considered if pat ient s
do not improve quickly.
I n addit ion, chronic st eroid use has a number of complicat ions t hat may
cont ribut e t o problems in t he I CU pat ient w it h SLE w hen used. O n a chronic
basis, st eroids are associat ed w it h hypert ension, somet imes ref ract ory t o
t reat ment . St eroids also suppress adrenal f unct ion, t hereby making t he pat ient
suscept ible t o acut e st ressors like surgery or shock. This condit ion can manif est
by ref ract ory hypot ension and needs t o be addressed preoperat ively and during
t he I CU course w it h maint enance st eroids. St eroids lead t o hyperglycemia, w hich
has recent ly been show n t o be an independent predict or of mort alit y in I CU
pat ient s; t hus, st rict glucose cont rol in t hese pat ient s is mandat ory. St eroids
also cont ribut e t o t he onset of pept ic ulcers in t he I CU and poor w ound healing.
Theref ore, t here needs t o be st rict at t ent ion t o bot h prophylaxis f or ulcer
disease and f requent reposit ioning t o prevent skin breakdow n and decubit us
ulcer f ormat ion.
The pat ient w it h SLE may experience problems relat ed t o t he disease it self w hen
hospit alized in t he I CU. The mult isyst em nat ure of t he disease of t en creat es
problems t hat make caring f or t hese pat ient s diff icult . Pat ient s can exhibit
pulmonary or cardiovascular disease. They can have pulmonary inf ilt rat es,
eff usions, and f ibrosis t hat necessit at e mechanical vent ilat ion. Pericardit is and
valvular dysf unct ion can lead t o shock and poor cardiac out put . Pat ient s may
become hyper-coagulable, making t hem suscept ible t o cent ral neurologic
t hrombosis or myocardial ischemia. Renal insuff iciency exacerbat ed by t he renal
st ressors of t he I CU may cause progressive renal f ailure t hat makes f luid and
elect rolyt e management in t he I CU diff icult and may require dialysis. Bone
marrow suppression may creat e anemia t hat w ill alt er oxygen-carrying capacit y
or t hrombocyt openia t hat can lead t o bleeding. Bot h of t hese condit ions may
require ongoing t ransf usions.
Suggested Readings
Davidson A, Diamond B. Aut oimmune diseases. N Engl J Med 2001; 354: 340–
350.
Kasper DL, Braunw ald E, Fauci AS, et al. , eds. Harrison's Principles of
I nt ernal Medicine, 16t h ed. New York: McG raw -Hill, 2005: 1515–1518.

Avoiding Common ICU Errors, 1st Edition (2007)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Avoiding Common ICU Errors, 1st Edition (2007)

Uploaded by

Copyright:

Available Formats

Authors:

Marcucci, Lisa; Martinez, Elizabeth A. ; Haut, Elliott R. ; Slonim,

Elizabeth A. Martinez MD, MHS

Anthony D. Slonim MD, DrPH

Edward T. Horn PharmD

Bradford D. Winters MD, PhD

Amisha V. Barochia MD, CCMD

M. Craig Barrett PharmD

Sean M. Berenholtz MD, MHS, FCCM

Daniel R. Brown MD, PhD, FCCM

Christina L. Cafeo RN, MSN

Molly B. Campion MS, CCC-SLP

Brendan G . Carr MD, MA

Sara E. Cosgrove MD, MS

Peter F. Cronholm MD, MSCE

J. Christopher DiG iacomo MD

Sherita Hill G olden MD, MHS

Michael D. G rossman MD, FACS

Nancy Sokal Hagerman MD

Nadia N. Hansel MD, MPH

Michael Joel Haut MD

Eugenie S. Heitmiller MD, FAAP

Deborah B. Hobson BSN

Edward T. Horn PharmD, BCPS

David G . Hunt RN, BSN

Elizabeth A. Hunt MD, MPH

Lee Ann R. Lau MD

Noah Lechtzin MD, MHS

John J. Lewin III PharmD, BCPS

Ying Wei Lum MD

Shelley Sylvester Magill MD

Elizabeth A. Martinez MD, MHS

Susanna Lovell Matsen MD

Christian Merlo MD, MPH

William G . Merz PhD

Robert K. Michaels MD, MPH

Jennifer Miles-T homas MD

Stephen M. Milner MD, FACS

T imothy M. Moore MD, PhD

Dana Y. Nakamura O T, CLT, CLMC

Julius Cuong Pham MD

Peter J. Pronovost MD, PhD

Jose Manuel Rodriguez-Paz MD

Tuhin K. Roy MD, PhD

Nirav G opal Shah MD

Richard Wong She MHB (Hons), MBChB, FRACS (Plastics)

Carrie A. Sims MD, MS

Vijay Anand Singh MD

Anthony D. Slonim MD, DrPH

Christopher J. Sonnenday MD, MHS

Dimitris Stefanidis MD, PhD

Joseph B. Straton MD, MSCE

Sandra M. Swoboda RN, MSN

G eorge C. Velmahos MD, PhD, MSEd

Kathleen A. Williams RN, MSN CRNP

Bradford D. Winters MD, PhD

Faramarz Zarfeshanfard RPh

I n t he six years since t he I nst it ut e of Medicine released it s landmark report “To

Day 1: Place t ransdermal clonidine. Addit ionally, administ er 100% of oral

Hughes JR: Clonidine, depression, and smoking cessat ion. JAMA

Signs And Symptoms

Signs and Symptoms

Lat ronico N, Shehu I , Seghelini E. Neuromuscular sequelae of crit ical illness.