Professional Documents
Culture Documents
pharmacy
journals.sagepub.com/home/hpx ISSN: 1545-1569
FMS
Nasal Glucagon 697
Thomas J. Borden, Terri L. Levien, and Danial E. Baker
Lefamulin 704
Liliya Kozhokar, Terri L. Levien, and Danial E. Baker
Original Articles
Short-Term Readmission Following Community-Acquired Pneumonia: A Cross-Sectional Study 712
Dung Thien Nguyen, Sang Thanh Huynh, and Ho Nhu Nguyen
Opioid Use in Vaso-Occlusive Crisis During Intravenous Opioid Drug Shortage 721
Katherine V. Katsivalis, Jeremy Kosacz, and Jennifer Austin Szwak
Safety and Efficacy Analysis of Apixaban Compared to Heparins in Hospitalized Non-Critically Ill COVID-19 Patients 727
Daniel Appiah, Nicholas J. Quinn, Emily G. Messing, and Keith T. Veltri
Assessment of an In-Sourced Patient Assistance Program at a Health System in West Virginia 734
Calvin Parmiter, David Cecere, Staci Czyzewski, and Michelle Gibson
Heparin Resistance in SARS-CoV-2 Infected Patients with Venous Thromboembolism 737
Terence Chau, Merlyn Joseph, Diana M. Solomon, Bryan Lee, and Lauren A. Igneri
Spontaneous Adverse Event Reporting by COVID-19 Vaccinated Healthcare Professionals Through an 744
Electronic Form Implemented by the Hospital Pharmacy
João Paulo Garcia Lopes da Cruz, Cristina da Conceição Ribeiro de Carvalho, Paula Alexandre da Cruz Silva,
Luis Filipe Campos Guerreiro, Tatiana Vedes Bento, Leila Vanessa Carmo Cardoso Martins Costa,
Raquel Filipa Miguel Margarido Duarte Simões, Rui Pedro Pinheiro Gonçalves Marques, Ana Cristina Castro Fernandes,
Luís Miguel Costa de Mendonça Galaio, Ana Isabel B. Correia, Ema Maria S. Leite Resende, and João Manuel Braz Gonçalves
The Impact of Job Instruction Training Within Pharmacy 752
Faria Munir, Christopher Loucks, and Petal Bartlett
Effect of Paralytic Agents on Post-Intubation Sedation in the Emergency Department 759
Jessica A. Pankey, Scott Christofferson, Ginger Barrick, Brandon A. Knettel, and Christine Knettel
case reports
Omadacycline for a Carbapenem-Resistant Enterobacter cloacae-Associated Wound Infection 767
Dylan C. VanDuyn, Saloni Chadha, Lauren A. Paul, Andrea R. Dressler, Mario V. Beccari, and Rajinder P.S. Bajwa
Mirabegron Associated Angioedema: A Case Report 771
Matthew T. Zuchowski, Jaylan M. Yuksel, and John Novi
Coronavirus Disease 2019-Related Extensive Thrombosis in a Patient Receiving Therapeutic Anticoagulation 774
With Dabigatran
Robert C. Ross, Andrew L. Hendrickson, Miranda P. Boraas, Abbie N. Rosen, and Andrew J. Franck
Calcium and Phosphate Solubility Curve Equation for Determining Precipitation Limits in Compounding 779
Parenteral Nutrition
Collin Anderson, Larry Eggert, Kristie Fitzgerald, Daniel Jackson, and Fred Farr
Carvedilol-Induced Thrombocytopenia: A Case Report 786
Angelina R. Raimonde and Anthony K. Dennis
SAGE
Deep Backfile
Deep Backfile for SAGE’s peer-reviewed scholarly journals are available in a wide
range of subject areas including business, humanities, social science, and science,
technology and medicine.
• Deep Backfile of over 620 journals • Coverage back to volume 1, issue 1, through
• More than 931,000 articles the last issue of 1998
• Lease and purchase of one entire deep • Content ownership of purchased content
backfile or individual titles available for • Journal Backfile articles posted in PDF
individual libraries and library consortia format with xml-based headers, abstracts
and references
For more information, contact your Library Sales Manager or email librarymktg@sagepub.com
journals.sagepub.com
Hospital Pharmacy
Editor in Chief
Joyce A. Generali, RPh, MS, FASHP
Clinical Professor, Emeritus
Assistant Editors
Danial E. Baker, PharmD, FASHP, FASCP
Professor of Pharmacy Practice, College of Pharmacy
Michael R. Cohen, RPh, MS, ScD, FASHP
President, Institute for Safe Medication Practices
Consulting Editor
Dennis J. Cada, PharmD, FASHP, FASCP
Executive Editor, The Formulary
Editor Emeritus
Neil M. Davis, MS, PharmD, FASHP
President, Safe Medication Practices Consulting, Inc.
Contents
Director’s Forum
Review of Medication Error Sources Associated With Inpatient Subcutaneous Insulin: Recommendations 689
for Safe and Cost-Effective Dispensing Practices
Connor McKay, Dan Schenkat, Kristin Murphy, and Elizabeth Hess
FMS
Lefamulin 704
Liliya Kozhokar, Terri L. Levien, and Danial E. Baker
Original Articles
Opioid Use in Vaso-Occlusive Crisis During Intravenous Opioid Drug Shortage 721
Katherine V. Katsivalis, Jeremy Kosacz, and Jennifer Austin Szwak
Safety and Efficacy Analysis of Apixaban Compared to Heparins in Hospitalized Non-Critically Ill COVID-19 Patients 727
Daniel Appiah, Nicholas J. Quinn, Emily G. Messing, and Keith T. Veltri
Assessment of an In-Sourced Patient Assistance Program at a Health System in West Virginia 734
Calvin Parmiter, David Cecere, Staci Czyzewski, and Michelle Gibson
Spontaneous Adverse Event Reporting by COVID-19 Vaccinated Healthcare Professionals Through an 744
Electronic Form Implemented by the Hospital Pharmacy
João Paulo Garcia Lopes da Cruz, Cristina da Conceição Ribeiro de Carvalho, Paula Alexandre da Cruz Silva,
Luis Filipe Campos Guerreiro, Tatiana Vedes Bento, Leila Vanessa Carmo Cardoso Martins Costa,
Raquel Filipa Miguel Margarido Duarte Simões, Rui Pedro Pinheiro Gonçalves Marques, Ana Cristina Castro Fernandes,
Luís Miguel Costa de Mendonça Galaio, Ana Isabel B. Correia, Ema Maria S. Leite Resende, and João Manuel Braz Gonçalves
Coronavirus Disease 2019-Related Extensive Thrombosis in a Patient Receiving Therapeutic Anticoagulation 774
With Dabigatran
Robert C. Ross, Andrew L. Hendrickson, Miranda P. Boraas, Abbie N. Rosen, and Andrew J. Franck
Calcium and Phosphate Solubility Curve Equation for Determining Precipitation Limits in Compounding 779
Parenteral Nutrition
Collin Anderson, Larry Eggert, Kristie Fitzgerald, Daniel Jackson, and Fred Farr
Manuscripts should be submitted online at http://mc.manuscriptcentral.com/hpx. For complete author guidelines, please
visit http://journals.sagepub.com/home/hpx and click the “Submit Paper” tab. Questions regarding the submission process
may be directed to Joyce Generali, RPh, MS, FASHP, at jgeneral@ku.edu, Ph: (913) 220-5665.
Hospital Pharmacy (ISSN 0018-5787) (J849) publishes six times annually—in February, April, June, August, October, and
December–by SAGE Publishing, 2455 Teller Road, Thousand Oaks, CA 91320. Periodicals postage paid at Thousand Oaks,
California, and at additional mailing offices. POSTMASTER: Send address change to Hospital Pharmacy, c/o SAGE
Publishing, 2455 Teller Road, Thousand Oaks, CA 91320.
Copyright 2022 by SAGE Publishing. All rights reserved. No portion of the contents may be reproduced in any form
without written permission from the publisher.
Subscription Information: All subscription inquiries, orders, back issues, claims, and renewals should be addressed to
SAGE Publishing, 2455 Teller Road, Thousand Oaks, CA 91320; telephone: (800) 818-SAGE (7243) and (805) 499-0721;
fax: (805) 375-1700; e-mail: journals@sagepub.com; website: journals.sagepub.com. Subscription Price: Institutions:
$329; Individuals: $187. For all customers outside the Americas, please visit http://www.sagepub.co.uk/customerCare.nav
for information. Claims: Claims for undelivered or damaged copies must be made no later than six months following month
of publication. The publisher will supply replacement issues when losses have been sustained in transit and when the
reserve stock will permit.
Copyright Permission: To request permission for republishing, reproducing, or distributing material from this journal,
please visit the desired article on the SAGE Journals website (journals.sagepub.com) and click “Permissions.” For addi-
tional information, please see www.sagepub.com/journals-permissions.
Advertising and Reprints: Current advertising rates and specifications may be obtained by contacting the advertising
coordinator in the Thousand Oaks office at (805) 410-7772 or by sending an e-mail to advertising@sagepub.com. To order
reprints, please e-mail reprint@sagepub.com. Acceptance of advertising in this journal in no way implies endorsement of
the advertised product or service by SAGE, the journal’s affiliated society(ies), or the journal editor(s). No endorsement is
intended or implied. SAGE reserves the right to reject any advertising it deems as inappropriate for this journal.
Change of Address: Six weeks’ advance notice must be given when notifying of change of address. Please visit us at
addressupdates.sagepub.com/support/tickets/new to complete the SAGE Journals Address Update form.
1095771
research-article2022
HPXXXX10.1177/00185787221095771Hospital PharmacyMcKay et al
Director’s Forum
Hospital Pharmacy
Abstract
Background: The challenge of addressing variation in practice is underlined by variation in institutional operational and
financial limitations, which ultimately directs institutional insulin dispensing strategy. Insulin therapy is multi-modal, and
inpatient pharmacies may be responsible for simultaneously managing up to 8 formulations. While available as a tool for
glycemic management in the inpatient setting, intravenous (IV) insulin and insulin pumps are out of scope for this review.
Considerations when designing subcutaneous insulin dispensing processes include medication safety, infection control,
pharmacy workflow, nursing workflow, drug cost, hardware cost, and hospital policy. Objective: This review provides
an overview of common inpatient subcutaneous insulin dispensing practices, describes the regulations directing current
practice, and discusses the benefits and risks of each dispensing practice. Finally, the review provides recommendations
for subcutaneous insulin dispensing practices with respect to institutional limitations that minimize patient safety risk and
consider the variable costs that practices incur. Methods: Common formulations will be presented along with medication
safety considerations and potential medication administration errors to preface strategies for management. Conclusion:
Nearing the centennial anniversary of the invention of insulin, the practice of dispensing subcutaneous insulin products has
remained a core responsibility of the hospital pharmacy as a primary treatment option for Diabetes Mellitus. Despite the
continued presence of this product as a universal component of the hospital formulary, subcutaneous insulin dispensing
practices remain varied institution-to-institution.
Keywords
cost, dispensing, insulin, safety
persistent hyperglycemia. Similar to insulin NPH, long act- Medication Administration Errors
ing insulins may be readily prepared by pharmacy for ready-
to-use format in patient-specific doses. Errors with long Wrong Dose
acting insulin may be more challenging to identify due to the Wrong dose errors may be defined as the intended patient
pharmacokinetics of the medication. Errors can lead to receiving an intended dose of insulin at an amount differing
delayed hypoglycemia and harm if not identified and treated from the written order, including errors of omission.18 Given
appropriately. the varying number of health professionals involved in pre-
paring patient-specific subcutaneous insulin doses, dosing
Concentrated Insulin errors may occur when staff are not familiar with standard
concentrations or routine doses. For example, a U-100 vial
In patients with severe insulin resistance (ie, requiring more may be mistaken to only contain 100 units of insulin, or a
than 200 units of insulin per day to maintain normoglyce- 10 mL vial may be mistaken to contain only 10 units of insu-
mia), concentrated insulins are typically employed for type II lin. There have been numerous reports of adverse drug
diabetic management.9 While onset is similar to short acting events associated with wrong dose errors from U-500 con-
insulin, duration of action for concentrated insulins is largely centrated insulin.3 Additional clinical interpretation of slid-
dose-dependent, with larger doses generally providing basal ing scale orders may lead to doses deviating from the
coverage. The only formulation commonly available is insu- original treatment plan. In any case, lack of product or soft-
lin regular U-500 (Humulin R U-500). Concentrated formu- ware familiarity may be contributory to knowledge-based
lations other than U-500 are unlikely to be on formulary in errors. Further dosing errors may be introduced with memory-
the inpatient setting and are out of the scope of the recom- based errors from incorrectly drawn doses, interruptions, or
mendations provided below. If concentrated pens are used, communication between staff.
typically, they are utilized as patient home medications.
Concentrated insulins are high-alert medications and should
be managed accordingly as case reports of overdose have Wrong Patient
been reported in the literature.14 Given the infrequency of Wrong patient errors occur as an unordered drug error for the
prescribing, dispensing either patient-specific pens or phar- patient receiving the subcutaneous insulin dose and an error
macy-drawn, patient-specific doses are preferred to mini- of omission for the intended patient.18 While barcode medi-
mize the risk of preparation errors (eg, 1 unit on the U-100 cation administration (BCMA) is a commonplace feature in
insulin syringe is equivalent to 5 units of U-500 insulin) or many hospitals, inadequate compliance or lack of patient-
administration error. Stock concentrated insulin should not specific labeling may lead to unintended administration.
be stored on patient care units, and each dose should be pre- Wrong patient errors may also occur in the prescribing and
pared or verified by pharmacy given literature suggesting a transcribing process when orders are incorrectly input for the
higher incidence of hypoglycemic events.15 In the absence of wrong patient yet are correctly dispensed and administered.
patient-specific pens, the pharmacy should dispense U-500
doses from vial stock in a patient-specific U-500 syringe.
Errors with concentrated insulin can be devastating since the
Wrong Drug
insulin is 5 times more concentrated than typical insulins. Wrong drug errors represent the administration of an incor-
Due to the pharmacokinetics, monitoring and treatment of an rect insulin product. Wrong drug errors may be introduced in
error may extend over 12 hours. the prescribing, transcribing, dispensing, or administration
processes.18 Attention to selecting the correct insulin product
when ordering or preparing insulin doses should be taken to
Insulin Pens
minimize the risk of administering the wrong product. Access
Insulin pens of various formulations are utilized in the inpa- to a variety of floor stock insulin products may amplify the
tient setting to reduce pharmacy and nursing time to prepare risk of wrong drug errors in the setting of subcutaneous insu-
and administer insulin. All formulations of short acting, lin administration.
intermediate acting, long acting, and concentrated insulins
are available in pen dosage form. While assumed, there is
Monitoring Error
evidence to support that insulin vials and pens confer the
same clinical efficacy in achieving glycemic control in the While inappropriate monitoring may lead to uncontrolled
inpatient setting.16 The advantages and disadvantages of blood glucose levels resulting in increased morbidity, moni-
insulin pens will be discussed further below. Errors with toring errors may also lead to improper management of insu-
insulin pens are less common due to the human factored lin administration. The American Diabetes Association
design of the pen. However, reports with wrong patient errors recommends that blood glucose be monitored in accordance
have occurred and require blood-borne pathogen monitoring with meal schedule. While alterations in dietary status, renal
to confirm no harm to the patient.17 function, acuity, or introduction of drug interactions should
692 Hospital Pharmacy 57(6)
ensure that blood glucose is monitored 4-to-6 times daily.1 Pharmacy Dispensing
Error in monitoring may lead to suboptimal therapy or over-
treatment as exogenous insulin requirements fluctuate. As previously mentioned, the utilization of vials or pens is a
differentiating factor across institutions. Product may be dis-
pensed as patient-specific in the original container with mul-
Wrong Time tiple doses intended for a single patient or as a patient-specific,
Wrong time errors are doses that are administered greater drawn-up syringe intended for a single patient. Patient-specific
than 60 minutes from the intended administration time. Not syringes containing long acting or intermediate acting insulin
including instances of documentation omission, wrong time and patient-specific insulin pens dispensed by a central phar-
errors occur in the transcribing, dispensing, administration, macy are used to reduce the risk of inaccurate dosing, provide
and monitoring process from pharmacy delay, with the pres- a ready-to-use format, decrease pharmacy and nursing time in
ence of more acute patient needs, or with patient-related fac- preparation, decrease needle-stick injuries, and prevent cross-
tors (eg, the patient is sleeping, is in a procedure, patient contamination.19-22 Cross-contamination may occur through
transfer, or has not yet eaten).18 Orders may also be incor- direct or indirect contact with potentially contaminated sur-
rectly timed, leading to wrong time errors. faces or equipment such as when a multi-dose vial is re-entered
with a contaminated needle or when an insulin pen is used for
multiple patients. For concentrated insulins, the use of pens
Wrong Route may confer an additional safety benefit by removing U-500
Since insulin may be given intravenously or subcutaneously, syringes from circulation with proper inventory management
there is additional potential for wrong route errors in the pre- processes in place to prevent mix-up with other potential insu-
scribing, transcribing, or administration processes.19 A cor- lin pen products. However, patient-specific pens are more
rectional insulin order may be incorrectly input for an expensive than stocking vials, are prone to duplicate dispens-
intravenous route when subcutaneous administration is ing, have the potential for premature needle withdrawal lead-
intended. Comparatively, the medication may be incorrectly ing to underdosing, and must be primed prior to first
administered when ordered correctly. Either method of error administration, leading to waste.23-27 In a survey of 474 inpa-
may confer suboptimal treatment or overdose requiring cor- tient institutions, 30% (142) reported inadvertently using insu-
rectional administration of additional insulin or closer moni- lin pens on multiple patients despite CDC recommendation
toring for hypoglycemia. that insulin pens be patient-specific.28,29 Comparatively, sin-
gle-use, patient-specific syringes are prepared in a sterile envi-
ronment and verified in the pharmacy to ensure product and
Insulin Dispensing Practices dose accuracy.
Variation in inpatient subcutaneous insulin dispensing prac-
tice may be delineated into 2 categories: product selection
Floor Stock
and workflow design. Product selection variation originates
from institutional clinical practice and formulary determina- MDVs stocked in an ADC throughout the hospital are utilized
tion. Product pricing and provider prescribing preference to minimize waste at the expense of additional work to ensure
will primarily dictate bolus and basal insulin therapy choice patient safety. Floor stock vials or pens may be labeled and
to treat hyperglycemia. The impact of product selection may assigned as patient-specific with multiple doses for a single
result in the utilization of vials versus pens. However, the patient or remain as non-patient-specific ADC stock invento-
majority of variation in practice will be driven by the institu- ried per unit with multiple doses for multiple patients. As
tion’s workflow design between sterile compounding, auto- MDVs stored in ADCs are prone to human error, operational
mated dispensing cabinet (ADC) restocking, and nurse processes outlining the proper identification of the vial and
medication administration. The design of the dispensing pro- patient must be well-designed with proper procedure to
cess is done with resource availability (eg, labor, hardware, ensure sterile administration and minimize cross-contamina-
software, and supplies) in mind. The following list defines tion. A 2-nurse double-check is only recommended for insulin
the key variabilities that must be considered when outlining intended for intravenous administration and is no longer rec-
any subcutaneous insulin dispensing practices: ommended for subcutaneous insulin.30 Subcutaneous Insulin
doses prepared by a nurse may be verified in several ways:
1. Source of product: vial or pen
2. Product that is dispensed: vial, pen, or pharmacy- 1. The prepared dose is labeled with a patient-specific
prepared syringe label that has a barcode generated by the ADC printer.
3. Location of product dispensed: pharmacy or ADC 2. The prepared dose is labeled with a generic barcode
4. Personnel labeling the product: pharmacy or nurse label stocked in or around the ADC.
5. Recipient(s) of the dispensed product: multiple 3. The patient-specific vial is scanned at the patient
patients or a single patient. bedside.
McKay et al 693
4. The floor stock vial is scanned in the medication Table 1. Cost Comparison of Available Insulin Products.31
room with a mobile scanner. The mobile scanner is
Insulin Product 3 mL Vial 3 mL Pen 10 mL Vial
then used to scan the patient prior to administration.
5. The product is not scanned or labeled. Regular $49.56 $62.46 $165.20
100 units/mL
When examining workflow for utilization of MDVs for Lispro $98.88 $127.29 $329.60
nurse-drawn preparation, the hospital pharmacy is typically 100 units/mL
Aspart $104.16 $134.13 $347.20
responsible for monitoring expiration dating, removing
100 units/mL
expired vials, and restocking to a unit-specific inventory
Glulisine $102.21 $131.64 $340.70
count. Insulin MDVs have an expiration date of either 28 days, 100 units/mL
42 days, or as instructed by the manufacturer package insert NPH $49.56 $62.46 $165.20
once the protective seal has been punctured or upon storage at 100 units/mL
room temperature.8 Expiration reporting functionality may be Glargine – $102.06 $340.30
automated through the ADC. However, some institutions 100 units/mL
store MDVs in ADC refrigerators with the expectation for Glargine-yfgn – $35.52 $118.40
manual documentation of expiration dating by nursing or 100 units/mL
pharmacy when puncturing the vial. With nursing documen- Detemir – $110.94 $369.80
tation, there is a reduced waste by ensuring the product is 100 units/mL
Degludec – $122.01 $406.70
short-dated upon initial use. Alternatively, pharmacy may
100 units/mL
manually pre-date the vials stored in the ADC refrigerator and
U-500 $267.66 $344.52 $1784.40a
re-label if the refrigerated product remains unused. Both 500 units/mL
manual processes are prone to error of omission.
mL = milliliter.
a
Product available as a 20 mL vial.
Cost
A review of available insulin formulations with associated confident with processes to prevent cross-contamination and
costs is listed below (Table 1). Costs are organized by regulatory noncompliance.
Average Wholesale Price (AWP) pricing for 3 mL vials, 3 mL
pens, and 10 mL vials for comparison.
In 2019, the average length of stay (ALOS) for hospital- Risk Reduction Strategies
ized patients in the United States was 5.5 days.32 For an illus-
Floor Stock
tration of cost difference, total short acting insulin cost across
ALOS for floor stock vials, patient-specific vials, and Storing MDVs as floor stock may be useful with operational
patient-specific pens will be extrapolated for insulin regular and technological limitations. When MDVs are stored as
at a dose of 12 units daily using each formulation. Utilizing floor stock rather than patient-specific pens or patient-spe-
floor stock 10 mL vials or 3 mL vials results in a total AWP cific vials, there are several considerations for establishing
acquisition cost of $10.90, while a patient-specific 3 mL vial safe handling practices. Compared to storage in central phar-
would result in a total AWP acquisition cost of $49.56, macy or medication room refrigerators, MDVs should ide-
assuming vial wasting following discharge. Comparatively, ally be stored in an ADC to prevent delays in care and wrong
short acting, patient-specific pen usage would result in a drug errors. MDVs may be used for multiple patients when
total AWP cost of $62.46, assuming pen wasting following stored under the described conditions using proper aseptic
discharge. While AWP pricing does not reflect pharmacy technique and smaller volume vials (ie, 3 mL vials) to limit
acquisition costs in practice, the relationship in pricing repeated vial access.5 When MDVs are stored in ADCs,
between products holds true for contractual pricing through ISMP states that MDVs should be segregated in locked-lid-
Wholesale Acquisition Cost or Group Purchasing Organi ded pockets to prevent the possibility of selecting the wrong
zations. Additionally, the demonstrative analysis performed insulin product for nurse preparation.
aligns with the literature. In prior budget impact analyses, Additionally, barcode scanners and automatic label print-
utilizing short acting 3 mL floor stock vials has considerably ers can be implemented to correctly identify insulin product,
reduced acquisition cost and waste.27,33 When considering prepare patient-specific syringes away from patient care
cost for short acting insulin products, preference should be areas, and correctly identify the patient for medication
shown for 3 mL vial products with larger volume vials and administration. Patient labels are ideally employed in con-
pens reserved for pharmacy dispensed products (ie, interme- junction with BCMA when possible.4 Floor stock should be
diate acting, long acting, and concentrated insulin products). limited to short acting insulin as these products carry a
As described previously, introducing floor stock product greater risk for wrong time errors. Subsequently, minimizing
carries organizational risk, and pharmacy leaders must be the availability of long acting, intermediate acting, and
694 Hospital Pharmacy 57(6)
Advantages Disadvantages
Pharmacy dispensing of • Low risk of wrong product error • Highest risk for wrong time error
patient-specific vials • Lower risk of wrong-patient error • Highest risk of wrong dose error
or pens • Low operational pharmacy support required • Highest risk of cross-contamination
Pharmacy dispensing • Low risk of wrong dose error • High risk for wrong time error
of patient-specific • Low risk of wrong product error • Highest operational pharmacy support required
pre-drawn syringes • Lowest risk of wrong patient error
• Lowest risk of waste
• Lowest risk of cross-contamination
• Lowest technological pharmacy support required
ADC patient-specific • Low risk of wrong time error • High technological pharmacy support required
vials or pens • Low risk of cross contamination • Highest risk of waste
• Low risk of wrong product error • Moderate risk of wrong dose error
• Low risk of wrong patient error
Floor stock vials or pens • Low risk of wrong time error • Highest technological pharmacy support required
• Lower risk of waste • Moderate risk of wrong patient error
• Lowest operational pharmacy support required • Highest risk of wrong dose error
• Highest risk of wrong product error
• Highest risk of cross-contamination
concentrated floor stock insulin reduces excess waste from product and wrong dose errors. However, there are opera-
vial loss, cross-contamination from multi-patient use, and tional concerns to providing short acting insulin as patient-
risk of wrong drug errors. Advantages and disadvantages for specific doses that incurs a risk of wrong time errors. Given
floor stock dispensing are listed in Table 2. that long acting, intermediate acting, and concentrated insu-
lins are routinely scheduled, these doses should be prepared
as patient-specific doses utilizing product barcode scanning
Patient-Specific Vials
in the pharmacy and dispensed to the floor in a ready-to-use
Utilizing patient-specific MDVs is a strategy to minimize the format rather than as floor stock to limit the potential for
risk of cross-contamination or to mitigate operational limita- wrong dose, wrong product, or wrong patient error.
tions. Per the CDC, MDVs should be stored and punctured in Advantages and disadvantages for patient-specific dose dis-
a location away from patient care areas (eg, operating rooms, pensing are listed in Table 2.
procedure carts, patient bays, and patient rooms). MDVs that
do enter these areas shall be designated for that patient for
Recommendations for Safe insulin use
continued use.5 In institutions that do not have product bar-
coding or patient-specific labeling capability, designated Short acting insulin may be dispensed in several ways, abid-
vials may be dispensed from the central pharmacy or assigned ing by the proper risk reduction strategies in place. Floor
from floor stock for patient use. Concerns with this approach stock vials are common, reduce departmental costs, and
include excess waste when a patient is discharged or when a improve ease of use for sliding scale orders. The authors
vial is lost and the potential for wrong time errors when preferably recommend that institutions provide patient-spe-
patient-specific vials are dispensed from the central phar- cific, short acting insulin vials to comply with CDC guid-
macy. To limit the potential for wrong drug errors, product ance. Alternatively, due to the cost of vials, the authors also
availability for patient-specific vials assigned from floor support institutions utilizing floor stock short acting insulin
stock should be limited to short acting insulin segregated in vials, assuming infection risks, wrong dose, and wrong drug
locked-lidded pockets with BCMA scanning when possible. errors are minimized through the use of technology. If staff
Advantages and disadvantages for patient-specific vial dis- are to use floor stock short acting insulin vials, institutions
pensing are listed in Table 2. must be able to provide appropriately labeled syringes before
entering the patient room.
The authors recommend that long acting and intermedi-
Patient-Specific Dose ate acting insulin be prepared in a patient-specific, ready-to-
ISMP and TJC state that insulin doses should be provided in use format by the pharmacy. Floor stock vials of long acting
the most ready-to-use format for administration.3,8 Pharmacy and intermediate acting insulin are not recommended given
dispensing of patient-specific doses reduces the risk of wrong the potential for product mix-up and wrong drug error.
McKay et al 695
The pharmacist product verification process and pharmacy on pharmacy capability to absorb additional sterile com-
access to technology-assisted workflow (TAWF) confirm pounding preparation and ADC inventory management
the right drug for the right patient. Administration errors capabilities. Thus, strategies for insulin dispensing should
with long acting and intermediate acting insulin may be be formulated by class. In this review of subcutaneous
more difficult to recognize and intervene to preclude harm. insulin dispensing, recommendations are provided to
While this may mean more workload for the pharmacy, it refine institutional processes and emphasize patient safety
avoids calculation errors by nursing, who are tasked with for each class of insulin.
many other important tasks. The authors recommend that
the pharmacy batch prepare all standing patient-specific Declaration of Conflicting Interests
long acting and intermediate acting insulin orders together The author(s) declared no potential conflicts of interest with respect
for ease of workload and determine a standard administra- to the research, authorship, and/or publication of this article.
tion time to successfully accomplish this.
At a minimum, the pharmacy must always prepare con- Funding
centrated insulin in patient-specific doses to minimize the
The author(s) received no financial support for the research, author-
need for manipulation and calculation by nursing. Vials ship, and/or publication of this article.
should never be stored on the patient care unit as floor stock.
Given the risk of morbidity with wrong dose errors and infre-
ORCID iD
quency of use, insulin pens for U-500 are preferred. Some
institutions may utilize insulin pens due to the low opera- Connor McKay https://orcid.org/0000-0002-0313-4878
tional pharmacy support required, patient and nurse satisfac-
tion associated with administration, and minimal difference References
in cost relative to vials.21 The authors preferably recommend 1. American Diabetes Association. 15. Diabetes care in the hos-
that insulin pens be used only for U-500 concentrated insulin pital: standards of medical care in diabetes – 2021. Diabetes
to reduce the risk of product mix-up. For example, in 2016, Care. 2021;44(Suppl 1):S211-S220.
Brown and Hertig,28 reported in a survey of 474 institutions 2. Centres for Disease control and Prevention. National diabetes
that approximately 30% of the survey respondents accounted statistics report 2020: estimates of diabetes and its burden in
that insulin pens had been used on more than 1 patient at the United States. 2020. Accessed December 14, 2020. https://
www.cdc.gov/diabetes/pdfs/data/statistics/national-diabetes-
least once in their institution. In institutions that do not pos-
statistics-report.pdf
sess the operational and technological capabilities to imple- 3. Institute for Safe Medication Practices (ISMP). ISMP guidelines
ment components of the described subcutaneous insulin for optimizing safe subcutaneous insulin use in adults. 2017.
dispensing best practices, patient-specific insulin pens may Accessed December 7, 2020. https://www.ismp.org/guidelines/
be an acceptable, provisional alternative for institutional use subcutaneous-insulin
with the intent to align with dispensing best practices. 4. Setra A, Jani Y. A longitudinal assessment of the quality
Recommendations are summarized below in Table 3. of insulin prescribing with different prescribing systems.
Pharmacy. 2021;9(1):53.
5. Pennsylvania Patient Safety Authority. Focus on high-alert
Conclusion medications. Pa Patient Saf Advis. 2004;1(3):6.
Subcutaneous insulin is one of the most commonly ordered 6. US Pharmacopeia. MEDMARX data report. Technical appen-
dix 12: products most frequently involved in harmful medica-
medications in the hospital, yet there remains little consen-
tion errors (Categories E-I), CY 2006, 2008.
sus on best practices for dispensing insulin products. Waste 7. The Centers for Disease Control and Prevention. Questions
and patient safety remain active concerns for this high- About Multi-Dose Vials. Department of Health and Human
alert medication. While variation in practice is conditional Services; 2019.
to institutional resources, processes may be refined to 8. Joint Commission Resources. Accreditation Standards Books
ensure optimal usage. Institutions may use either MDVs or for Hospitals, Ambulatory Care, and Behavioral Health Care.
pens exclusively or a blend of pens and MDVs depending The Joint Commission; 2021.
696 Hospital Pharmacy 57(6)
9. Munguia C, Correa R. Regular insulin. StatPearls; 2021. 23. Najmi U, Haque WZ, Ansari U, et al. Inpatient insulin pen
10. Islam N, Khanna NR, Zito PM. Lispro insulin. StatPearls; 2021. implementation, waste, and potential cost savings: a commu-
11. Rubin R, Khanna NR, McIver LA. Aspart insulin. StatPearls; nity hospital experience. J Diabetes Sci Technol. 2021;15(4):
2021. 741-747.
12. Saleem F, Sharma A. NPH insulin. StatPearls; 2021. 24. Kelton KA, Perk S, Loveland S, Perez-Nieves M, Fu H, Peng
13. Cunningham AM, Freeman AM. Glargine insulin. StatPearls; X. Economic outcomes with the conversion of insulin delivery
2021. methods in hospitals. J Med Econ. 2017;20(5):533-540.
14. Monroe PS, Heck WD, Lavsa SM. Changes to medication-use 25. Edmondson G, Criswell J, Krueger L, Eby EL. Economic
processes after overdose of U-500 regular insulin. Am J Health impact of converting from 10-mL insulin vials to 3-mL vials
Syst Pharm. 2012;69(23):2089-2093. and pens in a hospital setting. Am J Health Syst Pharm.
15. Reutrakul S, Wroblewski K, Brown RL. Clinical use of U-500 2014;71(17):1485-1489.
regular insulin: review and meta-analysis. J Diabetes Sci 26. Joubert M, Haddouche A, Morera J, Rod A, Reznik Y. Potential
Technol. 2012;6(2):412-420. insulin underdelivery from prefilled and reusable insulin pens
16. Horne J, Bond R, Sarangarm P. Comparison of inpatient gly- in cases of premature needle withdrawal: a laboratory evalua-
cemic control with insulin vials versus insulin pens in general tion. Diabetes Technol Ther. 2015;17(10):712-716.
medicine patients. Hosp Pharm. 2015;50(6):514-521. 27. Lee LJ, Smolen LJ, Klein TM, et al. Budget impact analysis of
17. Institute for Safe Medication Practices. Cross contamination insulin therapies and associated delivery systems. Am J Health
with insulin pens. ISMP Med Safe Alert. 2008;13(6):1-2. Syst Pharm. 2012;69(11):958-965.
18. Tariq RA, Vashisht R, Sinha A, Scherbak Y. Medication dis- 28. Brown KE, Hertig JB. Determining current insulin pen use
pensing errors and prevention. StatPearls; 2021. practices and errors in the inpatient setting. Jt Comm J Qual
19. Berdot S, Roudot M, Schramm C, Katsahian S, Durieux P, Patient Saf. 2016;42(12):568-a7.
Sabatier B. Interventions to reduce nurses’ medication admin- 29. The Centers for Disease Control and Prevention. Insulin Pens
istration errors in inpatient settings: a systematic review and Must Never be Used for More Than One Person. Department
meta-analysis. Int J Nurs Stud. 2016;53:342-350. of Health and Human Services; 2011.
20. Ward LG, Aton SS. Impact of an interchange program
30. Modic MB, Albert NM, Sun Z, et al. Does an insulin double-
to support use of insulin pens. Am J Health Syst Pharm. checking procedure improve patient safety? J Nurs Adm.
2011;68(14):1349-1352. 2016;46(3):154-160.
21. Shogbon AO, Ngo D, Jacob B, Kimble LP, Ryan G. Nurses’ 31. Medi-Span Price Rx [Online Database]. Medi-Span. Wolters
perceptions and satisfaction with the use of insulin pen devices Kluwer; 2021.
compared with insulin vial and syringes in an inpatient setting. 32. OECD. Length of Hospital Stay (Indicator). Organisation for
Diabetes Technol Ther. 2014;16(11):742-746. Economic Co-operation and Development; 2021.
22. Smallwood C, Lamarche D, Chevrier A. Examining
33. Elizabeth E, Smolen L, Pitts A, Krueger LA, Grimm D.
factors that impact inpatient management of diabetes and Economic impact of converting from pen and 10-ml vial to
the role of insulin pen devices. Can J Diabetes. 2017;41(1): 3-ml vial for insulin delivery in a hospital setting. Hosp Pharm.
102-107. 2014;49(11):1033-1038.
888914
research-article2019
HPXXXX10.1177/0018578719888914Hospital PharmacyBorden et al
FMS
Hospital Pharmacy
Nasal Glucagon
2022, Vol. 57(6) 697–703
© The Author(s) 2019
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/0018578719888914
https://doi.org/10.1177/0018578719888914
journals.sagepub.com/home/hpx
Each month, subscribers to The Formulary Monograph Service receive 5 to 6 well-documented monographs on drugs that are
newly released or are in late phase 3 trials.The monographs are targeted to Pharmacy & Therapeutics Committees. Subscribers
also receive monthly 1-page summary monographs on agents that are useful for agendas and pharmacy/nursing in-services. A
comprehensive target drug utilization evaluation/medication use evaluation (DUE/MUE) is also provided each month. With a
subscription, the monographs are available online to subscribers. Monographs can be customized to meet the needs of a facil-
ity.Through the cooperation of The Formulary, Hospital Pharmacy publishes selected reviews in this column. For more informa-
tion about The Formulary Monograph Service, contact Wolters Kluwer customer service at 866-397-3433.
Generic Name: GLUCAGON NASAL POWDER same 29 amino acids as both recombinant injectable and
Proprietary Name: Baqsimi (Eli Lilly) endogenous human glucagon.5
Approval Rating: 3S
Therapeutic Class: Antidotes, Hypoglycemia
Similar Drugs: Glucagon Injection
Pharmacokinetics
Sound-/Look-Alike Names: Glimepiride, Glipizide, GlucaGen, Glucagon is rapidly absorbed after intranasal administration.
Glucophage, Glucotrol, Glyburide, Glycotrol Peak plasma levels (Cmax) are reached within 15 minutes in
adults and 15 to 20 minutes in children. The apparent volume
of distribution is approximately 885 L. Median half-life is
Indications
approximately 35 minutes in adults and 21 to 31 minutes in
Nasal glucagon is Food and Drug Administration (FDA)- children. Glucagon is degraded in the liver, kidneys, and
approved for the treatment of severe hypoglycemia in patients plasma.1
4 years of age and older with diabetes mellitus.1 In a crossover study comparing intranasal glucagon 3 mg
Glucagon injection is approved for the treatment of severe and intramuscular (IM) glucagon 1 mg in adults with type 1
hypoglycemia in patients with diabetes mellitus.2,3 diabetes, time to maximum change in blood glucose concen-
Glucagon injection is also approved as a diagnostic aid in tration from baseline was approximately 5 minutes slower
radiological examination of the stomach, duodenum, small with intranasal glucagon; however, pharmacologic levels of
bowel, and colon to temporarily diminish intestinal motility.2,3 glucagon were detected within 5 minutes of administration
by both routes. Median time to Cmax was 20 minutes after
intranasal administration and 15 minutes after IM adminis-
Clinical Pharmacology
tration (P < .001). Mean Cmax was lower after intranasal
Glucagon is a hormone normally synthesized in the alpha administration (3155 pg/mL) than after IM administration
cells in the islets of Langerhans in the pancreas. Glucagon (3672 pg/mL; P = .003). In an analysis of administration
primarily influences glycogenolysis in hepatocytes, and suf- when nadir glucose was less than 50 mg/dL, the mean time
ficient hepatic glycogen stores are required for efficacy. required to achieve blood glucose normalization (at least 70
Repeated use of glucagon over a short period of time becomes mg/dL, or at least a 20 mg/dL increase from glucose nadir)
less efficacious.1-4 was 13 minutes for IM glucagon and 16 minutes for intrana-
Recombinant glucagon is a polypeptide hormone identi- sal glucagon (P < .001).6
cal to endogenous human glucagon that raises blood glucose
levels by glycogenolysis. For the injectable pharmaceutical
product available as a powder for reconstitution, glucagon is 1
Washington State University, Spokane, USA
synthesized in a nonpathogenic laboratory strain of
Corresponding Author:
Escherichia coli that has been genetically altered3 or is pro-
Terri L. Levien, Clinical Professor, Pharmacotherapy Department, College
duced by expression of recombinant DNA in a Saccharomyces of Pharmacy and Pharmaceutical Sciences, Washington State University,
cerevisiae vector with subsequent purification.2 The gluca- 412 E. Spokane Falls Blvd, Spokane, WA 99202-2131, USA.
gon used in the nasal formulation is a synthetic form with the Email: levient@wsu.edu
698 Hospital Pharmacy 57(6)
Similar results were observed in a study comparing intra- remaining 15 patients, 14 experienced 1 or more hypo-
nasal glucagon 2 mg and 3 mg with IM glucagon 1 mg in glycemic events and received treatment with nasal gluca-
children and adolescents 4 to 16 years of age with type 1 gon; the 14 patients were included in the efficacy and
diabetes. Both routes produced a rise in glucose of at least 25 safety analyses. Mean patient age was 10.2 years; 64.3%
mg/dL within 20 minutes of dosing. Plasma glucagon levels were male; 100% were white; mean duration of diabetes
rose within 5 minutes of administration by either route, was 6.3 years; and 71.4% were using an insulin pump,
reaching peak glucagon concentrations in a mean of 15 to 20 with a mean total daily insulin dose of 42.3 units.
minutes with intranasal administration and 17 minutes with Intervention: Patients received nasal glucagon 3 mg
IM administration. Peak glucagon levels ranged from 2952 administered as a single dose for the treatment of a mod-
to 5832 pg/mL following intranasal administration and from erate to severe hypoglycemic event. Blood glucose was
4382 to 6343 pg/mL following IM administration.7 measured just before or immediately after treatment and
The presence of nasal congestion or concomitant use of a again at 15, 30, and 45 minutes after dosing. If the patient
nasal decongestant had no impact on the pharmacokinetics did not respond within 15 minutes after nasal glucagon
of glucagon following nasal administration.1,8 administration, the caregiver was instructed to seek emer-
gency medical assistance. During the study, administra-
tion of nasal glucagon was allowed whenever a patient
Comparative Efficacy
experienced a moderate or severe hypoglycemic event,
Indication: Hypoglycemia in Type 1 Diabetes for up to a maximum of 4 events.
Mellitus Results: During the study, the 14 patients experienced 33
moderate hypoglycemic events. All patients had neuro-
Guidelines glycopenic symptoms with a blood glucose level of 70
Guideline: American Diabetes Association: standards of mg/dL or less (range, 42-70 mg/dL) at the time of nasal
medical care in diabetes – 2019 glucagon administration.
Reference: American Diabetes Association9
Comments: Glucagon is used in hypoglycemic emergen- Primary End Point(s)
cies to rapidly raise plasma blood glucose levels in dia- •• Proportion of patients with treatment response
betic patients. Glucagon should be prescribed in (defined as awakening or returning to normal status,
individuals at increased risk of hypoglycemia (blood glu- as judged by caregiver, within 30 minutes following
cose less than 54 mg/dL), so it is available if needed. nasal glucagon administration): All hypoglycemic
Glucagon is indicated especially for those unable or events (100%) resolved, with patients returning to
unwilling to consume carbohydrates by mouth. In addi- normal status within 30 minutes after nasal adminis-
tion, supplemental administration of carbohydrates is rec- tration of glucagon. More than half (54.5%) of
ommended after administration of glucagon. Caregivers hypoglycemic events resolved and patients returned
and individuals in close contact with the patient should to normal status within 10 minutes of nasal gluca-
receive instruction on the use of glucagon and its admin- gon administration. Time to return to normal status
istration. The intranasal form of glucagon is not specifi- after nasal administration was less than 5 minutes in
cally mentioned in the guidelines. 21.2% of patients, 5 to less than 10 minutes in
33.3%, 10 to less than 15 minutes in 12.1%, 15 to
Studies less than 20 minutes in 15.2%, 20 to less than 25
Drug: Nasal Glucagon minutes in 15.2%, and 25 to less than 30 minutes in
Reference: Deeb LC, et al, 20185 3%.
Study Design: Phase 3, prospective, open-label, multi-
center, nonrandomized study Secondary End Point(s)
Study Funding: Eli Lilly & Company •• Blood glucose levels improved over time, from a
Patients: 26 children or adolescents (4 to younger than baseline mean of 55 mg/dL to greater than 70 mg/dL
18 years) with type 1 diabetes mellitus of more than 1 at 15 minutes after nasal glucagon administration;
year’s duration and experiencing moderate or severe levels continued to increase over 45 minutes.
hypoglycemic events. Patients had to be living with 1 or •• Very few of the caregivers had previous experience
more caregivers and in good general health. Exclusion with injectable glucagon. Caregivers reported that the
criteria were pheochromocytoma or insulinoma; cardio- nasal form was easy to very easy to administer in
vascular, gastrointestinal (GI), liver, or kidney disease; or 93.9% of hypoglycemic events. Overall, caregivers
use of systemic beta-blockers, indomethacin, warfarin, or were relatively satisfied, satisfied, or very satisfied
anticholinergic medications. Of the 26 patients enrolled, with the nasal formulation. The glucagon dose was
11 were excluded due to study site noncompliance; of the administered within 30 seconds of the hypoglycemic
Borden et al 699
episode in 60.6% of events and within 2 minutes in minutes in all evaluable events. Of the 157 evaluable
100% of events. hypoglycemic episodes (145 moderate and 12 severe)
treated during the study, 151 (96.2%) resolved within
Comments: The study was conducted at 2 medical centers 30 minutes, and all (100%) severe hypoglycemic epi-
in the United States. Hypoglycemia awareness at baseline sodes resolved within 15 minutes.
was evaluated using the Clarke Hypoglycemia Awareness
Questionnaire. Caregivers and patients were trained in nasal Secondary End Point(s)
glucagon use. Caregivers were given the User-Friendliness
•• Mean blood glucose level improved progressively,
Assessment Questionnaire to describe nasal glucagon ease
from 47.9 mg/dL at the time of nasal glucagon admin-
of use. Patients were given a Nasal Score Questionnaire
istration to 84.4 mg/dL after 15 minutes; levels con-
post episode. Caregiver reports of ease of use of the nasal
tinued to increase further. In 4 of the 6 events that did
glucagon formulation were similar to a smaller study con-
not resolve within 30 minutes of nasal glucagon
ducted to evaluate the ability of primary caregivers and
administration, blood glucose levels were greater than
acquaintances without medical training to administer nasal
70 mg/dL within 30 minutes but did not return to nor-
glucagon versus injectable glucagon during simulated
mal status because of headache and/or nasal irritation;
severe hypoglycemic episodes. The study found that instruc-
in the other 2 events, patients eventually returned to
tion time was faster, retention of information was more
normal status within 45 minutes.
complete, and time to deliver the medication to the patient
•• Caregivers rated the nasal glucagon drug instructions as
and success in delivering the medication were superior in
easy to understand in 91% of events and rated the drug as
the nasal glucagon groups compared with the IM groups for
easy to administer in 80.5% of events. Overall, caregivers
both laypersons and experienced caregivers.10
were satisfied with nasal glucagon in 94.4% of events.
Limitations: This was a small study that did not use ran-
domization, concealed allocation, blinding, or controls.
Comments: Patients were selected from 9 centers in the
The study population was primarily male and all white.
United States and Canada. Hypoglycemia awareness at
Reference: Seaquist ER, et al, 201811,12
baseline was evaluated using the Clarke Hypoglycemia
Study Design: Phase 3, prospective, multicenter, single-
Awareness Questionnaire. Caregivers were trained in
arm, nonrandomized, open-label study
nasal glucagon use. Caregivers were given the User-
Study Funding: Eli Lilly & Company, Locemia Solutions
Friendliness Assessment Questionnaire to describe nasal
Patients: 129 adults (18-75 years of age) with type 1 diabe-
glucagon ease of use. Patients were given a Nasal Score
tes mellitus of more than 1 year’s duration and body mass
Questionnaire post episode. The study demonstrated ease
index (BMI) of 18.5 to 35 kg/m2; patients were otherwise of use and preference for nasal glucagon by caregivers.
healthy. Exclusion criteria were pheochromocytoma or Limitations: This was a small, single-arm study with no
insulinoma, or use of systemic beta-blockers, indometha- comparator group. Participants were predominantly white.
cin, warfarin, or anticholinergic medications. Of the 101
patients who completed the study, 69 patients experiencing Drug: Nasal Glucagon vs Glucagon for Injection
1 or more hypoglycemic episode(s) requiring nasal gluca- Reference: Rickels MR, et al, 20166,13
gon and with evaluable response data were included in the Study Design: Phase 3, randomized, open-label, noninfe-
efficacy analysis. In the main safety analysis population riority, crossover study
(patients who experienced at least 1 hypoglycemic episode Study Funding: Leona M. and Harry B. Helmsley
and received at least 1 nasal glucagon dose; n = 74), mean Charitable Trust, National Center for Research Resources,
patient age was 46.2 years, and 52.7% were female. National Center for Advancing Translational Sciences,
Intervention: Patients received nasal glucagon 3 mg National Institutes of Health
administered as a single dose following the onset of a Patients: 77 adults (18-64 years of age) with type 1 diabetes
hypoglycemic episode. mellitus of at least 2 years’ duration and weighing 50 kg or
more, with BMI between 20 and 35 kg/m2. Exclusion criteria
Results
included severe hypoglycemic episode in the month before
Primary End Point(s)
enrollment; pheochromocytoma or insulinoma; history of
•• Proportion of patients with treatment response (defined seizure disorder; cardiovascular, GI, liver, or kidney disease;
as awakening or returning to normal status within 30 use of systemic beta-blockers; donation of 225 mL or more
minutes following nasal glucagon administration): Of of blood within 8 weeks of enrollment; or consumption of 3
69 patients, 66 (95.7%) awakened or achieved normal or more alcoholic beverages daily. The efficacy analysis
status within 30 minutes of administration of nasal glu- included 75 patients who successfully completed both dos-
cagon in at least 1 evaluable hypoglycemic event; 64 ing visits. Baseline characteristics were similar between the
(92.8%) awakened or achieved normal status within 30 nasal and injection groups.
700 Hospital Pharmacy 57(6)
Table 1. Pooled Adverse Reactions (≥2%) in Adult Patients Table 5. Comparison of the Adverse Events Reported With
With Type 1 and Type 2 Diabetes Receiving Nasal Glucagon in Nasal Glucagon and Injectable Glucagon Dosage Forms.1-3
Clinical Trials.1
Nasal Injectable
Nasal glucagon 3 mg Adverse reaction glucagon glucagon
Adverse reaction (N = 153)
Headache X X
Nausea 26.1% Nausea X X
Headache 18.3% Vomiting X X
Vomiting 15% Upper respiratory tract irritation X
Upper respiratory tract irritationa 12.4% Necrolytic migratory erythema X
Injection-site reaction X
a
Rhinorrhea, nasal discomfort, congestion, cough, and epistaxis.
Table 6. Comparison of Nasal Glucagon and Injectable Glucagon pharmacokinetic profiles to its injectable counterpart. The
Drug Interactions.1-3 nasal device is associated with fewer steps and only requires
Nasal Injectable insertion into 1 nostril for delivery. Outcomes and safety in
glucagon glucagon diabetic patients, especially children, can be greatly
improved by prescription of nasal glucagon rescue kits,
Beta-blockers X X particularly in cases when an inexperienced user must
Anticholinergics X intervene on the patient’s behalf.
Indomethacin X X
Warfarin X X
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect
to the research, authorship, and/or publication of this article.
emergency assistance should be called immediately after
administering the dose. The tip of the device should be
inserted into the nostril and then the plunger pressed in until Funding
the green line is no longer showing. No inhaling is required. The author(s) received no financial support for the research, author-
If the patient responds to treatment, they should be given oral ship, and/or publication of this article.
carbohydrates to restore liver glycogen and prevent recur-
rence of hypoglycemia. If there has been no response after 15 ORCID iDs
minutes, an additional 3 mg dose of nasal glucagon from a Terri L. Levien https://orcid.org/0000-0001-5114-6067
new device may be administered.1 Danial E. Baker https://orcid.org/0000-0002-4605-3357
References
Product Availability and Storage
1. Baqsimi (glucagon nasal powder) [prescribing information].
Glucagon nasal powder was approved on July 24, 2019.14 It is Indianapolis, IN: Lilly USA; 2019.
available as an intranasal device containing one 3 mg dose of 2. GlucaGen (glucagon for injection) [prescribing information].
glucagon as a preservative-free, white powder. The product is Plainsboro, NJ: Novo Nordisk; 2018.
available in cartons that contain 1 or 2 intranasal devices.1 3. Glucagon for injection [prescribing information]. Indianapolis,
Nasal glucagon should be stored at temperatures not IN: Lilly USA; 2018.
4. Oh SH, Darwiche H, Cho JH, Shupe T, Petersen BE.
exceeding 30°C (86°F) in the shrink-wrapped tube pro-
Characterization of a novel functional protein in the pancreatic
vided.1 Removal of the shrink wrapping from the tube prior
islet: islet homeostasis protein regulation of glucagon synthesis
to use may expose the product to moisture and cause it to not in α cells. Pancreas. 2012;41(1):22-30.
work as expected. The shrink wrap should be kept on the 5. Deeb LC, Dulude H, Guzman CB, et al. A phase 3 multicenter,
tube until administration of the product is necessary.1 open-label, prospective study designed to evaluate the effec-
tiveness and ease of use of nasal glucagon in the treatment of
moderate and severe hypoglycemia in children and adolescents
Drug Safety/Risk Evaluation and with type 1 diabetes in the home or school setting. Pediatr
Mitigation Strategy (REMS) Diabetes. 2018;19(5):1007-1013.
6. Rickels MR, Ruedy KJ, Foster NC, et al. Intranasal glucagon
No REMS program has been established for nasal glucagon.14
for treatment of insulin-induced hypoglycemia in adults with
type 1 diabetes: a randomized crossover noninferiority study.
Conclusion Diabetes Care. 2016;39(2):264-270.
7. Sherr JL, Ruedy KJ, Foster NC, et al. Glucagon nasal pow-
Nasal glucagon is FDA-approved for the treatment of der: a promising alternative to intramuscular glucagon in youth
severe hypoglycemia in patients 4 years of age and older with type 1 diabetes. Diabetes Care. 2016;39(4):555-562.
with diabetes mellitus. The standard of care for moderate or 8. Guzman CB, Dulude H, Piché C, et al. Effects of common cold
severe hypoglycemia in patients with diabetes mellitus is and concomitant administration of nasal decongestant on the
IM or subcutaneous administration of glucagon. Glucagon pharmacokinetics and pharmacodynamics of nasal glucagon
emergency kits are carried by many diabetic patients, espe- in otherwise healthy participants: a randomized clinical trial.
cially those treated with insulin. Correct administration of Diabetes Obes Metab. 2018;20(3):646-653.
9. American Diabetes Association. Standards of medical care in
glucagon by a layperson or otherwise inexperienced
diabetes—2019. Diabetes Care. 2019;42(suppl 1):S1-S204.
bystander can be crucial in an emergency, and glucagon for
http://care.diabetesjournals.org/content/42/Supplement_1.
injection kits can be intimidating or difficult to use for the Accessed April 4, 2019.
general public. Instructions and pictures can be insufficient 10. Yale JF, Dulude H, Egeth M, et al. Faster use and fewer failures
to guide in reconstituting and administering the drug. A with needle-free nasal glucagon versus injectable glucagon
nasal glucagon dosage form offers an alternative that is in severe hypoglycemia rescue: a simulation study. Diabetes
easier for laypersons to use and has similar efficacy and Technol Ther. 2017;19(7):423-432.
Borden et al 703
FMS
Hospital Pharmacy
Lefamulin
2022, Vol. 57(6) 704–711
© The Author(s) 2019
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/0018578719897071
https://doi.org/10.1177/0018578719897071
journals.sagepub.com/home/hpx
Abstract
Each month, subscribers to The Formulary Monograph Service receive 5 to 6 well-documented monographs on drugs that
are newly released or are in late phase 3 trials. The monographs are targeted to Pharmacy & Therapeutics Committees.
Subscribers also receive monthly 1-page summary monographs on agents that are useful for agendas and pharmacy/nursing
in-services. A comprehensive target drug utilization evaluation/medication use evaluation (DUE/MUE) is also provided each
month. With a subscription, the monographs are available online to subscribers. Monographs can be customized to meet the
needs of a facility. Through the cooperation of The Formulary, Hospital Pharmacy publishes selected reviews in this column.
For more information about The Formulary Monograph Service, contact Wolters Kluwer customer service at 866-397-3433.
Keywords
anti-infectives, drug information, formulary management/P & T, infectious diseases
Generic Name: Lefamulin tract, sexually transmitted, and skin and skin structure infec-
Proprietary Name: Xenleta (Nabriva Therapeutics US, Inc) tions. The antibacterial spectrum comprises the gram-positive
Approval Rating: 1P bacteria S pneumoniae, S aureus (including methicillin-resis-
Therapeutic Class: Antibiotics, Pleuromutilins tant Staphylococcus aureus [MRSA] and community-acquired
Similar Drugs: Retapamulin MRSA), Streptococcus agalactiae, Streptococcus pyogenes,
Sound- or Look-Alike Names: Leflunomide, Xenical Streptococcus anginosus, Streptococcus mitis, Streptococcus
salivarius, and Enterococcus faecium (including vancomycin-
resistant enterococci); the gram-negative bacteria H influen-
Indications
zae, Haemophilus parainfluenzae, Moraxella catarrhalis, and
Lefamulin is indicated for the treatment of adults with com- Neisseria gonorrhoeae; and the atypical pathogens L pneu-
munity-acquired bacterial pneumonia (CABP) caused by the mophila, C pneumoniae, Chlamydia trachomatis, Mycoplasma
following susceptible microorganisms: Streptococcus pneu- genitalium, and M pneumoniae.1,5-12 Lefamulin is bactericidal
moniae, Staphylococcus aureus (methicillin-susceptible in vitro against S pneumoniae, H influenzae, and M pneu-
Staphylococcus aureus [MSSA] isolates), Haemophilus moniae (including macrolide-resistant strains) and bacterio-
influenzae, Legionella pneumophila, Mycoplasma pneu- static against S aureus and S pyogenes at clinically relevant
moniae, and Chlamydophila pneumoniae.1 concentrations. It is not active against Enterobacteriaceae and
Lefamulin is also undergoing evaluation for use in the Pseudomonas aeruginosa.1,12
treatment of health care–acquired bacterial pneumonia/ven- In cross-resistance studies, lefamulin remained active
tilator-associated bacterial pneumonia, osteomyelitis, pros- against some clinical isolates resistant to macrolides, tetra-
thetic joint infections, acute bacterial skin and skin structure cyclines, beta-lactams, quinolones, sulfamethoxazole/trim-
infections, and sexually transmitted diseases.2,3 ethoprim, mupirocin, and glycopeptides.1,11 Resistance due
to spontaneous mutations has been observed infrequently.
Resistance mechanisms affecting lefamulin include specific
Clinical Pharmacology protection or modification of the ribosomal target by ABC-F
Lefamulin, a pleuromutilin derivative, is a semisynthetic proteins, Cfr methyltransferase, or mutations of ribosomal
antibacterial agent for oral and intravenous (IV) administra- proteins L3 and L4.1,13 Cfr methyltransferase has the
tion that inhibits bacterial protein synthesis by selectively
binding to the peptidyl transferase center of the 50S subunit 1Washington State University, Spokane, USA
of the bacterial ribosome, preventing correct positioning of
the 3′ CCA ends of tRNAs (transfer ribonucleic acids) for Corresponding Author:
Terri L. Levien, Pharmacotherapy Department, College of Pharmacy and
peptide transfer.1,4,5 Pharmaceutical Sciences, Washington State University, 412 E. Spokane
Lefamulin’s in vitro antibacterial profile includes a broad Falls Blvd, Spokane, WA 99202-2131, USA.
spectrum of common bacterial pathogens causing respiratory Email: levient@wsu.edu
Kozhokar et al 705
Table 1. Pharmacokinetics of Single or Multiple Dosesa of Mean plasma protein binding of lefamulin ranges from
Lefamulin in Patients With Community-Acquired Bacterial 94.8% at 2.35 mcg/mL to 97.1% at 0.25 mcg/mL in healthy
Pneumonia.1 adults.1 The mean steady-state volume of the distribution of
Mean value after Mean value at lefamulin is 86.1 L (range, 34.2-153 L) in patients with CABP
Parameter Route single dose steady state after IV administration of lefamulin. Following single IV
administration of lefamulin 150 mg to healthy participants,
Cmax, mcg/mL IV 3.5 3.6
the highest lefamulin epithelial lining fluid (ELF) concentra-
Oral 2.24 2.24
tions were observed at the end of infusion. Mean ELF and
Cmin, mcg/mL IV 0.398 0.573
plasma AUC0-8 were 3.87 and 5.27 mcg h/mL, respectively.
Oral 0.593 0.765
AUC0-24h, mcg h/mL IV 27 28.6
Following a single IV administration, the estimated ratio of
Oral 30.7 32.7 ELF AUC to unbound plasma AUC is approximately 15.1
Mean total body clearance of lefamulin is 11.9 L/h (range,
Note. IV = intravenous; AUC = area under the curve. 2.94-30 L/h) in patients with CABP after IV administration.
a150 mg IV infused over 60 minutes (every 12 hours for multiple dosing);
Mean elimination half-life of lefamulin is approximately 8
600 mg orally 1 hour before or 2 hours after a meal (every 12 hours for
multiple dosing). hours (range, 3-20 hours) in patients with CABP.1
Lefamulin is primarily metabolized by CYP3A4. In
healthy adult participants administered radiolabeled lefam-
potential to mediate cross-resistance between lefamulin and ulin IV and oral doses, the mean total radioactivity excreted
phenicols, lincosamides, oxazolidinones, and streptogramin in feces was 77.3% (4.2%-9.1% unchanged) and 88.5%
A antibacterials.1 (7.8%-24.8% unchanged), respectively, and mean total
In vitro, lefamulin has demonstrated synergy with doxy- radioactivity excreted in urine was 15.5% (9.6%-14.1%
cycline against S aureus. No antagonism was observed unchanged) and 5.3% (percentage unchanged not deter-
between lefamulin and other antibacterials (eg, amikacin, mined), respectively.1
azithromycin, aztreonam, ceftriaxone, levofloxacin, line- Pharmacokinetics did not differ based on age, sex, race,
zolid, meropenem, penicillin, tigecycline, sulfamethoxazole/ weight, or renal impairment, including patients undergoing
trimethoprim, vancomycin).1,14 hemodialysis.1 Lefamulin pharmacokinetics have been eval-
uated in noninfected participants with normal hepatic func-
Pharmacokinetics tion and with moderate (Child-Pugh class B) or severe
(Child-Pugh class C) hepatic impairment following adminis-
Mean pharmacokinetic parameters observed after lefamulin tration of the injection form. Following IV administration,
single doses of 150 mg IV and 600 mg orally in the fasted half-life was prolonged in participants with severe hepatic
state were comparable.1,15 Mean oral bioavailability of impairment compared with participants with normal hepatic
lefamulin is approximately 25%, and peak plasma concentra- function (17.5 vs 11.5 hours). Protein binding of lefamulin
tion (Cmax) occurred 0.88 to 2 hours after administration to was also reduced in participants with hepatic impairment,
healthy participants.1 Bioequivalence was not established resulting in unbound active lefamulin concentrations that
between oral lefamulin in the fed state and IV or oral admin- increased with the degree of hepatic impairment. Unbound
istration in the fasted state; exposure was reduced and time to lefamulin plasma AUC0-∞ was increased 3-fold in partici-
peak concentration was prolonged (median, 4.5 hours) when pants with severe hepatic impairment compared with partici-
administered with food.1,15,16 Coadministration of a single pants with normal hepatic function. The effects on lefamulin
oral dose of lefamulin 600 mg with a high-fat (approximately pharmacokinetics following oral administration in partici-
50% of total calories from fat), high-calorie (approximately pants with hepatic impairment have not been assessed; the
800-1000 calories) breakfast slightly reduced bioavailability. oral formulation is not recommended in patients with moder-
The mean relative reduction for oral lefamulin (fasted vs fed) ate or severe hepatic impairment.1
was on average 22.9% (90% confidence interval [CI], 12.2%-
32.3%) for Cmax and 18.43% (90% CI, 11.7%-24.7%) for
area under the curve (AUC0-∞). Following a single IV dose, Comparative Efficacy
the AUC of lefamulin increased approximately dose propor- Indication: Community-Acquired Pneumonia
tionally, whereas Cmax increased less than dose proportion-
ally over a dose range of 25 to 400 mg. Following a single Guidelines
oral dose, lefamulin AUC increased more than dose propor- Guideline: Diagnosis and treatment of adults with com-
tionally over a dose range of 500 to 750 mg. Mean lefamulin munity-acquired pneumonia: An official clinical practice
AUC0-24h and Cmax were 73% and 30% higher, respectively, guideline of the American Thoracic Society and Infectious
in patients with CABP compared with healthy participants.1 Diseases Society of America
The pharmacokinetics of oral and IV lefamulin in patients Reference: American Thoracic Society/Infectious Diseases
with CABP are summarized in Table 1.1 Society of America17
706 Hospital Pharmacy 57(6)
Comments: For healthy outpatient adults with commu- CABP caused by a pathogen resistant to any of the study
nity-acquired pneumonia without comorbidities (eg, drugs or attributable to etiologies other than community-
chronic heart, lung, liver, or renal disease; diabetes mel- acquired bacterial pathogens; noninfectious cause of pul-
litus; alcoholism; malignancy; asplenia) or risk factors monary infiltrates; confirmed or suspected pleural
for antibiotic-resistant pathogens (MRSA, P aerugi- empyema; or need for mechanical ventilation. Patients
nosa), treatment with amoxicillin, doxycycline, or a mac- were mostly male (60%) and white (87%). Median
rolide (azithromycin, clarithromycin, clarithromycin patient age was 62 years (range, 19-91 years); more
extended release) is recommended; macrolide therapy patients in the lefamulin group were 65 years or older
should be considered only in areas with less than 25% (47.8% vs 39.3%) and 75 years or older (21% vs 15.3%).
pneumococcal resistance to macrolides. For outpatient Approximately 53% of patients had creatinine clearance
adults with comorbidities, the guidelines recommend (CrCl) less than 90 mL/min. Approximately 72% of
combination therapy with amoxicillin/clavulanate or patients were classified as PORT risk class III and 28%
cephalosporin plus a macrolide (azithromycin, clarithro- as PORT risk class IV or V. Common comorbid condi-
mycin) or doxycycline, or monotherapy with a respira- tions included hypertension (41%), asthma or chronic
tory fluoroquinolone (levofloxacin, moxifloxacin, obstructive pulmonary disease (COPD; 17%), and diabe-
gemifloxacin). Treatment strategies for inpatients with tes mellitus (13%). The most common pathogens at base-
community-acquired pneumonia are based on severity line were S pneumoniae, H influenzae, M catarrhalis, M
and risk for drug resistance. For inpatient adults with pneumoniae, L pneumophila, and C pneumoniae.
nonsevere community-acquired pneumonia without risk Intervention: Patients were randomized (1:1) to receive
factors for resistance, combination therapy with a beta- lefamulin 150 mg IV every 12 hours (n = 276) or moxi-
lactam and macrolide (ampicillin/sulbactam, cefotaxime, floxacin 400 mg every 24 hours (n = 275). Moxifloxacin-
ceftriaxone, or ceftaroline plus azithromycin or clarithro- treated patients received alternating doses of a placebo to
mycin) or monotherapy with a respiratory fluoroquino- maintain blinding. After 3 days (6 doses), patients could
lone (levofloxacin, moxifloxacin) is recommended; if be switched to oral study drug if prespecified improve-
contraindications to both macrolides and fluoroquino- ment criteria were met (lefamulin 600 mg every 12 hours
lones exist, combination therapy with a beta-lactam and or moxifloxacin 400 mg every 24 hours). If MRSA was
doxycycline is an option. For inpatients with severe com- suspected at screening, patients in the moxifloxacin group
munity-acquired pneumonia without risk factors for received adjunctive linezolid and patients in the lefamulin
resistance, a beta-lactam plus a macrolide or a beta-lac- group received linezolid placebo. Treatment duration
tam plus a fluoroquinolone is recommended. The guide- ranged from 5 to 10 days. In the initial protocol, patients
lines note that further research is needed regarding the with CABP due to MRSA or L pneumophila and patients
use of newer agents, including lefamulin, for the treat- with S pneumoniae with bacteremia received 10 days of
ment of community-acquired pneumonia in adults. active treatment, whereas all other lefamulin-treated
patients received 5 days of active therapy and all other
Studies
moxifloxacin-treated patients received 7 days of active
Drug: Lefamulin vs Moxifloxacin
therapy. After a protocol amendment, therapy duration
Reference: File TM, et al, 2019 (LEAP 1 trial)1,18,19
changed to 7 days for both lefamulin- and moxifloxacin-
Study Design: Phase 3, randomized, double-blind, dou-
treated patients, except for patients with MRSA who
ble-dummy, active-controlled, parallel-group, multi-
received 10 days of active therapy. The median duration
center, noninferiority study
of IV treatment was 7 days in the lefamulin group and 6
Study Funding: Nabriva Therapeutics
days in the moxifloxacin group, and the median duration
Patients: 551 patients (at least 18 years of age) with
of oral therapy was 4 days in both groups. A switch from
acute CABP, with radiographic findings suggestive of
IV to oral therapy occurred in 38.1% of lefamulin-treated
pneumonia, Pneumonia Outcomes Research Team
patients and in 44.3% of moxifloxacin-treated patients.
(PORT) risk class of at least III, and at least 3 of 4 symp-
Results
toms consistent with the disease (new or increased cough,
Primary End Point(s)
purulent sputum production, chest pain, dyspnea).
Exclusion criteria were receipt of more than 1 dose of a •• Rate of early clinical response (ECR; defined as an
short-acting oral or IV antibacterial for CABP within 72 improvement in 2 or more CABP signs/symptoms, no
hours before randomization; need for concomitant sys- worsening in any signs/symptoms, and no use of a con-
temic antibacterial therapy potentially effective against comitant, nonstudy antibiotic for CABP) in the inten-
CABP pathogens; hospitalization for 2 or more days tion-to-treat (ITT) population at 96 (±24) hours after
within 90 days before symptom onset, or residence in a the first study drug dose (Food and Drug Administration
nursing home or long-term health care facility within 30 [FDA]-required end point): 87.3% of patients with
days before symptom onset; confirmed or suspected lefamulin compared with 90.2% with moxifloxacin
Kozhokar et al 707
(treatment difference, −2.9% [95% CI, −8.5%-2.8%]); lefamulin and moxifloxacin treatment groups. The sub-
noninferiority of lefamulin was demonstrated. population analysis suggested that moxifloxacin was
•• Rate of investigator assessment of clinical response more effective than lefamulin in patients younger than 65
(IACR; defined as the resolution or improvement of years based on rates of both ECR (93.4% with moxifloxa-
CABP such that no additional antibiotic therapy was cin vs 84.7% with lefamulin) and IACR (88% vs 75.9%),
administered for CABP) at test of cure (5-10 days as well as in patients with minor disease severity accord-
after the last dose of the study drug) in the modified ing to American Thoracic Society criteria (93.8% vs
intention-to-treat (mITT) and clinically evaluable 75.9% for ECR and 89.6% vs 62.3% for IACR); in
(CE) populations (European Medicines Agency patients younger than 65 years, response rates were sig-
[EMA] coprimary end point): nificantly lower with lefamulin than with moxifloxacin in
|| mITT population: 81.7% of patients treated the presence of minor disease severity compared with
with lefamulin compared with 84.2% treated those not meeting criteria for minor disease severity.
with moxifloxacin (treatment difference, Limitations: This study included a small number of
−2.6% [95% CI, −8.9%-3.9%]); noninferiority patients with severe disease (PORT risk class V); enroll-
of lefamulin was demonstrated. ment from North America, Western Europe, and Latin
|| CE population: 86.9% of patients treated with America was low, precluding a sensitivity analysis by
lefamulin compared with 89.4% treated with geographic region.
moxifloxacin (treatment difference, −2.5%
[95% CI, −8.4%-3.4%]); noninferiority of
Reference: Alexander E, et al, 2019 (LEAP 2 trial)1,20-22
lefamulin was demonstrated.
Study Design: Phase 3, randomized, double-blind, dou-
ble-dummy, multicenter, noninferiority study
Secondary End Point(s) Study Funding: Nabriva Therapeutics
Patients: 738 patients (at least 18 years of age) with
•• Response (ECR) and success (IACR) by baseline
CABP and at least 3 of 4 symptoms consistent with the
pathogen in the microbiological ITT population did
disease (dyspnea, new or increased cough, purulent spu-
not differ substantially between the treatment groups,
tum production, chest pain) and a PORT risk class of II
but comparisons were limited by small numbers of
to IV. Exclusion criteria were receipt of more than 1 dose
patients in the subgroups by pathogen. The ECR
of a short-acting oral or IV antibacterial for CABP within
ranged from 84.2% (M pneumoniae) to 100% (S
72 hours before randomization; hospitalization for 2 or
aureus) for lefamulin and from 85.7% (L pneumoph-
more days within the past 90 days; confirmed or sus-
ila) to 100% (M catarrhalis) for moxifloxacin. The
pected MRSA; risk of major cardiac events or dysfunc-
IACR ranged from 72.7% (C pneumoniae) to 84.9% (S
tion; significant hepatic, immunologic, or hematologic
pneumoniae) for lefamulin and from 68.4% (C pneu-
diseases; or severe renal impairment (estimated CrCl
moniae) to 100% (M catarrhalis) for moxifloxacin.
less than 30 mL/min). Patients were mostly white (74%),
and 52% were male. Median patient age was 59 years
Other End Point(s) (range, 19-97 years); approximately 16% were 75 years
or older. Approximately 50% had CrCl of less than 90
•• The most common treatment-emergent adverse events
mL/min. Approximately 50% of patients were classified
were hypokalemia, nausea, insomnia, and infusion-
as PORT risk class II and 49% as PORT risk class III or
site pain in the lefamulin group (each in 2.9% of
IV. Common comorbid conditions included hyperten-
patients) and diarrhea in the moxifloxacin group
sion (36%), asthma or COPD (16%), and diabetes mel-
(7.7%). Rates of study drug discontinuation due to
litus (13%).
treatment-emergent adverse events were 2.9% for
Intervention: Patients were randomized (1:1) to receive
lefamulin and 4.4% for moxifloxacin.
lefamulin 600 mg orally every 12 hours for 5 days (n =
370) or moxifloxacin 400 mg orally every 24 hours for 7
Comments: Most patients were enrolled in Eastern
days (n = 368).
Europe (79%); less than 1% were enrolled in North
Results: The same FDA- and EMA-defined primary end
America. Baseline pathogens were similar between the 2
points from Leap 1 were evaluated.
groups. Noninferiority of lefamulin was concluded if the
Primary End Point(s)
lower limit of the 2-sided 95% CI for the treatment differ-
ence in ECR responder rates was greater than −12.5% and •• Rate of ECR in the ITT population at 96 (±24) hours
if the lower limit of the 2-sided 95% CI for the treatment after the first dose of the study drug: 90.8% of patients
difference in IACR success rates was greater than −10% with lefamulin compared with 90.8% with moxifloxacin
for both the mITT and CE populations. The ECR and (treatment difference, 0.1% [95% CI, −4.4%-4.5%]);
IACR rates by PORT risk class did not differ between the noninferiority of lefamulin was demonstrated.
708 Hospital Pharmacy 57(6)
•• Rate of IACR at test of cure (5-10 days after the last Lefamulin is also contraindicated with the use of sensitive
dose) in mITT and CE populations: CYP3A4 substrates that prolong the QT interval (eg, pimo-
|| mITT population: 87.5% of patients with lefamu- zide) because of the potential for increased plasma concen-
lin and 89.1% with moxifloxacin (treatment dif- tration of these drugs leading to QT prolongation and cases
ference, −1.6% [95% CI, −6.3%-3.1%]). of torsades de pointes.1
|| CE population: 89.7% of patients with lefamulin
Table 2. Adverse Reactions (≥2% Incidence) in Patients Receiving Lefamulin (Lefamulin Evaluation Against Pneumonia 1 Study Safety
Population).1
Table 3. Adverse Reactions (≥2% Incidence) in Patients Receiving Lefamulin (Lefamulin Evaluation Against Pneumonia 2 Study Safety
Population).1
Oral dosing
including QT prolongation, a woman should pump and dis- which may increase the risk of adverse reactions.
card her milk for the duration of treatment with lefamulin Concomitant use of oral lefamulin with strong CYP3A4
and for 2 days after the final dose.1 inhibitors or P-gp inhibitors should be avoided; patients
Safety and efficacy of lefamulin have not been established receiving oral lefamulin with moderate CYP3A inhibitors
in patients younger than 18 years.1 Initial pediatric studies or P-gp inhibitors should be monitored for lefamulin
are expected to be completed in 2024.23 adverse effects.1
Concomitant use of lefamulin tablets with sensitive
Adverse Reactions CYP3A4 substrates increases AUC and Cmax of the CYP3A4
substrates, which may increase the risk of toxicities associ-
The most common adverse reactions reported with lefamulin
ated with cardiac conduction. Concomitant use of oral
include diarrhea, administration-site reactions, hepatic
lefamulin with CYP3A substrates known to prolong the QT
enzyme elevation, nausea, hypokalemia, insomnia, and
interval is contraindicated. Close monitoring for adverse
headache.1 Tables 2 and 3 summarize the frequency of reac-
effects is required when using oral lefamulin concomitantly
tions with the injection and oral formulations.
with alprazolam, diltiazem, verapamil, simvastatin, or var-
denafil. Concomitant use of injectable lefamulin with
Drug Interactions CYP3A substrates does not affect exposure to the CYP3A
Concomitant use of oral or IV lefamulin with strong CYP3A4 substrates.1
inducers or P-gp inducers decreases lefamulin AUC and The pharmacodynamic interaction potential to prolong
Cmax, which may reduce the efficacy of lefamulin. the QT interval between lefamulin and other drugs that effect
Concomitant use of lefamulin with strong CYP3A4 inducers cardiac conduction is unknown. Concomitant use of inject-
or P-gp inducers should be avoided unless the benefit out- able or oral lefamulin with drugs known to affect cardiac
weighs the risk.1 conduction (eg, class IA and III antiarrhythmics, antipsy-
Concomitant use of oral lefamulin with strong CYP3A chotics, erythromycin, moxifloxacin, tricyclic antidepres-
inhibitors or P-gp inhibitors increases lefamulin AUC, sants) should be avoided.1
710 Hospital Pharmacy 57(6)
extensively drug-resistant Neisseria gonorrhoeae. Antimicrob 16. Zhang L, Wicha WW, Bhavnani SM, Rubino CM. Prediction
Agents Chemother. 2017;61(11):e01497-17. of lefamulin epithelial lining fluid penetration after intrave-
7. Paukner S, Sader HS, Ivezic-Schoenfeld Z, Jones RN. nous and oral administration using phase 1 data and popula-
Antimicrobial activity of the pleuromutilin antibiotic BC-3781 tion pharmacokinetics methods. J Antimicrob Chemother.
against bacterial pathogens isolated in the SENTRY anti- 2019;74(suppl 3):iii27-iii34.
microbial surveillance program in 2010. Antimicrob Agents 17. Metlay JP, Waterer GW, Long AC, et al. Diagnosis and treat-
Chemother. 2013;57(9):4489-4495. ment of adults with community-acquired pneumonia: an
8. Paukner S, Gruss A, Jensen JS. In vitro activity of lefamulin official clinical practice guideline of the American Thoracic
against sexually transmitted bacterial pathogens. Antimicrob Society and Infectious Diseases Society of America. Am J
Agents Chemother. 2018;62(5):e02380-17. Respir Crit Care Med. 2019;200(7):e45-e67.
9. Paukner S, Gelone SP, Arends SJR, Flamm RK, Sader HS. 18. File TM Jr, Goldberg L, Das A, et al. Efficacy and safety of
Antibacterial activity of lefamulin against pathogens most intravenous-to-oral lefamulin, a pleuromutilin antibiotic, for
commonly causing community-acquired bacterial pneumonia: the treatment of community-acquired bacterial pneumonia: the
SENTRY antimicrobial surveillance program (2015-2016). phase III Lefamulin Evaluation Against Pneumonia (LEAP 1)
Antimicrob Agents Chemother. 2019;63(4):e02161-18. trial. Clin Infect Dis. 2019;69:1856-1867.
10. Sader HS, Biedenbach DJ, Paukner S, Ivezic-Schoenfeld Z, 19. Nabriva Therapeutics AG. Study to compare lefamulin to moxi-
Jones RN. Antimicrobial activity of the investigational pleu- floxacin (with or without linezolid) for the treatment of adults
romutilin compound BC-3781 tested against gram-positive with pneumonia (LEAP). NLM Identifier: NCT02559310.
organisms commonly associated with acute bacterial skin ClinicalTrials.gov. https://clinicaltrials.gov/ct2/show/NCT02
and skin structure infections. Antimicrob Agents Chemother. 559310?term=lefamulin&rank=3. Updated March 29, 2018.
2012;56(3):1619-1623. Accessed September 5, 2019.
11. Sader HS, Paukner S, Ivezic-Schoenfeld Z, Biedenbach DJ, 20. Alexander E, Goldberg L, Das A, et al. Oral lefamulin is
Schmitz FJ, Jones RN. Antimicrobial activity of the novel pleu- safe and effective in the treatment of adults with community-
romutilin antibiotic BC-3781 against organisms responsible for acquired bacterial pneumonia (CABP): results of Lefamulin
community-acquired respiratory tract infections (CARTIs). J Evaluation Against Pneumonia (LEAP 2) study. Open Forum
Antimicrob Chemother. 2012;67(5):1170-1175. Infect Dis. 2018;5(suppl 1):S761.
12. Waites KB, Crabb DM, Duffy LB, Jensen JS, Liu Y,
21. Alexander E, Goldberg L, Das A, et al. Oral lefamulin is
Paukner S. In vitro activities of lefamulin and other anti- safe and effective in the treatment of adults with community-
microbial agents against macrolide-susceptible and mac- acquired bacterial pneumonia (CABP): results of Lefamulin
rolide-resistant Mycoplasma pneumoniae from the United Evaluation Against Pneumonia (LEAP 2) study [poster].
States, Europe, and China. Antimicrob Agents Chemother. 22nd Annual Making a Difference in Infectious Diseases
2017;61(2):e02008-16. Meeting; May 8-11, 2019; Orlando, FL. Poster 135E. https://
13. Mendes RE, Paukner S, Doyle TB, Gelone SP, Flamm RK, www.nabriva.com/Portals/0/Nabriva/Posters/MAD-ID/
Sader HS. Low prevalence of gram-positive isolates show- MAD-ID%202019%20LEAP2%20Poster,%20Final.pdf.
ing elevated lefamulin MIC results during the SENTRY Accessed October 2, 2019.
surveillance program for 2015-2016 and characterization 22. Nabriva Therapeutics AG. Study to compare lefamulin to
of resistance mechanisms. Antimicrob Agents Chemother. moxifloxacin for the treatment of adults with pneumonia
2019;63(4):e02158-18. (LEAP2). NLM Identifier: NCT02813694. ClinicalTrials.gov.
14. Veve MP, Wagner JL. Lefamulin: review of a promising novel https://clinicaltrials.gov/ct2/show/NCT02813694?term=lefam
pleuromutilin antibiotic. Pharmacotherapy. 2018;38(9): ulin&rank=2. Accessed September 5, 2019.
935-946. 23. Cox EM. NDA approval letter: Xenleta (lefamulin)
15. Wicha WW, Prince WT, Lell C, Heilmayer W, Gelone
(NDA 211672 211673). Food and Drug Administration.
SP. Pharmacokinetics and tolerability of lefamulin follow- https://www.accessdata.fda.gov/drugsatfda_docs/
ing intravenous and oral dosing. J Antimicrob Chemother. appletter/2019/211672Orig1s000,%20211673Orig1s000ltr.
2019;74(suppl 3):iii19-iii26. pdf. Published August 19, 2019. Accessed September 4, 2019.
1078815
research-article2022
HPXXXX10.1177/00185787221078815Hospital PharmacyNguyen et al
Original Article
Hospital Pharmacy
Abstract
Background: Community-acquired pneumonia continues to be a major cause of morbidity and mortality. Hospital
readmissions following community-acquired pneumonia are linked to significant cost of care and medical burdens. This
study aimed to determine the incidence and reasons for readmission as well as to assess factors associated with short-term
hospital readmission among community-acquired pneumonia patients. Methods: A retrospective, cross-sectional study
was conducted on 582 medical records of community-acquired pneumonia inpatients from December 2018 to December
2019 at an 800-bed tertiary hospital in Ho Chi Minh City, Vietnam. We collected data on patient characteristics, pneumonia
severity at hospital admission, microbiology and antibiotic resistance, appropriateness of empiric antibiotic therapies, and the
readmissions information. Multivariable logistic regression analyses were performed to identify factors associated with 30-day
hospital readmission. Results: Of the 582 hospitalized community-acquired pneumonia patients, 11.9% were readmitted to
the hospital within 30 days. About half of the cases (43.5%) were due to pneumonia. Multidrug-resistant bacteria accounted
for 43.2% of the pathogen isolates. A high Charlson comorbidity index (aOR, 1.40; CI 95%, 1.08-1.82) and multidrug-resistant
infection (aOR, 2.63; CI 95%, 1.05-6.56) were associated with higher odds of all-cause readmission. Conclusions: Hospital
readmissions within 30 days occurred frequently among community-acquired pneumonia inpatients, and the most common
reason for readmission recorded was pneumonia-related. Monitoring closely patients with multimorbidity or multidrug-
resistant infections may improve treatment outcomes.
Keywords
community-acquired pneumonia, readmission, antibiotic, antibiotic resistance.
the impacts of multidrug-resistant (MDR) infections on read- We tracked the 30-day readmission information by screen-
mission outcome among CAP patients. Meanwhile, antimi- ing medical records as each patient had a unique patient code
crobial resistance is a major healthcare problem worldwide for every admission at the hospital. If a patient was admitted
and also in Vietnam, given that a high prevalence of resis- to the hospital more than 1 time within 30 days after dis-
tance to common antimicrobials is seen in both nosocomial charge, only the first readmission was considered as an index
and community-acquired infections.10–12 readmission.
Because pneumonia hospital readmission is one of the
factors associated with a rise in healthcare expenditure and
Measurements
medical resources,13,14 identifying patients who are at risk for
readmissions is important to both healthcare professionals Eligible medical records were reviewed to extract data on
and policy makers. patient demographics, microbiology and antibiotic resis-
Therefore, this study aimed to determine the following: tance, antibiotic therapies and their appropriateness, and
(1) incidence of 30-day hospital readmission, (2) primary readmission information.
reason for readmission (pneumonia vs non-pneumonia Clinical status and laboratory test results at the time of
related) and (3) patient specific risk factors associated with admission were collected to evaluate the severity of pneumo-
readmissions among hospitalized CAP patients. nia using the CURB-65 score (score of 0 or 1: mild pneumo-
nia; score of 2: moderate pneumonia; and score of 3 or above:
severe pneumonia).15 Comorbidity burden was assessed
Methods using the Charlson comorbidity index (CCI),16 which assigns
Ethics approval was obtained from the Institutional Review a weighted score to each comorbid condition depending on
Board of Nguyen Trai Hospital (Approval reference number: the risk associated with one-year mortality.
1064/QĐ-BVNT; Date of approval: November 20, 2020). Information on microbiology and susceptibility tests was
The procedures used in this study adhere to the tenets of the collected using samples from lower respiratory tract (sputum
Declaration of Helsinki. or bronchoalveolar lavage fluid) or blood samples which
were acquired within 48 hours of hospital admissions. Gram
staining, conventional bacterial culture, and antimicrobial
Study Design and Setting susceptibility testing for typical bacteria were performed.
A retrospective, cross-sectional study was conducted on Disk diffusion testing was used to determine the antimicro-
adults 18 years of age or older, who were hospitalized with a bial susceptibility of the isolated pathogens. Tests for the
CAP diagnosis between December 2018 and December detection of atypical bacteria and viruses were not available
2019. This study was carried out at an 800-bed tertiary hos- at the hospital’s laboratory. A pathogen was considered to be
pital at Ho Chi Minh City, Vietnam. The hospital currently MDR if it was resistant to at least 3 antimicrobial agents
has 23 wards and provides general healthcare services for belonging to 3 different antimicrobial classes.17
approximately 1000 outpatients, 600 inpatients, and 200 We assessed the appropriateness of the empiric antibiotic
emergency department visits per day. selection using the 2007 ATS/IDSA consensus guidelines4
and hospital guidelines for CAP. The 2019 IDSA CAP guide-
line recommendations were not applied at the time of the
Participants study, as it takes time for new recommendations to be imple-
All patients hospitalized with a primary diagnosis of CAP mented into clinical practice.
based on International Classification of Diseases, Tenth Clinical instability at hospital discharge was defined
Revision (ICD-10) codes were included in the study if they using the 2007 IDSA/ATS criteria.4 In brief, a patient with
were at least 18 years of age and hospitalized for a minimum any of the following criteria within 24 hours before dis-
of 5 days with a diagnosis of CAP (ICD-10 codes J12-J18) charge was considered to be unstable: (i) temperature >
between December 2018 and December 2019. 37.8°C, (ii) heart rate > 100 beats per min, (iii) systolic
We excluded patients who (i) were hospitalized ≥ 2 days blood pressure < 90 mmHg, (iv) respiratory rate > 24
during the prior 90 days; (ii) had unresolved or worsening breaths per min, (v) oxygen saturation <90% or pO2 <
symptoms of pneumonia at discharge; (iii) were transferred to 60 mmHg, (vi) abnormal mental status or (vii) inability to
another healthcare center; (iii) discharged themselves against eat. For those whose arterial blood gas results were not
medical advice; (iv) died in the hospital; (v) were known to available, oxygenation was considered to be unstable if the
have active pulmonary tuberculosis at the time of admission patients experienced dyspnea (at room air), required oxygen
(ICD-10 codes A15-A16), and (vi) were diagnosed with HIV support within 24 hours prior to discharge or had a SpO2
infection (ICD-10 codes B20-B24, R75, Z21). value lower than 90%.18
714 Hospital Pharmacy 57(6)
A hospital readmission was considered pneumonia-related discharge, 10 were transferred to another healthcare center,
if the patient had a primary diagnosis of pneumonia based on and 43 discharged themselves against medical advice. After
ICD-10 codes J12-J18. Patients were categorized as having a excluding these patients, the final number of index hospital
pneumonia-unrelated readmission if their primary diagnoses admissions for CAP was 582 (Figure 1).
were different from pneumonia.
Demographic and Clinical Characteristics of the
Statistical Methods Study Population
We compared the characteristics between patients who had More than 3 quarters of the study population (76.1%) were
all-cause hospital readmissions and those who did not. For over the age of 65 years. Females accounted for more than
these analyses, continuous variables were compared using a half of the study population (54.6%). CURB-65 score calcu-
t-test (normally distributed data) or Mann–Whitney test lation was performed on 579 patients to determine the pneu-
(non-normally distributed data), and categorical variables monia severity, with more than half (62.2%) having mild
were compared using a Chi-square test or Fisher’s exact test. pneumonia. The median CCI was 1 (1-2), with 74.1% of the
Multivariable logistic regression analyses were performed to patients having a CCI ≥ 1. Hypertension and coronary heart
identify factors associated with all-cause readmissions. We disease were the most frequent comorbidities observed in the
included the following factors in the analyses (considering study. Median duration of antibiotic therapy and length of
their potential association with hospital readmissions hospital stay were 13 days and 14 days, respectively. There
described in prior studies and based on clinical practice): (i) were 579 out of 582 patients treated with an empiric antibi-
age; (ii) sex (male/female); (iii) CURB-65 score; (iv) CCI (v) otic therapy, the other 3 patients had delayed antibiotic treat-
MDR infection (yes/no); (vi) duration of antibiotic therapy; ment until the antimicrobial susceptibility testing results
(vii) appropriate empiric antibiotic therapy (yes/no); (viii) were available. Appropriate use of empiric antibiotic therapy
number of clinical instabilities at discharge. was observed in 51.6% of the patients. There were 106
All analyses were performed using SPSS version 20.0 patients (18.2%) discharged with at least one clinical insta-
with the significance level of 5%. bility factor. Detailed information about characteristics of the
study population is provided in Table 1.
Results
Characteristics of Pathogens Causing CAP
Between December 2018 and December 2019, there were
684 hospitalized CAP patients discharged alive. Of those, 49 Among 582 patients hospitalized for CAP, bacterial culture
had unresolved or worsening symptoms of pneumonia at identification was indicated in 301 patients within 48 hours
Nguyen et al 715
Table 1. Demographic and Clinical Characteristics of the Study Reasons for Hospital Readmission
Population (n = 582).
Sixty-nine out of 582 CAP patients (11.9%) were readmitted
Characteristics Frequency (%) to the hospital within 30 days after discharge. The most sig-
Sex nificant cause of hospital readmission was pneumonia
Male 264 (45.4) (36.2%), followed by cardiovascular causes (29.0%).
Age Reasons for rehospitalizations were illustrated in Table 3.
≥65 years 443 (76.1)
Median 76 (65-83) Factors Associated With Hospital Readmission
Comorbidities
Within 30 days
Hypertension 483 (83.0)
Coronary heart disease 373 (64.1) A higher percentage of patients above the age of 65 years was
Cerebrovascular disease 96 (16.5) observed in the 30-day readmission group compared to the
Congestive heart failure 102 (17.5) non-readmisison group (82.6% vs 75.2%, P > .05). Patients
Diabetes mellitus 195 (33.5) who had readmission within 30 days had higher prevalence
Chronic kidney disease 42 (7.2) of comorbidities, specifically congestive heart failure (34.8%
Chronic obstructive pulmonary 81 (13.9) vs 15.2%, P < .001) and chronic kidney disease (18.8% vs
disease 5.7%, P < .001). CCI was also higher in those with readmis-
Malignancy 25 (4.3)
sions than those without (2 vs 1, P < .001). The percentage of
No. of comorbidities 5 (3-6)
MDR bacteria isolates was significantly higher in the read-
Charlson comorbidities index
mission group (24.3% vs 11.4%, P = .037). Appropriate use
≥1 431 (74.1)
of empiric antibiotic therapy was observed with a lower
Median 1 (0-2)
Severity (CURB-65)a
prevalence in readmitted than non-readmitted patients
0-1 360 (62.2) (34.8% vs 53.9%, P = .003). Characteristics of patients with
2 175 (30.1) and without 30-day readmission are shown in Table 4.
≥3 44 (7.6) In the multivariable logistic regression analyses shown in
Median 1 (1-2) Table 5, a high CCI (aOR, 1.40; 95% CI, 1.08-1.82) and
Empiric antibioticsc MDR infection (aOR, 2.63; 95% CI 1.05-6.56) were associ-
β-lactam or β-lactam/β-lactamase 500 (85.9) ated with higher odds of all-cause readmissions.
inhibitor
Fluoroquinolone 310 (53.3) Discussion
Otherd 17 (2.9)
Duration of antibiotic therapy (days) 13 (9-14) Overall, 582 CAP inpatients were included in this retrospec-
Appropriate use of empiric antibiotic 299 (51.6) tive, cross-sectional study. Of which, 69 patients (11.9%)
therapyb were readmitted within 30 days of discharge. The number of
Length of stay (days) 14 (11-18) readmissions due to pneumonia-related causes were 25
Clinical instabilities (36.2%). Results from multivariable logistic regression anal-
≥1 106 (18.2) yses showed that a high CCI and infection caused by a MDR
Median 0 (0-0) pathogen were associated with the increased likelihood of
Values are given as frequency (percentage) or median (interquartile
all-cause readmissions within 30 days among CAP patients.
range).
Characteristics of Pathogens Causing CAP
a
n = 579 (3 patients whose CURB-65 score could not be measured).
b
n = 579 (3 patients did not receive empiric antibiotic therapy).
c
Some patients were prescribed with combination therapies, so the totals Among the pathogens isolated, Klebsiella pneumonia (35.8%)
add to more than 100 percent.
d
Other antibiotics included aminoglycoside (8), macrolide (4),
and Staphylococcus aureus (20.0%) were the most common.
glycopeptide (2), lincosamide (2), oxazolidinone (1). Notably, major CAP pathogens such as Streptococcus pneu-
moniae, Haemophilus influenzae, and Moraxella catarrhalis
accounted for a very small percentage (1.1%-2.1%). Some
of admission. There were 95 bacterial isolates, 41 (43.2%) patients might have been exposed to antibiotics, or self-med-
of which were MDR bacteria (Table 2). The majority of bac- icated with antibiotics in the outpatient settings prior to the
teria strains identified in this study were Gram-negative index hospital admissions,19 which decreased the detection
bacteria, with Klebsiella pneumoniae being the most preva- rates of community-acquired bacteria such as Streptococcus
lent (35.8%). Staphylococcus aureus was the most com- pneumoniae and Haemophillus influenzae in culture-based
monly isolated Gram-positive pathogens (20.0%). tests.20–22 Moreover, the sputum culture method had low
716 Hospital Pharmacy 57(6)
MDR: multidrug-resistant.
Table 5. Multivariable logistic regression analysis of factors sensitivity to detect Streptococcus pneumoniae,3,23 while
associated with all-cause hospital readmissions among CAP Staphylococcus aureus and aerobic Gram-negative bacteria
patients. were easier to be detected so their proportions might be con-
siderably larger than the actual results.24
Variable P OR 95 % CI
Age .392 1.02 0.98-1.05
Male sex .109 1.85 0.87-3.90
Reasons for Admissions
CURB-65 score .626 0.86 0.48-1.56 The all-cause readmission rate following hospital treatment
Charlson comorbidity index score .010 1.40 1.08-1.82 for CAP was 11.9%, in comparison to 7.3% to 15.5% in pre-
Multidrug-resistant infection .039 2.63 1.05-6.56 vious studies.6,25–29 The 30-day rehospitalization incidence
Duration of antibiotic therapy .112 1.06 0.99-1.14 from our study was more similar to those conducted on the
Appropriate use of empiric .061 0.46 0.21-1.04 elderly population (12.3%-15.5%)6,29 than those conducted
antibiotic therapy
on adults aged ≥ 18 years (7.3%-12%).25–28 Pneumonia-
Clinical instabilities at discharge .101 0.47 0.19-1.16
related readmission rate within 30 days of discharge was
718 Hospital Pharmacy 57(6)
higher in our study (36.2%) compared to other studies pathogens,37 while antibiotic underprescribing might not
(16.9%-35.8%).25,27–29 Most patients in our study had at least resolve pneumonia and therefore leads to other adverse out-
1 underlying condition along with pneumonia, including comes.38,39 Furthermore, most pathogens causing CAP could
chronic lung diseases, congestive heart failure, cerebrovas- not be identified at the time of hospitalization especially in
cular diseases, or diabetes. These conditions have long been developing countries, where molecular methods such as
known as factors that could increase the risk of polymerase chain reaction or urinary antigen test are not
pneumonia.30,31 always available. With a low culture positivity rate in CAP,
the choice of antibiotic treatment mostly depends on guide-
line-based recommendations. Our finding reassured the
Factors Associated With Hospital Readmission need to adhere to current guidelines to improve outcomes in
Within 30 days CAP patients.
Our study confirmed the finding from prior studies that Our study did not find the association of other factors,
comorbidities were associated with readmission in CAP namely age, sex, pneumonia severity, duration of antibiotic ther-
patients.6,32 In fact, the study population largely consisted of apy, and clinical instabilities at discharge with hospital read
elderly patients who had at least 1 comorbidity, specifically mission, while prior studies showed opposite results.6,26,29,40,41
cardiopulmonary diseases. People who had underlying The most significant differences of these studies from ours were
chronic lung conditions such as chronic obstructive pulmo- (i) their outcome of interest was 60-day readmission41; (ii) their
nary diseases were also prone to recurrent CAP as they are study design was a prospective study,26 and (iii) their study sam-
usually colonized with Streptococcus pneumoniae and ples had different baseline characteristics with respect to sex
Haemophilus influenzae in the trachea and main stem bron- and age.6,29,40
chi. Along with the impaired cough and mucociliary clear-
ance, these bacteria could easily cause pneumonia in the
alveoli.33 In addition, for elderly patients that had several
Strengths and Limitations
comorbidities which were not adequately controlled during There are several limitations of this study that must be
the index hospital admission, this could slow the recovery noted. Firstly, this study was conducted at a single center, so
after a pneumonia or symptoms may continue to worsen after the results may not be generalized to other institutions in
discharge and lead to unplanned readmissions. different countries. Secondly, the study design was retro-
In our study, the percentage of MDR bacteria among the spective and data were collected by reviewing medical
isolates collected was considerably high (43.2% in total). As records. Therefore, some data would be unavoidably miss-
expected, the multivariable regression analysis showed that ing and bias could have occurred. For example, arterial
infection with a MDR organism was associated with rehospi- blood gas results were not usually done on stable patients at
talization within 30 days (aOR, 2.63; 95% CI, 1.05-6.56). discharge. Lastly, we did not take rehospitalizations at other
Infection caused by antibiotic resistant bacteria has been hospitals into account so the actual incidence of hospital
proved to be an independent predictor of readmission fol- readmission could be higher. However, results from this
lowing pneumonia in a study conducted by Andruska et al.34 study could reflect the characteristics of CAP patients and
Results from the study by Jang and Ahn27 also showed that pathogens as well as the quality of hospital care of this par-
patients who were readmitted to the hospital after the first ticular geographic area. Furthermore, patients who were dis-
treatment of CAP had a higher though not statistically sig- charged at hospice state or who transferred to another
nificant incidence of infection caused by a MDR organism hospital were excluded from this study. Also, most eligible
than their non-readmitted counterparts. patients had healthcare insurance registered at this hospital,
Adherence to empiric antibiotic regimens recommended and they should visit this hospital so that their medical
in the treatment guidelines was associated with lower mor- expenses could be covered. In this context, the number of
tality rate, shorter length of stay in CAP patients in prior loss samples was insignificant.
studies.5,35,36 In the study conducted by Nour et al.,5 nonad- One of the strengths of our study was that we included
herence to guideline-based antibiotic treatment was associ- every qualified patient in a 1-year period at one of the high-
ated with a higher risk of all-cause 30-day readmission. In profile hospitals in the country. Therefore, the results could
our study, the overall adherence rate to empiric antibiotic represent the general incidence of hospital readmissions fol-
recommendations was moderate (51.6%), and it was lower lowing CAP as well as the practice of treating CAP in such a
in patients who had 30-day hospital readmissions. Although setting of low-income countries. Another strength of the
we did not find that appropriateness of antibiotic selection study is that we investigated patient and pathogen character-
was associated with 30-day readmission in the multivariable istics, as well as the appropriateness of antibiotics for the
regression analyses, this is still an important consideration treatment of CAP, which would help medical professionals
as overprescribing of antibiotics could increase the likeli- and policy makers improve the practice of treating CAP and
hood of recurrent infections caused by antibiotic resistant develop future guidelines.
Nguyen et al 719
18. DeMeulenaere S. Pulse oximetry: uses and limitations. J Nurs 30. Ramirez JA, Wiemken TL, Peyrani P, et al. Adults hospital-
Pract. 2007;3(5):312–317. doi:10.1016/j.nurpra.2007.02.021 ized with pneumonia in the United States: incidence, epidemi-
19. Nguyen KV, Thi Do NT, Chandna A, et al. Antibiotic use and ology, and mortality. Clin Infect Dis. 2017;65(11):1806–1812.
resistance in emerging economies: a situation analysis for Viet doi:10.1093/cid/cix647
Nam. BMC Public Health. 2013;13(1):1158. doi:10.1186/1471- 31. Torres A, Peetermans WE, Viegi G, Blasi F. Risk factors for
2458-13-1158 community-acquired pneumonia in adults in Europe: a litera-
20. Harris AM, Bramley AM, Jain S, et al. Influence of antibiot- ture review. Thorax. 2013;68(11):1057–1065. doi:10.1136/
ics on the detection of bacteria by culture-based and culture- thoraxjnl-2013-204282
independent diagnostic tests in patients hospitalized with 32. Adamuz J, Viasus D, Campreciós-Rodríguez P, et al. A pro-
community-acquired pneumonia. Open Forum Infect Dis. spective cohort study of healthcare visits and rehospitalizations
2017;4(1):ofx014. doi:10.1093/ofid/ofx014 after discharge of patients with community-acquired pneumo-
21. Spencer RC, Philp JR. Effect of previous antimicrobial
nia. Respirology. 2011;16(7):1119–1126. doi:10.1111/j.1440-
therapy on bacteriological findings in patients with primary 1843.2011.02017.x
pneumonia. Lancet. 1973;2(7825):349–350. doi:10.1016/ 33. Geppert EF. Recurrent pneumonia. Chest. 1990;98(3):739–
s0140-6736(73)93192-9 745. doi:10.1378/chest.98.3.739
22. Rhodes J, Hyder JA, Peruski LF, et al. Antibiotic use in 34. Andruska A, Micek ST, Shindo Y, et al. Pneumonia patho-
Thailand: quantifying impact on blood culture yield and gen characterization is an independent determinant of hospi-
estimates of pneumococcal bacteremia incidence. Am J Trop tal readmission. Chest. 2015;148(1):103–111. doi:10.1378/
Med Hyg. 2010;83(2):301–306. doi:10.4269/ajtmh.2010.09- chest.14-2129
0584 35. Dambrava PG, Torres A, Vallès X, et al. Adherence to
23. Johansson N, Kalin M, Giske CG, Hedlund J. Quantitative guidelines’ empirical antibiotic recommendations and
detection of Streptococcus pneumoniae from sputum samples community-acquired pneumonia outcome. Eur Respir J.
with real-time quantitative polymerase chain reaction for etio- 2008;32(4):892–901. doi:10.1183/09031936.00163407
logic diagnosis of community-acquired pneumonia. Diagn 36. Lui G, To HK, Lee N, et al. Adherence to treatment guideline
Microbiol Infect Dis. 2008;60(3):255–261. doi:10.1016/j.diag- improves patient outcomes in a prospective cohort of adults
microbio.2007.10.011 hospitalized for community-acquired pneumonia. Open Forum
24. Bartlett JG. Diagnostic tests for agents of community-acquired Infect Dis. 2020;7(5):ofaa146. doi:10.1093/ofid/ofaa146
pneumonia. Clin Infect Dis. 2011;52 (supplement 4):S296– 37. Llor C, Bjerrum L. Antimicrobial resistance: risk asso-
S304. doi:10.1093/cid/cir045 ciated with antibiotic overuse and initiatives to reduce
25. Capelastegui A, España Yandiola PP, Quintana JM, et al.
the problem. Ther Adv Drug Saf. 2014;5(6):229–241.
Predictors of short-term rehospitalization following discharge doi:10.1177/2042098614554919
of patients hospitalized with community-acquired pneumonia. 38. Petersen I, Johnson AM, Islam A, Duckworth G, Livermore
Chest. 2009;136(4):1079–1085. doi:10.1378/chest.08-2950 DM, Hayward AC. Protective effect of antibiotics against
26. Halm EA, Fine MJ, Kapoor WN, Singer DE, Marrie TJ, Siu serious complications of common respiratory tract infec-
AL. Instability on hospital discharge and the risk of adverse tions: retrospective cohort study with the UK general practice
outcomes in patients with pneumonia. Arch Intern Med. research database. BMJ. 2007;335(7627):982. doi:10.1136/
2002;162(11):1278–1284. doi:10.1001/archinte.162.11.1278 bmj.39345.405243.BE
27. Jang JG, Ahn JH. Reasons and risk factors for readmission 39. Winchester CC, Macfarlane TV, Thomas M, Price D.
following hospitalization for community-acquired pneumo- Antibiotic prescribing and outcomes of lower respiratory tract
nia in South Korea. Tuberc Respir Dis. 2020;83(2):147–156. infection in UK primary care. Chest. 2009;135(5):1163–1172.
doi:10.4046/trd.2019.0073 doi:10.1378/chest.07-2940
28. Jasti H, Mortensen EM, Obrosky DS, Kapoor WN, Fine MJ. 40. Neupane B, Walter SD, Krueger P, Marrie T, Loeb M. Predictors
Causes and risk factors for rehospitalization of patients hospi- of inhospital mortality and re-hospitalization in older adults with
talized with community-acquired pneumonia. Clin Infect Dis. community-acquired pneumonia: a prospective cohort study.
2008;46(4):550–556. doi:10.1086/526526 BMC Geriatr. 2010;10:22. doi:10.1186/1471-2318-10-22
29. Mather JF, Fortunato GJ, Ash JL, Davis MJ, Kumar A.
41. Dagan E, Novack V, Porath A. Adverse outcomes in patients
Prediction of pneumonia 30-day readmissions: a single- with community acquired pneumonia discharged with clinical
center attempt to increase model performance. Respir Care. instability from Internal Medicine Department. Scand J Infect
2014;59(2):199–208. doi:10.4187/respcare.02563 Dis. 2006;38(10):860–866. doi:10.1080/00365540600684397
1095894
research-article2022
HPXXXX10.1177/00185787221095894Hospital PharmacyKatsivalis et al
Original Article
Hospital Pharmacy
Abstract
Introduction: In October 2017, the Food and Drug Administration announced a shortage of intravenous (IV) opioid
medications. Patients with sickle cell disease (SCD) are particularly vulnerable, as high amounts of IV opioid medications
are standard therapy during vaso-occlusive crises (VOC). Our institution responded to the crises by implementing IV to
oral (PO) conversions of opioid therapies and encouraging multimodal pain management with non-opioid medications.
Objectives: The primary objective was the assessment of IV opioid medication utilization before and during the shortage.
Secondary objectives included total opioid consumption, length of stay, and prescribing of non-opioid analgesics. Methods:
This single-institution retrospective study included patients >18 years of age admitted to adult medicine teams with VOC
during February 2017 or February 2018. The amount of opioid medication administered to patients during both periods
was assessed, and quantities were then converted to PO morphine milligram equivalents and compared between years. The
number of patients receiving scheduled non-opioid medications were also compared. Length of stay and readmissions were
compared between years. Results: Between 2017 and 2018, IV opioid use for VOC decreased by 52% on inpatient services,
and there was a 34% reduction in overall opioid use. Oral opioid use more than doubled during the period (10.2% in 2017 vs
39.1% in 2018, P < .01). LOS between 2017 and 2018 (6 vs 6 days, P = .4774) and total number of ED visits (27 vs 8, P = .276)
were similar. There were significantly fewer 30-day readmissions in 2018 versus 2017 (15 vs 28, P = .025). Conclusion:
The implementation of IV opioid restrictions resulted in a decrease in IV opioid use in treatment of VOC in patients with
SCD without causing increases in length of stay or readmissions. Oral opioids should be considered an option for VOC
management in patients with SCD.
Keywords
analgesics, clinical services, disease management, intravenous therapy, pain management, drug/medical use evaluation
administration for opioid medications, reserving the IV route Table 1. Patient Characteristics.
for those who cannot take medications by mouth, have gas-
2017 (n = 47) 2018 (n = 42) P-value
trointestinal malabsorption, or who require immediate pain
control.5 Male sex, n (%) 21 (44.7) 18 (41.9) .0863
Although the consensus statement made by the SMH pro- Age (years), 23 [16-35] 24 [16-33] .6711
vides guidance for the preferential use of PO opioids in the median [IQR]
hospital setting, it does not specify management strategies Service .267
for the sickle cell population.5 Previous studies have shown Non-rounding, 18 (38) 21 (50)
n (%)
that utilization of a standard care pathway for all sickle cell
Rounding, n (%) 29 (62) 21 (50)
patients leads to an increased use of PCAs but also a decrease
in total length of stay.6 Conti and colleagues7 proposed an
oral morphine regimen in an emergency department (ED)
All opioid doses were converted to oral morphine milligram
setting where IV meperidine had previously been used. They
equivalents (MME).9 Doses administered via PCA were
reviewed the ED visits of 9 patients prior to the implementa-
determined by nursing documentation in the medical record
tion of the oral morphine protocol and after and noted a sta-
every 12 hours. The primary endpoint was the utilization of
tistically significant decrease in total ED visits, length of
IV opioid medications before and during the IV opioid
stay, and number of admissions after the implementation of
shortage. Secondary endpoints included total opioid con-
the oral protocol. Literature evaluating the use of oral opioid
sumption, length of stay, and prescribing of scheduled, non-
agents for inpatient management of VOC are limited. One
opioid analgesics. A subgroup analysis was performed to
case series of 3 patients reported successful transition of
assess patients admitted to housestaff services with round-
patients from PCAs to oral opioids during admissions for
ing pharmacists (rounding service) compared to hospitalist
VOC pain.8
services where pharmacists review patients but do not attend
During the shortage, our Pharmacy and Therapeutics
formal patient rounds (non-rounding service).
Committee approved IV opioid limitations on ordering that
This study received a formal Determination of Quality
was implemented at the time of order verification. All IV
Improvement status according to University of Chicago
push doses were limited to one-time orders if patients had an
Medicine’s institutional policy. As such, this was deemed
active diet order or were able to take oral medications. For
non-human subjects research and was therefore not reviewed
patients who were not able to take medications by mouth, IV
by the Institutional Review Board. Statistical analysis was
opioid orders were given a 24-hour limit. After 24 hours,
performed using StataCorp. 2013 (College Station, TX:
these orders could be renewed if patients were still not able
StataCorp LP). Descriptive statistics were summarized using
to take medications by mouth. All PCA-administered medi-
median and interquartile range. Chi square test and Fisher’s
cations were given a 24-hour limit. For patients with sickle
exact test were used for categorical data, and Wilcoxon Rank
cell VOC, the primary pharmacist in conjunction with the
Sum was used for continuous data.
acute pain service would determine if the patient needed
additional time on the PCA, could transition to oral medica-
tions, or required an alternative agent such as methadone,
Results
ketamine infusion, or lidocaine infusion. This evaluation
sought to determine the impact of the IV opioid shortage on During the 2 study periods, 89 patients (2017: N = 47; 2018:
adult patients requiring inpatient management of VOC. N = 42) were admitted for treatment of VOC. Patient charac-
teristics are described in Table 1. After implementing restric-
tions on IV opioid usage during the shortage, there was a 34%
Methods reduction in overall opioid use for treatment of VOC (Table
All patients 18 years and older with sickle cell disease dis- 2). IV opioid use for VOC decreased by 54% and 52% in the
charged from our institution with a primary admission diagno- ED and on inpatient services, respectively. Oral opioid use
sis of VOC in February of 2017 and 2018 were included. All more than doubled during the period. Total opioid administra-
admissions to the children’s hospital or where VOC was not tion per patient was similar between the 2 years (P = .651);
the primary reason for admission were excluded. These time however, significantly less IV opioid was used (P = 0.023)
frames were selected to provide similar periods before (2017) and more oral opioids were administered (P = .012) following
and during (2018) the IV opioid shortage. Patients were identi- the implementation of restrictions on IV opioids (Table 3). Of
fied by ICD-10 codes, and a retrospective chart review was total opioid usage, 79% was IV and 13% was oral in 2017
performed to determine the amount of opioid medications and compared to 57% IV and 43% oral in 2018. Transdermal fen-
non-opioid analgesia received by each patient. tanyl was a small portion of total opioid use in each time
The medication administration report was reviewed to period. During the 2017 time period, 2 patients received fen-
calculate the total dose of opioids received during the hospi- tanyl patches consistent with their home regimens which
talization and assess prescribing of non-opioid analgesics. accounted for 91% of the total fentanyl use during that period.
Katsivalis et al 723
2017 2018
Median [IQR] Median [IQR] % Change P-value
Total opioids, mg 1616 [881-6364] 1650 [733-3775] +2.1% 0.651
IV opioids, mg† 1170 [386-5288] 573 [214-1702] −51% 0.023
Oral opioids, mg† 180 [8-520] 719 [144-1560] +399% 0.012
Figure 2. Intravenous opioid administration by service type during inpatient stay.
Table 5. Healthcare Utilization During and Following Admission for Vaso-Occlusive Crisis.
The decision to transition to oral opioids during a hospi- receive any oral opioids. The IV opioid shortage and subse-
talization is often challenging and can be impacted by patient quent restrictions allowed for increased multidisciplinary
readiness, provider comfort with pain management, or the efforts to evaluate IV opioid use frequently, discuss early
desire to allow for increased patient control of opioid dosing transition to equianalgesic doses of oral opioids, and opti-
through PCAs.10 Prior to IV opioid restrictions, almost a mize non-opioid analgesic use. Our analysis shows that
quarter of patients treated for VOC at our institution did not patients receiving treatment during the time of the opioid
Katsivalis et al 725
Figure 3. Oral opioid administration by service type during inpatient stay.
shortage and restrictions received slightly higher total opioid occurred after the prescriber already evaluated the patient
doses; therefore, the use of oral agents did not preclude and discussed a pain management plan for the day. Not being
patients from getting higher opioid doses. The need for physically present for pain management assessments and
reduced IV opioid use also likely increased prescriber com- discussions can limit the impact of recommendations because
fort with opioid conversions and use of scheduled and as- prescribers may not want to change the plan once it has been
needed short-acting opioids. established with the patient.
The use of oral opioids prior to discharge may have also There are several limitations to our study. The 2 study
allowed for prescribers to identify appropriate opioid dis- populations were not randomly assigned; however, we did
charge regimens, as has been described in previous literature.8 select the same month in 2 different years to better match
We found a significant reduction in hospital readmissions weather patterns and viral illnesses that may trigger VOCs in
within 30 days, which may demonstrate that by identifying patients with SCD. Additionally, we were not able to deter-
effective oral regimens, patients had better pain control fol- mine if patients sought care at other local EDs or hospitals
lowing discharge. Alternatively, patients may have chosen to for VOC pain. Finally, the IV opioid restrictions limited use
seek care at other institutions; however, we expect that most to 24-hours but could be extended based on patient needs.
institutions had similar restrictions on IV opioid use due to We did, however, notice that IV opioids were often reordered
the national shortage. after the initial 24-hour time restriction and verified by our
The restrictions were more successful at reducing IV opi- central pharmacy team who were not part of the team dis-
oid use when a pharmacist was present during rounds. The cussing optimization of pain management. While this
reason for this is likely multifactorial. By being present on decreased the impact of our restrictions on our IV opioid sup-
rounds, the pharmacist could verbally reinforce the IV opioid ply, we did not feel it was safe practice to enforce a hard
restrictions and provide recommendations for equianalgesic 24-hour rule on IV opioids.
oral options as part of the discussion of patient management.
The pharmacist was also available to discuss the restrictions
Conclusion
on IV opioids directly with the patient and explain the pro-
cess of transitioning to an equivalent oral regimen. This is The implementation of IV opioid restrictions resulted in a
similar to findings that pharmacist-driven opioid stewardship decrease in IV opioid use in treatment of VOC in patients
can decrease IV opioid use and increase oral opioid use.11 with SCD without causing increases in length of stay or read-
Alternatively, on non-rounding services, pharmacist recom- missions. Oral opioids should be considered as an option for
mendations may have been conveyed via page or have VOC management in patients with SCD.
726 Hospital Pharmacy 57(6)
Original Article
Hospital Pharmacy
Abstract
Purpose: Heparin-based regimens are recommended for anticoagulation in hospitalized patients with COVID-19 though
a study reported similar mortality with apixaban in critically ill hospitalized COVID-19 patients. Our pilot study sought to
determine the differences in all-cause mortality, venous thromboembolism (VTE), and bleeding events between apixaban
and therapeutic heparin-based regimens in hospitalized non-critically ill COVID-19 patients. Methods: We conducted a
retrospective analysis of non-critically ill COVID-19 patients aged ≥ 18 years admitted to 3 campuses of Montefiore Medical
Center during the first (March 2020 to May 2020) and second (January 2021 to February 2021) COVID-19 surges, who
received within 48 hours of admission and continued for ≥72 hours a therapeutic dose of low-molecular-weight heparin
(LMWH), unfractionated heparin (UFH), or any apixaban dose for VTE prophylaxis. Outcomes data analyzed included
mortality, suspected or imaging-confirmed VTE, and bleeding using a defined criteria. Results: Overall, 162 patients met
eligibility for analysis. Baseline characteristics were similar between the 2 groups except liver and renal functions. Mortality
occurred in 10 (13.3%) patients on apixaban and 23 (26.4%) patients on a heparin-based regimen (P = .059). Confirmed VTE
events were not different between the groups (8% vs 13.8%, P = .359), but higher incidence of bleeding occurred in heparin-
based group (4% vs 52.9%, P < .001). Conclusion: There were no differences in mortality or confirmed VTE between
apixaban and heparin-based regimens except for more bleeding events with the heparins. This study highlights the utility of
apixaban in COVID-19.
Keywords
anticoagulants, COVID, disease management, respiratory
assessment score unless there are absolute contraindica- positive for COVID-19 within 72 hours of floor admission,
tions.3,6-8,10 These guidelines favor the use of heparins against and if they were initiated on either apixaban or a therapeu-
direct oral anticoagulants (DOACs) due to the increased risk of tic heparin-based regimen for at least 72 hours within
drug-drug interactions with DOACs and the potential declining 48 hours of admission. We excluded patients who within
patient status may worsen bleeding risk.11 Additionally, 48 hours of floor admission were either intubated or
DOACs are more difficult to reverse and monitor. transferred to the intensive care unit (ICU), identified as
Despite these guideline recommendations, anticoagula- palliative care, or had a code status of “do-not-resuscitate/
tion protocols for COVID-19 developed by various institu- do-not-intubate.” Patients were also excluded if they did
tions also incorporated the use of DOACs in addition to UFH not continue their outpatient anticoagulation regimen upon
and LMWH to their thromboprophylaxis management strat- admission or presented with confirmed VTE upon admis-
egy. The addition of DOACs was based primarily on their sion. Apixaban was dosed either 5 mg by mouth twice daily
safety and efficacy in other hypercoagulable disease states, or 2.5 mg twice daily based on renal function (Scr > 2.5 mg/
the convenience of oral administration as well as lack of dL or CrCl < 30 mL/min) per institutional guideline. The
required monitoring and less pharmacokinetic (PK) and heparin-based regimens were dosed as either low-molecu-
pharmacodynamic (PD) variability .12,13 lar-weight heparin (LMWH) enoxaparin 1 mg/kg SQ every
Billett et al14 observed in their study that the DOAC, apix- 12 hours (daily for CrCl < 30 mL/min) or enoxaparin
aban, was associated with a reduction in mortality in hospital- 1.5 mg/kg SQ. Any intravenous infusion of unfractionated
ized COVID-19 patients when compared to no anticoagulation. heparin (UFH) was administered to a therapeutic aPTT of
Paranjpe et al15 also suggested that systemic anticoagulation 46 to 70 following institutional protocol for VTE of 80 units/
confers a survival benefit in COVID-19 patients regardless of kg bolus followed by 18 units/kg/h infusion.
the anticoagulant used, though this study had significant limi-
tations. This conclusion was similar to Kabir16 who showed
Data Sources
early, high-dose anticoagulation in non-critically ill COVID-
19 patients improved survival, including with apixaban. All patients eligible for the study were identified in our elec-
Some notable clinical trials on anticoagulation in COVID-19 tronic medical record using Business Universe (SAP
such as REMAP-CAP,17 ACTIV4,18 and ATTACC19 com- BusinessObjects®). Data obtained included index date of
pared therapeutic heparin-based regimens to the usual care admission as well as the type and duration of anticoagulation
thromboprophylaxis in hospitalized COVID-19 patients used. Patients’ baseline demographics, pertinent laboratory
without including DOACs as a study arm. Combined results markers relevant to coagulopathy, occurrence of bleeding
from these trials reported that in non-critically ill patients and/or venous thromboembolic (VTE) event, and mortality
therapeutic anticoagulation with heparin increased survival to were obtained through our electronic medical record.
discharge.20 This leaves the question of the impact of DOACs
in non-critically ill COVID-19 patients still unanswered. To
Outcomes
build on previous literature and address confounders in previ-
ous studies, we examined the safety and efficacy of apixaban The primary outcome of this study was the difference in
compared to therapeutic heparin-based regimens on outcomes all-cause mortality between an apixaban and therapeutic
in non-critically ill patients hospitalized with COVID-19. heparin-based regimens from the index date to event occur-
rence, discharge, or transfer to the ICU. Secondary outcomes
assessed were the incidence of venous thromboembolism
Methods (VTE) and the incidence of major bleeding between apixa-
ban and heparin-based regimens in hospitalized non-criti-
Study Design cally ill COVID-19 patients. VTE was determined through
This pilot study was a retrospective cohort study of adult either diagnostic imaging (ultrasound and/or CT scan) or by
inpatients admitted to the 3 main campuses of Montefiore a surrogate indicator (D-dimer and/or fibrinogen elevation)
Medical Center, all located in the Bronx, New York. Data with clinical symptomatology of VTE. Bleeding was deter-
were collected at 2 time periods: the first COVID-19 surge mined if the patient was only initiated on pantoprazole IV
from March 2020 to May 2020, and the second COVID-19 drip or IV push every 12 hours because of signs and symp-
surge from January 2021 to February 2021. Approval for the toms of gastrointestinal bleeding after anticoagulant initia-
study was obtained through the institutional review board tion, or bleeding defined by the ISTH21,22 criteria as: Fatal
(IRB #2020-12313). bleeding, and/or symptomatic bleeding in a critical area or
organ, such as intracranial, intraspinal, intraocular, retro-
peritoneal, intraarticular, or pericardial, or intramuscular
Patients with compartment syndrome, and/or bleeding causing a fall
Patients ≥18 years of age were included in this study if they in hemoglobin level ≥2 g/dL (1.24 mmol/L) or leading to
were admitted to a general medical floor, if they tested transfusion of ≥2 units of whole blood or red cells.
Appiah et al 729
Note. AST = aspartate aminotransferase; ALT = alanine aminotransferase; COPD = chronic obstructive pulmonary disease; IQR = interquartile range;
SD = standard deviation; T2DM = type 2 diabetes mellitus; ULN = upper limit of normal; VTE = venous thromboembolism.
a
Apixaban: Low dose (n = 36), high dose (n = 39).
b
Heparin-based regimen: UFH (n = 42), LMWH (n = 45).
significant difference among the subgroups (χ2 = 4.778, similar to that of Billet et al14 which also demonstrated simi-
P = .189) (Supplemental Table 2). lar in mortality between apixaban and enoxaparin anticoagu-
lation . Our data from Table 1 shows baseline characteristics
were similar between the groups except for those admitted
Secondary Outcomes with acute kidney injury or AST/ALT ≥ 3 ULN. Significant
The incidence of confirmed VTE was 8% in the apixaban differences in the agents used in these patients with organ
group compared to 13.8% in the heparin-based group dysfunction reflects their real-world use as the heparins espe-
(P = .359). Suspected VTE was 8% in the apixaban group cially UFH as opposed to apixaban are more likely to be
compared to 24.1% in the heparin-based group (P = .001) selected for therapeutic anticoagulation in patients with renal
(Figure 2). or hepatic impairment due to risk for excessive anticoagula-
Bleeding occurred in 4% and 52.9% of apixaban and hep- tion with DOACs in these populations and the difficulty of
arin-based regimens respectively (P < .001). Hemoglobin anticoagulant reversal.
drop ≥2 g/dL in 24 hours after anticoagulant initiation was Despite the fact that most patients in the apixaban
significantly higher in patients on heparin-based regimen group received remdesivir (Table 1), this agent has been
compared to apixaban (P ≤ .001) (Table 2). Hemoglobin found not to have any impact on overall mortality in
drop was then major contributor to overall bleeding rates COVID-19 patients23 and as such less likely to affect the
especially in the heparin-based regimen. results of our primary outcome. Although steroids have
been found to reduce overall mortality in COVID-19,24
steroid use and their exposure time were similar between
Discussion the groups (Table 1). Thus, the impact of steroids on over-
Early in the COVID-19 pandemic, DOACs were included as all mortality is expected to be similar in both groups.
an anticoagulant option in our institution’s COVID-19 man- Additionally, data collection was performed at 2 different
agement protocols. Apixaban was chosen as the agent of time periods during the pandemic, during the first COVID
choice in this class because of its better safety profile in surge when standard of care was still evolving with lots of
patients with impaired renal function and its availability on experimental agents being used in COVID-19 manage-
the hospital formulary. ment and during the second COVID-19 surge where
The primary outcome from this study showed that there greater evidence was available to guide care. This ensured
was no significant difference in mortality between patients that the results of the primary outcome were unduly
on apixaban and a heparin-based regimen. These results are impacted by standard of care practices in these eras.
Appiah et al 731
Figure 2. Confirmed and suspected VTE events between apixaban and heparin-based regimen.
For VTE measurement, we chose to report both diagnos- showed 4 out of 6 on UFH having therapeutic aPPT with
tic confirmed VTE and clinically suspected VTE to reflect the remaining 2 subtherapeutic underscoring the unreliabil-
clinical practice. Even though the baseline levels of fibrino- ity of aPPT measurements in COVID-19 for UFH. Despite
gen, a risk factor for developing blood clots, were signifi- DOACs not currently being recommended as a preferred
cantly higher in the apixaban group compared to the anticoagulant for the prevention or management of COVID-
heparin-based group (Table 1), confirmed VTE was lower 19 coagulopathy in hospitalized patients, our results did not
in the apixaban group, though not statistically significant. show any increased risk of VTE with the use of apixaban in
However, suspected VTE was observed at a significantly non-critically ill hospitalized patients.
higher rate in the heparin-based group than in the apixaban Bleeding events were also significantly lower in the apix-
group. Though baseline d-dimer levels were not statisti- aban group compared to the heparin-based group, with more
cally different between the groups (Table 1), d-dimers significant bleeding in those patients on unfractionated hepa-
greater than 3 μg/mL were seen more often in the patients rin (Supplemental Table 4). A hemoglobin drop greater than
on heparin-based regimen at baseline underscoring a pos- 2 mg/dL within 24 hours was significantly higher in the hep-
sible ongoing clotting cascade in the heparin-based group. arin-based regimen group compared to the apixaban group.
Additionally, d-dimers and fibrinogen levels were still high Blood transfusion ≥2 units or the use of pantoprazole for
at the time of suspected VTE in the heparin-based group gastrointestinal bleeding from anticoagulant use were all
(Supplemental Table 3). The aPPT at time of VTE event higher in the heparin-based group. This implied a higher
732 Hospital Pharmacy 57(6)
incidence of hemorrhage in the heparin-based group could be the true impact of anticoagulants when extrapolated to this
due excessive anticoagulation. We also confirmed aPPT at patient population should be done with caution. Lastly,
the time of a bleeding event for unfractionated heparin (1/3 patients were only followed for the duration of their inpatient
therapeutic aPPT, 2/3 supratherapeutic aPPT). stay, limiting our ability to assess any outcomes that occurred
The randomized control trial ACTION compared therapeu- post-discharge.
tic anticoagulation of rivaroxaban or a heparin transitioned to
rivaroxaban, versus thromboprophylaxis with a heparin. They
Conclusion
found no improvement in clinical outcomes or reduction in
death with the therapeutic strategy (0.86 [95% CI: 0.59-1.22], Our study adds to the existing but limited literature in a
P = .40).25 Additionally, because the ACTION trial assessed 2 racially and ethnically diverse population, evaluating opti-
anticoagulation strategies rather than comparing anticoagu- mal anticoagulation strategies in patients with COVID-19
lants, a crossover was only considered if a patient changed induced coagulopathy. Mortality and VTE events between
from prophylactic to therapeutic dose (or vice versa) and not apixaban and therapeutic heparin-based regimens are not
between different drugs within the same study group per study significantly different when used in non-critically ill hospi-
protocol. In our study, no patients receiving heparin thrombo- talized patients. Apixaban, however, appears to have a lower
prophylaxis were included, based on results of prior litera- incidence of bleeding compared to heparin-based regimens.
ture.15,17,18 Also in contrast to our study, compared to More data are needed to better determine this benefit-to-risk
rivaroxaban, apixaban has been found to have a lower overall profile of apixaban with respect to noncritically ill COVID-
risk of bleeding26,27 and a more recent single-arm study of 19 patients requiring hospitalization.
apixaban use in COVID-19 patients admitted to the ICU found
apixaban to be safe and efficacious with no major bleeding Authors’ Note
events or thrombotic events.28 As the most preferred antico- Work by Daniel Appiah was conducted at Montefiore Medical
agulation approach during hospitalization is still unclear and Center, Bronx, NY.
many off-label and non-evidence-based strategies have been
used in institutions to date, our results may provide evidence Declaration of Conflicting Interests
for further investigation in this clinical setting. The author(s) declared no potential conflicts of interest with respect
Most similar to our study, Billett et al used logistic regres- to the research, authorship, and/or publication of this article.
sion to show a decrease in mortality with prophylactic use of
apixaban (OR = 0.46, P < .001) and enoxaparin (OR = 0.49, Funding
P < .001) compared to other anticoagulant strategies.
The author(s) received no financial support for the research, author-
Additionally, therapeutic apixaban was associated with
ship, and/or publication of this article.
decreased mortality (OR = 0.57, P < .006), similar to the pro-
phylactic regimens. However, these results were taken from
ORCID iD
the entire cohort including both general medicine and inten-
sive care patients.14 Nicholas J. Quinn https://orcid.org/0000-0002-9703-5845
Our study has several limitations. First being a retrospec-
tive study, treatment groups were not allocated randomly Supplemental Material
creating a selection bias. This bias is seen in the twice as Supplemental material for this article is available online.
many patients in the heparin group having AKI. Patients with
transaminitis were also mostly found in the heparin-based References
group. Also, as the study took place during 2 different surges 1. Tang N, Bai H, Chen X, Gong J, Li D, Sun Z. Anticoagulant
of COVID-19, standard treatments and the variants likely treatment is associated with decreased mortality in severe coro-
changed. We did not have an advantage to monitor the hepa- navirus disease 2019 patients with coagulopathy. J Thromb
rin effect with anti-Xa, a more predictable marker of heparin Haemost. 2020;18(5):1094-1099. doi:10.1111/jth.14817
response in COVID-19 patients as aPPT values may be 2. Cui S, Chen S, Li X, Liu S, Wang F. Prevalence of venous
altered from the normal baseline in COVID-19. Some of the thromboembolism in patients with severe novel coronavi-
patients had comorbidities that required anticoagulation, and rus pneumonia. J Thromb Haemost. 2020;18(6):1421-1424.
this could have an effect on the outcome. We did not account doi:10.1111/jth.14830
for other medications used by the participants which may 3. Lodigiani C, Iapichino G, Carenzo L, et al. Venous and arte-
rial thromboembolic complications in COVID-19 patients
have pharmacokinetically or dynamically interacted with the
admitted to an academic hospital in Milan, Italy. Thromb Res.
oral anticoagulant or increased the risk of bleeding. This 2020;191:9-14. doi:10.1016/j.thromres.2020.04.024
could have impacted the bleeding or VTE incidence in our 4. Klok FA, Kruip MJHA, van der Meer NJM, et al. Confirmation
patients. Additionally, suspected VTE may have been con- of the high cumulative incidence of thrombotic complica-
founded by observation bias from providers. Participants tions in critically ill ICU patients with COVID-19: an updated
who received various anticoagulants through substitution analysis. Thromb Res. 2020;191:148-150. doi:10.1016/j.throm
throughout the hospital stay were also excluded. Therefore, res.2020.04.041
Appiah et al 733
5. Llitjos JF, Leclerc M, Chochois C, et al. High incidence community-acquired pneumonia) study. Ann Am Thorac Soc.
of venous thromboembolic events in anticoagulated severe 2020;17(7):879-891. doi:10.1513/AnnalsATS.202003-192SD
COVID-19 patients. J Thromb Haemost. 2020;18(7):1743- 18. Nct. Anti-thrombotics for adults hospitalized with COVID-19
1746. doi:10.1111/jth.14869 (ACTIV-4). 2020. Accessed February 12, 2021. https://clini-
6. Spyropoulos AC, Levy JH, Ageno W, et al. Scientific and caltrials.gov/show/NCT04505774
standardization committee communication: clinical guidance 19. Houston BL, Lawler PR, Goligher EC, et al. Anti-thrombotic
on the diagnosis, prevention and treatment of venous thrombo- therapy to ameliorate complications of COVID-19 (ATTACC):
embolism in hospitalized patients with COVID-19. J Thromb Study design and methodology for an international, adap-
Haemost. 2020;18(8):1859-1865. doi:10.1111/jth.14929 tive Bayesian randomized controlled trial. Clin Trials.
7. The National Institute for Health. Coronavirus disease 2019 2020;17(5):491-500. doi:10.1177/1740774520943846
(COVID-19) treatment guidelines. National Institutes of Health. 20. Lawler PR, Goligher EC, Berger JS, et al. Therapeutic antico-
2021. Accessed August 16, 2020. https://www.covid19treat- agulation with heparin in noncritically ill patients with COVID-
mentguidelines.nih.gov/ 19. New Engl J Med. 2021;385(9):790-802. doi:10.1056/
8. Barnes GD, Burnett A, Allen A, et al. Thromboembolism and NEJMoa2105911
anticoagulant therapy during the COVID-19 pandemic: interim 21. Kaatz S, Ahmad D, Spyropoulos AC, Schulman S. Definition
clinical guidance from the anticoagulation forum. J Thromb of clinically relevant non-major bleeding in studies of antico-
Thrombolysis. 2020;50(1):72-81. doi:10.1007/s11239-020- agulants in atrial fibrillation and venous thromboembolic dis-
02138-z ease in non-surgical patients: communication from the SSC
9. Thachil J, Tang N, Gando S, et al. ISTH interim guidance of the ISTH. J Thromb Haemost. 2015;13(11):2119-2126.
on recognition and management of coagulopathy in COVID- doi:10.1111/jth.13140
19. J Thromb Haemost. 2020;18(5):1023-1026. doi:10.1111/ 22. Schulman S, Kearon C. Definition of major bleeding in clini-
jth.14810 cal investigations of antihemostatic medicinal products in non-
10. Moores LK, Tritschler T, Brosnahan S, et al. Prevention,
surgical patients. J Thromb Haemost. 2005;3(4):692-694. doi:
diagnosis, and treatment of VTE in patients with coronavirus 10.1111/j.1538-7836.2005.01204
disease 2019. Chest. 2020;158(3):1143-1163. doi:10.1016/j. 23. Beigel JH, Tomashek KM, Dodd LE, et al. Remdesivir for
chest.2020.05.559 the treatment of COVID-19: final Report. New Engl J Med.
11. Cuker A, Tseng EK, Nieuwlaat R, et al. American Society of 2020;383(19):1813-1826. doi:10.1056/nejmoa2007764
Hematology 2021 guidelines on the use of anticoagulation for 24. RECOVERY Collaborative Group, Horby P, Lim WS,
thromboprophylaxis in patients with COVID-19. Blood Adv. Emberson JR, et al. Dexamethasone in hospitalized patients
2021;5(3):872-888. doi:10.1182/bloodadvances.2020003763 with COVID-19. New Engl J Med. 2021;384(8):693-704.
12. Carrier M, Abou-Nassar K, Mallick R, et al. Apixaban to prevent doi:10.1056/NEJMoa2021436
venous thromboembolism in patients with cancer. New Engl J 25. Lopes RD, de Barros e Silva PGM, Furtado RHM, et al.;
Med. 2019;380(8):711-719. doi:10.1056/NEJMoa1814468 RHM. Therapeutic versus prophylactic anticoagulation for
13. Cohoon KP, De Sanctis Y, Haskell L, McBane RD, Spiro patients admitted to hospital with COVID-19 and elevated
TE. Rivaroxaban for thromboprophylaxis among patients D-dimer concentration (ACTION): an open-label, multicentre,
recently hospitalized for acute infectious diseases: a sub- randomised, controlled trial. Lancet. 2021;397(10291):2253-
group analysis of the MAGELLAN study. J Thromb Haemost. 2263. doi:10.1016/S0140-6736(21)01203-4
2018;16(7):1278-1287. doi:10.1111/jth.14146 26. Noseworthy PA, Yao X, Abraham NS, Sangaralingham LR,
14. Billett HH, Reyes-Gil M, Szymanski J, et al. Anticoagulation McBane RD, Shah ND. Direct comparison of dabigatran,
in COVID-19: Effect of Enoxaparin, heparin, and apixaban rivaroxaban, and apixaban for effectiveness and safety in
on mortality. Thromb Haemost. 2020;120(12):1691-1699. nonvalvular atrial fibrillation. Chest. 2016;150(6):1302-1312.
doi:10.1055/s-0040-1720978 doi:10.1016/j.chest.2016.07.013
15. Paranjpe I, Fuster V, Lala A, et al. Association of treatment dose 27. Proietti M, Romanazzi I, Romiti GF, Farcomeni A, Lip
anticoagulation with In-hospital survival among hospitalized GYH. Real-world use of apixaban for stroke prevention in
patients with COVID-19. J Am Coll Cardiol. 2020;76(1):122- atrial fibrillation: a systematic review and meta-analysis.
124. doi:10.1016/j.jacc.2020.05.001 Stroke. 2018;49(1):98-106. doi:10.1161/STROKEAHA.117
16. Kabir AA. Anticoagulation is the answer in treating noncriti- .018395
cal COVID-19 patients. Open Med. 2021;16(1):1486-1492. 28. Wenzler E, Engineer MH, Yaqoob M, Benken ST. Safety and
doi:10.1515/med-2021-0354 efficacy of apixaban for therapeutic anticoagulation in critically
17. Angus DC, Berry S, Lewis RJ, et al. The REMAP-CAP (ran- ill ICU patients with severe COVID-19 respiratory disease. TH
domized, embedded, multifactorial adaptive platform trial for Open. 2020;4(4):e376-e382. doi:10.1055/s-0040-1720962
1108718
research-article2022
HPXXXX10.1177/00185787221108718Hospital PharmacyParmiter et al
Original Article
Hospital Pharmacy
Abstract
Background: Lack of insurance coverage in conjunction with the high cost of prescription drugs plays a large role in
patient’s obtaining and properly taking their medications. Many drug manufacturers provide assistance opportunities for
indigent patients lacking the financial means to obtain the medication. It is hypothesized that providing a dedicated insourced
team to coordinate this process can successfully increase medication access for indigent patient populations. Methods:
Patients with little or no means for payment whom the patient assistance team was contacted to obtain a medication were
included. Information collected included the medication requested, the drug manufacturer, the cost of the medication, the
dose required, the number of doses approved, and whether the request was approved or denied. This data was analyzed for
approval rate, most common medications requested, total cost of medications provided, as well as which system hospitals
are utilizing the program most frequently. Results: In total, 153 patients were assessed to obtain medication assistance
through the program. With an overall approval rate of 59.5%, 91 patients were provided medication through this program.
58% of patients received insurance approval and were therefore denied no-cost drug. The program was able to obtain 283
fills of no-cost medications for patients with a total value of $2 060 633.83. Conclusion: A dedicated patient assistance
program is effective in obtaining financial assistance for patients to obtain high-cost medications for which they otherwise
could not provide payment. With the continued expansion of biologic medication development, the necessity for programs
to aid indigent populations increases exponentially.
Keywords
management, pharmacoeconomics, purchasing, technicians
these programs. Outsourcing this program came with an Table 1. Value of No-Cost Medication Obtained for Patients
administrative fee billed to the institution in addition to a Through the Patient Assistance Program at Each System Hospital
perceived limited scope of services. It was then hypothe- During the First Full Year of the Program.
sized that providing this assistance locally could allow for Hospital Dollars saved
program expansion and improved coordination of care
through greater access within the electronic health record to Camden Clark Medical Center $104 790.86
patient information and provider communication. Reynolds Memorial Hospital $18 247.54
St. Joseph’s Memorial Hospital $56 958.90
United Hospital Center $65 000.00
Methods Wheeling Hospital $65 000.00
West Virginia University Hospitals $1 750 636.53
One clinical pharmacy technician FTE, due to the back-
Grand total $2 060 633.83
ground pharmacy experience and medication knowledge to
allow for educated conversations with patients and providers
during the enrollment process, was designated to launch our
Table 2. Number of Patients Provided No-Cost Medication
local patient assistance program. Initial training for this tech- Through the Patient Assistance Program and the Number of
nician involved discussions with an external company and Medication Fills Completed for the Patients During the First Full
financial counselors within our local cancer center to deter- Year of the Program.
mine the process for the program within our electronic health
record. The team also met with health-systems familiar with Patients Prescriptions
Hospital assisted filled
the infrastructure necessary for maintaining a program such
as this. Requests for medication assistance are initially routed Camden Clark Medical Center 4 12
through our local patient financial services and prior authori- Reynolds Memorial Hospital 2 5
zation teams who forward these requests along to our clinical St. Joseph’s Memorial Hospital 2 6
pharmacy technician. United Hospital Center 1 2
Patient and medication request information is gathered Wheeling Hospital 1 2
from the provider and patient or his/her representative and West Virginia University Hospitals 81 256
then entered into a third-party software containing the neces- Grand total 91 283
sary documents required to apply for the manufacturer-spon-
sored programs. Much of this work may be completed for the
patient while still admitted to the hospital during his/her ini- obtaining a medication for that patient. Data collected
tial diagnosis. Patient privacy is ensured through a business included medication requested, dose requested, approved/
associate agreement (BAA), requiring maintained security of denied, number of doses approved, and denial reason if
PHI and HIPAA compliance, entered annually with the third- applicable. Requests were then separated by each system
party company. Patient consent is obtained at time of pro- hospital to determine program utilization throughout the
gram application. Once the application is complete the WVU Medicine enterprise.
manufacturer responds with either an approval or denial.
Approval duration varies by manufacturer and medication
Results
but is generally 1 year from the time of initial approval, at
which point the patient must be re-enrolled in the program. Between January 1st, 2020 and December 31st, 2020, 153
Delivery and medication administration is then coordinated patients requested medications through the patient assistance
on a recurring basis for doses delivered to the patient’s home program. Of these 153 patients, 91 (59.5%) were approved
or prior to each appointment for facility-administered medi- and able to obtain a medication through the program. Of the
cations. Further appeals may be conducted for any denials 62 patients who were not approved for medication assistance
that may occur upon initial enrollment or re-enrollment if the through the program, 36 (58.1%) were rejected related to
patient’s situation or manufacturer’s requirements changed insurance coverage of the requested medications. The remain-
after the initial approval. ing 26 (16.9%) patients were not approved for no-cost medi-
The primary aim of this analysis was to assess the rates of cation for other reasons including, but not limited to, loss of
approval of medications requested through our insourced follow up. These unapproved patients were often due to loss
patient assistance program from January 1, 2020 through of follow up. Patients also may have not been approved if the
December 31, 2020. Secondary objectives include deter- medication was prescribed for an unapproved indication for
mining the value of medications provided to patients and which the manufacturer could not supply the medication.
assessing reason for denial. The analysis was a retrospec- For the 91 patients approved for medication through
tive, IRB-approved analysis of patient data collected by our the program, there were 283 unique fills of medication
patient assistance program. Patients were included if the (Table 2) providing a total patient savings of $2 060 633.83
patient assistance program was contacted regarding (Table 1). Patients were referred from 6 different hospitals
736 Hospital Pharmacy 57(6)
Table 3. The 5 Medications Most Commonly Obtained for inflation may also lead to more patients requiring financial
Patients Through the Patient Assistance Program During the First assistance to continue to afford their healthcare which may
Full Year of the Program. inflate this number even further. Additionally, through the
Medication N (%) continued expansion of biologic and other high-cost medica-
tion development, the necessity for a program such as this
Ocrelizumab (Ocrevus®) 25 (13.9) increases exponentially. Through the first half of 2021, 11
Pembrolizumab (Keytruda®) 17 (9.4) new medications with available patient assistance were
Immune Globulin (Privigen®) 12 (6.7)
added to the health system formulary.
Rituximab (Rituxan®) 11 (6.1)
Bevacizumab (Avastin®) 10 (5.6)
Conclusions
Our in-sourced patient assistance program is effective in
within our, then 18-member, health system. The majority, 81
obtaining financial assistance for patients to obtain high-cost
(89%) patients, were referred from the flagship academic
medications. Growth and expansion of the patient assistance
medical center. The most common (14%) no-cost medica-
team will be key to its continued. The program’s ability to
tion obtained for patients was ocrelizumab (Ocrevus®) fol-
provide these services to all patients in need throughout our
lowed by other high-cost monoclonal antibodies and
health system hinges upon the resources we are able to allo-
immune globulin (Table 3).
cate to handle requests promptly and efficiently. Programs
such as ours have opportunity throughout the entire country
Discussion but their greatest impact can be seen in poverty-stricken
areas with large income disparities.
Though not specifically targeted by the program, patients
Further research should be done to determine any links
referred to and assisted by this program are commonly
between obtaining no-cost medications and readmission rates
ordered medications for debilitating and life-threatening
of patients as gathering clinical outcomes are not currently
diseases such as multiple sclerosis and various malignan-
obtained by our team and determining these may be crucial to
cies. Many of these patients now have a limited ability to
ensuring continued success and support for the program as it
work as a result of their disease. Concerns regarding their
matures. Additionally, further analyzing treatment adherence
ability to provide for their family, continue to pay their bills,
rates for patients receiving no-cost medications compared to
and keep a roof over their head are only amplified by the
those who did not may prove beneficial. Finally, a link to
fear of not being able to afford treatment for their disease.
treatment outcomes should also be researched. Results of
Our staff feel an overwhelming sense of purpose in assisting
these further studies could allow for additional support for
these patients who often express gratitude toward our clini-
implementation of programs such as these.
cal pharmacy technician who works tirelessly obtaining
medications and financial assistance for them. The reward-
ing nature of assisting patients through this program was ORCID iD
one of the key barriers to overcome when initially transi- Calvin Parmiter https://orcid.org/0000-0002-7082-7930
tioning this program from its previous department to live
within the pharmacy department. References
With continued expansion of the health system into more 1. Parekh N, McClellan M, Shrank WH. Payment reform, medi-
disproportionately low-income, rural areas of West Virginia, cation use, and costs: can we afford to leave out drugs? J Gen
there is a consistent increase in volume of patients with need Intern Med. 2019;34(3):473-476. doi:10.1007/s11606-018-
for patient assistance programs to access medication therapy. 4794-y.
This patient population represents a vulnerable group who 2. Okunrintemi V, Khera R, Spatz ES, et al. Association of
frequently have high readmission rates.4 By assisting these income disparities with patient-reported healthcare experience.
patients through provision of medication access, adherence J Gen Intern Med. 2019;34(6):884-892. doi:10.1007/s11606-
to therapy may be improved, in turn, potentially reducing 019-04848-4
3. United States Census Bureau. Median household income (in
readmission rates.
2020 dollars), 2016–2020.
Through calculation of an estimated per-bed savings 4. Murray F, Allen M, Clark CM, Daly CJ, Jacobs DM. Socio-
based on data from the hospitals with patients assisted demographic and -economic factors associated with 30-day
through the program, an additional estimated $1 M per year readmission for conditions targeted by the hospital readmis-
could be saved across the health system with the addition of sions reduction program: a population-based study. BMC Public
new sites. Healthcare costs rising at rates higher than that of Health. 2021;21:1922. doi:10.1186/s12889-021-11987-z
1111743
research-article2022
HPXXXX10.1177/00185787221111743Hospital PharmacyChau et al
Original Article
Hospital Pharmacy
Thromboembolism journals.sagepub.com/home/hpx
Abstract
Introduction: Heparin resistance has been reported in coronavirus disease 2019 (COVID-19) patients receiving intravenous
unfractionated heparin (IV UFH). Anti-Xa monitoring of IV UFH has been suggested over activated partial thromboplastin
times due to laboratory interference from elevated factor VIII and fibrinogen levels in COVID-19 patients. Information on
heparin resistance with anti-Xa monitoring in COVID-19 patients with confirmed venous thromboembolism (VTE) is lacking.
Methods: In this retrospective cohort study of patients with radiographically confirmed VTE, IV UFH dosage requirements
in COVID-19 positive patients were compared with COVID-19 negative patients. The primary endpoint was the IV UFH
dose needed to achieve a therapeutic anti-Xa level. Secondary endpoints included time to therapeutic anti-Xa, number
of dose adjustments to achieve therapeutic anti-Xa, and bleeding. Results: Sixty-four patients with confirmed VTE were
included (20 patients COVID-19 positive, 44 patients COVID-19 negative). Eighty-five percent (17 of 20) of COVID-19
positive patients achieved anti-Xa ≥ 0.3 units/mL with the first anti-Xa level drawn post-IV UFH infusion initiation. The
median UFH dose needed to achieve first therapeutic anti-Xa was similar between COVID-19 positive and COVID-19
negative patients (median [IQR]: 18 units/kg/hour [18-18] vs 18 units/kg/hour [18-18], P = .423). The median number of dose
adjustments and time to achieve therapeutic anti-Xa were also similar between the 2 groups. The frequency of patients
receiving IV UFH of more 35 000 units/day did not differ between the 2 groups. Two cases of clinically significant heparin
resistance in the COVID-19 positive group were identified. Conclusions: During the first wave of COVID-19, heparin
dose and time to therapeutic anticoagulation appeared to be similar between COVID-19 positive and COVID-19 negative
patients monitored by anti-Xa at our institution. More studies are required to evaluate clinically significant heparin resistance
in the context of the wide range of viral variants which developed, and beyond the population observed in this single center
retrospective study.
Keywords
anticoagulants, blood, COVID, critical care
patients showed these patients had normal antithrombin val- are checked daily. For patients unable to achieve anti-Xa
ues, but elevated levels of factor VIII and fibrinogen. ≥ 0.3 units/mL within 24 hours of IV UFH protocol initia-
Therefore, it is suggested anti-Xa activity assays be utilized tion, they were further evaluated for potential clinically sig-
over aPTT monitoring in COVID-19 patients.7,8 nificant heparin resistance. Heparin resistance was defined
We conducted a retrospective cohort study to determine if as either: (1) requirement >35 000 units of IV UFH per day
COVID-19 positive patients required more IV UFH to to achieve anti-Xa ≥ 0.3 units/mL; or (2) >35 000 units of IV
achieve therapeutic anticoagulation with anti-Xa monitoring UFH per day and unable to achieve anti-Xa ≥ 0.3 units/mL.
compared to COVID-19 negative patients. We hypothesized
the heparin dosage needed to achieve therapeutic anti-Xa
Results
levels between COVID-19 positive and COVID-19 negative
patients would not differ significantly when using anti-Xa A total of 321 patients were screened for eligibility. We
monitoring. included 64 patients with confirmed VTE in the analysis (20
patients COVID-19 positive, 44 patients COVID-19 nega-
tive). COVID-19 positive patients were more likely to be
Methods admitted or transferred to the intensive care unit (85% vs
This was an IRB-approved, retrospective medication use 52.3%, P = .014), receive mechanical ventilation (65% vs
evaluation of IV UFH in adult patients with radiographi- 31.8%, P = .013), and experience acute kidney injury (25%
cally confirmed VTE during the first wave of the COVID- vs 2.3%, P = .010). There were no significant differences
19 pandemic (March 2020 through June 2020). Exclusion between age, weight, body mass index, or the use of UFH
criteria included: conditions which may alter heparin phar- loading dose between the 2 groups (Table 1). The median
macokinetics including pregnancy and targeted tempera- duration of hospitalization was longer in COVID-19 positive
ture management; acquired conditions which may increase patients (median [interquartile range (IQR)]: 18.5 days [7.3-
risk for heparin resistance, including extracorporeal mem- 27] vs 7 days [4-17.8], P = .006).
brane oxygenation; less than 3 anti-Xa levels drawn or Eighty-five percent (17 of 20) of COVID-19 positive
duration of IV UFH infusion of less than 24 hours; or patients achieved anti-Xa ≥ 0.3 units/mL with the first anti-
received a direct oral anticoagulant (DOAC) prior to IV Xa level drawn post-IV UFH infusion initiation. In COVID-
UFH initiation (to avoid DOAC influence on the heparin 19 negative patients, 90.9% (40 of 44) achieved this value
anti-Xa assay). Baseline and peak inflammatory markers, with the first anti-Xa lab draw. For the primary outcome
such as C-reactive protein, ferritin, fibrinogen, D-dimer, (Table 2), the median UFH dose needed to achieve first anti-
and lactate dehydrogenase were collected in the COVID-19 Xa ≥ 0.3 units/mL was similar between COVID-19 positive
positive patients. and COVID-19 negative patients (median [IQR]: 18 units/
The primary outcome was to compare the IV UFH dose kg/hour [18-18] vs 18 units/kg/hour [18-18], P = .423). The
needed to achieve an anti-Xa ≥ 0.3 units/mL between median number of dose adjustments and time to achieve anti-
COVID-19 positive and COVID-19 negative patients. Xa ≥ 0.3 units/mL were also similar between the 2 groups.
Anti-Xa levels of <0.3, 0.3 to 0.7, and >0.7 units/mL were All patients achieved anti-Xa ≥ 0.3 units/mL with IV
considered to be subtherapeutic, therapeutic, and suprathera- UFH. A total of 23 patients (7 COVID-19 positive and 16
peutic, respectively. Secondary outcomes included: time to COVID-19 negative) required more than 35 000 units/day to
anti-Xa ≥ 0.3 units/mL; number of dose adjustments needed achieve or maintain anti-Xa ≥ 0.3 units/mL. There was no
to achieve anti-Xa ≥ 0.3 units/mL; IV UFH dose needed to significant difference between the COVID-19 positive and
achieve therapeutic anti-Xa (0.3-0.7 units/mL); number of COVID-19 negative patients in the frequency of IV UFH
dose adjustments needed to achieve therapeutic anti-Xa con- requirements greater than 35 000 units/day (35% vs 36.5%,
centrations; time to therapeutic anti-Xa concentrations; and P = .916). The median weight for these patients were 99.1 kg
bleeding. (IQR: 84.3-115.2). The median IV UFH dose required for
Data was analyzed with the Chi-Square/Fisher’s exact test first anti-Xa ≥ 0.3 units/mL achievement was 18 units/kg/
and Mann-Whitney U-test where appropriate. All laboratory hour (IQR: 18-18). The median time to anti-Xa ≥ 0.3 units/
analyses were performed using the same analyzer (ACL TOP mL achievement was 6.7 hours (IQR: 6.2-12.5). There was
750 CTS; Werfen) and reagents (HemosIL Liquid Anti-Xa; no statistically significant difference between median weight,
Werfen). median IV UFH dose for first anti-Xa ≥ 0.3 units/mL, or
Our institution’s VTE protocol has an optional IV UFH median time to anti-Xa ≥ 0.3 units/mL between the COVID-
bolus of 80 units/kg with an initial starting dose of 18 units/ 19 positive and COVID-19 negative patients.
kg/hour. Doses were based upon actual body weight and Three patients (2 COVID-19 positive patients and 1
dose-capping was at the provider discretion during the study COVID-19 negative patient) did not achieve anti-Xa
timeframe. Anti-Xa levels are drawn 6 hours after initiation ≥ 0.3 units/mL within 24 hours of UFH initiation. Of these 3
of the UFH infusion and every 6 hours thereafter until 2 con- patients, we identified 2 cases of potentially clinically sig-
secutive anti-Xa levels are therapeutic, then anti-Xa levels nificant heparin resistance. Both cases were COVID-19
Chau et al 739
Note. Median (IQR) unless otherwise specified. BMI = body mass index; ICU = intensive care unit; UFH = unfractionated heparin; AST = aspartate
transaminase; ALT = alanine transaminase; PT = prothrombin time; INR = international normalized ratio; aPTT = activated partial thromboplastin time.
a
Data available for 19 COVID-19 positive patients, 36 COVID-19 negative patients.
b
Data available for 18 COVID-19 positive patients, 36 COVID-19 negative patients.
c
Data available for 19 COVID-19 positive patients, 43 COVID-19 negative patients.
d
Data available for 14 COVID-19 positive patients, 11 COVID-19 negative patients.
e
Data available for 18 COVID-19 positive patients.
Table 2. Results.
COVID-19 COVID-19
positive n = 20 negative n = 44 P-value
Primary outcome
UFH dose for first anti-Xa ≥ 0.3 units/mL, units/kg/h 18 (18-18) 18 (18-18) .423
UFH dose ranges for first anti-Xa ≥ 0.3 units/mL, units/kg/h, n (%)
18 17 (85) 40 (90.9) .448
>18-22 1 (5) 3 (6.8)
>22-26 1 (5) 1 (2.3)
>26-30 0 (0) 0 (0)
>30 1 (5) 0 (0)
Other outcomes
Time to first anti-Xa ≥ 0.3 units/mL, h 6.13 (5.92-9.44) 6.25 (6.03-7.29) .317
Number of dose adjustments to achieve ≥0.3 units/mL 0 (0-0) 0 (0-0) .430
Frequency of dose adjustments to achieve ≥0.3 units/mL, n (%)
0 17 (85) 40 (90.9) .173
1 1 (5) 3 (6.8)
2 0 (0) 1 (2.3)
>2 2 (10) 0 (0)
First anti-Xa level post-UFH infusion ≥0.3 units/mL, n (%) 17 (85) 40 (90.9) .483
UFH dose for first therapeutic anti-Xa level range, unit/kg/h 15 (12-18) 15.5 (12.3-18) .865
Number of dose adjustments to achieve first therapeutic anti-Xa level range 1.5 (0.25-3) 1 (0-2) .190
Time to therapeutic target range attainment, h 21.1 (7.8-32.7) 15.8 (9.1-27.9) .690
High dose heparin infusions
Received >35 000 units/day to achieve or maintain anti-Xa ≥ 0.3 units/mL 7 (35) 16 (36.4) .916
Clinically significant heparin resistance* 2 (10) 0 (0) .094
positive, exhibited prolonged time to achieve anti-Xa 35 000 units/day of UFH per day. Details of these 3 patients
≥ 0.3 units/mL, and required over 50 000 units of IV UFH are described in Table 3.
per day. The third patient was determined not to be heparin Initial anti-Xa levels were frequently supratherapeutic
resistant as prolonged time to achieve anti-Xa ≥ 0.3 units/mL (>0.7 units/mL) and occurred in 64.1% of our patients (70%
was attributed to IV access issues and required less than in COVID-19 positive and 61.4% in COVID-19 negative
Chau et al 741
*In patients unable to achieve anti-Xa > 0.3 units/mL within 24 hours; heparin resistance definition: (1) requirement >35 000 units of IV UFH per day to
achieve anti-Xa ≥ 0.3 units/mL; or (2) >35 000 units of IV UFH per day and unable to achieve anti-Xa ≥ 0.3 units/mL)
patients, P = .504). There were no differences in the UFH resistance, the patient may either be transitioned to a direct
dose needed to achieve therapeutic anti-Xa range concentra- thrombin inhibitor or antithrombin may be increased by
tions (0.3-0.7 units/mL) between the COVID-19 positive and supplementation.
COVID-19 negative patients (median [IQR]: 15 units/kg/ For IV UFH patients with elevated levels of factor VIII
hour [12-18] vs 15.5 units/kg/hour [12.3-18], P = .865). The and fibrinogen, such as those seen with substantial systemic
median number of dose adjustments needed to achieve thera- inflammation, aPTT values will be falsely low but anti-Xa
peutic anti-Xa range was also similar (P = .190). Time to values will not be affected. These cases have been referred to
therapeutic target range attainment was longer in the COVID- as heparin pseudo-resistance.6,11 Management of these cases
19 group, but not statistically significant (median [IQR]: include changing the laboratory monitoring method of IV
21.1 hours [7.8-32.7] vs 15.8 hours [9.1-27.9], P = .690). UFH from aPTT to anti-Xa. Alternatively, anticoagulation
Hemoglobin decrease ≥2 g/dL from baseline occurred with low molecular weight heparin would be an acceptable
more frequently in COVID-19 patients (80% vs 47.7%, option. Switching to a direct thrombin inhibitor would not be
P = .028), but there was no difference in number of patients ideal as aPTT is typically used to monitor therapy.12
with hemoglobin <7 g/dL compared to COVID-19 negative Any COVID-19 patient with suspected heparin resistance
patients (35% vs 29.5%, P = .633). Protamine administration needs to be evaluated in the context of the laboratory test
occurred once in the COVID-19 negative group. No patients used to monitor IV UFH and the target anticoagulation goal.
received ≥2 units of packed red blood cells during or imme- Small observational, retrospective studies in COVID-19
diately after IV UFH infusion. Full International Society of patients with heparin resistance have shown normal anti-
Thrombosis and Hemostasis Criteria for major bleeding and thrombin values, but elevated levels of factor VIII and fibrin-
clinically relevant non-major bleeding was unable to be col- ogen. Therefore, it is suggested anti-Xa activity assays be
lected due to the retrospective nature of the study. utilized over aPTT monitoring in COVID-19 patients as
aPTT will be falsely low due to elevation of those factors.7,8
In our overall patient cohort, dosage requirements and
Discussion time to anti-Xa ≥ 0.3 units/mL with an UFH anti-Xa moni-
Heparin resistance has been traditionally defined as: (1) the toring strategy for VTE treatment were similar between
need for IV UFH doses of >35 000 units/day to achieve tar- COVID-19 positive and COVID-19 negative patients.
get anticoagulation; or (2) the need for more than 500 units/ While the incidence of heparin resistance outside of cardiac
kg to achieve target activated clotting time in patients under- surgery is unknown, small retrospective studies reported
going cardiopulmonary bypass.9,10 The mechanisms behind potential heparin resistance in 75% to 80% of examined
heparin resistance include: increased heparin-binding pro- COVID-19 patients.8,13 Notably, we identified 10% of our
teins; increased heparin clearance; antithrombin deficiency; COVID-19 positive patients with potential clinically sig-
and elevated levels of factor VIII and fibrinogen.10 nificant heparin resistance with anti-Xa monitoring. Our
In a patient with increased heparin-binding proteins, results suggest concerns for apparent heparin resistance in
increased clearance, or antithrombin deficiency, the aPTT COVID-19 patients may potentially be mitigated with anti-
and anti-Xa values will be lower than expected while receiv- Xa monitoring of IV UFH. While the COVID-19 group pre-
ing IV UFH. Management of heparin resistance due to sumably had higher inflammatory markers, higher rates of
increased heparin-binding proteins and clearance include acute renal failure and hypertriglyceridemia were also pres-
increasing the heparin dosage or consideration of an alterna- ent. Renal failure and elevated triglycerides (>360 mg/dL)
tive anticoagulation (ie, direct thrombin inhibitor). If anti- may impact anti-Xa levels through decreased heparin clear-
thrombin deficiency is identified as the cause of heparin ance and assay interference, respectively.14 In this study,
742 Hospital Pharmacy 57(6)
only one patient received IV UFH with hypertriglyceride- were minimized both treatment groups were on IV UFH
mia concomitantly. The significance of the interaction during the same timeframe. Another strength is the inclusion
between these factors on anti-Xa achievement would need of patients only with radiographically confirmed VTE. Our
to be further evaluated in larger studies. study timeframe spanned the first wave of the COVID-19
There are several limitations to this retrospective, single- pandemic when there was much discussion in both the
center study. The small sample size could increase the risk of public and medical media about thromboembolic events,
a Type II error. One reason for the small number of COVID- whether macrovascular or microvascular, contributing to
19 positive patients was our strict inclusion criteria of radio- the complexity and severity of illness of patients diagnosed
graphically confirmed VTE with IV UFH treatment. Previous with COVID-19. During this first wave, some institutions
studies reporting heparin resistance in COVID-19 positive adopted practices for empiric therapeutic-dose and inter-
patients included patients on UFH or low molecular weight mediate-dose prophylactic anticoagulation. Our institution
heparin with or without confirmed thrombosis.7,8,13 Second, did not adopt these practices due to inadequate supporting
we included patients on UFH infusions for at least 24 hours. evidence in the literature. Much data has since been pub-
We may not have captured all heparin resistant patients as lished on the subject of therapeutic-dose and intermediate-
those with initial low anti-Xa levels may have transitioned to prophylactic dose anticoagulation in COVID-19 in the
alternative anticoagulants. Third, inflammatory markers 2 years the pandemic has impacted the United States.15-19
such as interleukin-6, C-reactive protein, ferritin, and While the authors’ note the importance of VTE prophy-
D-dimer are not routinely obtained in our COVID-19 nega- laxis, our study is different as it only included patients with
tive patients thus we were unable to make comparisons confirmed VTE and an indication for therapeutic-dose
between the 2 study groups. It is assumed the COVID-19 anticoagulation. At this time, empiric therapeutic antico-
positive group would have higher levels of inflammation agulation in critically ill COVID-19 patients is not recom-
than COVID-19 negative patients. Fourth, the impact of mended.20,21 The majority of our COVID-19 patients were
COVID-19 variants on heparin resistance and thrombotic critically ill and our results reflect a more realistic clinical
complications is largely unknown. This study reviewed situation where therapeutic anticoagulation would be uti-
COVID-19 patients during the first wave in the northeastern lized in the intensive care unit. To date, there is a lack of
United States (March through June 2020). The results of this large studies examining the use of IV UFH for the treat-
study may not be generalizable to the current variant. Fifth, ment of VTE in COVID-19 patients as these studies uti-
although doses and time to anti-Xa ≥ 0.3 units/mL achieve- lized empiric anticoagulation or primarily low molecular
ment was evaluated in our study, ongoing IV UFH doses weight heparins, such as enoxaparin.15-19
required to maintain this therapeutic target was not evalu-
ated. Heparin resistance was primarily screened during the Conclusion
acute phase of IV UFH initiation and there is a possibility of
the development of heparin resistance later in the hospital Previous studies have suggested anti-Xa over aPTT monitor-
stay. One final consideration is the definition of heparin ing of IV UFH may be advantageous in COVID-19 patients.
resistance. While doses of >35 000 units/day of IV UFH has Our study revealed heparin dosage and time to therapeutic
been suggested,9,10 this definition has not been validated in anticoagulation appear to be similar between COVID-19
clinical trials. Additionally, patient body weight will signifi- positive and COVID-19 negative patients with anti-Xa mon-
cantly impact heparin resistance rates using this definition. itoring. COVID-19 patients may be successfully anticoagu-
Patients in our study had median weights of 92.1 and 84.4 kg lated with IV UFH using an anti-Xa monitoring strategy;
in the COVID-19 positive and negative groups, respectively. however concern for clinically significant heparin resistance
Generally accepted doses for initial treatment of VTE in COVID-19 patients monitored with anti-Xa levels may
(18 units/kg/hour) could easily exceed 35 000 units/day of IV still exist based on our findings. Future studies examining
UFH with these body weights. With strict application of this the use of anti-Xa for IV UFH in COVID-19 patients are
definition alone, 23 patients (median weight of 99.1 kg) needed to strengthen the recommendation to use an anti-Xa
would meet heparin resistance criteria (7 COVID-19 positive monitoring strategy over aPTT and to further define heparin
and 16 COVID-19 negative). By applying the definition of resistance in this population.
heparin resistance on patients unable to achieve anti-Xa
≥ 0.3 units/mL within 24 hours of IV UFH initiation, we Declaration of Conflicting Interests
identified 2 potentially clinically significant cases of heparin The author(s) declared no potential conflicts of interest with respect
resistance in COVID-19 positive patients with active VTE to the research, authorship, and/or publication of this article.
where delayed time to therapeutic goal may be detrimental.
Strengths of our study are the use of a control group and Funding
limiting therapeutic anticoagulation to IV UFH only. Any The author(s) received no financial support for the research, author-
concerns about variation in heparin potency in the IV UFH ship, and/or publication of this article.
Chau et al 743
Original Article
Hospital Pharmacy
Abstract
Aim: Implementation of a web-form based pharmacovigilance plan for the spontaneous notification of adverse events to
the Comirnaty® COVID-19 vaccine during its administration to hospital healthcare professionals. Methods: An electronic
pharmacovigilance form was developed containing 8 pre-defined event options, an open answer option for the description
of other events and/or symptoms, and a question about the overall intensity of symptoms. The adverse events reports were
standardized according to physiological and pathological condition. Results: A total of 4119 adverse events notifications were
obtained with a 45% rate of electronic notification. The most clinically relevant events reported were:tachycardia (n = 19),
dyspnea (n = 7), chest pain (n = 6), facial/labial edema (n = 6), lipothymia (n = 5), bronchospasm (n = 2), herpetic infection
(n = 2), vasculitis (n = 2), arrhythmia (n = 1), difficult to control arterial hypertension (n = 1), gastritis (n = 1), and spontaneous
abortion (n = 1). Regarding the intensity of symptoms (n = 2928), 70.0% were reported as mild, 25.8% as moderate, and 4.27%
as severe, with higher intensity in the second dose compared to first dose. The highest frequency of severe events were
reported in the groups from 40 to 59 years in both vaccination periods. During the vaccination process, no hospitalizations
and no deaths were notified and/or recorded. Conclusion: In this real world study, comparing with Comirnaty clinical trials
program, it was observed a higher frequency of adenomegaly and gastrointestinal disorders. Noteworthy, the notification of
a case of miscarriage. The use of hospital pharmacy pharmacovigilance electronic forms, seemed to be relevant to notification
adherence and to obtain a greater and faster knowledge of COVID-19 vaccine safety profile
Keywords
COVID, vaccines, medication safety, adverse drug reactions reporting/monitoring, pharmacists, education
Key Messages 1
Pharmacy Service of Centro Hospitalar Universitário Lisboa Norte,
Hospital de Santa Maria, Lisbon, Portugal
WHAT IS ALREADY KNOWN ON THIS SUBJECT 2
iMed.ULisboa – Research Investigation for Medicines, Faculty of
Pharmacy, University of Lisbon, Lisbon, Portugal
3
•• Emergence of the COVID-19 pandemic lead to rap- Information and Safety of Medicines, Pharmacovigilance Unit of Pharmacy
Service of Centro Hospitalar Universitário Lisboa Norte, Hospital de
idly development of vaccines against SARS-CoV-2. Santa Maria, Lisbon, Portugal
•• The information related to COVID-19 vaccines safety 4
Occupational Health Medical Service of Centro Hospitalar Universitário
are mainly based on clinical trials data and there is a Lisboa Norte, Hospital de Santa Maria, Lisbon, Portugal
lack of real world data. Corresponding Author:
•• Electronic pharmacovigilance forms are important João Paulo Garcia Lopes da Cruz, Pharmacy Service of Centro Hospitalar
Universitário Lisboa Norte/Hospital de Santa Maria, Av. Egas Moniz, Nº1,
fast resources for obtaining safety data, but there are
Lisboa 1649-028, Portugal.
limit data in context of COVID-19 vaccines. Email: jplopescruz@chln.min-saude.pt
Cruz et al 745
WHAT THIS STUDY ADDS passive pharmacovigilance plans, namely through spontane-
ous notification of adverse events with further analysis of the
•• Although this real-world study shows an overall causal relationship between event and drug use.5
adverse event profile similar to Comirnaty’s clinical Current pharmacovigilance systems are able to identify
trials, it also reveals an increase in the frequency of many of the main safety issues with medications.6 In the con-
adenomegaly, gastrointestinal disturbances, and one text of passive pharmacovigilance (encouraging healthcare
case of miscarriage. professionals and others to notify safety issues), in which noti-
•• The hospital pharmacy activity, namely through the fication depends entirely on the initiative and motivation of
development of electronic pharmacovigilance tools, individuals to do so, the main objective is the early detection
could contribute to fast obtain real data on safety of of new adverse drug reactions (ADRs), rare and serious ADRs.
COVID-19 vaccines. It has the advantage of being a comprehensive and cost-effec-
tive methodology, however, as the weak point is elevated
underreporting with one notification estimate of 6% of total
Introduction ADRs.6 In this methodology, several actions can be consid-
With the emergence of the COVID-19 pandemic, there was a ered: sending and filling out reporting forms, which can be
need to rapidly investigate and carry out clinical trials allow- collected regularly, individually or grouped, by phone, email
ing the subsequent commercialization of effective vaccines or web-based (electronic forms). Ease in reporting an adverse
against SARS-CoV-2. event is a key factor in obtaining a good notification rate.7
Comirnaty® (Pfizer-BioNTech) vaccine was the first vac- The development and use of national electronic forms has
cine against SARS-CoV-2 (developed using messenger RNA been a growing option for medicines agencies from several
technology) to be approved with conditional authorization countries around the world, as a way of increasing and accel-
by the European Medicines Agency (EMA) based on a Phase erating the spontaneous reporting of adverse reactions. The
1/2/3 multicenter, multinational, randomized, placebo-con- implementation of pharmacovigilance plans by the Hospital
trolled, observer blind clinical trial that evaluated its efficacy Pharmacy has also been shown to be effective in detecting
and safety in approximately 44 000 participants who received adverse reactions both in the hospital and outpatient settings,
2 doses of the vaccine.1 with special emphasis in the context of the COVID-19
Concerning the safety analysis, during the course of the pandemic.8-10
aforementioned clinical trial, in approximately 21 744 par- The aim of this study was to analyze the implementation
ticipants aged 16 years or more submitted to at least one dose and the results of a web-form based pharmacovigilance plan
of the vaccine, the symptoms mostly identified were of mild for the spontaneous notification of adverse events to the
to moderate intensity. Mostly of these events were associated Comirnaty® COVID-19 vaccine during its administration to
with pain at the injection site (>80%), fatigue (>60%), a population of healthcare professionals from a central uni-
headache (>50%), myalgia and chills (>30%), arthralgia versity hospital, within the scope of routine Hospital
(>20%), pyrexia and swelling at the injection site (>10%), Pharmacy activity, in collaboration with the Occupational
and whose frequency was more associated with older age.1-3 Health Service.
Also, information on the use of this vaccine in pregnancy is
still limited. Animal studies do not indicate direct or indirect
Methods
adverse effects with respect to pregnancy, embryonal/fetal
development, parturition, or postnatal development. It is not In Portugal and also at Centro Hospitalar Universitário
known whether it is excreted in breast milk. Therefore, its Lisboa Norte (CHULN), the vaccination process for health-
administration in pregnancy, until further evidence, should care professionals began on December 27th, 2020 and until
only be considered when the potential benefits outweigh any February 24th, 2021, a total of 4700 individuals have been
potential risks for the pregnant woman, fetus and/or infant.3,4 vaccinated, of which 4392 (93.4%) completed vaccination (2
Thus, because clinical trials take place in a relatively short doses). Data was gathered from December 28th, 2020 until
period and because the enrolled population generally excludes April 1st, 2021.
certain individuals with specific physiological conditions, This is a single-center cross-sectional study, with data
such as pregnancy, pediatric age and/or important comorbidi- collected over a period of 4 months. All eligible healthcare
ties, it is crucial to keep on drug safety surveillance, namely professionals (those who had the disease COVID-19 for less
through its use in real life context, since it will be used in a than 6 months or those who had signs and symptoms sugges-
large number of individuals of several ages, physiological tive of the disease on the day of vaccination were excluded)
conditions and comorbidities, that the rare (≥1/10 000 to within the scope of the vaccination plan of healthcare pro-
<1/1000) and very rare events (<1/10 000) may arise. fessionals at the CHULN, were able to participate in this
This post-authorization drug safety assessment can be study. The Comirnaty® vaccine was outlined to be adminis-
carried out either through phase IV clinical trials, develop- tered in 2 doses 21 days apart (per national government
ment of safety record databases, or development of active or guidance).4
746 Hospital Pharmacy 57(6)
Vaccines were mostly (95%) prepared by the Hospital At least one hospital pharmacist was always present at the
Pharmacy in a laminar flow cabinet, in individual pre-filled vaccination center. During the observation period (30 minutes
syringes and administered at a vaccination center by the after administration of the vaccine), this pharmacist presented
nursing team. The administration was clinically supervised the electronic pharmacovigilance form developed to the vac-
by the Occupational Health Service medical team and an cinees, reinforcing for the importance of its voluntary and
allergist/immunologist with proper medical equipment for individual ADR reporting, after receiving an email with a link
any assistance in case of severe allergic adverse reaction, direct access to the form (in compliance with data protection
supported by the emergency department medical team. regulations and professional secrecy). He also explained the
An electronic pharmacovigilance form was developed by most frequently expected possible adverse events.
the Hospital Pharmacy Drug Safety Unit and was sent by The data were collected by a hospital pharmacist by
e-mail to CHULN healthcare professionals for spontaneous downloading the answers from the e-form sent. The obtained
notification of adverse events. This e-form was performed adverse events reports were standardized according to physi-
with 8 pre-defined event options (according to the highest ological and pathological condition, and submitted system-
frequency observed in development clinical trials), and an atically to the pharmacovigilance department of the
open answer option for the description of other events and/or Portuguese Medicines Regulatory Authority (INFARMED)
symptoms not previously described in the form (Box 1). for further analysis and incorporation into the European
During the course of the study, a question about the overall Database of Pharmacovigilance (EudraVigilance) of
intensity of symptoms in a pre-defined scale (mild, moder- COVID-19 vaccines. Adverse events of greater clinical rel-
ate, or severe) was added. evance and/or seriousness were additionally reported directly
Cruz et al 747
Figure 1. Flow diagram of vaccinated healthcare professionals, submitted electronic forms and ADR reporting, according to first and
second dose of Comirnaty® vaccine.
to INFARMED web-form (Portal RAM), in accordance with labial edema (n = 6), lipothymia (n = 5), bronchospasm
this institution guidance.11 (n = 2), herpetic infection (n = 2), vasculitis (n = 2), arrhyth-
Statistical descriptive analysis was performed using R/ mia (n = 1), difficult to control arterial hypertension (n = 1),
RStudio software (version 4.0.2). gastritis (n = 1), and spontaneous abortion (n = 1).
The adverse events reported according to the first and sec-
ond dose with frequency n > 5 are shown in Figure 2.
Results The question regarding the overall intensity of symp-
About 4096 spontaneous notifications were made by health- toms was only incorporated later into the form, as men-
care professionals through filling the developed electronic tioned earlier. Data was obtained for 2928 notifications
form (notification rate of 45%), and 23 directly to the (71.1% of the initial sample), with 771 (26.3%) concerning
Occupational Health Service medical team (notification rate the first dose and 2157 (73.7%) the second dose. Of these,
of 0.25%), according with the study flow diagram (Figure 1). 70.0% were reported as having mild intensity (87.3%–first
Thus, within the scope of the implemented pharmacovigi- dose, 63.8%–second dose), 25.8% as moderate (11.8%–
lance plan, a total of 4119 notifications were obtained, with first dose, 30.7%–second dose), and only a low percentage
1723 referring to the first dose and 2393 to the second dose, (4.27%) reported having symptoms of severe intensity
and 3 not indicating which dose they referred to. About (0.91%–first dose, 5.47%–second dose) (Supplemental
14.6% (n = 251) and 8.36% (n = 200) reported having no Table 1).
symptoms in the first and second doses, respectively. The Figure 3 shows the events most frequently reported
The ages of the notifiers varied between 20 and 83 years according to the degree of intensity and dose administered,
old (average of 40.9 years old), with around 78.1% belonging generally showing an increase in the degree of intensity in
to the female gender. the second dose compared to the first dose and Table 2 details
In Table 1 a description of the adverse events reported the events reported with severe intensity.
with a frequency greater than 1% is shown, after excluding In the analysis by age group, the highest frequency of
notifications that did not report the respective dose and those severe events occurred in the groups from 40 to 59 years in
that reported having no symptoms. both vaccine administrations. Of notice, the age group
Regarding the most clinically relevant events reported, between 60 and 69 years of age did not report any event of
regardless of their frequency, the following stand out: tachy- severe intensity after the first dose, and after the second dose,
cardia (n = 19), dyspnea (n = 7), chest pain (n = 6), facial/ the percentage was the lowest (7.2%) among all (Table 3).
748 Hospital Pharmacy 57(6)
Table 1. Adverse Events Reported by Healthcare Professionals During the vaccination process, no hospitalizations were
with a Frequency Greater Than 1%. recorded and no deaths were notified and/or recorded.
Reported adverse events N %
Discussion
Injection site (pain, redness, swelling, itching) 3350 91.40
From the analysis of the received notifications, it is high-
Head, joint, muscle, body pain 1621 44.23
lighted, in general, the higher frequency of notifications
Fatigue, drowsiness 1180 32.20
related to the second administered dose (54.5%). Reporting
Chills 751 20.49
Febrile 431 11.76
by gender (female: 78.1%) was in line with the WHO
Nausea, vomiting, diarrhea, abdominal pain 376 10.26 Working Paper in “Gender equity in the health workforce:
Adenomegaly 123 3.36 Analysis of 104 countries” that women account for 70% of
Rash 45 1.23 all workers in the health and social sector.12 At our hospital,
this rate is even higher (76.9%).
Figure 2. Percentage of adverse events reported according to the first and second dose of the Comirnaty® vaccine.
Table 2. Frequency and Percentage of Reported Adverse Events, in the Notifications That Mentioned the Degree of Intensity as Severe.
Severe reported adverse events N %
First dose
Othersa 7 0.91
Head, joint, muscle, body pain 5 0.65
Injection site (pain, redness, swelling, itching) 4 0.52
Chills 2 0.26
Fatigue, drowsiness 2 0.26
Second dose
Head, joint, muscle, body pain 110 5.10
Fatigue, drowsiness 95 4.40
Injection site (pain, redness, swelling, itching) 94 4.36
Chills 76 3.52
Febrile 76 3.52
Nausea, vomiting, diarrhea, abdominal pain 40 1.85
Adenomegaly 15 0.70
Othersb 11 0.51
Tachycardia 3 0.14
a
Adenomegaly; Arrhythmias; Dyspnea; Febrile; Paresthesias; Rash; Tachycardia (each, N = 1; 0.13%).
b
Bronchospasm; Dyspnea; Facial/labial edema; Hands and feet pain; Insomnia; Muscle weakness; Nasal obstruction; O2 desaturation; Paresthesias; Rash;
Rhinitis, rhinorrhea (each, N = 1; 0.05%)
Cruz et al 749
Figure 3. Most frequently reported adverse events according to the degree of intensity and dose administered.
Table 3. Frequency and Percentage of Severe Adverse Events In the second dose, events were, in general, more fre-
Notifications by Age Group. quently reported as moderate (30.7%) and severe (5.47%),
compared to the first dose (11.8% and 0.91%, respectively;
Nº Severe reported
Age group (years) Nº Reports adverse events % P < .01). The population aged with 60 years or older reported
7.2% of severe events in the second dose.
First dose Additional data (in process, n = 682) about the reports of
20-29 162 3 1.86 the intensity of symptoms related with booster dose (third
40-49 216 10 4.63 dose) show that 58.5% (n = 357) were reported as mild,
50-59 136 7 5.15 31.8% (n = 190) as moderate, 10.0% (n = 61) as severe and
Second dose 0.3% (n = 2) as very severe (this degree of severity was
20-29 428 87 20.3
reported only with the booster dose).
30-39 639 136 21.29
The presence of adenomegaly (or lymphadenopathy) was
40-49 491 175 35.64
reported more frequently compared to those occurred in clin-
50-59 377 96 25.47
ical trials (3.4%; n = 123% vs 0.3%; n = 64, respectively),
60-69 194 14 7.21
with n = 16 reported as severe in our study.1
Of note, it was the notification of gastrointestinal disor-
Regarding age, it was found that the highest percentage of ders (nausea, vomiting, diarrhea, abdominal pain) which, in
severe notifications occurred in the 40 to 49 age group being aggregate, corresponded to a frequency greater than 10%
less reported in age groups above 60 years-old (both doses), (10.3%; n = 376). About n = 40 were reported as having been
in line with what was observed in clinical trials in which the of severe intensity; all after the second dose.
incidence of systemic reactogenicity was higher between 16 Severe intensity of fatigue was reported at approximately
and 55 years.1 3.3% (n = 97/2928) compared to approximately 4% reported
The most frequently reported events were similar to those in phase 2/3 clinical trials, with the majority of cases occur-
previously reported in phase 2/3 trials, that showed a higher ring at the second dose. (4.4%; n = 95/2157).1
percentage of events reported after the second dose com- One case of miscarriage has been reported at 6 weeks’
pared to the first dose (70.3% and 29.7%, respectively).3,13 gestation, 1 month after the second administration, in a
We also noticed in this study that the frequency of injection woman with no history of previous abortions. Of note, there
site reactions was similar for the 2 administered doses (93.9% was a positive antiphospholipid antibody analysis 2 weeks
and 89.7%, respectively). after the abortion. This case was later notified by the Hospital
750 Hospital Pharmacy 57(6)
Pharmacy Drug Safety Unit to the Portuguese Medicines The development of pharmacovigilance activity through
Regulatory Authority (INFARMED) through the national the use of electronic forms by the Hospital Pharmacy, seemed
electronic pharmacovigilance web-form (Portal RAM – relevant to obtain a higher rate of adherence to spontaneous
INFARMED). According to studies carried out, namely on adverse event reporting and, consequently, a greater and
data collected in the UK population, the percentage of abor- faster knowledge of Comirnaty® vaccine safety profile.
tions in the first trimester in vaccinated women was not
higher than the percentage of abortions in unvaccinated Acknowledgments
women. A causal relationship between abortion and the Not applicable.
Comirnaty® vaccine has not been established.14
The present electronic reporting methodology was applied Contributorship Statement
in a time period very close to administration (up to approxi- The specific contribution of authors in the paper was:
mately 7 days after vaccine administration) and is therefore Planning, conception and design, interpretation of data, paper
not sensitive to the occurrence of adverse events later in writing: João Paulo Garcia Lopes da Cruz.
time. Thus, it is our intention to carry out a new evaluation, Conduct the study, reporting, introduce the data: Cristina da
using the same methodology, at 6 months and 1 year after Conceição Ribeiro de Carvalho, Paula Alexandre da Cruz Silva,
Luis Filipe Campos Guerreiro.
vaccine administration. Also, as a limitation, reported ADRs
Acquisition of data: Tatiana Vedes Bento, Leila Vanessa Carmo
in the present study were not clinically checked by a pharma- Cardoso Martins Costa, Raquel Filipa Miguel Margarido Duarte
covigilance health care professional, namely its cause effect Simões, Rui Pedro Pinheiro Gonçalves Marques, Ana Cristina
relationship. Castro Fernandes, Luís Miguel Costa de Mendonça Galaio, Ana
Finally, it should be noted that, in the context of adher- Isabel B. Correia.
ence to notification through the use of this electronic form, Statistical analysis and interpretation of data: Cristina da
the rate reached of about 45% is quite significant, compared Conceição Ribeiro de Carvalho, João Paulo Garcia Lopes da Cruz,
to previously published studies.15 In a study that reported the Paula Alexandre da Cruz Silva.
first French experience of reporting ADRs directly by indi- Medical validation of reporting data: Ana Isabel B. Correia, Luís
viduals/patients after a mass immunization campaign with Miguel Costa de Mendonça Galaio.
vaccines against pandemic Influenza A (H1N1), out of 4746 Medical revision of the paper: Ema Maria S. Leite Resende.
Pharmaceutical revision of the paper: João Manuel Braz
notifications received during the study period, 1006 (21 .2%)
Gonçalves.
came directly from vaccinated individuals.16
Although some countries have already implemented legis- Declaration of Conflicting Interests
lation that allows direct notification by the patient or target
The author(s) declared no potential conflicts of interest with respect
individual of the drug (case of vaccines), this notification rate to the research, authorship, and/or publication of this article.
is very variable, including countries such as Portugal, Malta,
and Hungary where this service was introduced more recently Funding
and, consequently, the notification rate is still low.15,17 In one The author(s) received no financial support for the research, author-
of these studies, out of 50 countries included, direct patient ship, and/or publication of this article.
reporting systems exist in 44 and represent 9% of total report-
ing, with a total of 27 countries having a patient-specific Ethics Approval Statement
reporting form and 31 providing a form to fill out online in This study was not submitted to the Ethics Committee as it was part
order to improve compliance with the notification.15 of normal pharmacovigilance activities, with all adverse events
reported through spontaneous notifications made by health profes-
Conclusion sionals. This study was developed with the formal knowledge of the
Institutional Board.
In the present study, performed within the scope of the vac-
cination process for healthcare professionals, the profile of
ORCID iD
reported adverse events related to the Comirnaty® vaccine
was similar to its previously carried out clinical development João Paulo Garcia Lopes da Cruz https://orcid.org/0000-0002-
program, except for the higher frequency of adenomegaly 3454-7229
and gastrointestinal disorders (nausea, vomiting, diarrhea,
abdominal pain). Noteworthy in this study, the notification of Supplemental Material
a case of miscarriage. Supplemental material for this article is available online.
Overall, the profile of reported adverse events was mild to
moderate in intensity, being less frequent and milder in older References
adults than in younger adults, and more frequent and severe 1. Polack FP, Thomas SJ, Kitchin N, et al. Safety and efficacy
after the second dose than after the first dose, although injec- of the BNT162b2 mRNA COVID-19 vaccine. N Engl J Med.
tion site reactions were similar at the 2 doses. 2020;383(27):2603-2615. doi:10.1056/NEJMoa2034577
Cruz et al 751
2. Food and Drug Administration (FDA). Fact Sheet for 11. Autoridade Nacional do Medicamento e Produtos de Saúde
Healthcare Providers Administering Vaccine (Vaccination (INFARMED). Relatório de farmacovigilância. Monitorização
Providers). Emergency Use Authorization (EUA) of the Pfizer- da segurança das vacinas contra a COVID-19 em Portugal.
BioNTech COVID-19 Vaccine to Prevent Coronavirus Disease 2021. Accessed July 6, 2022. https://www.infarmed.pt/docu-
2019 (COVID-19). Accessed June 19, 2021. https://www.fda. ments/15786/4268692/Relat%C3%B3rio+de+Farmacovi
gov/media/144413/download gil%C3%A2ncia+-+Monitoriza%C3%A7%C3%A3o+da
3. Comirnaty. Summary of Product Characteristics. Accessed +seguran%C3%A7a+das+vacinas+contra+a+COVID+
June 7, 2021. https://www.ema.europa.eu/en/documents/prod- 19+em+Portugal+atualizado/709e77f5-ab06-092d-338b-
uct-information/comirnaty-epar-product-information_en.pdf 44c99586f796
4. da Direcção-Geral S. Norma no 021/2020 de 23/12/2020 12. Boniol M, McIsaac M, Xu L, Wuliji T, Diallo K, Campbell J.
- Campanha de Vacinação contra a COVID-19: Vacina Gender Equity in the Health Workforce: Analysis of 104 coun-
COMIRNATY. 2021. Accessed July 6, 2022. https://covid19. tries. Working Paper 1. Geneva: World Health Organization.
min-saude.pt/wp-content/uploads/2022/02/Norma_021_2020_ 2019. https://apps.who.int/iris/bitstream/handle/10665/311314/
act_03112021_pdf-1152kb.pdf WHO-HIS-HWF-Gender-WP1-2019.1-eng.pdf
5. European Medicines Agency (EMA). Guideline on Good 13. European Medicines Agency (EMA). Comirnaty - EPAR
Pharmacovigilance Practices (GVP). Accessed July 6, 2022. Public Assessment Report. 2021. Accessed June 7, 2021.
https://www.ema.europa.eu/en/human-regulatory/post-authori- https://www.ema.europa.eu/en/documents/assessment-report/
sation/pharmacovigilance/good-pharmacovigilance-practices comirnaty-epar-public-assessment-report_en.pdf
6. Inácio P, Cavaco A, Airaksinen M. The value of patient report- 14. Male V. Are COVID-19 vaccines safe in pregnancy? Nat
ing to the pharmacovigilance system: a systematic review. Br J Rev Immunol. 2021;21(4):200-201. doi:10.1038/s41577-021-
Clin Pharmacol. 2017;83(2):227-246. doi:10.1111/bcp.13098 00525-y
7. World Health Organization, ed. A Practical Handbook on the 15. Margraff F, Bertram D. Adverse drug reaction report-
Pharmacovigilance of Antiretroviral Medicines. World Health ing by patients: an overview of fifty countries. Drug Saf.
Organization; 2009. 2014;37(6):409-419. doi:10.1007/s40264-014-0162-y
8. Pérez-Ricart A, Gea-Rodríguez E, Roca-Montañana A, Gil- 16. Durrieu G, Palmaro A, Pourcel L, et al. First French expe-
Máñez E, Pérez-Feliu A. Integrating pharmacovigilance into rience of ADR reporting by patients after a mass immu-
the routine of pharmacy department: experience of nine years. nization campaign with Influenza A (H1n1) pandemic
Farm Hosp. 2019;43(4):128-133. doi:10.7399/fh.11169 vaccines: a comparison of reports submitted by patients and
9. Toklu HZ, Mensah E. Why do we need pharmacists in phar- healthcare professionals. Drug Saf. 2012;35(10):845-854.
macovigilance systems? Online J Public Health Inform. doi:10.2165/11631800-000000000-00000
2016;8(2):e193. doi:10.5210/ojphi.v8i2.6802 17. Inch J, Watson MC, Anakwe-Umeh S. Patient versus health-
10. La Gatta A, Ghidinelli F. The future role of pharmacists and care professional spontaneous adverse drug reaction report-
pharmacovigilance after the advent of the COVID-19 pandemic. ing: a systematic review. Drug Saf. 2012;35(10):807-818.
Vacunas. 2021;22(2):127-128. doi:10.1016/j.vacun.2020.12.003 doi:10.2165/11631650-000000000-00000
1113951
research-article2022
HPXXXX10.1177/00185787221113951Hospital PharmacyMunir et al
Original Article
Hospital Pharmacy
Abstract
Purpose: To provide pharmacy leaders and clinicians with the basic principles of a training method called job instruction (JI)
and its application within pharmacy as a strategy to teach routine pharmacy work. Summary: Job instruction is a reliable
way to teach a person to quickly learn how to perform a work standard correctly, safely, and efficiently. Job instruction
is a proven and successful technique for improving time to train, productivity, employee turnover and quality of work. In
healthcare, JI training is not well-known or widely utilized. Job Instruction was the strategy of choice to stabilize the variable
training practices within The University of Kansas Health System pharmacy enterprise, and a prospective, single center, multi-
cohort study was conducted to study its impact. Conclusion: The preliminary results of this study provide initial insight
into the applicability of job instruction in pharmacy. The study indicated the benefits of JI training within different teams to
standardize training practices, decrease error and safety event rates, and increase satisfaction of team members with on-
the-job training. The results of this study will be used to further analyze and optimize training practices in The University of
Kansas Health System pharmacy enterprise.
Keywords
staff development, pharmacy administration, education, pharmacy, teaching, pharmacy service, hospital, quality Improvement
standardization of the job eliminates practice variation, into the pharmacy staff training model was the focus of this
reduces errors, and increases productivity. An example tem- pilot study. The goal of this study is not intended to teach
plate of a job breakdown form is illustrated in Figure 2. readers how to structure and implement JI training. The
In healthcare, job instruction can help bridge the gap intent is to report an example of JI implementation that will
in providing more effective training for new staff while help pharmacy health system leaders and clinicians under-
also continually updating best practices for veteran staff.5 stand how applying JI concepts to routine clinical and opera-
For example, Virginia Mason Medical Center (VMMC), a tional tasks within a pharmacy department can lead to
336-bed hospital, employed JI methodologies to train positive impacts of change such as reduction in error rates
staff on hand hygiene job breakdowns. Baseline audits and increase in training satisfaction of staff.
prior to JI at VMMC showed that only 83.5% of staff
were washing their hands properly and on a consistent Job Instruction Evaluation at The
basis. After piloting JI training in 467 nurses at VMMC,
the reliability of hand hygiene in this pilot population
University of Kansas Health System
went above 98%. The impact of JI training at TUKHS was evaluated as a pro-
The University of Kansas Health System (TUKHS) saw spective, single center, multi- cohort study within the inpa-
an opportunity to incorporate JI in its training processes to tient and outpatient pharmacy departments. The University
address variability, increase efficiency and productivity, and of Kansas hospital in Kansas City where this study took
decrease on the job errors. Integration of JI methodologies place is an academic medical center that staffs over 900
754 Hospital Pharmacy 57(6)
1 Access Restocks 1.1 Normal Restock 1.1 Populates the Restock lists
1.2 Choose Restock 1.2 Various Restock options
Simulation Considerations
Table 1. TUKHS Pharmacy Training Modules Adapted from the Toyota Talent Development Process.8
Abbreviations: TUKHS, The University of Kansas Health System; TWI, Training Within Industry.
Level 1: Reaction The degree to which participants find the training favorable, engaging, and relevant to their jobs
Level 2: Learning The degree to which participants acquire the intended knowledge, skills, attitude, confidence, and
commitment based on their participation in the training
Level 3: Behavior The degree to which participants apply what they learned during training when they are back on the job
Level 4: Results The degree to which target outcomes occur as a result of the training
inpatient beds. The study period was from May 1st, 2020 to engaged with the content and acquired the intended knowl-
June 1st, 2021. There was a formal review by the Institutional edge and skill. In this study, level 1 and 2 results were
Review Board and this study was granted exemption status. assessed independently by the master trainer (and not
Training Within Industry training was launched by inter- included within the scope of this study) and level 3 results
ested pharmacy areas to train current staff on JI methodol- were assessed by a survey questionnaire that was distrib-
ogy; training consisted of a 9-week cohort course with a uted to pharmacy staff prior to JI training and post-JI train-
targeted 1-week JI certification training for pharmacy lead- ing. Level 4 is dedicated to measuring the direct results of
ership and frontline trainers per cohort. Job Instruction fun- the intended training (key performance indicators), which
damentals of job breakdown sheets and the 4-step instruction are established prior to the training being initiated. One
method were utilized during a dedicated JI training week. focus of this pilot study was to track and report pre-speci-
Table 1 showcases the schedule of the 9-week cohort course, fied key performance indicators for each job breakdown to
included the 1-week JI certification training embedded see what impact a standardized approach to training had on
within module 3 of the course. The course was taught by a these job functions.
master trainer, which was a pharmacy department staff posi-
tion specifically created at TUKHS to develop and imple-
Overall Cohort Results
ment TWI-based training and education. Pharmacy staff
members who finished the 9-week cohort course would be Since the initiation of TWI cohorts and JI training within the
deemed certified JI trainers and go on to train pharmacy TUKHS pharmacy enterprise, there have been 8 total cohorts
staff members, in their respective pharmacy departments, on that have completed their the 9-week training modules. A total
various job breakdowns. number of 67 pharmacy team members completed JI certifica-
The primary objective was to evaluate if applying JI tion since initiation of the program through the duration of this
training would impact 3 key areas of interest: the reduction study, which included pharmacists, pharmacy residents, and
of error rates, increase in applied learning skills from train- pharmacy technicians (see Table 3). The fundamental skills of
ing, and increase in job satisfaction of staff. Data was creating job breakdowns are taught to JI trainers during the
obtained from conducted audits, survey responses, auto- cohort process, and cohorts will continue to write, validate,
matic dispensing cabinet stockout reports, and an institu- and train to different job breakdowns as time goes on.
tional safety incident reporting portal. The Kirkpatrick The following 3 charts (Figures 3-5) showcase results from
evaluation model was used to analyze the results of JI train- a survey questionnaire that was distributed to pharmacy staff
ing on pharmacy department staff (see Table 2). This evalu- in the areas that completed JI training. The survey was given
ation methodology is a 4 level evaluation tool that it utilized to participants prior to the implementation of job instruction in
to create, deliver, measure and validate training standards.9 their area and then the same survey was given post-training to
Levels 1, 2, and 3 focus on the impact of learning on the compare differences in response. There were a total of 43 indi-
participants; these levels assess whether participants vidual responses collected in the pre-survey and 43 individual
756 Hospital Pharmacy 57(6)
Table 3. Number of TUKHS Pharmacy Staff Members that are Job Instruction Trained.
Pharmacy Department
responses in the post-training survey. The data show that, with on a particular task. The following outcome results illustrate
newly certified trainers delivering the training, there was an the impact of JI in various areas within TUKHS pharmacy.
increase in not only applied learning skills but also staff satis- Post-JI training level 4 key indicator data was collected at
faction in their training experience post JI implementation. various timepoints (up to 6 months post-JI training). Each JI
cohort got to choose which key performance measures they
wanted to track for each job breakdown they created.
The Kirkpatrick Evaluation Model Examples of level 4 results from the Kirkpatrick evaluation
Outcome Categorization model from our cohorts included:
As a relatively new concept in the healthcare space, job
instruction has shown significant improvements in the way •• Intravenous (IV) room hood set-up audits (IV room
we are training our pharmacy teams and the benefits that technician cohort)
standardized training can provide. There were many posi- || 80% reduction in aseptic defects in setting up a stan-
tive impacts of change seen in reduction in error rates, dardized IV laminar or vertical flow hood 4 months
increase in applied learning skills from training, increase in post-training of IV room staff. Defect was defined
job satisfaction of staff, and the time to train a staff member as missing any step in the IV hood set-up protocol.
Munir et al 757
100
90
80
70
60
50
40
30
20
10
0
Presented the Operation (Important Steps, Key Points, Modeled the Operation 3+ times
Reasons)
|| The fundamental skills of creating job breakdowns Future program direction includes a continued emphasis
are taught to JI trainers during the cohort process, on increased accountability of all levels of management
and cohorts will continue to write, validate, and involved in training. This will be done through increased
train to different job breakdowns as time goes on. stakeholder engagement and understanding of TWI and JI
For example: the IV room technician cohort has training prior to cohort implementation. Cohort specific met-
created 13 job breakdowns that have gone through rics will also continue to be tracked to showcase the impact
the validation phase and are now part of the train- of standardized training practices and increased employee
ing plans of IV room pharmacy technicians. engagement.
Original Article
Hospital Pharmacy
Department journals.sagepub.com/home/hpx
Abstract
Purpose: This study aimed to evaluate the frequency at which postintubation sedation is administered following use of long-
acting paralytic agents compared to short-acting paralytic agents during rapid sequence intubation performed in the emergency
department. Methods: This retrospective, single-center study of intubated patients in the emergency department analyzed
the difference in time to administration of additional sedation following use of a short-acting paralytic (succinylcholine)
compared to use of a long-acting paralytic (rocuronium or vecuronium). A total of 387 patients were available for analysis.
The primary outcome was additional sedation given within 15 minutes following administration of a paralytic agent. The
secondary outcome sought to evaluate the incidence of hyperkalemia due to paralytic agents by comparing potassium level
before and after paralytic administration. Results: 46.9% of patients who received a short-acting paralytic agent received
additional sedation within 15 minutes, compared to 40.9% of patients who received a long-acting paralytic agent. The
Chi-square analysis comparing the short and long-acting paralytic groups showed no statistically significant difference (χ²
[1, N = 387] = 1.24, P = .266) in the frequency of additional sedation administered. Excluding patients who did not receive any
additional sedation, the mean time from paralytic administration to additional sedation in all patients was 20.03 ± 18 minutes.
No statistically significant difference was detected between groups regarding changes in potassium level. Conclusion:
The use of long-acting paralytic agents was not associated with increased time to administration of sedation compared to
shortacting paralytic agents. There is an opportunity to reduce the time to sedation administration for intubated patients
receiving both short- and long-acting paralytic agents.
Keywords
CPR, emergency medicine, analgesics, respiratory, critical care, anesthetics
agent, post-intubation sedation agent, and patient demo- 8 were otherwise ineligible for analysis (Figure 1). Of the
graphics. Data regarding the time of paralytic administration 387 patients available for analysis, 128 patients (33.1%)
and time of sedative medication administration were also received a short-acting paralytic (succinylcholine) and 259
collected, which are recorded and entered into the electronic (66.9%) received a long-acting paralytic (rocuronium = 238,
record by the ED nurse. vecuronium = 21). Patients frequently received etomidate
The time from the administration of paralytic to the as the induction agent during RSI (89.1%). The mean age
administration of an additional sedative was measured in was 63 ± 16 years with a median of 65 years (IQR 54, 76).
minutes. A patient was considered to have received post- Two hundred fifteen (55.6%) participants were male. There
intubation sedation if it was recorded that they received any were no statistically significant baseline differences
of the following within 90 minutes of paralytic administra- between groups (Table 1).
tion: propofol, midazolam, dexmedetomidine, ketamine,
lorazepam, or diazepam. For the purposes of our analysis,
analgesic medications (eg, morphine or fentanyl) were not
Primary Outcome
included as sedation agents. Of the 128 total patients who received a short-acting para-
The primary outcome was successful administration of a lytic, 60 (46.9%) received additional sedation within 15 min-
sedating agent within 15 minutes of administration of the par- utes. Of the 259 total patients who received a long-acting
alytic agent. The secondary outcome was the change in potas- paralytic, 106 (40.9%) received additional sedation within
sium level, which compared the potassium level prior to 15 minutes (Figure 2). The Chi-square analysis comparing
paralytic administration to the level within 2 hours following the short and long-acting paralytic groups showed no statisti-
paralytic administration. Baseline characteristics included cally significant difference (χ² [1, N = 387] = 1.24, P = .266)
age, weight, sex, and race/ethnicity. Laboratory variables in the frequency of additional sedation administered within
included potassium and serum creatinine, which were col- 15 minutes of paralytic.
lected prior to or within 2 hours following intubation. In total, 166 out of 387 patients (42.9%) received addi-
tional sedation within 15 minutes of paralytic administra-
tion. Three hundred four (78.6%) of total patients received
Data Analysis additional sedation within 90 minutes following paralytic
We used simple frequencies to summarize patient character- administration, while 83 patients (21.4%) received no
istics, clinical data, and study outcomes, including the pri- additional sedation. Excluding patients who did not receive
mary outcome of administration of post-intubation sedation any additional sedation, the mean time from paralytic
within 15 minutes of paralytic administration and the second- administration to the administration of additional sedation
ary outcome of change in potassium level. We conducted was 20.03 ± 18 minutes. The median time was 14 minutes
2-sample t-tests and Pearson Chi-Square tests to assess (IQR 8,25).
potential differences in patient characteristics for those who
met the primary outcomes versus those who did not. We also Secondary Outcome
used a Pearson Chi-Square test to assess potential differences
in the primary outcome (yes/no) among patients who Only 25 patients had potassium levels recorded both prior
received long acting versus short acting paralytics. Finally, to paralytic administration and within 2 hours of paralytic
we conducted a 2-sample t-test to compare changes in potas- administration. Six of these patients had received a short-
sium level in the long acting and short acting groups. acting paralytic and 19 had received a long-acting paralytic.
The sample size estimate was based on an estimated 20% In the short-acting group, there was a mean change of
between-group difference in the primary outcome, according −0.117 ± 1.73 in potassium level. In the long-acting group,
to previous literature regarding ED sedation practices.5-8,10 A there was a mean increase in potassium level of
desired sample of 450 patients was estimated, assuming a 0.026 ± 0.9938. This difference was not statistically sig-
power of 80% and an alpha level set at 0.05. Of the 545 par- nificant (t[23] = −0.256, P = .800), but this finding should
ticipants who underwent intubation during the study period, be interpreted with caution given the low sub-sample size.
387 eligible patients were included in this analysis.
Discussion
Results Many patients who are intubated in the ED are in critical
condition and often in pain or distress. Adequate sedation for
Characteristics of Study Subjects intubated patients relieves pain and anxiety.6,11 Patients who
During the study period, a total of 545 intubations were have received paralytic agents may falsely appear comfort-
performed by ED physicians. One hundred fifty-eight cases able while unable to communicate or even gesture. Paralysis
were excluded from analysis: 83 expired while in the ED, without sedation may cause significant distress, tachycardia,
67 did not receive induction and/or paralytic agents, and hypertension, and can ultimately lead to worse patient
762 Hospital Pharmacy 57(6)
outcomes.7,11 Of the patients who did receive post-intubation In this retrospective analysis, there was no difference
sedation, the median time to sedation was 14 minutes. detected in the frequency at which patients received sedation
However, of the total study population, less than half of in the 15 minutes following paralytic administration whether
patients received sedation within 15 minutes of paralytic they had received a short-acting or long-acting agent. Patients
administration. While the limitations of this study make it who received rocuronium or vecuronium for RSI were given
difficult to generate a precise estimate of the frequency of post-intubation sedation at a similar frequency compared to
non-sedation, the comparison of the 2 groups can determine patients who received succinylcholine. This result does not
whether a difference exists when a patient is treated with a support the hypothesis that long-acting paralytic use has
short versus long-acting paralytic. implications on post-intubation sedation. As the overall
Pankey et al 763
Note. Comparisons across groups based on primary outcome utilized 2-sample t-tests for continuous variables and Chi-square tests for categorical variables.
frequency of patients receiving additional sedation within minutes without being sedated. This was not exclusive to
15 minutes was lower than desired, this comparison offers patients who had received long-acting paralytics. When
that there is an opportunity to improve sedation provision in administering the induction agent during RSI, the emergency
all intubated patients, regardless of the type of paralytic used. department staff should have an additional sedative already
Because sedatives commonly used for induction do not have available to be given within 10 minutes after intubation.
a duration of action longer than 15 minutes, protocols that Many barriers can prevent prompt sedation following RSI
encourage prompt administration of sedation following RSI and paralytic use. Pharmacists play a vital role ensuring post-
would be beneficial. intubation sedation, therefore decreasing time to adminis-
Etomidate is the most frequently used induction agent. Its tration. The pharmacist’s knowledge of duration, dosage
brief duration of action is usually considered a positive char- calculations, and ability to procure drugs can help the team to
acteristic, but it may contribute to patients being paralyzed ensure timely administration. Beyond logistical factors
and not adequately sedated.11 In this analysis, the mean time involving staffing and documentation, patient-specific fac-
to additional sedation following paralytic administration was tors such as hypotension or altered mental status may
approximately 20 minutes. As this is longer than the duration decrease the likelihood of receiving post-intubation sedation.
of action of etomidate, even patients who did receive addi- In patients with pre- and post-intubation hypotension, physi-
tional sedation could have been paralyzed for several cians may be hesitant to prescribe sedation due to fear of
764 Hospital Pharmacy 57(6)
Figure 2. Time to additional sedation in patients who received a short-acting paralytic (left) and long-acting paralytic (right).
Percentages show the proportion of patients who received sedation within 15 minutes following paralytic administration, more than
15 minutes following paralytic administration, and patients who received no additional sedation.
Note. Patients who received sedation at an interval greater than 90 minutes following paralytic administration were considered to have not received any
additional sedation.
worsening hypotension.5,12 Physicians may also decide to restrict its use in select patients at risk for hyperkalemia.4
withhold additional sedation in patients with poor neurologic Those most at risk for hyperkalemia are trauma patients with
status due to limited patient response. In patients with altered crush and burn injuries, and this population was not preva-
mental status, it is possible that sedation is delayed due to the lent in our study. In this study, the data set showed a mean
caretaker’s belief that the patient is unable to perceive pain. decrease in potassium level in the group that had received
It is important to recognize that altered mental status does succinylcholine. However, there were only 6 patients who
not eliminate the need for sedation following intubation.5,7,12 had received succinylcholine available for analysis. In the
Adequately assessing which patients are incapable of feeling long-acting paralytic group, there was a mean increase in
pain is very difficult in the emergency setting, especially in potassium level. The difference between groups was not sta-
patients given a paralytic. There is insufficient evidence on tistically significant due to the small sub-sample size. More
the need for sedation in this subpopulation. Because it is research is needed to better understand the reasons for choos-
unclear how altered a person must be to not perceive pain or ing long-acting paralytics over short-acting paralytics,
distress, erring on the side of more sedation is important to including both patient-specific factors and physician prefer-
reduce discomfort of the patient.7 Lastly, the induction agent ence. To better quantify the impact of paralytic type on labo-
of choice may influence the time to sedation. With mid- ratory values, future studies are needed to assess these values
azolam being used in 1.6% of patients who received a short- both prior to and following intubation in a larger sample of
acting paralytic and 5.0% of patients who received a patients.
long-acting paralytic, it is possible that this expanded the Many aspects of our study increase its application to prac-
time needed to give appropriate sedation due to its longer tice in the community. Our study had less stringent exclusion
duration of action. However, to best ensure sedation for all criteria compared to previous studies, and patients with car-
patients, a standard protocol for all induction agents is likely diac or traumatic arrest were included in our analysis.
a better approach to ensure timely sedation. Additionally, because our study was conducted at a commu-
The secondary outcome of this analysis examined the nity hospital, the results are likely a better reflection of all
change in potassium level, measured before and after para- sedation practices than results of studies conducted utilizing
lytic administration in a small sub-sample of patients. Most data from the NEAR database, which represents a set of EDs
patients did not have both pre- and post-paralytic potassium that are engaged in airway research. Further, our endpoint of
levels available, limiting the utility of this analysis. 15 minutes was informed by the duration of action of etomi-
Succinylcholine is known to increase potassium, which may date, which is by far most frequently used in practice.
Pankey et al 765
blocker. Pediatr Emerg Care. 2009;25(6):393-396. doi:10.1097/ use in the emergency department. Emerg Med J. 2013;30:893-
PEC.0b013e3181a7923b 895. doi:10.1136/emermed-2012-201812
9. Bonomo JB, Butler AS, Lindsell CJ, Venkat A. Inadequate 11. Oglesby AJ. Should etomidate be the induction agent of choice
provision of postintubation anxiolysis and analgesia in the for rapid sequence intubation in the emergency department?
ED. Am J Emerg Med. 2008;26(4):469-472. doi:10.1016/j. Emerg Med J. 2004;21:655-659. doi:10.1136/emj.2003.009043
ajem.2007.05.024 12. Groth CM, Acquisto NM, Khadem T. Current practices and
10. Watt JM, Amini A, Traylor BR, Amini R, Sakles JC, Patanwala safety of medication use during rapid sequence intubation. J
AE. Effect of paralytic type on time to post-intubation sedative Crit Care. 2018;45:65-70. doi:10.1016/j.jcrc.2018.01.017
1095767
case-report2022
HPXXXX10.1177/00185787221095767Hospital PharmacyVanDuyn et al
Case Report
Hospital Pharmacy
Abstract
Introduction: Carbapenem-resistant Enterobacteriaceae (CRE) are problematic pathogens because infections caused by
these organisms are associated with significant morbidity and mortality. These organisms often harbor multiple resistance
mechanisms, which makes it difficult to treat their associated infections. Treatment typically consists of intravenous
antibiotics that are selected based on the specific susceptibility pattern for the pathogen. Data on the use of oral antibiotics
for the treatment of infections caused by CRE are sparse. Case Presentation: In this case, a 62-year-old female presented
with a chronic left leg wound infection. She previously underwent surgical debridement and skin grafting, which were
unsuccessful. She was initially prescribed minocycline for the infection, but the wound got re-infected. At this time, the
wound had significant surrounding erythema, drainage, and slough. A wound culture was obtained and demonstrated growth
of carbapenem-resistant Enterobacter cloacae and methicillin-resistant Staphylococcus aureus. The patient was initiated on oral
omadacycline, and she responded with resolution of the cellulitis and wound drainage. Conclusion: This case demonstrates
that omadacycline may be beneficial as an oral medication for the treatment of complicated acute bacterial skin and skin
structure infections caused by carbapenem-resistant Enterobacter cloacae.
Keywords
anti-infectives, infectious diseases, outcomes research, pharmacists, education, skin
Discussion
To our knowledge, this is the first case report to demonstrate
successful use of oral omadacycline for the treatment of car-
bapenem-resistant Enterobacter cloacae in a complicated
Figure 1. Infected left leg wound prior to omadacycline
treatment. This demonstrates surrounding erythema, drainage, ABSSSI. Traditionally, infections caused by CRE require
and slough on the base of the ulcer. treatment with intravenous antibiotics.4 Although new anti-
biotics, such as ceftazidime-avibactam and plazomicin, have
been developed to treat carbapenemase-producing organ-
Enterobacter cloacae in a complicated acute bacterial skin isms, intravenous use of these medications may cause
and skin structure infection (ABSSSI). adverse reactions, administration difficulties, and increased
patient cost. Treatment with oral medications is often limited
by resistance genes encoded on plasmids.3 As shown in this
Case Presentation
case, this carbapenem-resistant Enterobacter cloacae was
Written informed consent was obtained from the patient for resistant to all of the oral antibiotics that are traditionally
publication of this case report and the accompanying images. used, so omadacycline was used for treatment.
The patient was a 62-year-old Caucasian female with a past Omadacycline received Qualified Infectious Disease
medical history of obesity, deep vein thrombosis of the left Product and Fast Track designations because of the need for
lower extremity on anticoagulation, venous insufficiency broad-spectrum antimicrobial medications with the ability to
involving both lower extremities, and necrotizing fasciitis of combat multidrug-resistant organisms. This medication
the left lower extremity status post debridement in April 2019. received US Food and Drug Administration (FDA) approval
After the initial debridement, her wound failed to heal appro- in October 2018 for the treatment of community-acquired
priately. Subsequently, the patient underwent surgical debride- bacterial pneumonia (CABP) and ABSSSIs, and it is avail-
ment and skin grafting on 3 different occasions. Unfortunately, able in intravenous and oral formulations.5 For ABSSSI and
the skin grafts failed, and the wound never healed. CABP, the US FDA identified the susceptibility breakpoint
The patient was seen for her chronic left leg wound infec- for Enterobacteriaceae as a minimum inhibitory concentra-
tion in February 2020 at the Wound Care Center. A wound tion (MIC) of ≤4 mg/L.6
culture was obtained and demonstrated growth of moderate Omadacycline is a semisynthetic aminomethylcycline
carbapenem-resistant Enterobacter cloacae and rare methi- derived from minocycline.7 Its novel modification allows
cillin-susceptible Staphylococcus aureus. She was initially it to overcome traditional tetracycline-based resistance
prescribed minocycline 200 mg by mouth twice daily for mechanisms and provides broad-spectrum coverage against
10 days for the infection and ondansetron as needed for nau- aerobic and anaerobic gram positive and gram negative
sea. She responded to minocycline with resolution of the cel- organisms, as well as atypical pathogens. It has also dem-
lulitis, but she had significant nausea despite the use of onstrated activity against multidrug-resistant organisms,
ondansetron. Once her infection improved, she was started such as CRE.
on local wound care and compression therapy. The patient’s The approval of omadacycline for ABSSSI was based on
wound had slow improvement, and the left leg wound even- a phase III non-inferiority trial, which compared it to line-
tually got re-infected in May 2020. As shown in Figure 1, the zolid.8 Patients in this study were initiated on the intrave-
wound had significant surrounding erythema, drainage, and nous formulation and allowed to transition to the oral
slough on the base of the ulcer. formulation after 3 days. Although this study showed non-
Another wound culture was obtained and demonstrated inferiority to linezolid for ABSSSIs, only 15.4% of patients
growth of moderate carbapenem-resistant Enterobacter clo- had a polymicrobial mixed gram positive and gram negative
acae and methicillin-resistant Staphylococcus aureus. As infection. In addition, patients with an infection caused
shown in Figure 2, this carbapenem-resistant Enterobacter solely by gram negative pathogens were excluded because
cloacae was resistant to all traditional oral antibiotic options. linezolid was used as the comparator.
VanDuyn et al 769
Figure 2. Left leg wound Enterobacter cloacae susceptibilities prior to omadacycline treatment. This demonstrates growth of
carbapenem-resistant Enterobacter cloacae.
Conclusion
A surveillance study assessed the activity of omadacycline
among clinical isolates from the US and Europe from 2016 to Omadacycline may be beneficial as an oral medication for
2018.9 Omadacycline was tested against 49,000 bacterial iso- the treatment of complicated ABSSSIs caused by carbape-
lates from 73 different medical centers. In this study, carbape- nem-resistant Enterobacter cloacae. Additional data are
nem-resistant Escherichia coli had an MIC50 and MIC90 of 1 needed to investigate and confirm our findings.
and 2 mg/L, respectively. However, carbapenem-resistant
Klebsiella pneumoniae had an MIC50 of 4 mg/L and an Authors’ Note
MIC90 of 8 mg/L. For species in the Enterobacter cloacae The data represented in this manuscript were presented as a
complex, omadacycline had an MIC50 and MIC90 of 2 and poster at the virtual ASHP Midyear Clinical Meeting 2021 on
4 mg/L, respectively. Against ceftazidime-non-susceptible December 8, 2021.
770 Hospital Pharmacy 57(6)
Case Report
Hospital Pharmacy
Abstract
Purpose: To review a case of angioedema associated with mirabegron. Summary: A 77-year-old woman presented to
the emergency department with swelling of the left anterior side of the tongue beginning 90 minutes prior to arrival. She
noticed tingling on her tongue while eating a hamburger, chips, and ice cream. The patient had no history of medication-
related allergies. Home medications include acetaminophen, aspirin, biotin, black cohosh, Co-enzyme Q-10, cranberry, fish
oil, multivitamin, alendronate, and mirabegron all taken orally. The patient reports being on mirabegron for 3 to 4 years but
recently decided to self-taper off the medication. Her initial laboratory test results were unremarkable. The patient was
diagnosed with left tongue angioedema, and she received methylprednisolone, epinephrine, famotidine, and 2 units of plasma.
After medication administration, the patient underwent a flexible nasopharyngeal laryngoscopy resulting in no significant
angioedema in the pharynx and hypopharynx with a small area of possible edema noted on the right base of tongue/lingual
tonsil. Patient’s symptoms improved with medication treatment and discontinuation of mirabegron. The use of Naranjo et
al. adverse-event probability scale revealed that mirabegron was the possible (score of 4) cause of the patient’s left tongue
angioedema. Conclusion: A 77-year-old woman developed drug induced tongue angioedema associated with mirabegron
after being compliant on this medication for 3 to 4 years. Patient was self-tapering her dose of mirabegron. This patient
responded well to medication treatment and discontinuation of drug. To our knowledge, this is the second case report of
mirabegron induced angioedema.
Keywords
mirabegron, angioedema, tongue
typically present in this type of angioedema. This reaction has Friday, mirabegron 50 mg twice weekly on Monday and
a gradual onset, unlike allergic angioedema. Symptoms will Friday, all taken orally. She also used polyvinyl alcohol 1.4%
subside after discontinuing the offending agent.4,5 Mirabegron, liquiform tear eye drops, 1 drop into both eyes as needed. Of
a beta-3-agonist, is indicated for overactive bladder and uri- these medications, acetaminophen,8 alendronate,9 aspirin10
nary incontinence.6 We describe a case of drug-induced have been rarely associated with angioedema. This case
tongue angioedema related to use of mirabegron. A MEDLINE report will focus on mirabegron as possible cause. The
search, without date limits, using the terms mirabegron and patient reports taking mirabegron for about 3 or 4 years and
angioedema or drug-induced angioedema identified 1 case up until recently she has been taking it as prescribed, once
report.7 daily. She has recently decided to self-taper off this medica-
tion and started to take it twice weekly. Due to the rapid onset
of her angioedema and her positive response to methylpred-
Case Report nisolone, famotidine, and epinephrine the patient was advised
A 77-year-old female with a body mass index of 19.01 kg/m2 to discontinue mirabegron on discharge. The patient was
arrived at the emergency department with complaints of left- scheduled to follow up with her primary care provider in
sided tongue swelling and tingling, tightness in the back of 1 week post discharge to discuss whether to continue mirabe-
her throat and difficulty speaking because of a lisp. Her gron considering this occurrence. It was decided to discon-
symptoms began after consuming a cheeseburger, potato tinue mirabegron. The Naranjo et al adverse-event probability
chips, and ice cream. Due to symptom progression, she pre- scale indicated a score of 4 which indicates possible causal-
sented to the emergency room 1.5 hours later. She has no pre- ity of adverse drug reaction (ADR).11
vious history of allergies, angioedema, or anaphylaxis. Her
past medical history was significant for osteoporosis and
overactive bladder, both of which are not chronic degenera-
Discussion
tive diseases. Vitals on admission include blood pressure: Mirabegron is indicated for overactive bladder and urinary
182/95 mmHg, pulse: 98 BPM, respiratory rate: 18 breaths/ incontinence. It works by activating beta-3 adrenergic recep-
minute, SpO2: 98% on room air. Body temperature was tors in the bladder resulting in relaxation of the detrusor
36.3°C. Given the patient was hemodynamically stable ana- smooth muscle during the urine storage phase and increasing
phylaxis was ruled out. On physical examination, patient’s bladder capacity. Data have shown that beta-adrenoceptors,
anterior tongue was significantly swollen concerning for predominately the beta-3, mediate detrusor smooth muscle
angioedema versus allergic reaction. The remainder of her tone, and promote the storage function of the human bladder.
physical exam was unremarkable. Complete blood count and Mirabegron is generally well tolerated with the most com-
differential, osmolality, and international normalized ratio mon side effects being increased blood pressure, dry mouth,
were all unremarkable. The basic metabolic panel was unre- urinary tract infection, dizziness, headache, and constipation.
markable except for slightly elevated calcium at 10.8 mg/dL Serious, but rare, ADRs include hypersensitivity reactions
and bicarbonate 31 mmol/L. COVID-19 and respiratory and head and neck angioedema (occurring in <1% of
pathogen were both unremarkable. Electrocardiogram patients).6 The initial mirabegron package insert in 6/201212
showed normal sinus rhythm without ST elevation or other did not have angioedema listed as a possible ADR. Due to
obvious signs of ischemia or arrhythmia. Patient received post marking information reported, the package insert was
intravenous methylprednisolone 125 mg, intravenous famoti- updated in 8/201513 to list angioedema as a rare but serious
dine 20 mg, and intramuscular epinephrine 0.3 mg shortly ADR. To date there is only 1 case report of mirabegron asso-
after admission to the emergency department. Patient also ciated angioedema. This case study examines a 53-year-old
received 2 units of plasma 2 minutes after receiving the medi- female who experienced dyspnea, dysphagia, and throat
cations listed. After medications and plasma were adminis- tightness for 3 to 4 weeks. Through careful history taking the
tered otolaryngology preformed a bedside flexible team was able to reveal that she had increased her dose of
pharyngo-laryngoscopy procedure with the following results: mirabegron from 25 to 50 mg 1 day prior to symptoms onset.
no significant angioedema in pharynx/hypopharynx, small Three days after discontinuing mirabegron she reported
area possible edema noted on the right base of the tongue/ symptom resolution.7 Similarly, we report a patient who
lingual tonsil. Patient’s symptoms began to improve about recently changed her mirabegron dose who experienced
90 minutes after medications were administered and the tongue angioedema. She was taking mirabegron 50 mg daily
patient was discharged, with a total length of stay being for 3 to 4 years and recently cut back to 50 mg on Monday
1 day, without any changes to her home medications. Her and Fridays. She presented Friday evening and had received
home medications included the following: acetaminophen her dose for the day. Based on the rapid onset of her tongue
500 mg twice daily, aspirin 81 mg daily, biotin 1000 μg daily, swelling and response to treatment we propose the patient
black cohosh compound 1 tablet daily, Co-enzyme Q-10 experienced drug-induced allergic angioedema. We propose
1000 mg daily, cranberry 300 mg daily, fish oil 1000 mg the mechanism causing her angioedema was a type I hyper-
daily, multivitamin 1 tablet daily, alendronate 70 mg every sensitivity mediated by mast cell degranulation and a rapid
Zuchowski et al 773
release of histamine. This is supported by the rapid onset and 2. Kaplan AP. Angioedema. World Allergy Organ J. 2008;1(6):
response to antihistamines and corticosteroids. 103-113.
3. Moreau ME, Garbacki N, Molinaro G, Brown NJ, Marceau F,
Adam A. The kallikrein-kinin system: current and future phar-
Conclusion macological targets. J Pharmacol Sci. 2005;99(1):6-38.
4. Agostoni A, Cicardi M. Drug-induced angioedema without
A 77-year-old female developed tongue angioedema after self-
urticaria. Drug Saf. 2;24(8):599-606.
tapering her mirabegron. Drug-induced allergic angioedema is 5. Kalambay J, Ghazanfar H, Martes Pena KA, Munshi RA,
a rare but is a medical emergency. This case report shows the Zhang G, Patel JY. Pathogenesis of drug induced non-aller-
importance of obtaining an accurate medication history. gic angioedema: a review of unusual etiologies. Cureus.
2017;9(8):e1598.
Author Contributions 6. Myrbetriq (Mirabegron) [Prescribing Information]. Astellas
Pharma US Inc;2021.
Matthew T. Zuchowski, Jaylan M. Yuksel, and John Novi contrib-
7. Davis H, Wallace T, Gilkerson C, Saunders E. Drug-induced
uted to drafting the article, critically revising the article, and final
angioedema: a rare side effect of mirabegron. W V Med J. 1-3.
approval of the article.
8. Deepthi A, Shaheen, Kumar H, Ashraf S, Deepak JH. Images
in medicine: an atypical presentation of unilateral tongue
Declaration of Conflicting Interests angioedema caused by acetaminophen. J Clin Diagn Res.
The author(s) declared no potential conflicts of interest with respect 2017;11(4):ZJ01-ZJ02.
to the research, authorship, and/or publication of this article. 9. Musette P, Kaufman JM, Rizzoli R, Cacoub P, Brandi ML,
Reginster JY. Cutaneous side effects of antiosteoporosis treat-
Funding ments. Ther Adv Musculoskelet Dis. 2011;3(1):31-41.
10. Marshall LL. Angioedema associated with aspirin and rofe-
The author(s) received no financial support for the research, author- coxib. Ann Pharmacother. 2005;39(5):944-948.
ship, and/or publication of this article. 11. Naranjo CA, Busto U, Sellers EM, et al. A method for estimat-
ing the probability of adverse drug reactions. Clin Pharmacol
ORCID iD Ther. 1981;30(2):239-245.
Matthew T. Zuchowski https://orcid.org/0000-0002-2598-016X 12. Approval Label, Myrbetriq. Orig-1 Drugs@FDA. 2012.
Accessed December 14,2021. https://www.accessdata.fda.gov/
drugsatfda_docs/label/2012/202611s000lbl.pdf
References 13.
Labeling-Package Insert. Suppl-6 Drugs@FDA. 2015.
1. Lewis LM. Angioedema: etiology, pathophysiology, current Accessed December 14,2021. https://www.accessdata.fda.gov/
and emerging therapies. J Emerg Med. 2013;45(5):789-796. drugsatfda_docs/label/2015/202611s006lbl.pdf
1108715
research-article2022
HPXXXX10.1177/00185787221108715Hospital PharmacyRoss et al
Case Report
Hospital Pharmacy
With Dabigatran
Abstract
Introduction and Objective: Coronavirus disease 2019 (COVID-19) is associated with respiratory failure and a
hypercoagulable state. Studies have shown the use of oral anticoagulants, specifically dabigatran, can significantly decrease
mortality from COVID-19. Dabigatran is an oral direct thrombin inhibitor commonly used for nonvalvular atrial fibrillation
and for the treatment or prevention of venous thromboembolism. The association of COVID-19-related extensive
thrombosis while receiving full therapeutic anticoagulation with dabigatran has not been well-established in current literature.
Case Report: We present a 73-year-old male patient with a history of persistent atrial fibrillation anticoagulated with
dabigatran presenting with an active COVID-19 infection admitted to the intensive care unit. On hospital day 7, he developed
extensive arterial and venous thromboembolisms. To our knowledge, this is the first published case of COVID-19-related
extensive thrombosis while receiving full therapeutic anticoagulation with dabigatran. Discussion: Guidelines recommend
prophylactic or therapeutic-dose anticoagulation with unfractionated heparin or low-molecular weight heparin for all patients
if no contraindications exist; however, recommendations for the use of therapeutic oral anticoagulants have not been well
established. Further studies are warranted to establish appropriate use of oral anticoagulants in the setting of COVID-19.
Conclusion: Evidence from this report suggests clinicians should closely monitor patients at risk for hypercoagulability
regardless of the anticoagulation therapy the patient may be receiving. Additionally, evidence from this case suggests a
possible inferiority in the anticoagulation ability of dabigatran in patients with active COVID-19.
Keywords
critical care, anticoagulants, COVID, pharmacokinetics
knowledge, is the first case report of COVID-19-related limb ischemia. The CTA showed moderate volume pulmo-
extensive thrombosis while receiving full therapeutic antico- nary emboli, extensive arterial thrombus extending from the
agulation with dabigatran. distal aorta through the majority of the right iliac arteries,
and branch artery occlusions of large portions of the left kid-
ney. Vascular surgery recommended transitioning anticoagu-
Case Report
lation to an IV unfractionated heparin (UFH) infusion given
The patient was a 73-year-old white male with a past medical the development of thrombus while anticoagulated on thera-
history of Guillain-Barre syndrome (diagnosed 17 months peutic dabigatran. Due to a high risk of repeat thrombosis,
prior to admission), hyperlipidemia, persistent NVAF (diag- surgical intervention was deferred. Since admission, he had
nosed 16 months prior to admission), and obesity (body mass not missed any doses of dabigatran.
index [BMI] on admission was 33.3 kg/m2 [115 kg]). The patient was transitioned to an IV UFH infusion
Additionally, he had an extensive surgical history (all of (titrated to a goal anti-factor Xa level of 0.3-0.7 International
which occurred greater than 1 year prior to admission) units/mL) from hospital day 7 until hospital day 12, at which
including a left tibia fracture repair, right knee replacement, point he was converted to therapeutic enoxaparin (dosed as
right rotator cuff repair, and left rotator cuff repair. He 1 mg/kg subcutaneously [SC] twice daily). On hospital day
reported no significant family history, had a previous smok- 16, he started warfarin and continued on therapeutic enoxa-
ing history (quit 10 years prior to presentation), and reported parin until his International Normalized Ratio (INR) was
no use of alcohol or illicit drugs. He reported good medica- within his goal range of 2.0 to 3.0. Due to the need for inva-
tion adherence to his home medications, which included sive mechanical ventilation and requirement for amiodarone
dabigatran 150 mg by mouth twice daily, and metoprolol suc- for refractory atrial fibrillation, warfarin was discontinued,
cinate 100 mg by mouth every morning and 50 mg by mouth and he was placed back on a therapeutic enoxaparin on hos-
every evening. He also reported taking a non-prescription pital day 27. He was later transitioned back to an IV UFH
fish oil supplement daily by mouth, dose unspecified. He infusion on day 30 after developing an acute kidney injury.
reported no known medication allergies or adverse drug On hospital day 32, after days of developing progressive mul-
reactions. tiorgan failure, life-sustaining treatment was withdrawn, and
The patient presented to the emergency department after the patient died.
approximately 2 weeks of rhinorrhea, congestion, and non-
productive cough. He complained of worsening fatigue and
cough in the days leading to admission. He was found to be
Discussion
COVID-19 positive by polymerase chain reaction (PCR) Reports of VTE and arterial thromboembolism (ATE) have
testing. Of note, he had not received the COVID-19 vaccina- been common since the start of the COVID-19 pandemic,
tion due to his history of Guillain-Barre per advice from his which have led to recommendations regarding the use of
primary care physician. In the emergency department, he anticoagulants for prophylaxis in this hypercoagulable
was started on non-invasive positive pressure mechanical state.2,3,5 In the beginning of the pandemic, recommendations
ventilation for hypoxia and an intravenous (IV) diltiazem using prophylactic-dose low-molecular-weight heparin
infusion for his atrial fibrillation with rapid ventricular rate (LMWH) or UFH were not sufficient and led to breakthrough
(RVR). He was admitted to the medical intensive care unit thrombosis.5 This prompted further studies to compare the
(MICU) for acute hypoxic respiratory failure secondary to use of therapeutic-dose versus intermediate dose versus pro-
COVID-19 pneumonia. He was continued on metoprolol for phylactic dose anticoagulation.5,6 Current anticoagulation
rate control and dabigatran to prevent stroke and systemic recommendations for patients with COVID-19 receiving
embolism. His IV diltiazem infusion was converted to diltia- intensive care unit level of care includes prophylactic-dose
zem extended-release oral formulation and the dose was UFH or LMWH for all patients as long as no contraindica-
adjusted as indicated for rate control. He was initiated on tions exist, according to the Society of Critical Care
remdesivir daily for 5 days (1 dose of 200 mg IV, followed by Medicine, Surviving Sepsis Campaign, and the National
4 doses of 100 mg IV) and dexamethasone daily for 10 days Institutes of Health (NIH).6,7
(1 dose of 6 mg IV, followed by 9 doses of 6 mg orally) for Additionally, the NIH recommends full-strength LMWH/
the management of COVID-19. UFH for non-critically ill patients over prophylactic-dose
Pertinent laboratory tests can be seen in Table 1 for the and is preferred over the use of DOACs.6 This recommenda-
first 8 days of the hospital stay. On hospital day 7, he reported tion is based off multiple studies showing a significant ben-
acute pain and paresthesia of the right leg from mid-tibia and efit using therapeutic dosing in patients with COVID-19. In
distally. Upon exam, his right foot was cold and pale com- these studies, therapeutic dosing could have included any
pared to his left side. Additionally, a pedal pulse could not be LMWH at the manufacturer’s recommended therapeutic
palpated or located with bedside doppler ultrasound. Vascular dosing or UFH titrated to therapeutic effect based on partial
surgery was consulted, and a stat computed tomography thromboplastin time (PTT) or anti-factor Xa monitoring per
angiography (CTA) was ordered due to concern for acute institutional protocol. The preferred agent was enoxaparin at
776 Hospital Pharmacy 57(6)
Lab test [reference range] Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 8
AST (U/L) [0-45] 35 44 32 33
ALT (U/L) [0-40] 30 58 47 46
CRP (mg/dL) [<0.3-0.5] 14
ANC (k/cmm) [1.8-7.8] 7.43 4.86 15.91
PTT (sec) [23.0-37.7] 16.4 >200 67.0
PT (sec) [12.2-14.8] 39.3
INR [0.9-1.2] 1.3
WBC (k/cmm) [4.6-10.8] 8.35 5.28 12.19 11.13 13.0 17.29 14.39 15.08
Hgb (g/dL) [13.9-18] 18.4 16.3 16.1 14.9 16.1 17.4 15.3 15.2
HCT (%) [41-52] 52.3 46.7 46.1 42.5 45.7 49.6 42.9 42.8
Plt (k/cmm) [130-440] 164 141 198 184 210 205 160 165
Glucose (mg/dL) [65-99] 119 148 142 146 132 117 119 172
BUN (mg/dL) [9-20] 27 24 24 24 24 22 23 24
Creatinine (mg/dL) [0.5-1.2] 1.1 0.8 0.7 0.7 0.9 1.0 0.8 0.9
Sodium (mmol/L) [135-145] 138 136 138 140 139 139 135 134
Chloride (mmol/L) [23-32] 102 104 105 107 105 102 101 99
CO2 (mmol/L) [23-32] 19 21 20 20 21 22 24 22
Potassium (mmol/L) [3.5-5.0] 4.0 4.7 3.9 3.9 4.0 4.0 4.3 4.4
Note. ALT = alanine aminotransferase; ANC = absolute neutrophil count; AST = aspartate aminotransferase; BUN = blood urea nitrogen; CO2 = serum
bicarbonate; CRP = C-reactive protein; HCT = hematocrit; Hgb = hemoglobin; INR = international normalized ratio; Plt = Platelets; PT = prothrombin time;
PTT = partial thromboplastin time; WBC = white blood cells.
a dose of 1 mg/kg SC twice daily. Further, typical prophylac- ulcer prophylaxis was not indicated.11 Another major con-
tic dosing was enoxaparin 40 mg SC daily or heparin cern with dabigatran is the storage of the medication.
5000 units SC 2 to 3 times daily based on institutional proto- Dabigatran must be stored in its original packaging to mini-
col. Regarding safety, a non-significant increase in non-fatal mize product breakdown from moisture and avoid a loss of
bleeding (defined based on International Society on potency.10 While our patient’s home storage conditions of
Thrombosis and Haemostasis [ISTH] criteria as fatal bleed- dabigatran are unknown, our facility uses manufacturer blis-
ing, symptomatic bleeding in a critical area or organ, or ter packs of dabigatran while inpatient to control for these
bleeding causing a drop in hemoglobin of at least 20 g/L) was storage conditions. The last major concern with the use of
observed for the therapeutic dosing group.8,9 DOACs involves their efficacy in the obese patient popula-
One concern with using DOACs compared to LMWH or tion, especially with our patient having a BMI of 33.3 kg/m2.
UFH in general is the increased drug interactions present; Dabigatran has been reported to have significantly reduced
however, most of these interactions lead to increased sys- plasma concentrations by about 20% in patients weighing
temic concentrations. Decreased systemic concentrations of more than 100 kg compared to patients weighing 50 to
dabigatran can occur with the concomitant use of medica- 100 kg10,12 While there is limited data on therapeutic failure
tions that increase the gastric pH or induce the P-glycoprotein of dabigatran in patients with a BMI 30 to 40 kg/m2, the
efflux pump.10 Medications that increase the gastric pH (eg, available case reports of patients with a BMI greater than
magnesium or aluminum-containing antacids) decrease the 40 kg/m2 supports the idea of increased body weight signifi-
absorption of dabigatran, whereas medications that induce cantly contributing to the diminished effect of dabigatran.
the P-glycoprotein efflux pump (eg, carbamazepine, phe- The definitive reason for the lack of therapeutic efficacy of
nytoin) may reduce the bioavailability of dabigatran by dabigatran in our patient case remains unclear, although the
increasing drug efflux back into the intestinal lumen for elevated BMI could have played a significant role.
excretion.10 From hospital admission to the development of While the findings in this case are noteworthy, this case
extensive thrombus, our patient did not receive any concom- report also has some important limitations. Most importantly,
itant medications that could have explained any change in several coagulation/inflammatory tests (D-dimer, C-reactive
the systemic concentration of dabigatran. Of note, the protein (CRP), ferritin, and fibrinogen) were not collected at
patient did not receive any stress ulcer prophylaxis. Our baseline and have limited data during his admission; how-
patient did not have risk factors present to indicate an ever, these inflammatory markers can also be elevated in the
increased risk of clinically important gastrointestinal bleed- absence of an active thrombosis and would not definitively
ing (eg, at least 48 hours of mechanical ventilation, platelets indicate a thrombus. Additionally, the PTT, prothrombin
<50 k/mm3, INR > 1.5, PTT > 2 times baseline), so stress time (PT), and INR were not continuously collected during
Ross et al 777
6. COVID-19 Treatment Guidelines Panel. Information on 12. Sebaaly J, Kelley D. Direct oral anticoagulants in obesity: an
COVID-19 treatment, prevention and research. National updated literature review. Ann Pharmacother. 2020;54:1144-
Institutes of Health. Updated: May 31, 2022. Accessed January 1158. doi:10.1177/1060028020923584
24, 2022. https://www.covid19treatmentguidelines.nih.gov/ 13. Thomas AV, Lin KP, Stillson JE, et al. A case series of
7. Surviving Sepsis Campaign Panel. COVID-19 guidelines. thromboelastography-guided anticoagulation in COVID-19
Society of Critical Care Medicine. 2020. Accessed January 24, patients with inherited and acquired hypercoagulable states.
2022. https://sccm.org/SurvivingSepsisCampaign/Guidelines/ Case Rep Med. 2021;2021:5568982. doi:10.1155/2021/556
COVID-19 8982
8. Spyropoulos AC, Goldin M, Giannis D, et al. Efficacy and 14. Hartmann J, Ergang A, Mason D, Dias JD. The role of TEG
safety of therapeutic-dose heparin vs standard prophylactic or analysis in patients with COVID-19-associated coagulopathy:
intermediate-dose heparins for thromboprophylaxis in high- a systematic review. Diagnostics. 2021;11:172. doi:10.3390/
risk hospitalized patients with COVID-19: the HEP-COVID diagnostics11020172
randomized clinical trial. JAMA Intern Med. 2021;181:1612- 15. Lopez P, Velez R, Rivera V, Rodriguez N, Yamamura Y.
1620. doi:10.1001/jamainternmed.2021.6203 Characteristics of p-glycoprotein (pgp) upregulated in chronic
9. The ATTACC, ACTIV-4a, and REMAP-CAP Investigators. cocaine users and HIV infected persons. Retrovirology.
Therapeutic anticoagulation with heparin in noncritically ill 2005;2:142.
patients with COVID-19. N Engl J Med. 2021;385:790-802. 16. Illanes-álvarez F, Márquez-Ruiz D, Márquez-Coello M,
doi:10.1056/NEJMoa2105911 Cuesta-Sancho S, Girón-González JA. Similarities and dif-
10. Pradaxa (dabigatran etexilate mesylate) [package insert].
ferences between HIV and SARS-CoV-2. Int J Med Sci.
Boehringer Ingelheim Pharmaceuticals, Inc.; 2011. 2021;18(3):846-851. doi:10.7150/ijms.50133
11. Barletta JF, Bruno JJ, Buckley MS, Cook DJ. Stress ulcer pro- 17. Bibas M, Biava G, Antinori A. HIV-associated venous thrombo-
phylaxis. Crit Care Med. 2016;44(7):1395-1405. doi:10.1097/ embolism. Mediterr J Hematol Infect Dis. 2011;3(1):e2011030.
CCM.0000000000001872 doi:10.4084/MJHID.2011.030
1104367
case-report2022
HPXXXX10.1177/00185787221104367Hospital PharmacyAnderson et al
Case Report
Hospital Pharmacy
Nutrition
Abstract
Clinicians have published research and reports on calcium and phosphate solubility within parenteral nutrition (PN) for
over 40 years. Foundational empirical laboratory investigation in the 1980s motivated by the prevalence of neonatal rickets
and osteomalacia in the Neonatal Intensive Care Unit (NICU) population led to precipitation curves that have guided PN
prescribing and compounding. Over subsequent decades, numerous publications have expanded the knowledge of factors
influencing calcium and phosphate solubility in formulating optimal and safe PN admixtures. Failure to adhere to known
principles has led to serious injury and death. Known solubility curves are derived from empiric analysis of a finite number
of conditions and concentrations, whereas custom PN orders vary widely and rarely match the admixture composition in
the data set used to derive the published precipitation curves. Various commercial platforms have been developed to aid
the pharmacist in assessing the potential for precipitation when evaluating a PN order. Some applications plot the calcium
and phosphate concentrations of the prescribed PN against known published graphs most similar to the order, allowing
the pharmacist to judge the risk of precipitation. Other approaches use intellectually protected trade secret algorithms to
determine calcium and phosphate solubility across a continuum of conditions. This publication reports equations that have
been used successfully for over 2 decades in our regional network of NICUs and shared with others to determine safe
prescribing limits for calcium and phosphate concentrations using an electronic PN prescribing program.
Keywords
nutrition, nutritional support, intravenous therapy, pediatrics
Prescribing and compounding parenteral nutrition PN admix- composition and pH are linked variables. Studies show that
tures is a high-risk process due to the complexity of the large the precipitation curves became more favorable, that is, cal-
number of components, the wide range of concentrations, and cium-phosphate solubility increases, with increased amino
the various additives prescribed to meet clinical demands of acid concentrations and decreased pH.6,7 Original research
therapy.1 To help guide professionals in the safe provision of studied various amino acid products at discrete concentra-
PN therapy, the American Society for Parenteral and Enteral tions within PN.3,6,8-11 Historically, the pharmacist would
Nutrition (A.S.P.E.N.) has published recommendations.2 compare an individual patient’s order against the most closely
These guidelines emphasize the need for all formulations to matched reference precipitation curves from published
be evaluated for risk of calcium and phosphate precipitation. sources. Over time software was developed that would auto-
Early work within PN formulations determined pH to be a matically plot the calcium and phosphate concentrations of a
predominant factor in calcium and phosphate solubility.3 At
lower pH, the monobasic phosphate ion (H2PO4−) prevails
1
allowing solubility with calcium 60-fold greater than the Primary Children’s Hospital, Salt Lake City, UT, USA
2
Intermountain Healthcare, Salt Lake City, UT, USA
dibasic phosphate ion (HPO42−), which predominates at 3
Salt Lake Community College, Salt Lake City, UT, USA
higher pH.4 In addition to the pH and calcium/phosphate con- 4
University of Utah Health, Salt Lake City, UT, USA
centrations, the amino acid concentration, amino acid prod-
Corresponding Author:
uct, form of calcium utilized, and temperature also influence Collin Anderson, Primary Children’s Hospital, 100 N Mario Capecchi
calcium and phosphate solubility.5 Since amino acid solutions Drive, Salt Lake City, UT 84113-1100, USA.
are the principle acidic component of PN, amino acid Email: collin.anderson@imail.org
780 Hospital Pharmacy 57(6)
Table 1. Assigned Amino Acid Factor Values by Source Product. A solution factor is assigned to each PN order:
Commercial amino acid product Assigned amino acid factor
Equation 4: Solution factor
Aminosyn PF, Premasol, 200
TrophAmine ( calcium gluconate in mEq/L)0.863 *
Aminosyn II, Clinisol, Travasol 100 Solution Factor = /
( inorganic phosphate in mmol/L )1.19
prescribed PN against known curves.12 Because an ordered ( final AA% )
PN rarely matches the published curves within the software,
the pharmacist must still make a judgment regarding the The solution factor value is compared to the precipitation
potential for precipitation of the ordered product. Industry has limit value. The calcium and phosphate within the PN admix-
developed certain intellectually protected algorithms based ture is considered compatible if the solution factor value is
on published precipitation curves.13 These unpublished algo- less than the precipitation limit value.
rithms are used to stop an order from being placed that would Example for 2-in-1 PN with y-site administration with fat
have a high potential for precipitation. We report equations emulsion
utilized in an electronic PN prescribing program to prospec- 1 kg neonate
tively determine if an order for PN is likely to precipitate. 3.5 g/kg of amino acid (Trophamine or Premasol)
40 mg cysteine per g of amino acids
3 g/kg of fat emulsion (20%) run over 24 hours (0.63 mL/
Methods hour) y-site administration with PN
Equations were developed according to mathematical prin- Calcium 3.6 mEq/kg
ciples to match the general hyperbolic shape of published Phosphates 1.8 mmol/kg
curves, incorporating those elements most strongly influenc- Volume of 2-in-1 PN 120 mL/kg
ing the buffering capacity and pH of a PN admixture. Precipitation limit = 200 + (40 * 200)/100 − (0.63 * 24 * 20
A numeric value, referred to as the precipitation limit, is * 200)/(3.5 * 1 * 1000) = 262.7
assigned to each PN order according to the following Solution factor = (300.863 * 151.19)/2.92 = 161.8
formulas: The solution factor value (161.8) is less than the precipita-
tion factor value (262.7), therefore the calcium and phos-
Equation 1: 2-in-1 PN that will be run separate from fat phate within the solution is determined to be compatible at
emulsion standard storage and administration temperatures.
Example for 3-in-1 PN
Precipitation limit = AAF + ( CYS*AAF ) /100 30 kg adolescent
1.6 g/kg of Travasol
Equation 2: 2-in-1 PN with y-site administration with fat 0.9 g/kg of fat emulsion (20%)
emulsion Calcium 1 mEq/kg
Phosphates 0.9 mmol/kg
Precipitation limit = AAF + ( CYS*AAF ) / Volume of 3-in-1 PN 75 mL/kg
Precipitation limit = 100 + (0 * 100)/100 – (0.9 * 100)/
100 − ( FIR*FIT*FPP*AAF ) / ( AA*WT*1000 )
(1.6 * 10) = 94.4
Solution factor = (13.330.863 * 121.19)/2.13 = 84.4
Equation 3: 3-in-1 PN The solution factor value (84.4) is less than the precipita-
tion factor value (94.4), therefore the calcium and phosphate
Precipitation limit = AAF + ( CYS*AAF ) / within the PN is determined to be compatible at standard
100 − ( FAT*AAF ) / ( AA*10 ) storage and administration temperatures.
Figure 5. TrophAmine 1.5% with 10% dextrose, 12% Intralipids, and cysteine (40 mg/g amino acids) parenteral nutrition.
Figure 9. TrophAmine 2% with 10% dextrose, 12% Intralipids, Figure 11. Predicted calcium and phosphate program limits
and cysteine (40 mg/g amino acids) parenteral nutrition. at varying cysteine concentrations, 3% TrophAmine 2-in-1
parenteral nutrition solution.
administration of PN. Equations have been developed to pre- Declaration of Conflicting Interests
dict calcium and phosphate solubility across a continuum of The author(s) declared the following potential conflicts of interest with
conditions. These equations have been successfully utilized respect to the research, authorship, and/or publication of this article:
within an electronic PN prescribing program for more than 2 Collin Anderson: Consultant for Baxter on a limited basis (<10 hours
decades across numerous facilities. per year) for a product which is not a part of this manuscript.
Anderson et al 785
Case Report
Hospital Pharmacy
Carvedilol-Induced Thrombocytopenia:
2022, Vol. 57(6) 786–789
© The Author(s) 2022
Article reuse guidelines:
A Case Report sagepub.com/journals-permissions
DOI: 10.1177/00185787221108719
https://doi.org/10.1177/00185787221108719
journals.sagepub.com/home/hpx
Abstract
Thrombocytopenia is a commonly encountered complication of hospitalized patients, and can be caused by many factors
including infections, surgery, and medications. Drug-induced thrombocytopenia (DITP) should be considered when a patient
presents with an unexpected occurrence of thrombocytopenia. Many drugs can induce thrombocytopenia either as a direct
effect on thrombopoiesis within the bone marrow or by drug-dependent antibody-mediated destruction of platelets within
the circulation. We present the case of a 44-year-old female who presented with 2 weeks of nausea, vomiting, and headaches
occurring with her hemodialysis sessions. On admission she was found to be thrombocytopenic with a platelet count of 35 K/µL.
Her last known normal platelet count was 299 K/µL from 2 months prior. A thorough work up of her thrombocytopenia was
found to be unremarkable. Her newly started carvedilol was thought to be the most likely cause based on the rapid platelet
recovery upon drug discontinuation and negative workup of other potential causes.
Keywords
carvedilol, thrombocytopenia, beta-blocker, platelets, heparin-induced thrombocytopenia
amputation, and recently diagnosed hypertension. Upon pre- approximately 2 weeks prior to the patient’s presentation.
sentation to the emergency department, she was found to Further work-up for ITP was unremarkable due to the lack
have thrombocytopenia of unknown origin that necessitated of platelet clumping or morphologic abnormalities. HIV
further workup. Her platelet count was 35 K/µL on admis- panel resulted as undetectable, as were hepatitis C virus
sion, and her last known normal platelet count was 299 K/µL antibody and hepatitis B surface antigen. Helicobacter
from 2 months prior. Her prior to admission medication list pylori antigen was also not detected. Thyroid-stimulating
included carvedilol 12.5 mg twice daily, which was started hormone (TSH) was 2.18. Her autoimmune profile (C reac-
approximately 2 weeks prior to admission, cyanocobalamin tive protein, rheumatoid factor, C3, C4, and antinuclear
1000 mcg IM once monthly, and insulin lispro via insulin antibodies screen) was unremarkable, and immunoglobulin
pump. She was admitted to an outside hospital 1 month prior G antibody was less than 14 and within normal limits. By
to admission for the amputation of her right lower extremity the third day of hospitalization, the patient’s platelets had
and received vancomycin and cefepime as pre- and post- more than doubled to 86, and continued to rise to 189 by
operative antibiotics. The patient had no personal or family discharge. Amlodipine was started prior to discharge to
history of autoimmune disorders, thrombocytopenia or replace carvedilol, and this was well tolerated.
hematologic malignancy.
Her complete blood count (CBC) with differential was
found to be within normal limits, except for the following: red
Discussion
blood cells (RBC) 3.98, hemoglobin 11.6, mean corpuscular Drug-induced thrombocytopenia typically occurs around 5
hemoglobin concentration (MCHC) 29.9, and red cell distribu- to 10 days after initial drug administration with a median
tion width (RDW) 16.7. Her basic metabolic panel (BMP) was platelet nadir of about <20 × 109/L2. Determining the
also within normal limits, except for abnormal serum creati- offending drug can be difficult since patients are typically on
nine and blood urea nitrogen (BUN) due to ESRD. GI/liver several medications at a time. Bakchoul et al2 discuss 5 clini-
labs, iron studies (despite iron saturation of 19), and vitamin cal criteria that may help in the establishment of probable
B12 level were all within normal limits. A peripheral smear and drug-induced thrombocytopenia: (1) exposure to the drug
morphology of the patient’s platelets was performed and occurred prior to thrombocytopenia; (2) recovery was com-
resulted in normal appearing platelets with normal granularity. plete and sustained upon discontinuation of that drug; (3) the
The initial differential diagnoses included drug-induced throm- drug was the only new drug started prior to the thrombocyto-
bocytopenia (DIT) versus heparin-induced thrombocytopenia penic episode or other drugs were continued or reintroduced
(HIT) versus thrombotic thrombocytopenic purpura (TTP) ver- after discontinuation of the suspected drug with sustained
sus immune thrombocytopenic purpura (ITP). The prospective normal platelet counts; (4) other causes of thrombocytopenia
work-up included hemolysis labs, HIV panel, HCV and HBV were ruled out, and (5) re-exposure to the suspected drug
panels, autoimmune profile, hematology consult, and HIT- resulted in recurrent thrombocytopenia. Similarly, the
antibody since the patient believed unfractionated heparin was Naranjo et al5 adverse drug reaction probability scale is one
used during her dialysis sessions. Given the patient’s normal of the most widely used causality assessment tools. This
bilirubin, haptoglobin, and only mildly elevated LDH, throm- scale consists of 10 questions that, when totaled, provide a
botic angiopathies, such as TTP and hemolytic uremic syn- final score that is associated with one of 4 categories of like-
drome (HUS), were ruled out. lihood that the drug was associated with the reaction: doubt-
A 4 T-score of 5 was calculated, which is categorized as ful, possible, probable, or definite. Treatment of drug-induced
intermediate risk. She received 2 points for having a platelet thrombocytopenia necessitates drug discontinuation and
fall greater than 50% and a platelet nadir of greater than monitoring of platelets, but in instances of significant dis-
20 000. She received one point for a platelet fall consistent ease, high doses of intravenous immunoglobulin can be
with days five through 10, but not clear as the patient did given to patients at high risk of bleeding or with evidence of
state she received heparin with her hemodialysis sessions. bleeding. However, only 2 case reports have been published
She received zero points for active thrombosis. Finally, she about using intravenous immunoglobulin in these instances.6,7
received 2 points for not having any alternative causes of Upon discontinuation of the suspected drug, platelet counts
thrombocytopenia. Per the HIT protocol at Rush University can start to improve after 4 to 5 half-lives of the drug and/or
Medical Center, a HIT-antibody was ordered on day 1 of hos- drug metabolite.2
pitalization, which resulted as negative with an optical den- The mechanism behind beta-blocker-associated throm-
sity of 0.191. No serotonin-release assay (SRA) was ordered bocytopenia is not fully understood, but its proposed mecha-
per the protocol, since HIT was deemed unlikely based on nism may be immunologic in nature owing to positive
the negative HIT-antibody. platelet-reactive antibody testing or reduced platelet pro-
On day 2 of hospitalization, carvedilol was held as a duction.8,9 Metoprolol, sotalol, labetalol, carvedilol, ateno-
possible cause of thrombocytopenia since this was the only lol, and propranolol are the 6 beta blockers that have been
new medication for the patient and had been started tested at the Wisconsin Blood Center for platelet-reactive
788 Hospital Pharmacy 57(6)
Day 1 Day 2 Carvedilol stopped Day 3 Day 4 Day 5 Day 6 Day 7 Day 8 Day 9
35 44 86 95 128 139 171 153 189
43
antibodies between 1995 and 2018.10 Only atenolol and pro- amlodipine for blood pressure management, which was well
pranolol yielded positive results. Furthermore, no standard tolerated. Carvedilol-induced thrombocytopenia is rare, and
testing assays, criteria for differentiating positive and nega- not well understood as there have only been 2 other cases
tive results, and data on sensitivity and specificity of these reported in the literature to date.
assays have been established.11
The Naranjo score for our patient was 6, suggesting a Declaration of Conflicting Interests
probable association between carvedilol and the patient’s The author(s) declared no potential conflicts of interest with
thrombocytopenia. After a thorough literature search on respect to the research, authorship, and/or publication of this
PubMed using the search terms “thrombocytopenia, beta article.
blockers, carvedilol,” 2 case reports were found regard-
ing thrombocytopenia with carvedilol use. Alex et al12 Funding
reported their patient was also recently started on
The author(s) received no financial support for the research, author-
carvedilol and experienced a platelet drop of about ship, and/or publication of this article.
151 000 within 19 days of therapy initiation. Once the
patient had been switched from carvedilol to metoprolol, Ethics
the platelet count improved in as little as 24 hours.
Similarly, their Naranjo score was 8, indicating carvedilol Informed consent was obtained from the study subject.
was the probable cause of their patient’s thrombocytope-
nia. James et al13 reported the case of a hospitalized ORCID iDs
patient who was initiated on ceftriaxone, doxycycline, Angelina R. Raimonde https://orcid.org/0000-0003-3265-1628
metformin, insulin, furosemide, carvedilol, losartan, and Anthony K. Dennis https://orcid.org/0000-0002-8146-4813
nebulizer treatments. By day 3, the patient had progres-
sive thrombocytopenia, with a negative work up for References
Dengue and chronic liver disease. Ceftriaxone, doxycy- 1. Stasi R. How to approach thrombocytopenia. Hematol Am Soc
cline and furosemide were held on day 5, however plate- Hematol Educ Program. 2012;2012(1):191-197. doi:10.1182/
lets continued to decline. It was not until carvedilol was asheducation-2012.1.191
held that the platelets began to recover over the next 24 to 2. George JN, Aster RH. Drug-induced thrombocytopenia: patho-
48 hours. Platelets had returned to normal at her 14 days genesis, evaluation, and management. Hematology Am Soc
follow up visit.14 Our patient also displayed rapid recov- Hematol Educ Program. 2009;153-158. doi:10.1182/ashedu-
ery of her platelets the day after discontinuing carvedilol, cation-2009.1.153
as shown in Table 1. Her platelets had doubled from day 3. Bakchoul T, Marini I. Drug-associated thrombocytopenia.
Hematology. 2018;2018:576-583. doi:10.1182/asheducation-
one to day three and continued to uptrend throughout the
2018.1.576
remainder of her inpatient stay. Carvedilol has a half-life 4. Arnold DM, Kukaswadia S, Nazi I, et al. A systematic evalua-
of about 6 hours, so platelet recovery within 24 hours tion of laboratory testing for drug-induced immune thrombocy-
seems appropriate. The carvedilol package insert reports topenia. J Thromb Haemost. 2013;11(1):169-176. doi:10.1111/
a 1% to 3% incidence of thrombocytopenia from the clin- jth.12052
ical trials. 5. Naranjo CA, Busto U, Sellers EM, et al. A method for estimat-
ing the probability of adverse drug reactions. Clin Pharmacol
Ther. 1981;30:239-245. doi:10.1038/clpt.1981.154
Conclusion 6. Ray JB, Brereton WF, Nullet FR. Intravenous immune
globulin for the treatment of presumed quinidine-
A 44-year-old female developed new onset thrombocytope- induced thrombocytopenia. DICP. 1990;24(7-8):693-695.
nia after the recent initiation of carvedilol for newly diag- doi:10.1177/106002809002400706
nosed hypertension. Carvedilol was determined to be the 7. Tvito A, Bakchoul T, Rowe JM, Greinacher A, Ganzel C. Severe
most likely cause based on the rapid platelet recovery upon and persistent heparin-induced thrombocytopenia despite
drug discontinuation, and negative workup of other potential fondaparinux treatment. Am J Hematol. 2015;90(7):675-678.
causes. Prior to hospital discharge, the patient was trialed on doi:10.1002/ajh.23971
Raimonde and Dennis 789
8. Magnusson B, Rödjer S. Alprenolol-induced throm- 11. George JN, Raskob GE, Shah SR, et al. Drug-induced throm-
bocytopenia. Acta Med Scand. 1980;207:231-233. bocytopenia: a systematic review of published case reports.
doi:10.1111/j.0954-6820.1980.tb09711.x Ann Intern Med. 1998;129:886-890. doi:10.7326/0003-4819-
9. Patrassi GM, Casonato A, Fabris F, Girolami A. 129-11_part_1-199812010-00009
Thrombocytopenia secondary to oxprenolol, a beta-blocking 12. Alex S, Chretien KC, Cho A, Aggarwal A. Probable carvedilol-
agent. Acta Haematol. 1982;68:75-76. doi:10.1159/000206955 induced thrombocytopenia. Am J Health Syst Pharm.
10. George JN. Platelets on the web. Drug-induced thrombocy- 2013;70:598-602. doi:10.2146/ajhp120383
topenia. 2015. Accessed September 18, 2020. https://www. 13. James J, Cyriac C, Chacko SS, et al. Carvedilol induced throm-
ouhsc.edu/platelets/ditp.html. bocytopenia. EJPMR. 2021;8(1):521-522.
SAGE Video streaming collections are developed
in partnership with leading academics, societies SAGE Video collections
and practitioners, including many of SAGE’s own
• Business & Management
authors and academic partners, to deliver cutting-
edge teaching and research-oriented video within • Counseling & Psychotherapy
the social, behavioral, and health sciences. • Criminology & Criminal Justice
sagepub.com/reviewerrewards
SAGE Premier
Online access to over 1,100 SAGE journals in business; humanities and social
sciences; and science, technology, and medicine...
The SAGE Premier package includes leading SAGE Premier includes the following:
international peer-reviewed titles, including high- • access to over 1,100 SAGE journal titles,
impact research journals published on behalf comprising over 1 million full-text articles;
of more than 400 scholarly and professional
• access via SAGE Journals online platform;
societies. Our journals cover a wide range of
subject areas—including business; humanities • complimentary access back to 1999,
and social sciences; and science, technology, where available, while a current
and medicine—allowing libraries to support the subscription is maintained; and
varied teaching and research needs of their • High percentage of journals ranked in the JCR.
students and faculty. See SAGE website for latest figures.
journals.sagepub.com
SAGE is home to over 1,000 prestigious and highly cited journals from a wide range of
disciplines, available electronically on the SAGE Journals online platform. Search and
browse all journals content—including journals in business, humanities and social sciences,
and science, technology, and medicine.