See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/271773677

Do we need a new definition of sepsis?

Article  in  Intensive Care Medicine · February 2015

DOI: 10.1007/s00134-015-3680-x

CITATIONS READS

30 1,156

3 authors:

Manu Shankar-Hari Clifford S Deutschman

King's College London Hofstra North Shore-LIJ School of Medicine, Feinstein Institute for Medical Research
113 PUBLICATIONS   4,093 CITATIONS    284 PUBLICATIONS   16,160 CITATIONS   

SEE PROFILE SEE PROFILE

Mervyn Singer
University College London
277 PUBLICATIONS   19,151 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Immune depletion in sepsis View project

Long term outcomes in sepsis survivors View project

All content following this page was uploaded by Manu Shankar-Hari on 24 February 2015.

The user has requested enhancement of the downloaded file.

Intensive Care Med
DOI 10.1007/s00134-015-3680-x WHAT’S NEW IN INTENSIVE CA RE

Manu Shankar-Hari
Clifford S. Deutschman
Do we need a new definition of sepsis?
Mervyn Singer

practitioners who treat ‘septic’ patients may struggle. A

Received: 9 November 2014
Accepted: 22 January 2015
! Springer-Verlag Berlin Heidelberg and ESICM 2015
M. Shankar-Hari
Department of Critical Care Medicine, Guy’s and St Thomas’ NHS
Foundation Trust, London SE17EH, UK
e-mail: manu.shankar-hari@kcl.ac.uk
Tel.: 44-207-188-3044
M. Shankar-Hari
Division of Asthma, Allergy and Lung Biology, King’s College
London, London, UK
C. S. Deutschman
Department of Pediatrics, North Shore-Long Island Jewish-Hofstra
School of Medicine, Steven and Alexandra Cohen Children’s
Medical Center, 269-01 76th Avenue, New Hyde Park, NY 11040,
USA
e-mail: cdeutschman@nshs.edu
Tel.: 001 (718) 470-3392
C. S. Deutschman
Feinstein Institute for Medical Research, Room 3140, 350
Community Drive, Manhasset, NY 11030, USA
M. Singer ())
Bloomsbury Institute of Intensive Care Medicine, University
College London, Cruciform Building, Gower St, London WC1E
6BT, UK
e-mail: m.singer@ucl.ac.uk
Tel.: 44-207-679-6714
Of all critical care conditions, sepsis has shaped health
policy, dominated the research agenda, and entered the
public lexicon with energetic, high profile educational
campaigns vowing to reduce the attendant mortality. Yet,
when asked to provide a concise definition, even
consensus conference convened in 1992 characterized
sepsis as a syndrome of presumed or confirmed infection
and at least two of four systemic inflammatory response
syndrome (SIRS) criteria [1]. However, though sensitive,
this collection of non-validated variables lacks specificity.
Indeed, most patients admitted to critical care units are
SIRS-positive while the presumption of infection is a
default diagnosis that is frequently adopted but often not
confirmed. The second consensus conference, sitting in
2001, compounded the specificity issue by maintaining
the 1992 illness construct yet further extending the list of
clinical and laboratory variables to describe the protean
manifestations of the host response to infection [2]. In
addition, a proliferation of other definitions created to suit
trial entry criteria, management guidelines, and retro-
spective analyses of databases have led to a widely
discrepant epidemiology that applies to both ‘sepsis’ and
‘septic shock’.
As a result, recently reported trends detailing sepsis
incidence and outcomes are of questionable validity.
Identifying septic patients on the basis of discharge
diagnosis codes is complicated by the likely influence of
awareness campaigns, high-profile research papers, and
reimbursement formulae [3]. Using claims-based data,
Rhee et al. noted a 170 % increase in sepsis between 2003
and 2011. Yet, during the same period, these investigators
noted a 22 % reduction in the large number of patients
admitted to hospitals in the USA with a primary diagnosis
of pneumonia [3]. Whereas global estimates suggest 19
million individuals are hospitalized with sepsis per annum
[4], recently completed multi-centre trials investigating
the impact of early goal-directed therapy [5, 6] have
struggled to recruit at the planned rate. Using the
Nationwide Inpatient Sample US hospital population
database and analysed by four separate data extraction
methods, the computed annual incidence of severe sepsis
differed 3.5-fold (ranging from 300 to 1,031 per 100,000
population) and this discrepancy was associated with a
twofold difference in mortality (ranging from 14.7 to
29.9 %) [7]. The Australian and New Zealand ICU
database analysis [8] identified a tenfold higher annual
incidence of patients with septic shock compared to an
equivalent number of Italian ICUs (Dr G. Bertolini, Gi-
ViTi, personal communication), with mortality rates of 22
versus 61 %, respectively.
So now, over a decade after the second consensus
conference, what factors mandate the need to revisit and
refine definitions of sepsis?
Firstly, our knowledge of the underlying pathophysi-
ology has expanded [9, 10], and the fundamental role of
the host response in producing self-harm is now well
established (Fig. 1). Indeed, sepsis was defined at the
2010 Merinoff symposium as ‘‘a life-threatening condi-
tion that arises when the body’s response to an infection
injures its own tissues and organs’’ [11]. However, our
understanding of the key determinants of the highly
complex host response remains rather rudimentary. In
particular, recognizing when an adaptive and protective
homeostatic/allostatic response becomes maladaptive and
thus deleterious is imperative for optimal timing of an
appropriate therapeutic intervention.
Second, the SIRS criteria are too non-specific to retain
value. The SIRS-based definition will identify most
patients with a simple infection, even a common cold, as
septic [2]. Sepsis thus needs to be clearly delineated
from a straightforward ‘uncomplicated’ infection as the
host response process that leads to significant organ
dysfunction.
Fig. 1 A 2015 concept of
sepsis pathophysiology Infec!on [PAMPs]
Danger signals Tissue injury [DAMPs]
Detec!on and transduc!on of danger
Third, as important as it is to consider the possibility of
organ dysfunction in patients with presumed/confirmed
infection, it is equally vital to address whether undiag-
nosed infection underlies new organ dysfunction.
Fourth, the emergence of resistant bacterial strains
mandates prudent use of antibiotics [12] yet some retro-
spective analyses strongly suggest that prompt initiation
of antibiotics reduces mortality [13, 14]. We need to
strike an appropriate balance.
Fifth, the ability to interrogate large electronic dat-
abases capturing sequential clinical, laboratory, and
imaging data should greatly facilitate the development of
new, more precise and validated definitions. These con-
structs should prove superior to the arbitrary
physiological and biochemical threshold values currently
used to identify organ dysfunction. The impact of co-
morbidities on sepsis-related outcomes and the true
attributable mortality from sepsis may also be better
understood [8, 15].
Sixth, we need to more fully appreciate the heteroge-
neity of sepsis and to temper the need to identify
homogenous cohorts. The inciting infection is heteroge-
neous with respect to types of pathogen, organism load,
site, and lag time. Patients are heterogeneous with respect
to their underlying age, gender, genetic makeup, comor-
bidities, concurrent medication, etc. The host response is
heterogeneous both within and between organ systems.
This host response is also affected by delivered inter-
ventions that vary considerably in differently resourced
healthcare systems and divergent local management
practices. Each of these factors may influence sepsis-
Pa"ern recogni!on receptors
(e.g. Toll-like receptors)
Local or systemic
Host response response effec!ve resolu!on
–> danger cleared
Local or systemic
response ineffec!ve
–> danger persists with
2° impact upon !ssues
Cell- and organ level
abnormali!es
death
 related outcomes and the response to a specific therapy.
We propose that this broad-based heterogeneity be
embraced and better characterized, as ‘pure’ homogenous
septic cohorts are unlikely to be identified in the near
future.
Precision medicine should become the norm when
evaluating new therapeutics and reassessing current
drugs. This approach should enhance the likelihood of
success in what has been a dismal therapeutic area to date.
These caveats apply equally to the use of novel biologics
as to large-scale randomised controlled trials of existing
therapies. What we have learned about sepsis argues for
adoption of strategic approaches where one size cannot fit
all.
A single universal definition using validated variables
should thus support a more accurate epidemiological
characterization, appropriate power for interventional
trials, and benchmarks for providers and hospitals. The
same process should be applied to other aspects of septic
pathobiology. For example, the definition of septic shock
currently revolves around variable blood pressure and/or
lactate levels, with loosely termed or undefined ‘adequacy
of fluid resuscitation’ and ‘persistent’ hypotension.
Defining sepsis must, however, be an ongoing iterative
process requiring minor or major revisions as new find-
ings come to light. In much the same way that software
enhancements move from version 1.0 to 1.1 or to 2.0
depending on the magnitude of change, so a new sep-
sis 3.0 definition must be refined into versions 3.1, 3.2,
and so on until an eventual complete overhaul generates
the development of sepsis 4.0.
Conflicts of interest The authors declare that they are co-chairs
(MS, CD) and a co-opted member (MSH) of the Sepsis Redefini-
tions Task Force co-supported by the European Society of Critical
Care Medicine and the Society for Critical Care Medicine. They
have no conflict of interest.

References
1. Bone RC, Sibbald WJ, Sprung CL
(1992) The ACCP-SCCM consensus
conference on sepsis and organ failure.
Chest 101:1481–1483
2. Levy MM, Fink MP, Marshall JC, for
the International Sepsis Committee et al 8. Kaukonen KM, Bailey M, Suzuki S,
(2003) 2001 SCCM/ESICM/ACCP/
ATS/SIS international sepsis definitions
conference. Intensive Care Med
29:530–538
3. Rhee C, Gohil S, Klompas M (2014)
Regulatory mandates for sepsis care—
reasons for caution. N Engl J Med
370:1673–1676
4. Adhikari NK, Fowler RA, Bhagwanjee
S, Rubenfeld GD (2010) Critical care
and the global burden of critical illness
in adults. Lancet 376:1339–1346
5. ProCESS Investigators, Yealy DM, JA
Kellum, Huang DT et al (2014) A
randomized trial of protocol-based care
for early septic shock. N Engl J Med
370:1683–1693
6. ARISE Investigators, Peake SL,
Delaney A, Bailey M et al (2014) Goal-
directed resuscitation for patients with 12. Nathan C, Cars O (2014) Antibiotic
early septic shock. N Engl J Med
371:1496–1506
7. Gaieski DF, Edwards JM, Kallan MJ,
Carr BG (2013) Benchmarking the
incidence and mortality of severe sepsis
in the United States. Crit Care Med
41:1167–1174
Pilcher D, Bellomo R (2014) Mortality
related to severe sepsis and septic shock 14. Kumar A, Roberts D, Wood K et al
among critically ill patients in Australia
and New Zealand, 2000–2012. JAMA
311:1308–1316
9. Xiao W, Mindrinos MN, Seok J, for the
Inflammation and Host Response to
Injury Large-Scale Collaborative
Research Progam et al (2011) A
genomic storm in critically injured
humans. J Exp Med 208:2581–2590
10. Hotchkiss RS, Monneret G, Payen D
(2013) Sepsis-induced
immunosuppression: from cellular
dysfunctions to immunotherapy. Nat
Rev Immunol 13:862–874
11. Czura CJ (2011) ‘‘Merinoff symposium
2010: sepsis’’—speaking with one
voice. Mol Med 17:2–3
resistance—problems, progress, and
prospects. N Engl J Med
371:1761–1763
13. Ferrer R, Martin-Loeches I, Phillips G
et al (2014) Empiric antibiotic
treatment reduces mortality in severe
sepsis and septic shock from the first
hour: results from a guideline-based
performance improvement program.
Crit Care Med 42:1749–1755
(2006) Duration of hypotension before
initiation of effective antimicrobial
therapy is the critical determinant of
survival in human septic shock. Crit
Care Med 34:1589–1596
15. Clermont G, Angus DC, Linde-Zwirble
WT, Griffin MF, Fine MJ, Pinsky MR
(2002) Does acute organ dysfunction
predict patient-centered outcomes?
Chest 121:1963–1971

View publication stats