Implications of the new sepsis definition on research and practice

Brian C. Peach

PII: S0883-9441(16)30878-4
DOI: doi:10.1016/j.jcrc.2016.11.032
Reference: YJCRC 52348

To appear in: Journal of Critical Care

Please cite this article as: Peach Brian C., Implications of the new sepsis denition on
research and practice, Journal of Critical Care (2016), doi:10.1016/j.jcrc.2016.11.032

This is a PDF le of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its nal form. Please note that during the production process
errors may be discovered which could aect the content, and all legal disclaimers that
apply to the journal pertain.
 ACCEPTED MANUSCRIPT

PT
RI
Implications of the New Sepsis Definition on Research and Practice

SC
Brian C. Peach 1

University of Florida

NU
1
1225 Center DR
PO Box 100197
Gainesville, FL 32610-0197
MA
Corresponding author: Brian C. Peach, bpeach01@ufl.edu
D
P TE
CE
AC
 ACCEPTED MANUSCRIPT

Abstract

Introduction: The Society of Critical-Care Medicine and the European Society of Intensive Care Medicine
recently announced a marked change in the sepsis definition. A task force of 19 sepsis clinicians and
researchers made the change based on advances in the understanding of the septic process.

PT
Summary of Change: The task force determined there were numerous justifications for a revision of the
sepsis definition. Reasons for making a change included that there was an overwhelming focus on
inflammation, the sepsis continuum model was misleading, the systemic inflammatory response

RI
syndrome (SIRS) criteria lacked adequate sensitivity and specificity, the multiple definitions for sepsis,
septic shock, and organ dysfunction led to discrepancies in reported incidence and mortality, and they

SC
felt the term severe sepsis was redundant (Singer et al., 2016)1. The systemic inflammatory response
criteria have been replaced by the sequential organ failure assessment (SOFA) score and the qSOFA
score in the newly operationalized definition (Singer et al., 2016)1.

NU
Implications on Research & Practice: There are multiple implications on research and practice, including
potential fiscal effects on hospitals and universities as they adopt the new definition. The change in
MA
definition may have a negative impact on sepsis research in the short-term, but may limit systematic
bias and improve the quality of sepsis research in the long-term.

Conclusions: While the new operationalized Sepsis-3 definition appears on the surface to be an
improvement over the previous iterations, it remains to be seen if clinical outcomes will improve and if
D

research studies will be more robust using the new criteria. Future research on the impact of this change
TE

on research and practice will be essential, to determine if the definition and associated clinical criteria
need further revision.
P

Key Words: sepsis, severe sepsis, septic shock, systemic inflammatory response syndrome (SIRS) criteria,
CE

Sequential [Sepsis-related] Organ Failure Assessment (SOFA) score, Logistic Organ Dysfunction System
(LODS) criteria
AC

Introduction

The Society of Critical-Care Medicine (SCCM) and the European Society of Intensive Care

Medicine (ESICM) recently announced a marked change in the sepsis definition (Singer et al., 2016)1. The

original definition (Sepsis-1) created in 1991 (Bone et al., 1992)2 was revised in 2001 (Sepsis-2) with

the addition of clinical criteria for inadequate tissue perfusion (Levy et al., 2003)3. Since the 2001

revision, there have been numerous advances in understanding the pathology, underlying biological

mechanisms, management, and epidemiology of sepsis, which prompted a review of the Sepsis- 2

definition (Singer et al., 2016)1. A task force of 19 sepsis clinicians and researchers convened in 2014 to

review the current evidence, and to revise the definition through meetings, Delphi processes, analysis of
 ACCEPTED MANUSCRIPT

electronic health record databases, and subsequent voting on criteria (Singer et al., 2016)1. After

meeting, the task force circulated their Sepsis-3 definition to 31 international professional societies,

and requested peer review and endorsement of the new definition (Singer et al., 2016)1. The new

PT
definition was announced in a February 23rd, 2016 Journal of the American Medical Association article

(Singer et al., 2016)1, and the implications of this significant change of research and practice will be

RI
discussed later in this article.

SC
Summary of Change

NU
Leading up to the SCCM/ESICM task force creation, several of its members advocated for a
MA
change to the Sepsis-2 definition based on their own teams research findings (Deutschman & Tracey,

2014; Vincent, Opal, Marshall, & Tracey, 2013) 4, 5. The task force determined there were numerous
D

justifications for a revision of the sepsis definition. Reasons for making a change included that there was
TE

an overwhelming focus on inflammation, the sepsis continuum model was misleading, the systemic

inflammatory response syndrome (SIRS) criteria lacked adequate sensitivity and specificity, the multiple
P
CE

definitions for sepsis, septic shock, and organ dysfunction led to discrepancies in reported incidence and

mortality, and they felt the term severe sepsis was redundant (Singer et al., 2016, p.801)1.
AC

Sepsis according to the Sepsis-1 definition was not a disease, but an unbridled cytokine

inflammatory response to an infectious trigger (Bone et al., 1992)2. At the core of the 1991 and 2001

sepsis, severe sepsis, and septic shock definitions is the SIRS criteria. The SIRS criteria included four

clinical signs which when altered were thought to represent an inflammatory response: temperature,

heart rate, respiratory rate, and white blood cell count (Bone et al., 1992)2. In order to be diagnosed

with sepsis, an individual must have at least 2 SIRS criteria and a confirmed or suspected infection (Bone

et al., 1992)2. More recent research however, has shown that sepsis involves activation of both pro-

inflammatory and anti-inflammatory responses (Hotchkiss, Monneret, & Payen, 2013)6. In addition,
 ACCEPTED MANUSCRIPT

there are also other endogenous factors that rapidly change the cardiovascular, neurological, endocrine,

and hematological systems in response to infection and the inflammatory response (Deutschman et al.,

2014; Singer, De Santis, Vitale, & Jeffcoate, 2004)4,7. Other factors like the source of infections,

PT
comorbidities, and iatrogenic interventions also have an impact on body response (Deutschman et al.,

2014) 4.

RI
SC
Under the Sepsis-1 and Sepsis-2 definitions, sepsis, severe sepsis, and septic shock were

seen as steps in a continuum. As a patients condition declined, they advanced from one stage to the

NU
next with unresolved septic shock (the final stage) resulting in death. Many patients however lacked

sufficient SIRS criteria under the previous definitions to receive a sepsis diagnosis, but had complicated
MA
courses similar to those with diagnoses resulting in mortality. In a study of 109,663 patients admitted

between 2000 and 2013 to 172 Australian or New Zealand intensive care units with infection and organ
D

failure, 13,278 patients (12.1%) had SIRS-negative sepsis (Kaukonen, Bailey, Pilcher, Cooper, & Bellomo,
TE

2015)8. These results raise the question of whether the SIRS criteria possess adequate construct validity.
P
CE

While individuals with at least 2 aberrations in the SIRS criteria are often septic, many are not.

Other conditions that could be mistaken for sepsis using these criteria are primary hypothermia,
AC

endocrine disorders (e.g. hyperthyroidism / hypothyroidism), autoimmune disorders (e.g. lupus, cystic

fibrosis), cardiac and pulmonary disorders (e.g. CHF, COPD), and uncomplicated infections without a

protracted inflammatory response (e.g. pneumonia, urinary tract infections). In a multicenter

prospective cohort study of 3,147 patients in 198 ICUs in 24 countries, 87% of patients on admission had

2 SIRS criteria and 93% had two during the admission, with out-of-normal-range respiratory and heart

rates seen in over 70% of patients (Sprung et al., 2006)9. There is a clear direct association between the

number of SIRS criteria and risk for organ failure and mortality (Sprung et al., 2006)9, but the presence of

2 criteria as part of a sepsis diagnosis appears to be overly sensitive and may lead to inflated sepsis

incidence data.
 ACCEPTED MANUSCRIPT

Another prospective cohort study of 2085 patients admitted to an Australian tertiary hospital

found that patients with >/=2 SIRS criteria had a sensitivity of 70.6% and a specificity of 37.5%, with a

positive predictive value of 63.7% and a negative predictive value of 45.1% in predicting

PT
microbiologically confirmed infections (Lai & Kruger, 2011)10. Even when weighting the SIRS criteria in

favor of the temperature and white blood cell count scores, patients with three or greater SIRS criteria

RI
had a marginally better positive predictive value of 64.5% and a negative predictive value of 44.6% (Lai

SC
et al., 2011)10. The evidence suggests that using unweighted or weighted SIRS criteria can lead to a high

NU
number of false positives and false negatives.

Another issue addressed in the Singer et al. (2016) paper1, is the variability in use of the terms
MA
sepsis, severe sepsis, and septic shock. At present, there is no diagnostic test for sepsis, nor one that

differentiates the different degrees of illness. There is also no current process to operationalize the
D

definitions, which according to Singer et al. (2016) leads to significant variations in the reported
TE

incidence and mortality rates 1. With the new definition and operationalized clinical criteria, the task
P

force aimed to increase specificity using criteria that identified infection, host response, and organ
CE

infection (Singer et al., 2016)1. Improved discriminative abilities, could lead to a better epidemiological
AC

understanding of the sepsis disorder. The task force eliminated the term severe sepsis, which they

labeled redundant, as nearly all patients with a sepsis diagnosis end up in an intensive care unit

(Singer et al., 2016, p.801) 1. Under the Sepsis-3 definition, there are two diagnostic categories, sepsis or

septic shock (Singer et al., 2016)1.

Recognizing that there are other body processes at play during sepsis besides the exaggerated

inflammatory response, the SCCM and ESICM task force decided to define sepsis as Life-threatening

organ dysfunction caused by a dysregulated host response to infection. with the following clinical

operationalized definition: organ dysfunction can be represented by an increase in the Sequential

[Sepsis-related] Organ Failure Assessment (SOFA) score of 2 points or more. (Singer et al., 2016, p. 801)
 ACCEPTED MANUSCRIPT

1
. Septic shock was defined as a subset of sepsis in which particularly profound circulatory, cellular, and

metabolic abnormalities are associated with a greater risk of mortality than with sepsis alone (Singer et

al., 2016, p. 801)1. Septic shock was operationalized by a vasopressor requirement to maintain a mean

PT
arterial pressure of 65mmHg or greater and serum lactate level greater than 2mmol/L (>18mg/dL) in the

absence of hypovolemia) (Singer et al., 2016, p.801) 1.

RI
SC
The SOFA scoring tool is not new, but is not a standard tool in many intensive care units. It is a

screening tool for organ dysfunction, with an increase in score of 2 points from baseline being

NU
associated with >10% increased risk of mortality (Singer et al., 2016)1. When accounting for a patients

baseline risk level, a SOFA score of 2 points or greater is associated with a two to twenty-five fold
MA
increased risk of mortality when compared to patients with a SOFA score less than 2 (Seymour et al.,

2016)11 . Variables needed to compute a SOFA score include patients pulmonary-function ratio
D

(PaO2/FiO2 mmHg), platelet counts (x103/), bilirubin (mg/dL), MAP score or vasopressor requirement,
TE

GCS score, creatinine (mg/dL), and urine output (ml/day) (Singer et al., 2016)1. Each variable is graded
P

on a 0-4 scale with predetermined ranges or medication dosages for each level (Singer et al., 2016)1.
CE

In addition to the SOFA score, Singer et al. (2016) encouraged usage of the qSOFA, a newly
AC

created simplified SOFA score that has been validated for use in non-ICU settings (e.g. in ambulances,

emergency rooms, and medical-surgical floors) (Singer et al., 2016)1. The qSOFA includes only 3

variables: systolic blood pressure (100 mm Hg or <), respirations (22/min or >), and a Glasgow Coma

Scale score (13 or <), but any combination of two variables demonstrated predictive validity similar to

that of the SOFA score when used in non-ICU settings (qSOFA AUROC= 0.81; 95% CI, 0.80-0.82 v. SOFA

in non-ICU settings: AUROC= 0.79, 95% CI, 0.78-0.80) (Singer et al., 2016, p.805)1.

A recent retrospective study of 148,907 patient encounters with suspected infection in

emergency departments, intensive care units, wards, step-down units, or post-anesthesia care units,
 ACCEPTED MANUSCRIPT

derived from 12 hospitals between 2010 and 2012, evaluated the validity of the SOFA, qSOFA, SIRS, and

Logistic Organ Dysfunction System (LODS) criteria (Seymour et al., 2016)11. In ICU encounters with

suspected infection, there was no statistically significant difference between the SOFA and the more

PT
variable-heavy LODS score in terms of predictive validity (Seymour et al., 2016)11. Both tools however

demonstrated stronger predictive validity when compared to the SIRS criteria and qSOFA score

RI
(Seymour et al., 2016) 11. In encounters with suspected infection in non-ICU settings, the predictive

SC
validity of qSOFA was superior to the SOFA and SIRS criteria (Seymour et al., 2016) 11.

NU
Implications on Research and Practice
MA
As with changes in healthcare policy, changes in definitions and clinical criteria carry financial

implications for schools, hospitals, and other healthcare institutions. For over two decades, medical,
D

nursing, allied health programs, and healthcare institutions have trained clinicians to recognize sepsis
TE

using the previous definition based on the SIRS criteria. A change in the definition will necessitate higher

education institutions to change their critical care content and textbooks, and healthcare facilities to
P
CE

spend the money to re-train their clinicians with no guarantee of better outcomes. The SOFA and qSOFA

scores like screening tools based on the SIRS criteria will be sensitive to iatrogenic interventions such as
AC

medications (e.g. sedation may affect blood pressure, heart rate, respirations, and the Glasgow Coma

Scale score; blood pressure medications can affect the blood pressure; and mechanical ventilation may

affect the respiratory rate). Given that there is no guarantee of better outcomes by using the SOFA and

qSOFA scores, as opposed to other adopted early warning tools (e.g. the Modified Early Warning System

score), hospitals may be slow to make a change.

There are however potential financial benefits for hospitals that invest in re-training employees

earlier. Early identification and treatment of sepsis with antibiotics and appropriate fluid resuscitation

has been shown to improve mortality, and is the hallmark of the 2012 Surviving Sepsis Campaign
 ACCEPTED MANUSCRIPT

guidelines (Dellinger et al., 2013)12. For every hour that antibiotics are delayed, mortality goes up by

7.6% (Kumar et al., 2006) 13. The use of the sequential organ failure assessment (SOFA) and qSOFA

scores, may result in earlier identification of septic patients, less resource utilization, and better

PT
outcomes. Improved outcomes spells better reimbursement amounts for hospitals from the Center for

Medicare and Medicaid Services (CMS) and insurance providers. In order to compute a SOFA score

RI
however, additional labs (i.e. PaO2, platelet count, creatinine level, and bilirubin level) must be sent,

SC
meaning that a new operationalized definition will bring with it associated costs for hospitals, insurance

NU
companies, and patients that may negate improvements in reimbursement. The benefits and

disadvantages of using the SOFA and qSOFA need to be investigated, and future research will be
MA
necessary to study the impact of this change on hospital budgets and reimbursement income.

In addition to the implications on clinical practice, there is likely to be a significant impact on

D

sepsis research. Under the previous definition, there were considerable issues with miscoding related to
TE

clinicians misclassification of the level of sepsis. For example, providers would often misclassify a
P

patient as having sepsis when they had severe sepsis, or severe sepsis when they met criteria for
CE

septic shock. By eliminating the severe sepsis category and operationalizing the sepsis and septic
AC

shock definitions, there will likely be less misclassification bias related to provider error, which will

result in more robust epidemiological research studies. There may however continue to be

misclassification issues as inadequately resuscitated patients are started on vasopressors, and coded

according to definition as having septic shock. How this issue will be addressed is yet to be

determined, but should be listed as a possible limitation in published manuscripts pertaining to sepsis.

A major limitation of this change in the short-term is the scarcity of available data for analysis

under the new definition. Routine monitoring of serum bilirubin levels and calculated pulmonary-

function ratios (pf-ratios), 2 data points required to compute a SOFA score, is not currently standard

practice in many intensive care units. Until collecting this data becomes standard, researchers will find it
 ACCEPTED MANUSCRIPT

difficult to study sepsis using the new definition. The utility of studies in progress that collect data

exclusively on the SIRS criteria may be questioned.

A benefit of using the SOFA score in the future will be the additional data available for

PT
researchers to mine. The SOFA score factors in lab values that were not requisite to determine patients

RI
sepsis stage. With this additional data, researchers may be able to perform secondary data analysis

SC
studies that were previously impossible because of insufficient data. In addition, the new CMS sepsis

bundle that must be adopted by hospitals for reimbursement will ensure that more detailed records are

NU
kept in clinical settings.
MA
Conclusion

While the new Sepsis-3 definition appears on the surface to be an improvement over the
D

previous iterations, it remains to be seen if clinical outcomes will improve and if research studies will be
TE

more robust using the new criteria. The Seymour et al. (2016) study demonstrated that the SOFA score
P

is more discriminating that the SIRS criteria in ICU settings, and is less expensive than the LODS score,
CE

which requires additional lab work to be collected for calculation. Additionally the qSOFA score

demonstrated superior predictive validity when compared to the more encompassing SOFA score and
AC

the SIRS criteria in non-ICU settings.

Hospitals and institutions of higher learning will need time to re-train their students to identify

sepsis and septic shock using SOFA and qSOFA. With that said, it may be some time before hospitals

transition away from SIRS, as budget constraints will make it challenging to change their monitoring

tools and re-train their staff. Failure to adopt these tools in a timely manner however may have

important financial implications, and more importantly result in poorer outcomes. Future research on

the impact of this change on research and practice will be essential, to determine if the definition and

associated clinical criteria need further revision.

 ACCEPTED MANUSCRIPT

Conflicts of interest: none

Funding source: This research did not receive any specific grant from funding agencies in the public,

commercial, or not-for-profit sectors

PT
Acknowledgements: I wish to thank Dr. Jeannie Cimiotti, PhD, RN, FAAN for her assistance in

RI
proofreading this manuscript.

SC
NU
MA
D
P TE
CE
AC
 ACCEPTED MANUSCRIPT

References

[1] Singer, M., Deutschman, C.S., Seymour, C.W., Shanker-Hari, M., Annane, D., Bauer, M., Angus, D.C.
The third international consensus definitions for sepsis and septic shock (Sepsis-3). JAMA 2016; 315(8),
801-10. DOI: 10.1001/jama.2016.0287

PT
[2] Bone, R.C., Balk, R.A., Cerra, F.B., Dellinger, R.P., Fein, A.M., Knaus, W.A., Sibbald, W.J. Definitions
for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM
Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care

RI
Medicine. Chest 1992; 101(6), 1644-55. PMID: 1303622
[3] Levy, M.M., Fink, M.P., Marshall, J.C., Abraham, E., Angus, D., Cook, D., Ramsay, G. 2001

SC
SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Intensive Care Med 2003;
29(4), 530-8. DOI: 10.1007/s00134-003-1662-x

NU
[4] Deutschman, C.S. & Tracey, K.J. Sepsis: current dogma and new perspectives. Immunity 2014; 40,
463-475. DOI: 10.1016/j.immuni.2014.04.001
[5] Vincent, J.L., Opal, S.M., Marshall, J.C., & Tracey, K.J. Sepsis definitions: time for a change. Lancet
2013; 381(9868), 774-5. DOI: 10.1016/S0140-6736(12)61815-7
MA
[6] Hotchkiss, R.S., Monneret, G., & Payen, D. Sepsis-induced immunosuppression: from cellular
dysfunctions to immunotherapy. Nat Rev Immunol 2013; 13(12), 862-874. DOI: 10.1038/nri3552
D

[7] Singer, M., De Santis, V., Vitale, D., & Jeffcoate, W. Multiorgan failure is an adaptive, endocrine-
mediated, metabolic response to overwhelming systemic inflammation, Lancet 2004; 364(9433), 545-8.
TE

DOI: 10.1016/S0140-6736(04)16815-3

[8] Kaukonen, K-M., Bailey,M., Pilcher, D., Cooper, D.J., & Bellomo,R. Systemic Inflammatory Response
P

Syndrome criteria in defining severe sepsis. N Engl J Med 2015; 372(17), 1629-1638. DOI:
10.1056/NEJMoa1415236
CE

[9] Sprung, C.L., Sakr, Y., Vincent, J-L., Le Gall, J-R., Reinhart, K., Raneiri, V.M.Payen, D. An evaluation of
systemic inflammatory response syndrome signs in the Sepsis Occurrence in Acutely ill Patients (SOAP)
AC

study. Intensive Care Med 2006; 32, 421-7. DOI: 10.1007/s00134-005-0039-8

[10] Lai, N.A. & Kruger, P. The predictive ability of a weighted systemic inflammatory response syndrome
score for microbiologically confirmed infection in hospitalised patients with suspected sepsis. Crit Care
Resusc 2011; 13(3), 146-150. PMID: 21880000

[11] Seymour, C.W., Liu, V.X., Iwashyna, T.J., Brunkhorst, F.M., Rea, T.D., Scherag, A., Angus, D.C.
Assessment of clinical criteria for sepsis. For the third international consensus definitions for sepsis and
septic shock (Sepsis-3). JAMA 2016; 315(8), 762-774. DOI: 10.1001/jama.2016.0288

[12] Dellinger, R.P., Levy, M.M., Rhodes, A., Annane, D., Gerlach, H., Opal, S.M., and the Surviving
Sepsis Campaign Guidelines Committee including the Pediatric Subgroup. Surviving Sepsis Campaign:
International guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med 2013;
41(2), 580-637. DOI: doi: 10.1007/s00134-012-2769-8

[13] Kumar, A., Roberts, D., Wood, K.E., Light, B., Parrillo, J.E., Sharma, S., Suppes, R., Feinstein, D.,
Zanotti, S., Taiberg, L., Gurka, D., Kumar, A., & Cheang, M. Duration of hypotension before initiation of
 ACCEPTED MANUSCRIPT

effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care
Med 2006; 34(6), 1589-1596. DOI: 10.1097/01.CCM.0000217961.75225.E9

PT
RI
SC
NU
MA
D
P TE
CE
AC