Professional Documents
Culture Documents
Souvik Maitra, MD, DNB, EDIC, Assistant Professor, Anirban Som, MD, Attending
Consultant, Sulagna Bhattacharjee, MD, DNB, Senior Resident
PII: S1198-743X(18)30294-5
DOI: 10.1016/j.cmi.2018.03.032
Reference: CMI 1263
Please cite this article as: Maitra S, Som A, Bhattacharjee S, Accuracy of quick Sequential Organ
Failure Assessment (qSOFA) score and systemic inflammatory response syndrome (SIRS) criteria for
predicting mortality in hospitalized patients with suspected infection: A meta-analysis of observational
studies, Clinical Microbiology and Infection (2018), doi: 10.1016/j.cmi.2018.03.032.
This is a PDF file 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 final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Accuracy of quick Sequential Organ Failure Assessment (qSOFA) score and systemic
inflammatory response syndrome (SIRS) criteria for predicting mortality in hospitalized patients
PT
Souvik Maitra MD, DNB, EDIC1, Anirban Som MD2, Sulagna Bhattacharjee MD, DNB3
1 Assistant Professor, Department of Anaesthesiology, Pain Medicine & Critical Care, All India
Institute of Medical Sciences, New Delhi, India
RI
2 Attending Consultant, Department of Anaesthesiology, MAMBS, Max Super Specialty
Hospital, Saket, New Delhi, India
SC
3 Senior Resident, Department of Anaesthesiology, Pain Medicine & Critical Care, All India
Institute of Medical Sciences, New Delhi, India
U
Correspondence to:
Dr Anirban Som, MD
AN
Attending Consultant,
Department of Anaesthesiology, MAMBS,
Max Super Specialty Hospital
M
Email: anirbansm@gmail.com
Key Words: Sepsis, qSOFA, SIRS, prognosis, mortality
TE
C EP
AC
1
ACCEPTED MANUSCRIPT
Abstract
Objective: To identify sensitivity, specificity and predictive accuracy of quick sequential organ
failure assessment (qSOFA) score and systemic inflammatory response syndrome (SIRS) criteria
PT
Methods: This meta-analysis followed MOOSE consensus statement for conducting and
reporting the results of systematic review. PubMed & EMBASE were searched for the
RI
observational studies which reported predictive utility of qSOFA score for predicting mortality in
SC
patients with suspected or proven infection with the following search words: ‘qSOFA’, ‘q-SOFA’,
‘quick- SOFA’, ‘Quick Sequential Organ Failure Assessment’, ‘quick SOFA’. Sensitivity,
U
specificity, area under receiver operating characteristic curves (ROC) with 95% confidence
interval of qSOFA and SIRS criteria for predicting in-hospital mortality was collected for each
AN
study and a 2x2 table was created for each study.
Results: Data of 406802 patients from 45 observational studies have been included in this meta-
M
analysis. Pooled sensitivity (95% CI) and specificity (95% CI) of qSOFA≥2 for predicting
D
mortality in patients who are not in intensive care unit (ICU) is 0.48(0.41- 0.55) and 0.83(0.78-
TE
0.87) respectively. Pooled sensitivity (95% CI) of qSOFA ≥2 for predicting mortality in patients
(both ICU and non- ICU setting) with suspected infection is 0.56(0.47- 0.65) and pooled
EP
Conclusion: qSOFA has been found to be a poorly sensitive predictive marker for in-hospital
C
developing another scoring system with higher sensitivity to identify high-risk patients with
infection.
2
ACCEPTED MANUSCRIPT
Sepsis and its consequences are a serious burden to the health care and society1. A meta-analysis
published in 2016 found that hospital mortality was 17% and 26% for patients with sepsis and
severe sepsis respectively during the years 1997- 20152. It is well known that early recognition
and timely therapy are the keys to successful outcome in patients with sepsis and septic shock3.
PT
A consensus definition of sepsis was developed in 19924 and again re-examined in 20015.
However, a need for a new definition of sepsis is well recognized, as the understanding of
RI
pathophysiology of sepsis has changed6. The third international definition of sepsis (Sepsis-3)
SC
coined the term quick sequential organ failure assessment (qSOFA) incorporating altered
mentation, systolic blood pressure of 100 mm Hg or less, and respiratory rate of 22/min or
U
greater, which provides simple bedside criteria to identify adult patients with suspected infection
who are likely to have poor outcomes7. This definition also states that patients who are out of the
AN
hospital, in the hospital ward or in the emergency department, tend to have a poor outcome when
any two of the above criteria are present in them. Seymour et al8 reported that presence of any
M
two of the qSOFA criteria predicts in-hospital mortality with an area under the receiver operating
D
characteristics curve (AUROC) of 0.81. This predictive value was similar to the full sequential
TE
organ failure assessment score (SOFA) in the out side intensive care unit (ICU) setting. On the
other hand, the systemic inflammatory response syndrome (SIRS) criteria were inferior to the
EP
SOFA score for predicting mortality in ICU patients with infection.8 Criteria in the qSOFA model
are simple, easy to use and could be used for making the practicing physician alert about the
C
possible poor outcome in such patients. However, to be used as a screening tool for identifying
AC
patients who are at risk of death from infection, qSOFA is expected to be highly sensitive for
Although both the derivation and validation cohorts of qSOFA criteria were from the United
States, since publication of the Sepsis-3 definition, a large number of studies from different
regions of the world have evaluated this score for prognostication in patients with suspected
3
ACCEPTED MANUSCRIPT
infection. Results of the individual studies reported a wide range of predictive accuracy of
qSOFA.
In this systematic review and meta-analysis, we have assessed the pooled predictive value of
PT
Methods
RI
studies in epidemiology) group in conducting and reporting of its results9.
SC
Data source
Two authors (AS & SM) independently searched PubMed EMBASE databases for the studies
U
which reported predictive utility of qSOFA score for predicting mortality in patients with
suspected or proven infection. Following search words were used to identify potentially eligible
AN
studies: ‘qSOFA’, ‘qSOFA sepsis’, ‘q-SOFA’, ‘quick- SOFA’, ‘Quick Sequential Organ Failure
Assessment’, ‘quickSOFA’. The details of Medline search strategy have been provided in the
M
online supplementary material. The initial search was conducted on 30th October 2017 and a
D
subsequent search was conducted on 6th March 2018. We included all studies available till that
TE
date. Observational studies where qSOFA was reported in a manner that it was possible to create
a 2x2 table have been included in this meta-analysis. Initially a pooled analysis has been done
EP
incorporating all the studies where predictive value of qSOFA has been reported. After that,
another pooled analysis has been done incorporating studies, which reported both qSOFA and
C
SIRS criteria for predicting mortality in such patients. References of the studies were also
AC
manually searched for potentially eligible studies. We did not search for the ongoing trials and
authors were not contacted for unpublished data. Studies for which full text was not available in
4
ACCEPTED MANUSCRIPT
Study eligibility
All prospective and retrospective studies that assessed the predictive accuracy of qSOFA for
predicting in-hospital or 30-day or 28-day mortality in adult patients having suspected or proven
infection and reported enough information to formulate a 2x2 contingency table for further
PT
analysis have been included in this meta-analysis.
Study Selection
RI
Two authors (SM and AS) independently read the abstract of the potentially eligible trials. Both
SC
of them selected eligible trials as per inclusion criteria. When a difference in opinion was
encountered, the opinion of the third author (SB) was considered to be final.
U
Data Collection
We collected the required data from the full text of the trials. Initially all data were tabulated in
AN
Microsoft Excel™ spreadsheets (Microsoft Corp., Redmond, WA). AS and SB extracted data
from the eligible studies and SM crosschecked those data independently. Statistical analyses were
M
Data Items
TE
The following data were collected from each study: name of the first author, year of publication,
total number of patients studied, patient population, place of study (emergency department or
EP
intensive care unit), sample size, time point of evaluation, cut off points for qSOFA and SIRS,
sensitivity, specificity, area under receiver operating characteristic curves (AUROC) with 95%
C
confidence interval (CI) for predicting in-hospital mortality. Where in-hospital mortality has not
AC
been reported, 30- day or 28-day mortality has been noted. We also collected data of qSOFA in
non- ICU patients from the studies where it has been reported. In the context of this meta-analysis
about predictive accuracy, sensitivity was defined as the percentage of patients with a positive
index test who died during hospitalization or within 28 or 30 days; while specificity was used to
describe the percentage of patients with a negative index test who survived the period of
5
ACCEPTED MANUSCRIPT
Quality assessment
Methodological qualities of the included studies have been systematically evaluated by QUADAS
2 guidelines10 and summarized in RevMan 5.3 (Review Manager (RevMan) [Computer program].
Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014).
PT
Statistical analysis
A 2x2 table with the four core values, i.e., the numbers of true positives, false positives, true
RI
negatives and false negatives was constructed for all the included studies. Studies, which did not
SC
report these four values, relevant values were calculated from the reported sensitivity, specificity,
and predictive values in Revman 5.3 (Review Manager (RevMan) [Computer program]. Version
U
5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014). Initially a
bivariate random effects regression model for diagnostic meta-analysis was used to obtain
AN
weighted summary estimates of the sensitivity and specificity of qSOFA and SIRS criteria. We
positive and negative likelihood ratios with the midas or metandi command in Stata (STATA SE
D
12.0, STATA Corp, College Station, TX, USA). Pooled diagnostic odds ratio (OR) was also
TE
calculated by random effects model by metandi command. A subgroup analysis was planned to
directly compare the predictive accuracies of qSOFA and SIRS from studies that reported the
EP
Quantification of heterogeneity11 was not done by I2 value, since this parameter is not useful in
C
analysis of accuracy studies. Therefore, reasons for heterogeneity among different studies were
AC
searched for using regression models with following covariates: sample size, clinical setting,
prevalence of disease, age of the patients, cut-off value of the index test. Publication bias was
tested by visual inspection of funnel plot and Deeks’ regression model12, although they are known
6
ACCEPTED MANUSCRIPT
Results
PubMed & EMBASE search as per predefined search words revealed 256 articles and 250 articles
respectively, and references from the relevant articles revealed 57 articles. After removal of
PT
duplications, 112 unique articles were assessed from abstract and full text for possible inclusion
in this meta-analysis. Finally, 56 articles fulfilled our inclusion criteria; whereas a total of 11
RI
articles13-23 were excluded, the details of which have been shown in the flow diagram illustrating
different stages of database searching and study selection as per PRISMA24 (Preferred Reporting
SC
Items for Systematic Reviews and Meta-analyses) checklist in figure 1.
U
Methodological qualities of the included studies have been depicted in e figure 1. As can be seen
from the diagram, most studies have a high risk of bias in patient selection because of the
AN
retrospective nature of their analyses, while an uncertain element of bias is introduced by the risk
applicability of the index tests, whereas the reference standard, i.e. mortality, is beyond any risk
D
of bias. In contrast, the selected patient populations do raise questions about applicability in a few
TE
studies.
Data of 406,802 patients from 45 observational studies8, 24-67 have been included in this meta-
EP
analysis where predictive ability of qSOFA for in-hospital/ 30-day mortality has been reported.
Primarily we collected data of in-hospital mortality; whereas 30-day or 28-day mortality was
C
predictive accuracy of qSOFA. Among these, in 13 studies, predictive ability of both qSOFA and
SIRS criteria for in-hospital mortality has been reported. Individual study characteristics have
been provided in e table 1 [online supplement]. All the studies have been published between 2016
and 2018. We considered any study, which had been designed or commenced before the
publication of Sepsis-3 definition7 (February 2016) as a retrospective cohort study. Only four
7
ACCEPTED MANUSCRIPT
studies included here are prospective in design32,45,59, 64. Reported predictive accuracy of qSOFA
for predicting in-hospital mortality/30-day mortality in the included studies has been depicted in e
PT
Data were available of 199705 patients from 38 studies regarding predictive accuracy of qSOFA
to predict in-hospital mortality. Pooled sensitivity (95% CI) and specificity (95% CI) of
RI
qSOFA≥2 for predicting mortality in non-ICU patients are 0.48(0.41- 0.55) and 0.83(0.78-
SC
0.87) respectively. A summary ROC curve has been provided in figure 2 and AUC with 95% CI
was found to be 0.71(0.67- 0.75). Forests plot for sensitivity and specificity at study level and
U
pooled analysis level has been provided in e figure 2.
AN
A meta-regression analysis did not find any relation between baseline risks of in-hospital
mortality (p=0.05) and sample size (p=0.58). However, mean age of the included patients has
been found to be a significant co-variate for predictive accuracy of qSOFA (I2=99%, p<0.01, e
M
figure 3). When the probability of death from suspected infection is 20%, a qSOFA ≥2 increases
D
possibility of death to 41% [Fagan’s nomogram, e figure 4]. No evidence of publication bias was
TE
found in Deek’s regression model (p=0.76) [e figure 5], although it is important to stress that this
When patients from both ICU and non- ICU settings are included in the meta-analysis, pooled
sensitivity (95% CI) of qSOFA ≥2 for predicting mortality in patients with suspected infection is
C
0.56(0.47- 0.65) and pooled specificity (95% CI) is 0.78(0.71-0.83). A summary ROC curve
AC
has been provided in e figure 6 and AUC with 95% CI was found to be 0.73(0.69- 0.77).
Twenty-one studies reported predictive accuracy of both qSOFA and SIRS criteria for predicting
in-hospital mortality. Pooled sensitivity (95% CI) of SIRS ≥2 for predicting mortality in
hospitalized patients with suspected infection is 0.82(0.73- 0.89) and pooled specificity (95%
8
ACCEPTED MANUSCRIPT
CI) is 0.36(0.24- 0.50) [figure 3]. Diagnostic OR (95% CI) of SIRS is 2.59 (2.09- 3.22) with a
positive LR (95% CI) of 1.30 (1.10- 1.50) for predicting in-hospital mortality. A summary of
Considering the studies where both qSOFA and SIRS were reported, pooled sensitivity (95% CI)
PT
of qSOFA ≥2 for predicting mortality in hospitalized patients with suspected infection is 0.60
(0.45- 0.73) and pooled specificity (95% CI) is 0.76 (0.64- 0.84) [figure 4]. In all summary ROC
RI
curves we have found a large area of 95% prediction interval, which is a measure of between-
SC
study heterogeneity. So that, the prediction interval can predict the sensitivity and specificity of
future studies.
U
Discussion AN
This meta-analysis of observational studies illustrates that qSOFA is a less sensitive predictive
marker than SIRS criteria for in-hospital mortality in hospitalized patients with suspected
M
infection. However, qSOFA is more specific than SIRS criteria in this regard. The most important
strength of this meta-analysis is its large number of patients’ database from various geographical
D
Global burden of sepsis is increasing worldwide with estimated load up to 560 cases per 100000
persons per year2. It is also the single most expensive health care condition in the United States
EP
accounting for 5.2% of total health care expenditure69. The Third International Consensus
dysregulated host response to infection” on the basis of updated knowledge about the
AC
pathophysiology of sepsis. As early diagnosis and management are the keys to successful
outcome in sepsis patients, Third International Consensus Definitions Task Force developed an
easy and reproducible criteria termed qSOFA to identify patients with suspected infection who
would require escalation of therapy and monitoring. Seymour et al8 found that qSOFA had a
predictive accuracy equivalent to that of the full SOFA score for in-hospital mortality in non-
9
ACCEPTED MANUSCRIPT
ICU patients. However, in patients who are in ICU, full SOFA score had better predictive validity
than qSOFA7.
Interestingly, this meta-analysis has found that qSOFA has a poorer sensitivity for predicting
mortality even in hospitalized patients who are not in ICU. As most of the ICU patients receive
PT
mechanical respiratory support and/or vasopressors, qSOFA is expected to be of little actual
clinical value in this setting. We have found an AUC of summary ROC curve of 0.72 and
RI
included studies in this meta-analysis reported a wide range of AUCROC of 0.55 to 0.82. But
SC
from a clinical point of view, AUCROC is of little value in this scenario, because the aim of
formulating qSOFA was to develop it as a screening tool. As a screening tool for identification of
high-risk patients, it is desirable to have at least a high sensitivity70. We have found that pooled
U
specificity of qSOFA is 0.79 for all patients, which is reasonably good.
AN
This meta-analysis included a large number of patients from different settings and geographical
areas; different hospitals have variable quality of care that would obviously have affected
M
patients’ ultimate outcome. This has been statistically reflected in this meta-analysis in terms of a
D
significant amount of heterogeneity. We have also found that age of the included patients is a
TE
As qSOFA is a simple criterion and does not include any laboratory investigation, its widespread
EP
use in different settings including emergency departments and pre-hospital areas is feasible.
Although blood lactate is a well-known prognostic marker in patients with sepsis71; benefits of
C
adding blood lactate to the original qSOFA is of questionable value8,33. The current Sepsis-3
AC
definition recommends that qSOFA can be used to identify patients who are at risk of death or
prolonged ICU stay72. However, this meta-analysis did not find any evidence that it can be used
to identify patients who are at risk of in-hospital death, since the overall sensitivity is only 0.54.
Being a predictive tool designed to predict mortality, the score should have a reasonably high
level of sensitivity, which qSOFA evidently lacks. As qSOFA is not a diagnostic criterion to
10
ACCEPTED MANUSCRIPT
On the contrary, the SIRS criteria, which have been criticized for a long time, have been found to
be very non-specific for predicting mortality in our analysis42. The SIRS criteria, which were
originally developed in 19924, have been found to be a predictor of mortality in various infectious
and non- infectious conditions such as trauma73 and acute myocardial infraction74. Presence of
PT
SIRS criteria denotes activation of innate immunity system irrespective of its cause and both
infectious and non-infectious triggers can activate immunologic response. As there is a wide
RI
range of activators of the innate immunity system, the SIRS criteria are known to be non-specific
SC
as marker of sepsis.
U
Limitations
The most important limitation of our analysis is the presence of large amount of heterogeneity in
AN
this meta- analysis. Though we have conducted a meta-regression analysis, we could not identify
any factor associated with heterogeneity except for mean age of the patient cohorts. It’s obvious
M
that as we have included a large number of patients from different settings, a significant amount
D
of heterogeneity will always exist. In the absence of individual patient data, it was impossible to
TE
adjust the various pre existing clinical risk factors of death in these patients. Most of the included
studies in this meta-analysis are retrospective reviews of patients’ database and they are not
EP
We analyzed predictive accuracy of qSOFA and SIRS criteria from different dataset. A meta-
C
regression analysis could have taken into account of qSOFA and SIRS from same study.
AC
Conclusion
In this meta-analysis of observational studies, qSOFA has been found to be a poorly sensitive
predictive marker for in-hospital mortality in hospitalized patients with suspected infection. As
the original aim of developing qSOFA score was to identify patients who are at risk of death or
11
ACCEPTED MANUSCRIPT
prolonged ICU stay, it is reasonable to recommend developing another scoring system with
higher sensitivity.
Author Contributions:
Drs Souvik Maitra and Anirban Som had full access to all of the data in the study and take
PT
responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: Souvik Maitra and Anirban Som.
Drafting of the manuscript: Sulagna Bhattacharjee and Souvik Maitra
RI
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Souvik Maitra and Anirban Som
Obtained funding: Not applicable
SC
Administrative, technical, or material support: Literature search was done at BB Dikshit
Library, All India Institute of Medical Sciences, New Delhi, India
Study supervision: Souvik Maitra
Conflict of interests: None
U
Source of support: Nil
AN
Figure legends
M
Figure 1: Flow diagram showing stages of database search and study selection
Figure 2: Summary ROC curve (bivariate model) for pooled analysis of sensitivity and specificity
D
of qSOFA≥2 for predicting in-hospital mortality [Data from non-ICU patients included here]
TE
Figure 3: Summary ROC curve (bivariate model) for pooled analysis of sensitivity and specificity
Figure 4: Summary ROC curve (bivariate model) for pooled analysis of sensitivity and specificity
of qSOFA≥2 for predicting in-hospital mortality [Data from the studies where predictive accuracy
C
Table legends
Table 1: Characteristics of q-SOFA and SIRS criteria for predicting mortality in hospitalized
12
ACCEPTED MANUSCRIPT
PT
RI
References:
1. Liu V, Escobar GJ, Greene JD, Soule J, Whippy A, Angus DC, et al. Hospital deaths in
SC
patients with sepsis from 2 independent cohorts. JAMA. 2014;312:90-2.
U
International Forum of Acute Care Trialists. Assessment of Global Incidence and
AN
Mortality of Hospital-treated Sepsis. Current Estimates and Limitations. Am J Respir
3. Kumar A, Roberts D, Wood KE, Light B, Parrillo JE, Sharma S, et al. Duration of
Conference: definitions for sepsis and organ failure and guidelines for the use of
EP
5. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, et al; International
C
6. Vincent JL, Opal SM, Marshall JC, Tracey KJ. Sepsis definitions: time for change.
Lancet. 2013;381:774-5.
13
ACCEPTED MANUSCRIPT
The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3).
JAMA. 2016;315:801-10.
8. Seymour CW, Liu VX, Iwashyna TJ, Brunkhorst FM, Rea TD, Scherag A, et al.
PT
Assessment of Clinical Criteria for Sepsis: For the Third International Consensus
RI
9. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-
SC
analysis of observational studies in epidemiology: a proposal for reporting. Meta-
U
2000;283:2008-12.
10. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al;
AN
QUADAS-2 Group. QUADAS-2: a revised tool for the quality assessment of diagnostic
11. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med.
D
2002;21:1539-58.
TE
12. Deeks JJ, Macaskill P, Irwig L. The performance of tests of publication bias and other
sample size effects in systematic reviews of diagnostic test accuracy was assessed. J
EP
13. Wang S, Li T, Li Y, Zhang J, Dai X. [Predictive value of four different scoring systems
C
for septic patient's outcome: a retrospective analysis with 311 patients]. Zhonghua Wei
AC
14. Wang Y, Wang D, Fu J, Liu Y. [Predictive value of SOFA, qSOFA score and traditional
evaluation index on sepsis prognosis]. Zhonghua Wei Zhong Bing Ji Jiu Yi Xue.
2017;29:700-704.
14
ACCEPTED MANUSCRIPT
15. Siddiqui S, Chua M, Kumaresh V, Choo R. A comparison of pre ICU admission SIRS,
EWS and q SOFA scores for predicting mortality and length of stay in ICU. J Crit Care.
2017;41:191-193.
16. April MD, Aguirre J, Tannenbaum LI, Moore T, Pingree A, Thaxton RE, et al. Sepsis
PT
Clinical Criteria in Emergency Department Patients Admitted to an Intensive Care Unit:
RI
Med. 2017;52:622-631.
SC
17. Innocenti F, Tozzi C, Donnini C, De Villa E, Counts A, Zanobetti M, et al. SOFA score
in septic patients: incremental prognostic value over age, comorbidities and parameters
U
of sepsis severity. Intern Emerg Med. 2017 Feb 10. [Epub ahead of print]
18. Dorsett M, Kroll M, Smith CS, Asaro P, Liang SY, Moy HP. qSOFA Has Poor
AN
Sensitivity for Prehospital Identification of Severe Sepsis and Septic Shock. Prehosp
Procalcitonin and Mr-Proadrenomedulin Combination with Sofa and Qsofa Scores for
TE
Sepsis Diagnosis and Prognosis: A Diagnostic Algorithm. Shock. 2017 Oct 11. [Epub
ahead of print]
EP
20. Patidar KR, Shaw J, Acharya C, Thacker LR, White MB, Ganapathy D, et al. No
21. Sager R, Wirz Y, Amin D, Amin A, Hausfater P, Huber A, et al. Are admission
15
ACCEPTED MANUSCRIPT
22. Jawa RS, Vosswinkel JA, McCormack JE, Huang EC, Thode HC Jr, Shapiro MJ, Singer
AJ. Risk assessment of the blunt trauma victim: The role of the quick Sequential Organ
23. Kwak H, Suh GJ, Kim T, Kwon WY, Kim KS, Jung YS, Ko JI, Shin SM. Prognostic
PT
performance of Emergency Severity Index (ESI) combined with qSOFA score. Am J
RI
24. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, et al. The
SC
PRISMA statement for reporting systematic reviews and meta-analyses of studies that
U
2009;6:e1000100.
25. Donnelly JP, Safford MM, Shapiro NI, Baddley JW, Wang HE. Application of the
AN
Third International Consensus Definitions for Sepsis (Sepsis-3) Classification: a
26. Chen YX, Wang JY, Guo SB. Use of CRB-65 and quick Sepsis-related Organ Failure
D
Assessment to predict site of care and mortality in pneumonia patients in the emergency
TE
27. Wang JY, Chen YX, Guo SB, Mei X, Yang P. Predictive performance of quick Sepsis-
EP
related Organ Failure Assessment for mortality and ICU admission in patients with
28. Ho KM, Lan NS. Combining quick Sequential Organ Failure Assessment with plasma
AC
score in predicting mortality of patients with and without suspected infection. J Crit
Care. 2017;38:1-5.
al; Hellenic Sepsis Study Group. Validation of the new Sepsis-3 definitions: proposal
16
ACCEPTED MANUSCRIPT
30. Williams JM, Greenslade JH, McKenzie JV, Chu K, Brown AFT, Lipman J. Systemic
PT
31. Singer AJ, Ng J, Thode HC Jr, Spiegel R, Weingart S. Quick SOFA Scores Predict
RI
Infection. Ann Emerg Med. 2017;69:475-479.
SC
32. Raith EP, Udy AA, Bailey M, McGloughlin S, MacIsaac C, Bellomo R, et al; Australian
and New Zealand Intensive Care Society (ANZICS) Centre for Outcomes and Resource
U
Evaluation (CORE). Prognostic Accuracy of the SOFA Score, SIRS Criteria, and
qSOFA Score for In-Hospital Mortality Among Adults With Suspected Infection
AN
Admitted to the Intensive Care Unit. JAMA. 2017;317:290-300.
33. Freund Y, Lemachatti N, Krastinova E, Van Laer M, Claessens YE, Avondo A, et al;
M
Sepsis-3 Criteria for In-Hospital Mortality Among Patients With Suspected Infection
TE
34. Burnham JP, Kollef MH. qSOFA score: Predictive validity in Enterobacteriaceae
EP
of Sepsis-3 criteria and quick SOFA in patients with cirrhosis and bacterial infections.
AC
patients with sepsis outside ICU: A post hoc subgroup analysis by the Japanese
Association for Acute Medicine Sepsis Registry Study Group. J Infect Chemother.
2017;23:757-762.
17
ACCEPTED MANUSCRIPT
37. Amland RC, Sutariya BB. Quick Sequential [Sepsis-Related] Organ Failure Assessment
(qSOFA) and St. John Sepsis Surveillance Agent to Detect Patients at Risk of Sepsis:
38. Kim MW, Lim JY, Oh SH. Mortality prediction using serum biomarkers and various
PT
clinical risk scales in community-acquired pneumonia. Scand J Clin Lab Invest.
2017;77:486-492.
RI
39. Park HK, Kim WY, Kim MC, Jung W, Ko BS. Quick sequential organ failure
SC
assessment compared to systemic inflammatory response syndrome for predicting sepsis
U
40. Ranzani OT, Prina E, Menéndez R, Ceccato A, Cilloniz C, Méndez R, et al. New Sepsis
41. Askim Å, Moser F, Gustad LT, Stene H, Gundersen M, Åsvold BO, et al. Poor
M
42. Churpek MM, Snyder A, Han X, Sokol S, Pettit N, Howell MD, et al. Quick Sepsis-
EP
Early Warning Scores for Detecting Clinical Deterioration in Infected Patients outside
C
43. de Groot B, Stolwijk F, Warmerdam M, Lucke JA, Singh GK, Abbas M, et al. The most
commonly used disease severity scores are inappropriate for risk stratification of older
44. Moskowitz A, Patel PV, Grossestreuer AV, Chase M, Shapiro NI, Berg K, et al; Center
for Resuscitation Science. Quick Sequential Organ Failure Assessment and Systemic
18
ACCEPTED MANUSCRIPT
45. Huson MAM, Katete C, Chunda L, Ngoma J, Wallrauch C, Heller T, van der Poll T, et
al. Application of the qSOFA score to predict mortality in patients with suspected
PT
infection in a resource-limited setting in Malawi. Infection. 2017;45:893-896.
46. Haydar S, Spanier M, Weems P, Wood S, Strout T. Comparison of QSOFA score and
RI
SIRS criteria as screening mechanisms for emergency department sepsis. Am J Emerg
SC
Med. 2017;35:1730-1733.
U
Septic Patients Using qSOFA Criteria at the SAMU Regulation: A Retrospective
qSOFA Criteria for Predicting 28-Day Mortality in Critically Ill Septic Patients During
M
the Early Period After Emergency Department Presentation. Ann Emerg Med.
D
2018;71:1-9
TE
49. Kim M, Ahn S, Kim WY, Sohn CH, Seo DW, Lee YS, et al. Predictive performance of
the quick Sequential Organ Failure Assessment score as a screening tool for sepsis,
EP
mortality, and intensive care unit admission in patients with febrile neutropenia. Support
50. Quinten VM, van Meurs M, Wolffensperger AE, Ter Maaten JC, Ligtenberg JJM.
AC
Sepsis patients in the emergency department: stratification using the Clinical Impression
Sequential Organ Failure Assessment score? Eur J Emerg Med. 2017 May 8. [Epub
ahead of print]
19
ACCEPTED MANUSCRIPT
in 497 cases with blood culture-positive sepsis in an emergency department. Int J Infect
Dis. 2017;58:52-57.
PT
J, Piñera P, Navarro-Bustos C, et al; representing the Infectious Disease Group of the
RI
criteria, qSOFA score and GYM score for 30-day-mortality in older non-severely
SC
dependent infected patients attended in the emergency department. Eur J Clin Microbiol
U
53. Forward E, Konecny P, Burston J, Adhikari S, Doolan H, Jensen T. Predictive validity
of the qSOFA criteria for sepsis in non-ICU inpatients. Intensive Care Med.
AN
2017;43:945-946.
54. Henning DJ, Puskarich MA, Self WH, Howell MD, Donnino MW, Yealy DM, et al. An
M
Emergency Department Validation of the SEP-3 Sepsis and Septic Shock Definitions
D
and Comparison With 1992 Consensus Definitions. Ann Emerg Med. 2017;70:544-
TE
552.e5.
55. Huson MA, Kalkman R, Grobusch MP, van der Poll T. Predictive value of the qSOFA
EP
score in patients with suspected infection in a resource limited setting in Gabon. Travel
56. Moore CC, Hazard R, Saulters KJ, Ainsworth J, Adakun SA, Amir A, et al. Derivation
AC
and validation of a universal vital assessment (UVA) score: a tool for predicting
2017;2:e000344.
57. Peake SL, Delaney A, Bailey M, Bellomo R; ARISE Investigators. Potential Impact of
the 2016 Consensus Definitions of Sepsis and Septic Shock on Future Sepsis Research.
20
ACCEPTED MANUSCRIPT
58. Guirgis FW, Puskarich MA, Smotherman C, Sterling SA, Gautam S, Moore FA, Jones
AE. Development of a Simple Sequential Organ Failure Assessment Score for Risk
PT
59. Szakmany T, Pugh R, Kopczynska M, Lundin RM, Sharif B, Morgan P, et al; Welsh
Digital Data Collection Platform collaborators. Defining sepsis on the wards: results of
RI
a multi-centre point-prevalence study comparing two sepsis definitions. Anaesthesia.
SC
2018;73:195-204.
60. Goulden R, Hoyle MC, Monis J, Railton D, Riley V, Martin P, Martina R, Nsutebu E.
U
qSOFA, SIRS and NEWS for predicting inhospital mortality and ICU admission in
emergency admissions treated as sepsis. Emerg Med J. 2018 Feb 21.[Epub ahead of
AN
print]
61. Müller M, Guignard V, Schefold JC, Leichtle AB, Exadaktylos AK, Pfortmueller CA.
M
62. Costa RT, Nassar AP Jr, Caruso P. Accuracy of SOFA, qSOFA, and SIRS scores for
mortality in cancer patients admitted to an intensive care unit with suspected infection. J
EP
of SOFA, qSOFA and SIRS for predicting mortality and organ failure among sepsis
AC
patients admitted to the intensive care unit in a middle-income country. J Crit Care.
2018;44:156-160.
Castillo J. Prognostic accuracy of qsofa according to the site of infection in older patient
attended in Emergency Department. Eur J Intern Med. 2018 Jan 4. [Epub ahead of print]
21
ACCEPTED MANUSCRIPT
SOFA for predicting mortality in patients with acute pyelonephritis associated with
upper urinary tract calculi. J Urol. 2017 Dec 29.[Epub ahead of print]
PT
Yamagishi Y, Suematsu H, Mikamo H. Could qSOFA and SOFA score be correctly
RI
2018;24:228-231.
SC
67. LeGuen M, Ballueer Y, McKay R, Eastwood G, Bellomo R, Jones D; Austin Health
RRT qSOFA investigators. Frequency and significance of qSOFA criteria during adult
U
rapid response team reviews: A prospective cohort study. Resuscitation. 2018;122:13-
18.
AN
68. Finkelsztein EJ, Jones DS, Ma KC, Pabón MA, Delgado T, Nakahira K, et al.
M
Comparison of qSOFA and SIRS for predicting adverse outcomes of patients with
suspicion of sepsis outside the intensive care unit. Crit Care. 2017;21:73.
D
69. Torio CM, Andrews RM. National inpatient hospital costs: the most expensive
TE
conditions by payer, 2011. Statistical Brief #160. Healthcare Cost and Utilization
70. Goetzinger KR, Odibo AO. Statistical analysis and interpretation of prenatal diagnostic
imaging studies, Part 1: evaluating the efficiency of screening and diagnostic tests. J
C
71. Filho RR, Rocha LL, Corrêa TD, Pessoa CM, Colombo G, Assuncao MS. Blood
72. Shankar-Hari M, Phillips GS, Levy ML, Seymour CW, Liu VX, Deutschman CS, et al;
Sepsis Definitions Task Force. Developing a New Definition and Assessing New
22
ACCEPTED MANUSCRIPT
Clinical Criteria for Septic Shock: For the Third International Consensus Definitions for
PT
Care. 2018;28:110-116.
RI
74. Huang WC, Chou RH, Chang CC, Hsu CY, Ku YC, Huang HF, et al. Systemic
SC
in Patients with Acute Myocardial Infarction. Acta Cardiol Sin. 2017;33:477-485
U
AN
M
D
TE
C EP
AC
23
ACCEPTED MANUSCRIPT
Table 1: Characteristics of q-SOFA and SIRS criteria for predicting mortality in hospitalized
Performance q-SOFA≥2
SOFA≥2 q-SOFA≥2
SOFA≥2a q-SOFA≥2
SOFA≥2b SIRS≥2
SIRS≥2
PT
of the score (n=406,802) (n=207,708) (n=350,164) (n=348,980)
RI
0.65) 0.55) 0.73)
SC
Specificity 0.78(0.71- 0.83(0.78- 0.76(0.64- 0.36(0.24- 0.50)
U
Diagnostic 2.5(2.1-2.9) 4.36(3.72- 4.63(3.57- 2.59 (2.09- 3.22)
AN
Odds ratio
5.13) 6.00)
summary ROC
0.77) 0.75) 0.78)
ratio (+)
2.92) 3.05)
TE
a Data from only non-ICU patients have been included in this analysis
C
b Studies where both prognostic accuracies of q-SOFA and SIRS have been reported from
AC
Identification
PT
(n = 506) (n=57)
RI
Records after duplicates removed
(n =112)
SC
Screening
U
Records screened Records excluded
(n =112) (n = 56)
AN
M
Full-text articles assessed n=11 full-text articles
for eligibility excluded because,
Eligibility
Studies included in
C
quantitative synthesis
(meta-analysis)
AC
(n = 45)
ACCEPTED MANUSCRIPT
PT
RI
SC
SROC with Prediction & Confidence Contours
1.0
U 18
AN
27
29
M
36
17 28
30
D
2 25
TE
22
8 38
Sensitivity
23
3
9
10
EP
2611
0.5 4
12
37
35
14
C
5 19
16
31
3320
AC
15
34 Observed Data
24
Summary Operating Point
6 SENS = 0.48 [0.41 − 0.55]
13 32
SPEC = 0.83 [0.78 − 0.87]
21
SROC Curve
AUC = 0.71 [0.67 − 0.75]
71 95% Confidence Contour
0.0
1.0 0.5 0.0
Specificity
ACCEPTED MANUSCRIPT
PT
RI
SC
1.0 20
U
11
16 23
7
AN
9
14 15
10
18
M
6
5
4
17 13
D
19
1 12
TE
Sensitivity
EP
8
0.5
C
AC
Observed Data
0.0
1.0 0.5 0.0
Specificity
ACCEPTED MANUSCRIPT
PT
RI
SC
1.0 20
U 11 19
AN
3
M
164
D
2 14
7
TE
Sensitivity
EP
15 8
0.5 5
9
C
6 12
10
18
AC
21 Observed Data
13
Summary Operating Point
17 SENS = 0.60 [0.46 − 0.73]
SPEC = 0.76 [0.64 − 0.84]
SROC Curve
AUC = 0.74 [0.70 − 0.78]
1 95% Confidence Contour
0.0
1.0 0.5 0.0
Specificity