A talk by Derek Angus at the 2017 meeting of the Scandinavian Society of Anaestesiology and Intensive Care Medicine. All available content from SSAI2017: https://scanfoam.org/ssai2017/ Content delivered in collaboration between scanFOAM, SSAI & SFAI.

1. The Third International
Consensus Definitions for
Sepsis and Septic Shock
(Sepsis-3)
Derek C Angus, MD MPH
The CRISMA Center
Department of Critical Care Medicine
University of Pittsburgh

2. Why?

3. Issues with the 1991 and 2001 Definitions
 SIRS – based
 “Severe Sepsis”
 Different criteria
yielding different results

4. Increased Understanding of Sepsis Pathobiology
 More than just rampant inflammation
 Key role of immunosuppression
 Contribution of non-immune mechanisms
 Possible adaptive nature of organ dysfunction – hibernation
 Re-appraisal of the nature of septic shock

5. Singer M, Deutschman CS,
Seymour CW, Shankar-Hari M et al.
Third International Consensus
Definitions for Sepsis and
Septic Shock (Sepsis-3)
JAMA 2016; 315: 801-10
The Document

6. 1. Beyond the remit of the task force to define infection
2. Sepsis is not simply infection + two or more SIRS criteria
3. The host response is of key importance
4. Sepsis represents bad infection where
bad = infection leading to organ dysfunction
5. “Severe sepsis” is not helpful and should be eliminated
CONSENSUS
Task Force Decisions

7. Sepsis is life-threatening organ dysfunction caused
by a dysregulated host response to infection
The Definition of Sepsis

8. Sepsis is life-threatening organ dysfunction
caused by a dysregulated host response to infection
The Definition of Sepsis
Key Distinctions
So … “sepsis” now = the old “severe sepsis”

9. Sepsis is life-threatening organ dysfunction caused
by a dysregulated host response to infection
The Definition of Sepsis
Key Distinctions
As opposed to the
“regulated host response”
that characterizes the non-septic response to infection

10. Seymour CW, Liu VX, Iwashyna TJ et
al.
Assessment of Clinical Criteria for
Sepsis
For the Third International Consensus
Definitions for Sepsis and Septic Shock
(Sepsis-3)
JAMA, 2016, 315 (8)

11. A life threatening organ dysfunction due to a
dysregulated host response to
infection.
What is sepsis?
1 2
who fares badly?
Among encounters with
suspected infection,

12. How to define ‘fares badly’?
There is no gold standard for
sepsis
Infected
Septic
 “Fares badly” is a proxy
 ‘Bad outcomes’ (e.g., death) are more
common among infected patients who
are septic than those who are not
Infected
Fares
badly

13. We did not..
 Study criteria for infection
 Build an alert or sniffer among non-infected patients
We did..
Infected Denominator (“cohort”)
‘Fares
badly’
Cohort
Cases

14. Our challenges
 What data to use?
 How to identify infection?
 What clinical criteria to study?
 How to define ‘fares badly’?

15. What data source to use?

16. External datasets
>700,000 encounters
170 academic, community hospitals in rural-urban locale
Prehospital, ED, ward
Community and hospital-acquired infections

17. How to identify infection?
Used electronic health records
First episode of cultures and antibiotics
 Excluded prophylactic antibiotics, intra-operative
Determined when infection first suspected
Time of infection
suspected
Admission to ED
Transfer to ICU
Discharge
Time window for candidate criteria
72 hrs6 hrs

18. What clinical criteria to study?

19. (Bad) outcomes more common in sepsis
…
 Clinical review committees
 Death in the hospital
 Prolonged stay in the ICU
 Discharge diagnosis of sepsis
 Positive microbiologic cultures
Infected
Really
sick

20. Patients in primary cohort
Variables Statistic
Total encounters 148,907
Confirmed bacteremia 6,875 (5)
Age, mean (SD) 61 (19)
Male, no. (%) 63,311 (43)
Onset of infection within 48 hrs, no. (%) 128,358 (86)
Location when infection suspected, no. (%)
Emergency department 65,934 (44)
Ward 49354 (33)
Intensive care 15,768 (11)

21. Distribution of existing criteria
20,130
22,357
15,852
6,777
1,406
0
10
20
30
40
50
SIRS criteria
Proportion(%)
245
458
627
778
873
862
803
692
588
467
397
310
251
170
135
102
174
0
10
20
30
40
50
SOFA score
329
604
377
696
925
682
774
868
552
549
488
287
283
164
354
0
10
20
30
40
50
Logistic organ dysfunction score
27,56016,0679,6905,0133,3261,9921,047
625
418
238
163
111
90
72
41
20
31
0
10
20
30
40
50
SOFA score
Outside the ICU (N=66,522)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
29,78915,0854,4297,8784,4261,6001,088
788
442
366
211
128
118
52
122
0
10
20
30
40
50
Logistic organ dysfunction score
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 160 1 2 3 4
261
996
2,454
2,857
1,364
0
10
20
30
40
50
SIRS criteria
Proportion(%)
0 1 2 3 4
No. of
patients
No. of
patients
ICU only (N=7,932)
These criteria are complex
and require laboratory tests

22. Developing new criteria
Focus on timeliness, ease of use
Studied 21 variables from Sepsis-2
Multivariable logistic regression for in-hospital mortality
Respiratory rate ≥ 22 bpm
Altered mentation
Systolic blood pressure ≤ 100 mmHg

23. Assessment of criteria
0.64 (0.62,
0.66)
<0.01
0.74 (0.73,
0.76)
<0.01 0.20
0.75 (0.73,
0.76)
0.01 <0.01 <0.01
0.66 (0.64,
0.68)
SIRS
SOFA
LODS
qSOFA
ICU encounters
N = 7,932
AUROC in-hospital
mortality
0.76 (0.75,
0.77)
<0.01
0.79 (0.78,
0.80)
<0.01 <0.01
0.81 (0.80,
0.82)
<0.01 <0.01 0.72
0.81 (0.80,
0.82)
SIRS
SOFA
LODS
Outside the ICU encounters
N = 66,522
AUROC in-hospital
mortality
qSOFA
SOFA and LODS superior
in the ICU
qSOFA similar to complex
scores outside the ICU

24. Assessment of criteria
1 2 3 4 5 6 7 8 9 10
0.1
1
10
100
1000
10000
Decile of baseline risk for in-hospital mortality
Foldchange,in-hospitalmortality
SIRS ³2 vs. SIRS <2
SOFA ³2 vs. SOFA <2
LODS ³2 vs. LODS <2
qSOFA ³2 vs. qSOFA <2
Baseline risk (%)
Median
Minimum
Maximum
0.2 0.8 1.4 1.9 2.4 2.9 3.4 3.9 4.6 6.2
0.2 0.6 1.1 1.6 2.1 2.6 3.1 3.6 4.2 5.2
0.6 1.0 1.6 2.1 2.6 3.1 3.6 4.2 5.2 17.7
qSOFA similar to complex
scores outside the ICU
Outside the ICU encounters
N = 66,522
Decile of baseline risk of in-hospital
mortality

25. qSOFA in external datasets
Adequate predictive validity (AUC range 0.7 to 0.8)
 Hospital acquired infections
 Ward and ICU encounters
 Prehospital records

26.  Not retained during qSOFA model build
 Serum lactate at various thresholds added to qSOFA
1 2 3 4 5 6 7 8 9 10
0.1
1
10
100
Decile of baseline risk for in-hospital mortality
Foldchange,in-hospitalmortality
qSOFA ³2 vs. qSOFA <2
(qSOFA + serum lactate) ³2
vs. (qSOFA + lactate) <2
Baselinerisk (%)
Median
Minimum
Maximum
1.0 1.9 2.7 3.4 4.2 5.0 5.8 6.7 7.9 10.4
0.6 1.4 2.3 3.0 3.8 4.6 5.4 6.2 7.3 8.9
1.4 2.3 3.0 3.8 4.6 5.4 6.3 7.3 8.9 30.3
Baseline risk
qSOFA + baseline
qSOFA + lactate ≥2.0
+ baseline
1- Specificity
Sensitivity
All KPNC encounters
N = 321,380
Sensitivity
1 - specificity
Decile of baseline risk of in-hospital
mortality
Serum lactate

27. Conclusions
 In the ICU, the SOFA and LODS have greater predictive
validity than qSOFA or SIRS
 Outside the ICU, the qSOFA has similar predictive
validity to more complex scores
Please visit www.qsofa.org

28. Caveats, implications, and next steps …
Retrospective, using charted info in EHR
 Largely in US healthcare setting
Finds ‘who does badly’ among infected patients
 A proxy for ‘septic’
No blood tests
If it were to hold up …
 ‘Prompt’ for more careful monitoring, transfer to ICU, etc.
But, like all RRT prompts, unclear if ‘activation’ improves
outcome …

29. qSOFA over time …Deaths
N=1,769
0 2 3
Survivors
N=1,769
A
B
qSOFA
0
1
2
3
missing
Time after onset of suspected infection, hours, in 6-hour epochs
(0 to 48 hours in 6-hour epochs)
Initial qSOFA
0 24 48
1
0 24 48 0 24 48 0 24 48
InitialqSO
FA
M
ax
qSO
FA
at24h
M
ax
qSO
FA
at48h
M
ean
qSO
FA
at48h
C
rude
trajectory
groups
0.5
0.6
0.7
0.8
0.9
1.0
qSOFA measure added to baseline model
AUROCforin-hospitalmortality
B

30. Initial qSOFA
0 2 of 3 1% mortality
1 1 in 4
¾ stay low <4% mortality
¼ climb to 2+ >16% mortality
2+ 1 in 14 >20% mortality
Kievlan et al (under review)

31. qSOFA vs. SIRS in LMICs
Rudd et al (under review)
~6,500 patients from 9 cohorts across Africa, Asia and Latin America

32. Am J Respir Crit Care Med 2015; 192:958-964
SIRS Sensitivity
SIRS is an appropriate response to infection –
or any other stimulus that activates inflammation

33. Severe Sepsis
 Confusing
 Most people say “sepsis” when they mean
“severe sepsis”
 Is “severe sepsis” really needed ?