Background:
Several studies were published to validate quick-SOFA (qSOFA), namely in comparison with Systemic Inflammatory Response Syndrome (SIRS) criteria. We performed a systematic review and a metaanalysis with the aim of comparing qSOFA and SIRS in patients outside the ICU.
Method:
We searched the MEDLINE, CINAHL, and Web of Science database, from February 23rd, 2016 until June 30th, 2017, to identify full-text English language studies published after the Sepsis-3 publication that compared qSOFA and SIRS as well as sensitivity or specificity to diagnose sepsis, the hospital and ICU length of stay (LOS) and hospital mortality. Data extraction from the selected studies followed the recommendations of the MOOSE group and the PRISMA statement.
Results:
From 4,022 citations, ten studies met the inclusion criteria. Pooling all the studies, a total of 229,480 patients were evaluated. The meta-analysis of sensitivity for the diagnosis of sepsis comparing qSOFA and SIRS was in favor of SIRS (1.32 [0.40-2.24], p<0.0001, I2=100%). One study described the specificity for the diagnosis of infection comparing SIRS (84.4% [76.2–90.6]) to qSOFA (97.3% [92.1–99.4]) and demonstrated a better specificity of qSOFA. The meta-analysis of AUROC of six studies comparing qSOFA and SIRS was in favor of qSOFA (0.03 [0.01-0.05], p=0.002, I2=48%) as a predictor of in-hospital mortality.
Conclusion:
The SIRS was significantly superior than qSOFA for sepsis diagnosis and qSOFA was slightly better than SIRS in predicting hospital mortality. The association of both criteria could provide a better model to initiate or escalate therapy in sepsis patients.
Commentary.
The Sepsis 3.0 Criteria for the diagnosis of sepsis were released in February 2016 with much fanfare. [1] In essence, the presence of 2 or more of the following criteria (qSOFA) were used to diagnose sepsis outside the ICU:
- Respiratory Rate > 20
- SBP < 100mmHg
- Altered mental status
and a SOFA score of 2 or more (assuming a baseline SOFA score of 0) was used to diagnose sepsis in the ICU. As a “clinical scientist” I believe that Sepsis 3.0 is a step backward and not forward. While the SIRS concept was not perfect,[2,3] it provided a useful clinical framework to screen for sepsis, with the realization that 12% of patients with sepsis would prove to be “SIRS negative”. As the meta-analysis of Serafim et al demonstrated (see Figure) SIRS outperformed qSOFA as a tool to diagnose sepsis (outside the ICU). While qSOFA may better predict mortality, this attribute is of little worth to the bedside clinician.
As far as I can tell, the SOFA score is not routinely measured in ICU patients (at least in the USA), therefore the use of this criteria to diagnose the development of sepsis in an ICU patient is problematic. Furthermore, it assumes that the patient has a baseline SOFA score of 0; this is usually not the case in many ICU patients who frequently have multiple co-morbidities.
In my Books, sepsis is best defined (at this time) as “Infection PLUS organ dysfunction” (any organ dysfunction)… Just Keep it Simple!. As diagnostic tests to identify sepsis improve (biomarkers, bacterial PCR, etc), it is likely that in the future “laboratory tests” will be included in the variables used to diagnose sepsis, resulting in diagnostic criteria that have greater sensitivity and accuracy.
In summary, Sepsis 3.0 was clearly not developed by clinicians for clinicians, and should be abandoned.
References
- Singer M, Deutschman CS, Seymour CW et al. The third international consensus definitions for sepsis and septic shock (Sepsis-3). JAMA 2016; 315:801-10.
- Kaukonen KM, Baily M, Pilcher D et al. Systemic inflammatory response syndrome criteria criteria in defining severe sepsis. N Engl J Med 2015; 372:1629-38.
- Vincent JL, Opal SM, Marshall JC et al. Sepsis definitions: time for change. Lancet 2013; 381:774-75.
- iSepsis- SEP-1: Conspiracy Theories and Fake News! - March 3, 2018
- iSepsis – Sepsis 3.0- Flogging a dead horse! - February 23, 2018
- iSepsis – Patients with sepsis have SCURVY - February 4, 2018
Is anyone using qSOFA and SOFA in clinical practice in EDs, on the floors, on rapid response teams, or during ICU rounds? Severe sepsis and SIRS still seem alive at least from a documentation police standpoint, And for better or worse, the terms I hear more often on ICU signout and when talking to other providers are: getting septic, pretty septic, and septic shock.
There is resistance to change. There is a resistance to new ideas. In my ICU, it is the junior faculty who are most willing to accommodate Sepsis 3.0 definitions into their practice. The necessity to do away with SIRS has been evident for many years given the fact that any healthy individual who walks up three flights of stairs will likely be SIRS positive. Moreover, any elderly patient with delirium and decreased oral intake due to a non-bacteremic UTI or URI would also have a high probability of meeting “not so sick” SIRS criteria. The new Sepsis definition includes organ dysfunction and therefore replaces the old “Severe Sepsis” definition. Hence, the meta-analysis reviewed above unfortunately compares apples to oranges. A more accurate evaluation would be one that compares old Severe Sepsis vs Sepsis 3.0. This, of course, has already been performed and published (JAMA. 2017;317(3):301-308 & JAMA. 2017;317(3):290-300). I am an avid subscriber to Dr. Marik’s position on CVP monitoring and non-hypoxemic lactate production. Despite the data, I still find some, more experienced colleagues holding on to CVP volume responsiveness and SIRS criteria. It takes a village to promote change, and just like with everything, it will take time for… Read more »
Dr. Marik is completely correct on this issue. However, the comment below about junior vs. senior faculty is out of line. Youth brings exuberance and sometimes gullibility. Age brings experience and sometimes stodginess. But there is no sense in throwing mud at one another, and people of any age are capable of seeking better answers to serious problems. As regards SOFA based sepsis diagnosis, it is needlessly complex. Dr. Marik is correct. If one accepts that the Sepsis-3 definition is somewhere in the ballpark, all one should need to diagnose it is an infection and an organ dysfunction. What I have to say that I do respect about the SOFA-based criteria is that it allows for lesser degrees of injury to 2 or more organs for diagnosis, which may prove useful for earlier diagnosis over time. However, SOFA for diagnosing sepsis in the ICU addresses just about 15% of the septic patients, since the best epidemiology available indicates that 86.8% of sepsis cases arise outside the hospital. The so-called prospective comparison of severe sepsis with the Sepsis-3 diagnostic criteria is spurious. A careful read of the methods shows that those authors defined severe sepsis as infection with an increased lactate.… Read more »
?nfect?on+organ dysfunction okey but related dysfunction that is heart failure +tonsillitis not sepsis
i become increasingly confused, Paul, in a way, with every new conversation regarding sepsis and septic shock. but, in a way, it’s confusion in a good way.
because every discussion involves, i believe, basically very good, very smart folk (whether its the authors of sepsis -3 , singer and deutchman who scott interviewed, or the many other minds struggling with this entity that kills more than stroke and AMI combined in 30 days) who are trying to wrassle, trying to tie down, a very elusive critter, a Fata Morgana.
but, i think, the attempt is obviously essential, and worthwhile, because each conversation, each dissection of the newest (and older) paradigms, protocols, beliefs, studies, gives a clearer vision and understanding of the entity we are trying to subdue.
at least i hope so.
thank you again, Paul.
tom
After reading your thoughts on lactate myths, do you think lactate > 2 in absence of liver/kidney function lab abnormality = organ dysfunction? For example, SIRS positive patient, with infectious source and lactate > 2 with otherwise normal measured parameters. Does this fit your definition of sepsis?
Jeremy: That is an excellent question that I had not thought about. So, I would say that if a patient with an infection has no evidence of organ dysfunction but had an elevated isolated lactate that by definition that would be a severe infection and not sepsis. However, I think it would be uncommon for a patient with an infection to have an elevated lactate in the absence of any organ dysfunction. As is clear from this question, the diagnosis of sepsis is not that clear with many grey areas. I plan to post the results of another study looking at qSOFA and the Sepsis 3.0 stuff quite soon.
[…] while qSOFA had higher specificity. See the following posts at PumCrit for further discussion: [https://emcrit.org/isepsis/isepsis-sepsis-3-0-much-nothing/] […]
[…] while qSOFA had higher specificity. See the following posts at PulmCrit for further discussion: [https://emcrit.org/isepsis/isepsis-sepsis-3-0-much-nothing/] […]
[…] (antar lpk, temp) men qSOFA predicerar mortalitet bättre (troligen variabeln medvetandesänkning). Jag håller inte alltid med Paul Marik men här tycker vi nog samma […]