Sepsis at the Intersection

It’s been a year since the “sepsis-3” definition was released at the Society of Critical Care Medicine (SCCM) meeting and concurrently published in the Journal of the American Medical Association (JAMA). This definition quickly found itself at the intersection of clinical care, Centers for Medicare & Medicaid Services (CMS) quality metric efforts, coding, and reimbursement. There was no shortage of voiced opinions maligning the new definition. Even the major journal of the Society of Critical Care Medicine (SCCM), which had endorsed sepsis-3, published editorials criticizing it.

Let’s sum up the criticisms we heard: the new definition has no prospective data, “my” society wasn’t invited to the discussion, people will die due to delayed recognition of sepsis, and providers/payors will have no standard definition to follow. There were also rumors, such as that CMS would be rejecting sepsis-3 and telling its contractors not to follow it, or that Coding Clinic would be supporting the old rather than the new definition of sepsis; however, neither of these conclusions came to fruition.

Years ago, Coding Clinic offered guidance that SIRS plus infection equaled a diagnosis of sepsis, but that was invalidated once ICD-10 arrived. More recently, in 2016, Coding Clinic addressed the situation by reiterating that code assignment is based on provider documentation, not clinical criteria. Though Coding Clinic did not agree to call the old “severe sepsis” now just “sepsis,” it is not the same as endorsing a SIRS-based definition of sepsis. More than anything, the Official Coding Guidelines and Coding Clinic have backed away from having coders perform clinical validation of provider diagnoses. In many hospitals, that job has now been added to the physician advisor’s list of duties.

CMS submitted a response to sepsis-3 that was published in JAMA in July 2016. In it, CMS promised to monitor further research with the new definition, but still maintained that the new definition would delay sepsis recognition and reverse the reduction in mortality from studies that used a SIRS-based sepsis definition. The sepsis-3 authors countered that “we are surprised by the concern … that the new definition will delay the diagnosis of sepsis and negatively affect survival.”

“We are also unaware of any prospective interventional studies using SIRS alone as an entry criterion,” they continued. “All the cited studies showing mortality benefits through quality improvement programs involve patients who had established organ dysfunction (i.e., “sepsis” in the new definition).”

In fact, one of the authors of the CMS response was Dr. Emanuel Rivers, who authored the first “early goal-directed therapy” study. His own study included patients with SIRS criteria plus either a systolic blood pressure of less than 90 mmHg after a fluid challenge, or a lactate greater than 4 mmol/L. That sounds a lot like organ dysfunction.

The January 2017 SCCM conference brought a 2017 update to the Surviving Sepsis Campaign Guidelines, as well as several companion articles. Including several of the sepsis-3 authors among a much broader panel, the document accepts the basic premise of sepsis-3 that sepsis is life-threatening organ dysfunction caused by a dysregulated host response to infection. SIRS is not mentioned even once in the new guidelines, and the inflammatory response is only mentioned in that its presence in the absence of infection does not warrant antimicrobial therapy. These guidelines do not try to massage the sepsis-3 definition, as the 2012 guidelines did to sepsis-2. What these guidelines do, however, is offer clinical guidance on best practices in the treatment of sepsis. These guidelines represent a unification of the sepsis expert community that the key question is not “who has sepsis,” but rather, “how do we treat those infected patients who have a high risk of death?” The specific clinical recommendations are well worth reviewing, but beyond the scope of this perspective article.

Prospective studies are underway to evaluate the best markers for identifying “septic” patients at high risk of death. Will the best ones be SOFA, qSOFA, SIRS, lactate, procalcitonin, or something else entirely? Being better at identifying high-risk patients early, which SIRS never was great at, is the key step in targeting the recommended therapies.

CMS’s sepsis core measure has always been based on patients with organ dysfunction or elevated lactate. SIRS is only one component, and not even a mandatory component, of CMS’s severe sepsis definition. Now the sepsis expert community is moving past SIRS, and it is time for CMS to do the same. CMS promised to monitor future studies utilizing sepsis-3. Good thing, because that will likely be all of them, from this point forward.

The sepsis rifts are beginning to heal, and sepsis-3 is here to stay – until sepsis-4 arrives. But how apart were we actually, since the goal has always been to identify the “severe sepsis” patients who benefit from evidence-based therapies? I, for one, eagerly await the arrival of prospectively validated studies on measures for sepsis recognition, and I don’t care what it is – as long as it works.