Reconciling Sepsis 2 and Sepsis 3: Are We Missing Something in Between?Clarissa Barnes, MD, FACP Sepsis. To anyone who treats patients or is involved in clinical documentation, quality, or denials and the revenue cycle, this term causes a lot of angst. It is a serious cause of morbidity and mortality to our patients and an ongoing source of increased cost for healthcare. (1,2) Years after the alarm was sounded about the need to better recognize and treat this diagnosis, confusion and debate on how to even define it persists. But what if the reasons there is no consensus is because neither is quite right and there is opportunity to better define Impending Sepsis. The fundamental disagreement comes down to Sepsis-2 vs Sepsis-3. Sepsis-2 is SIRS criteria plus a locus of infection and in that same documentation structure Severe Sepsis is sepsis that has progressed to include end organ damage and Septic Shock includes fluid-resistant hypotension. Sepsis-3 is a dysregulated host response to an infection with evidence of end-organ damage. Septic Shock is the same in Sepsis-2 with sepsis that has progressed to included fluid-resistant hypotension. So, Sepsis-2 Severe Shock is a pretty good match to Sepsis-3 Sepsis. And Septic Shock remains essentially the same. Why the disagreement? Part of the issue is philosophical. Is it really sepsis if there is no evidence of damage? Part of the disagreement is practical. Recognizing that sepsis (and missing sepsis) was harming patients in 2015, the Centers for Medicare and Medicaid Services (CMS) added a sepsis quality measure to the Hospital Inpatient Quality Reporting Program (IQRP) that was called SEP-1. (3) This program required hospitals to report compliance for sepsis patients for the 3- and 6-hour treatment bundle which included timely antibiotics, IV fluids, cultures, measuring lactic acid levels, the use of vasopressors for shock/fluid-refractory hypotension, and the repeat bedside evaluations/exams to assess how patients were responding to treatment. Inclusion in these measures uses the SIRS plus infection definition. It is impractical (and frankly ridiculous) to ask clinicians to use different definitions for different purposes and different payers and so Sepsis-2 often becomes the default definition in a system due to CMS’s SEP-1. And part of the issue is a fundamental misunderstanding of what the Sepsis-2 definition of Sepsis looks like in a clinical scenario. From those who tell me Sepsis 3 is THE definition of sepsis, they often cite that many patients may have SIRS criteria and an infection, but that does not make them septic. And they are correct. A pancreatitis patient with tachycardia and an elevated respiratory rate who has a concurrent urinary tract infection may not have sepsis. Let me be clear (because I hear this argument on occasion), just because a patient may meet the definition of Sepsis-2, does not mean they have Sepsis. In other words, not all SIRS plus infection is Sepsis. Those measures are necessary, but they aren’t sufficient. The missing piece is the judgement of the treating physician and that is the part that is harder to reflect on chart review. I was asked recently to do reviews of clinical denials for a system in New York (where the use of Sepsis-2 is mandated by law). Every chart I looked at had SIRS and had an infection, or at least the concern for one initially on presentation. However, it was often difficult to tell whether the patient truly had sepsis just from the notes. But it was often much clearer when the chart was evaluated to see what had been done. And that is the biggest opportunity for a better definition. In CDI our mission is precise language to accurately reflect what is happening to the patient and the medical decision-making process. Maybe there is a continued struggle between Sepsis-2 and Sepsis-3 because we fundamentally do not have the right language to reflect the clinical space between a localized infection and fully realized sepsis. There should be a way to reflect the difference between a patient who comes to the clinic and has a urinary tract infection and has SIRS criteria for an unrelated reason (such as anxiety causing tachycardia and tachypnea) and the patient who comes to the ER with Impending Sepsis due to a urinary source. There needs to be a better way to clarify because those patients are going to need to be treated very differently by the physician. One will get antibiotics for sure (because there is still an infection) but it will be a script to pick up an oral antibiotic from the local pharmacy. The other will get IV antibiotics, additional testing (lactic acid, blood cultures, etc.), and a higher level of monitoring and care as well as a clear need for hospitalization. And if that patient presents early enough, is recognized soon enough, and is treated, they may never have end-organ damage. That is the very definition of what recognizing sepsis early is about: preventing progression and ultimate harm. Personally, using the Sepsis-3 definition for what is realized sepsis makes sense. But it still misses the point of the clinical outcomes that are trying to be impacted by early recognition as someone is starting down that path. And, not having another term for those patients who are clearly between a localized infection and the point at which organs are damaged misses an opportunity to be precise about patient who are truly ill and require more treatment and monitoring and therefore resources because they are at higher risk. Impending Sepsis may be a term that helps bridge that gap or there may be a better term (and if you do, I’d love to hear it!).
Dr Barnes is the South Dakota Medicaid Medical Director. |