Sepsis–3: Seeing the Entire Picture*
It is important to begin, however, by reiterating that there is a key distinction between the two very different products of the conference: “definitions” and “clinical criteria.” Per the Merriam-Webster dictionary, a definition describes “the essence of a thing” or, more generically, “what something is.” The Sepsis-3 definition of sepsis is “life-threatening organ dysfunction caused by a dysregulated host response to infection,” whereas that for septic shock is “a subset of sepsis in which profound circulatory, cellular and metabolic abnormalities are associated with a greater risk of mortality than from sepsis alone.” These definitions reflect the most up-to-date scientific evidence available, but they are not especially useful clinically. To rectify this defect, Seymour et al (6) and Shankar-Hari et al (7) presented data-driven clinical criteria. Importantly, these criteria were derived by testing the ability of combination of clinical variables to identify those patients with suspected infection who ultimately arrived at an outcome consistent with the presence of sepsis—death or an ICU length of stay greater than or equal to 3 days for sepsis (6), mortality in excess of 40% for septic shock (7). Thus, the clinical criteria do not, in actuality, identify patients with sepsis or septic shock; the absence of a “gold standard” test that unequivocally identifies either condition makes that impossible, limiting analysis to the identification of proxy outcomes. Thus, it is important to note that, despite wording to the contrary, the work by Sterling et al (5) addresses only the clinical criteria—and not the definitions.
Sterling et al (5) tested the hypothesis that application of the Sepsis-3 criteria for septic shock would fail to detect patients who met earlier (Sepsis-1) criteria (8, 9). They conducted their analysis using two preexisting databases focused on emergency department (ED) management of patients presenting with the combination of suspected infection, greater than or equal to 2 systemic inflammatory response syndrome criteria and a systolic blood pressure less than 90 mm Hg after fluid resuscitation. The authors identified 470 such patients in their ED datasets, 96 subsequently died. Of these 470, 200 met Sepsis-3 clinical criteria for septic shock, that is, they were hypotensive and/or had a requirement for vasopressors despite resuscitation and a serum lactate level greater than 2 mmol/L. Included within these 200 were 57 of the 96 patients who died. Thus, use of Sepsis-3 criteria for septic shock failed to identify 37 of the patients (39%) who met the proxy outcome criteria, whereas Sepsis-1 criteria identified all 96 fatalities.