Evaluation of a novel 5-group classification system of sepsis by vasopressor use and initial serum lactate in the emergency department.

Abstract

Prognostication in sepsis is limited by disease heterogeneity, and measures to risk-stratify patients in the proximal phases of care lack simplicity and accuracy. Hyperlactatemia and vasopressor dependence are easily identifiable risk factors for poor outcomes. This study compares incidence and hospital outcomes in sepsis based on initial serum lactate level and vasopressor use in the emergency department (ED). In a retrospective analysis of a prospectively identified dual-center ED registry, patients with sepsis were categorized by ED vasopressor use and initial serum lactate level. Vasopressor-dependent patients were categorized as dysoxic shock (lactate >4.0mmol/L) and vasoplegic shock (≤4.0mmol/L). Patients not requiring vasopressors were categorized as cryptic shock major (lactate >4.0mmol/L), cryptic shock minor (>2.0 and ≤4.0mmol/L), and sepsis without lactate elevation (≤2.0mmol/L). Of 446 patients included, 4.9% (n=22) presented in dysoxic shock, 11.7% (n=52) in vasoplegic shock, 12.1% (n=54) in cryptic shock major, 30.9% (n=138) in cryptic shock minor, and 40.4% (n=180) in sepsis without lactate elevation. Group mortality rates at 28days were 50.0, 21.1, 18.5, 12.3, and 7.2%, respectively. After adjusting for potential confounders, odds ratios for mortality at 28days were 15.1 for dysoxic shock, 3.6 for vasoplegic shock, 3.8 for cryptic shock major, and 1.9 for cryptic shock minor, when compared to sepsis without lactate elevation. Lactate elevation is associated with increased mortality in both vasopressor dependent and normotensive infected patients presenting to the emergency department (ED). Cryptic shock mortality (normotension+lactate >4mmol/L) is equivalent to vasoplegic shock mortality (vasopressor requirement+lactate <4mmol/L) in our population. The odds of normotensive, infected patients decompensating is three to fourfold higher with hyperlactemia. The proposed Sepsis-3 definitions exclude an entire group of high-risk ED patients. A simple classification in the ED by vasopressor requirement and initial lactate level may identify high-risk subgroups of sepsis. This study may inform prognostication and triage decisions in the proximal phases of care.