Lactate and combined parameters for triaging sepsis patients into intensive care facilities

Purpose: To find predictors of intensive care unit (ICU) requirement within the first 48 hours in newly diagnosed sepsis patients presenting at the emergency department. Materials and methods: Analysis of a prospective observational cohort was performed. We recruited new sepsis patients at the emergency department, and collected baseline characteristics and parameters. Variables were compared between patients: those that required ICU within 48 hours and those that did not. Multivariate analysis was performed to identify independent predictors. Results: Out of 719 patients enrolled, 275 were confirmed to have sepsis. There were 107 patients (39%) that required ICU admission within 48 hours. Independent predictors for ICU requirement were: lower body temperature (P = .019), initial lactate (P = .02), 2-hour lactate clearance (P = .035), and the Sequential Organ Failure Assessment (SOFA) score without cardiovascular component (SOFA _{no CVS}) (P < .001). The optimal cutoff values for the two strongest predictors were: SOFA _{no CVS} ≥5 (adjusted OR, 5.3; 95% CI, 1.9-14.7) and initial lactate ≥1.9 mmol/L (adjusted OR, 3.3; 95% CI, 1.2-8.9). We also proposed a combined "LACTIC score" with higher predictive ability. Conclusions: We suggested a way to predict ICU requirement in sepsis patients and proposed a combined score that might be better than individual parameters. Further validation should be performed before using them clinically.