Early lactate dynamics for short-term mortality in emergency department sepsis.

BACKGROUND: Early risk stratification in emergency department (ED) sepsis is essential for guiding monitoring and escalation. Although admission lactate is prognostic, the value of early serial values and short-interval changes is uncertain. This study aimed to evaluate whether early serial measurements of lactate and high-sensitivity cardiac troponin T predict 48-hour mortality in adults with sepsis presenting to the emergency department. METHODS: This prospective observational study was conducted in the resuscitation area of a tertiary ED. Adults (≥ 18 years) meeting Sepsis-3 criteria and managed for at least 6 h were included. Arterial lactate and high-sensitivity cardiac troponin T were measured at presentation, 3 h, and 6 h. Predefined delta values reflected changes between time points. The primary outcome was 48-hour all-cause mortality. Discrimination was assessed using receiver operating characteristic (ROC) analysis; cut-offs were identified with Youden’s J, and odds ratios (ORs) with 95% confidence intervals (CIs) were calculated. RESULTS: A total of 118 patients were analyzed (median age 77 years, 44.9% female); 26 (22%) died within 48 h. Lactate was higher in non-survivors at all time points (0 h: 5.6 vs. 3.1 mmol/L; 3 h: 7.3 vs. 2.3; 6 h: 9.2 vs. 2.0; all p < 0.001). ROC analysis showed improvement: area under curve (AUC) for 0 h = 0.726, 3 h = 0.896, 6 h = 0.958. ΔL0–6 < 1.7 mmol/L predicted mortality with 96.2% sensitivity, 90.2% specificity, and OR 230.6 (95% CI 27.8–1909.0; p < 0.001). hs-cTnT showed moderate discrimination (AUC 0.647–0.728) but did not improve lactate models. CONCLUSIONS: In ED sepsis, serial lactate—particularly the 6-hour value and 0–6 h change—outperformed admission values for predicting 48-hour mortality. A 0–3–6 h lactate schedule with actionable thresholds may improve early bedside risk stratification.