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

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. 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. 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. 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.