Prognostic value of the lactate-to-albumin ratio for short- and long-term mortality in sepsis

Abstract

Abstract The lactate-to-albumin ratio (LAR) integrates acute metabolic stress with the systemic response reflected by serum albumin, but its association with longer-term outcomes in sepsis remains incompletely characterized. We analyzed 4,916 adults meeting Sepsis-3 criteria in the MIMIC-IV database. Baseline LAR was calculated from the first lactate and albumin measurements obtained within 24 h of intensive care unit admission. Associations with in-hospital, 28-day, 90-day, and 365-day mortality were evaluated using multivariable Cox models, restricted cubic splines, and time-dependent receiver operating characteristic analysis. Hospital non-survivors had a higher median LAR than survivors (0.974 vs. 0.564; P  < 0.001). Each one-unit increase in LAR was associated with higher mortality after adjustment for age, sex, body mass index, SOFA score, and APACHE III score (adjusted hazard ratios, 1.149–1.172; all P  < 0.001). Time-dependent areas under the curve were 0.653 (95% CI, 0.631–0.675) at 28 days, 0.684 (0.660–0.707) at 90 days, and 0.665 (0.635–0.694) at 365 days. Baseline LAR is an accessible marker of risk that may complement, but should not replace, established clinical assessment. External validation and analyses incorporating serial measurements and treatment timing are required before clinical thresholds are adopted.