Early longitudinal lactate-to-albumin ratio and mortality in critically ill adults: a temporally aligned multicohort study

The prognostic value of longitudinal lactate-to-albumin ratio (LAR) patterns in critically ill adults remains uncertain. In this retrospective multicohort study, we evaluated early LAR dynamics in MIMIC-IV and independently assessed them in eICU using temporally aligned exposure and landmark definitions. In MIMIC-IV, LAR was assessed during 0–<96 h, with mortality follow-up beginning at 96 h. Among 1,456 landmark-eligible patients, higher mean LAR was associated with 30-day and 365-day mortality (adjusted risk ratio per 1-SD increase, 1.22 [95% CI, 1.11–1.34] and 1.09 [1.02–1.17], respectively), but associations were similar for mean lactate. A descriptive three-class trajectory model identified stable-low, moderate-declining, and high-increasing patterns; neither non-reference trajectory was associated with increased mortality after adjustment. Trajectory structure changed under alternative temporal definitions. In eICU ( n = 1,181), independently estimated trajectories differed, although a stable-high class was associated with higher hospital and ICU mortality. Secondary prediction analyses showed only modest incremental performance for mean LAR and little contribution from trajectory class alone. In this selected population with sufficient serial measurements, longitudinal LAR provides prognostic information but does not establish superiority over lactate, a universal high-risk trajectory phenotype, or clinical utility.