Elevated serum osmolality predicts intensive care unit mortality independently in respiratory failure: Evidence from a real‐world clinical dataset

Abstract Background The prognostic value of admission serum osmolality in intensive care unit (ICU) patients with Respiratory failure (RF) remains unclear. Methods This study included 1927 patients with RF who were admitted to the ICU for the first time, based on the MIMIC‐IV database in the United States. Patients were stratified into three groups according to the tertile of admission serum osmolality: low (T1: 271–287 mOsm/kg), intermediate (T2: 294–302 mOsm/kg), and high (T3: 311–332 mOsm/kg). Primary outcomes were 30‐ and 90‐day all‐cause mortality. Cox models adjusted for 30 covariates; restricted cubic spline (RCS) and subgroup analyses explored dose–response and interactions. Results Kaplan–Meier curves differed significantly ( p < 0.05). Compared with T1, fully adjusted hazard ratios for T3 were 1.34 (95% CI 1.09–1.64) at 30 days and 1.29 (1.07–1.56) at 90 days ( p ‐trend ≤ 0.006). RCS showed an approximately linear rise above 298 mOsm/kg. Associations were stronger in patients with diabetes, sepsis, or cerebrovascular disease ( p < 0.05). Conclusions Elevated admission serum osmolality is an easily available, independent predictor of short‐ and medium‐term mortality in ICU RF, supporting early risk stratification and tailored fluid management.