Abstract
BACKGROUND: This study examines the Stress Hyperglycemia Ratio (SHR) as a predictor of mortality in acute brain injury (ABI) patients using the MIMIC-IV v3. 1 database. METHODS: In this retrospective cohort study of 2,423 ABI patients, SHR was calculated as SHR = [Admission blood glucose (mg/dL)] / [28.7 × HbA1c (%) - 46.7]. Mortality outcomes included ICU, in-hospital, 30, 60, 90, and 365-day mortality. Cox regression models adjusted for covariates assessed the association between SHR and mortality risk, with restricted cubic splines confirming linearity. Predictive performance was evaluated using ROC curves, incorporating SHR, Glasgow Coma Scale (GCS), and first-day ventilation status. RESULTS: SHR was significantly associated with mortality across all outcomes, showing a linear relationship. Adjusted hazard ratios (HR) for in-hospital and ICU mortality were 1.18 (95% CI: 1.06-1.32, p = 0.003) and 1.16 (95% CI: 1.02-1.32, p = 0.029), respectively. Dichotomized SHR indicated increased in-hospital mortality risk (HR: 1.44, 95% CI: 1.13-1.83, p = 0.003). Combining SHR with GCS and ventilation status improved predictive accuracy, achieving AUCs of 0.817 for ICU mortality and 0.788 for in-hospital mortality. Robustness was supported by E-values of 2.24 and 2.37 for in-hospital and ICU mortality. CONCLUSION: SHR independently predicts short- and long-term mortality in ABI patients, with enhanced utility when combined with GCS and ventilation status, supporting its role in clinical risk stratification.