Predicting in-hospital mortality in acute traumatic intracranial hematomas: derivation of a rapid assessment tool.

BACKGROUND: Early prediction of in-hospital mortality in patients with acute traumatic intracranial hemorrhage is essential for optimizing management strategies and guiding resource allocation. Although several prognostic models exist for traumatic brain injury, most are complex and not tailored to this specific subgroup. OBJECTIVES: This study aimed to identify independent predictors of in-hospital mortality in patients with acute traumatic intracranial hemorrhage and to develop a novel, simplified prognostic score (ISS-AG) for use in clinical practice. METHODS: We conducted a retrospective observational study including 161 patients with acute traumatic intracranial hematomas admitted to a tertiary trauma center. Clinical, radiological, and laboratory variables were collected. Survivors and non-survivors were compared using univariate analyses, and significant predictors were further assessed through multivariable logistic regression. Discrimination was evaluated with receiver operating characteristic (ROC) analysis. RESULTS: The overall in-hospital mortality rate was 39.8%. Non-survivors were older (median 66.5 vs. 44 years, p < 0.001) and had higher Injury Severity Scores (ISS) and glucose levels on admission. Multivariate analysis identified age (OR 1.036, 95% CI 1.013-1.061, p = 0.003), ISS (OR 1.077, 95% CI 1.024-1.132, p = 0.004), and serum glucose (OR 1.010, 95% CI 1.002-1.019, p = 0.019) as independent predictors of mortality. These variables were incorporated into a multivariable model that demonstrated excellent discrimination (AUC = 0.869), from which the simplified integer ISS-AG bedside score was derived (AUC = 0.814). A cut-off ≥ 4 predicted mortality with 89.1% sensitivity and 61.5% specificity. CONCLUSIONS: The ISS-AG score, integrating age, ISS, and glucose levels, represents a rapid and effective tool for predicting in-hospital mortality in patients with acute traumatic intracranial hemorrhage. This simple model may aid clinicians in risk stratification and early decision-making.