Age-dependent performance of conventional trauma scores in older trauma patients: A retrospective trauma registry study

Population aging has increased the burden of trauma among older adults. Accurate risk stratification remains challenging because age-related physiological vulnerability may reduce the predictive accuracy of conventional trauma scoring systems. Evidence regarding the clinical utility of these scores across older age strata remains limited. To evaluate the age-stratified performance of conventional trauma scoring systems for predicting in-hospital mortality among older trauma patients. This retrospective cohort study used trauma registry data from a Level I trauma center in Taiwan from January 2017 to June 2025. Geriatric patients were defined as those aged ≥ 65 years. The Revised Trauma Score (RTS), Injury Severity Score (ISS), Geriatric Trauma Outcome Score (GTOS), and Trauma and Injury Severity Score (TRISS) were evaluated for predicting in-hospital mortality using receiver operating characteristic (ROC) curve analyses. A simplified exploratory risk score was derived from multivariable logistic regression. Among 15,682 trauma patients, 6,653 (42.4%) were aged ≥ 65 years. Geriatric patients had higher injury severity and mortality than younger patients (4.3% vs. 1.5%, p < 0.001). In the geriatric cohort, GTOS showed the highest discrimination (AUC 87.5%), followed by 1-TRISS (86.1%), ISS (85.5%), and RTS (68.6%). Discrimination declined with advancing age across all scores. A simplified exploratory score showed graded mortality risk across risk strata. Conventional trauma scores showed complementary performance, but discrimination declined with advancing age. 1-TRISS may support risk confirmation and care prioritization, whereas GTOS may identify particularly high-risk patients for closer monitoring. The simplified exploratory risk score may offer an interpretable framework for age-adapted geriatric trauma risk assessment.