Prognostic Value of Trauma Scoring Systems for Mortality and Intensive Care Unit Admission in Multiple Trauma Patients

Background: Accurate assessment of trauma severity and prediction of intensive care needs are crucial for improving outcomes in polytrauma patients. This study compared the predictive accuracy of four widely used trauma scoring systems: Glasgow Coma Scale (GCS), Injury Severity Score (ISS), Revised Trauma Score (RTS), and Trauma and Injury Severity Score (TRISS). Methods: A prospective observational study was conducted at Shahid Rahnamoon University Hospital, Yazd, Iran, from August to November 2018. Consecutive multiple trauma patients admitted to surgical wards or the intensive care unit (ICU) were included, while those discharged directly from the emergency department were excluded. Demographic and clinical data were collected on admission, and all four trauma scores were calculated according to standard guidelines. Patients were followed during hospitalization and 30 days after admission. Prognostic accuracy was evaluated using sensitivity, specificity, and area under the curve of receiver operating characteristic curve (AUC) analysis. Results: A total of 769 patients were included (mean age 32.3 ± 21.7 years; 76.1% male). The overall mortality rate was 10.8%. All scoring systems demonstrated significant predictive value. GCS showed the highest prognostic accuracy for mortality (AUC = 0.912, 95% CI: 0.86–0.95) and ICU admission (AUC = 0.947, 95% CI: 0.91–0.97). In comparison, RTS yielded AUC = 0.861 (95% CI: 0.80–0.91) for mortality and 0.832 (95% CI: 0.78–0.87) for ICU admission; TRISS had AUC = 0.906 (95% CI: 0.86–0.94) and 0.884 (95% CI: 0.85–0.91), respectively; ISS performed with AUC = 0.858 (95% CI: 0.80–0.91) for mortality and 0.894 (95% CI: 0.86–0.92) for ICU admission. Conclusion: GCS remains the most accurate and clinically feasible scoring tool for predicting mortality and ICU admission in multiple trauma patients. Although ISS and TRISS provide additional insights, particularly for benchmarking, their complexity limits real-time application in emergency settings.