BackgroundEarly risk stratification in traumatic brain injury (TBI) is essential for timely triage, resource allocation, and clinical decision-making within the critical first hours of admission. This study aimed to compare the predictive performance of five established clinical scoring systems includes the Glasgow Coma Scale (GCS), Revised Trauma Score (RTS), Mechanism, GCS, Age, and Arterial Pressure (MGAP) score, Modified Early Warning Score (MEWS), and Rapid Emergency Medicine Score (REMS) for early mortality prediction in patients with blunt TBI.
MethodsThis single-center retrospective observational cohort study evaluated 444 patients aged 18 to 89 years with blunt TBI admitted to the intensive care unit of a tertiary trauma center in Tehran, Iran, between March 2022 and March 2025. The primary outcome was early mortality, defined as death within 24 hours of admission. Discriminative performance was assessed using the area under the receiver operating characteristic curve (AUC) with 95% confidence intervals (CI) derived via bootstrap resampling (1,000 iterations). Pairwise comparisons of AUCs were conducted, and optimal diagnostic cutoffs were identified using the Youden Index.
ResultsWithin 24 hours of hospital admission, early mortality occurred in 97 patients (21.8%), while 347 patients (78.2%) survived. The trauma-specific and neurological scoring systems demonstrated the highest discriminative capacities: RTS achieved the highest accuracy (AUC = 0.676, 95% CI: 0.617-0.737), followed closely by GCS (AUC = 0.669, 95% CI: 0.608-0.727) and MGAP (AUC = 0.657, 95% CI: 0.594-0.724). General physiological scores exhibited lower performance, with MEWS achieving an AUC of 0.651 (95% CI: 0.595-0.707) and REMS demonstrating the lowest discriminative ability (AUC = 0.601, 95% CI: 0.538-0.659). Pairwise analysis confirmed that RTS GCS and MGAP significantly outperformed REMS, though no statistically significant differences were observed among RTS, GCS, and MGAP themselves. All evaluated systems demonstrated only modest overall predictive performance (AUC < 0.70).
ConclusionTrauma-specific and neurologically oriented scoring systems (RTS, GCS, and MGAP) provide superior and comparable prognostic accuracy for 24-hour mortality in blunt TBI compared to general emergency scores like REMS. However, the absolute predictive power of all evaluated models remains modest. Traditional systems relying on static admission variables fail to capture the dynamic, multifactorial nature of secondary brain injury, highlighting the critical need for multidimensional prognostic tools incorporating physiological time-series data or machine learning algorithms.