Utilizing injury severity score, Glasgow coma scale, and revised trauma score for trauma-related in-hospital mortality and ICU admission prediction; originated from 7-year results of a nationwide multicenter registry

Objective: During the past few decades, many scoring systems have been developed to evaluate the severity of injury and predict the outcome in trauma patients. This study aimed to assess the capacity of three common trauma scoring systems: injury severity score (ISS), Glasgow coma scale (GCS), and revised trauma score (RTS) in predicting in-hospital mortality and ICU admission in patients with traumatic injury. Methods: This is a multicenter study of the hospital-based national trauma registry of Iran (NTRI), an ongoing registry-based trauma database. This study included trauma cases from 12 major trauma centers throughout the country admitted between July 2016 and November 2023. The inclusion criteria were all patients admitted to the emergency department due to trauma, hospitalized for at least 24 hours, deceased within the first 24 hours of admission, and patients transferred from the intensive care unit (ICU)s of other hospitals. Results: A total of 50,458 traumatic patients, with 38,740 (76.9%) being male, were included in this study. After adjustment for confounders, head, face, and neck injuries were associated with the highest odds of death (OR: 7.51, P-value<0.001), whereas abdominal injuries were associated with the highest odds of ICU admission (OR: 4.58, P-value<0.001). Each Unit increase in RTS score was accompanied by a 61% decrease in odds of death (OR: 0.39, P-value<0.001). The area under the ROC curve for predicting in-hospital mortality was 0.81 (0.79 to 0.82) in ISS, 0.78 (0.77 to 0.80) in GCS, and 0.75 (0.73 to 0.76) in RTS. There was a significant difference between RTS and GCS, as well as RTS and ISS for in-hospital mortality prediction (P-values< 0.001). The area under the ROC curve for the prediction of ICU admission was 0.75 (0.74 to 0.75) in ISS, 0.63 (0.62 to 0.63) in GCS, and 0.62 (0.61 to 0.63) in RTS. There was a statistically significant difference between ISS and GCS, as well as ISS and RTS, for ICU admission prediction (P-value<0.001). Conclusion: ISS is the best predictor of in-hospital mortality and ICU admission, compared to GCS and RTS.