Assessment of Trauma Scoring Systems for Outcome Prediction in Craniofacial Trauma Patients Requiring Intensive Care

BACKGROUND: Craniofacial trauma causes many admissions and severe complications. Different systems assess injury severity but lack predictive validity for determining mortality in critically ill craniofacial trauma patients. This study therefore evaluated various trauma scores and clinical variables to determine their predictive value for mortality. METHOD: In this retrospective, single-centre study, 54 adult patients with craniofacial trauma were admitted to the intensive care unit. Their injuries were graded on admission using various methods, including ISS, TRISS, RTS, CCL, APACHE-II, and GCS to assess severity. The ROC curve analysis was used to assess the accuracy of the scoring systems. Multivariate analysis was used to identify predictors of in-hospital mortality. Clinical variables included patient age, sex, mechanism of injury, and use of protective equipment. RESULTS: Of 54 patients, 9 (16.6%) died during hospitalisation. None had used protective safety equipment at the time of injury. Survivors stayed in the hospital and ICU for less time than nonsurvivors. Survivors had GCS scores of 13, while nonsurvivors had 8 (P=0.003). Median ISS (27 versus 59), APACHE II (18 versus 33), and TRISS (96.8 versus 21.7) scores also differed significantly between the 2 groups. TRISS and APACHE-II showed excellent discrimination, with AUC values of 0.872 and 0.863, respectively. RTS and GCS also performed well, while ISS and CCL demonstrated moderate predictive ability. FISS demonstrated no meaningful value. CONCLUSION: TRISS and APACHE II provide the most accurate mortality prediction in critically ill patients with craniofacial trauma. Physiological scores with anatomic assessments may better predict risk and support decisions for this high-risk group.