Abstract / Summary
IntroductionExisting CT-based prognostic models for traumatic brain injury (TBI) are not tailored to specific injury mechanisms and lack sensitivity to cerebral oedema changes. This study aimed to develop an assessment method for cerebral oedema and compare it with conventional CT scoring systems in predicting long-term outcomes in patients with TBI following road traffic accidents (RTAs).MethodsA retrospective analysis was conducted on 227 patients with TBI following RTAs admitted to a single center (2019.1-2023.12). A clinical prognostic model (IMPACT) was defined, comprising age, motor score, and pupillary light reflex. The A5 (+1) CT classification was defined based on the involvement of cerebral lobes by oedema and the resulting mass effect. The Marshall CT classification, Rotterdam CT score, Helsinki CT score, and A5 (+1) CT classification were added to the IMPACT model to assess their value in predicting outcomes using area under the curve (AUC).ResultsThe AUC for A5 (+1) CT classification reached 0.849, surpassing Marshall CT classification (0.720, p < 0.001), Rotterdam CT score (0.745, p = 0.004), and Helsinki CT score (0.746, p = 0.002). Adding A5 (+1) CT classification to IMPACT model increased the AUC by 0.076 (p = 0.001). Adding Rotterdam CT score increased the AUC by 0.038 (p = 0.028), while incorporating Marshall CT classification and Helsinki CT score yielded no significant changes (p = 0.138, p = 0.057).DiscussionThe A5 (+1) CT classification outperforms conventional CT scoring systems in assessing long-term outcomes for patients with TBI following RTAs. Its inclusion in the IMPACT model enhances predictive performance.