Objective: To assess whether the IMPACT score predicts mortality in patients with traumatic brain injury (TBI).
Design: Cohort study.
Setting: A total of 260 patients with moderate and severe TBI were enrolled from the Intensive Care Unit (ICU) of Hospital Regional Docente de Trujillo, a trauma referral hospital in northern Peru. The data collection period spanned from 2018 to 2023.
Patients: Patients with moderate and severe TBI admitted to the ICU.
Interventions: Patients with TBI in whom the predictive performance of the IMP ACT score for mortality was assessed.
Main Outcome Measure: In-hospital mortality.
Results: A total of 260 patients with moderate and severe TBI were included. When evaluating the IMPACT score and its models (core, extended, and laboratory), we found adequate predictive values expressed by their respective ROC AUCs (0.74, 0.83, and 0.84). Additionally, upon constructing a predictive model utilizing the IMPACT score laboratory model, an increased risk of mortality was observed (adjusted RR 1.21; 95% CI 1.12–1.31). Furthermore, the variables: mechanism of TBI (adjusted RR 5.6; 95% CI 3.03–10.85), Glasgow Coma Scale (GCS) (adjusted RR 1.35; 95% CI 1.15–1.58), and use of vasopressors (adjusted RR 2.2; 95% CI 1.14–4.42) behaved as independent predictors of in-hospital mortality.
