Accuracy of the revised Baux score for predicting in-hospital mortality of patients with burns: A retrospective cohort study from Ethiopia.

Abstract

Accurate prediction of mortality in patients with burns is essential for clinical decision-making and resource allocation. The revised Baux (rBaux) score is widely used; however, its performance in low-resource settings remains insufficiently validated. We aimed to evaluate the predictive accuracy of the rBaux score for in-hospital mortality in an Ethiopian burn center. A retrospective cohort study was conducted at the Addis Ababa Burn, Emergency, and Trauma (AaBET) Hospital. Adult patients with burn admitted between January 2020 and September 2025 GC were included. Demographic and clinical data were extracted from medical records. The rBaux score was calculated for each patient. Discriminatory performance was evaluated using receiver operating characteristic (ROC) curve analysis, and the optimal cutoff was determined using Youden's index. A total of 149 patients were included, with a median age of 25 years (IQR: 21-38 years) and a median total body surface area (TBSA) of 24% (IQR: 18-35%). The rBaux score demonstrated excellent discrimination (AUC = 0.923; 95% CI: 0.850-0.996; p < 0.001). The optimal cutoff was 81.5, with a sensitivity of 84.6% and specificity of 94.3%. In multivariable analysis, age (adjusted OR = 1.10), TBSA (adjusted OR = 1.18), and female sex (adjusted OR = 9.11) were independently associated with mortality. The rBaux score demonstrated excellent predictive accuracy in this setting. However, context-specific cutoff values and further validation are recommended to optimize its clinical applicability in resource-limited environments.