Prognostic Value of Mortality Scoring Systems in Patients With Severe Burns: Identifying Key Predictors of Mortality and Comparative Analysis Between Survivors and Non-Survivors.

BACKGROUND Burns remain a major public health challenge for modern healthcare systems due to high mortality and long-term consequences. This study evaluated the effectiveness of several scoring systems in predicting mortality and identified key prognostic factors in patients with severe burns. MATERIAL AND METHODS A retrospective analysis of 144 adult burn patients admitted to the intensive care unit was conducted. Mortality risk was assessed using total body surface area (TBSA), revised Baux, Abbreviated Burn Severity Index (ABSI), Belgium Outcome Burn Injury (BOBI), and Burn Mortality Prediction (BUMP) scores. Two analytical approaches were applied: a baseline clinical model including key independent variables (age, TBSA, inhalation injury), and separate logistic regression models for each scoring system. Predictive performance was evaluated using receiver operating characteristic (ROC) curves and compared using the DeLong test. RESULTS Overall mortality rate was 43.1%. Non-survivors were older (59.2 vs 46.5 years), had larger burned surface area (42.3% vs 35.3%), and more frequently presented with inhalation injury (59.7% vs 26.8%). The baseline model demonstrated excellent discriminative ability (AUROC=0.87, P<0.001). Among scoring systems, the revised BAUX achieved the highest AUROC (0.86, P<0.001), followed by BOBI and BUMP (both 0.83, P<0.001). However, pairwise comparisons showed no statistically significant differences between the best-performing scores. CONCLUSIONS The baseline clinical model and composite scoring systems demonstrated strong and comparable predictive performance. Simple clinical models based on key variables may provide an effective alternative for mortality risk assessment, while established scoring systems remain valuable due to their ease of use in routine practice. The findings highlight the need for early identification of high-risk patients and timely clinical management optimization.