Predicting Burn Mortality Using a Simple Novel Prediction Model.

Area of Science:

• Trauma Surgery
• Critical Care Medicine
• Burn Management

Background:

• Accurate prognostication is crucial for resource allocation and patient management in burn care.
• Existing burn-specific mortality prediction models are often complex and lack simplicity.
• There is a need for models incorporating both endogenous and exogenous patient factors.

Purpose of the Study:

• To develop a simple, burn-specific mortality prediction model.
• To identify key predictors of mortality in burn patients.
• To create a tool for bedside prognostication.

Main Methods:

• Prospective observational study of burn patients.
• Collection of demographic data, TBSA, and laboratory results (CBC, kidney function, ABG).
• Statistical analysis including logistic regression, ROC curves, and Hosmer-Lemeshow goodness of fit, with bootstrapping validation.

Main Results:

• Total Burn Surface Area (TBSA) and APACHE II score were significant independent predictors of mortality.
• TBSA (OR 1.094) and APACHE II score (OR 1.166) retained significance in logistic regression.
• The developed prediction model demonstrated good performance with an AUC of 0.778.

Conclusions:

• Mortality prediction in burn patients can be accurately achieved at the bedside.
• The combination of TBSA and APACHE II score provides a reliable prognostic tool.
• This simplified model aids in clinical decision-making and resource management for burn patients.