Prediction of Survival in Severe Burned Patients through Abbreviated Burn Severity Index

Abstract

Albana Aleksi, Monika Belba

Introduction: Many predictive models for mortality are developed to identify the most important factors which can influence the outcome and many prognostic scores are created and are validated in many studies according to the characteristics of each country

Objectives: This study aims to describe the characteristics of the patients admitted to our Intensive Care Unit(ICU) for the period in the study and to observe how the ABSI score can predict an outcome.

Patients And Methods: This study is an observational retrospective cohort study that observed adult burn patients (≥ 20 years old) hospitalized in the ICU of the Burn Service in UHCT during 2014-2019. Logistic regression analysis was used to determine the relationship between mortality and possible contributing variables of the ABSI score. The discriminative power of the revised ABSI score was assessed by receiver operating characteristics curve analysis Results: Inhalation injury and presence of full-thickness burn had a strong relationship with mortality. ROC analysis as a valuable tool to evaluate predictive model revealed that ABSI score assessed outcome with a high accuracy AUC=0.973.

Conclusions: The ABSI score is a great representation of the mortality hazard at admission to the burns center and may be utilized to complement the existing referral criteria for triage choices.

PDF