Is It time to revise Abbreviated Burns Severity Index (ABSI) after 37 years

Introduction: Abbreviated Burns Severity Index (ABSI) was introduced in 1982. With constant progress in burn care quality, alongside with the changing of the population pyramid, some modification has been suggested for this risk prediction tool. The aim of this study was to evaluate the accuracy of these new modifications in a large sample of burnt patients.Methods: All patients with acute burn who were hospitalized during 2010 to 2017 in the burn department of Imam Reza Hospital, Mashhad, Iran were included. Age, gender, burnt total body surface area (TBSA), inhalational injury, and burn degree were extracted from patients’ medical records. We evaluated the pattern of the odds ratio for all predictors of ABSI. Data were analyzed using SPSS 25.Results: Overall, 6098 patients were included. The area under the curve (AUC) for ABSI was 0.937. Logistic regression showed an exponential pattern for an odds ratio of age. So, we used the newly suggested scoring which had changed age scoring in modified ABSI model 1 and changes in age, gender and burnt TBSA in modified ABSI model 2. The scoring of other variables was kept unchanged. The modified ABSI models (1) and (2) had an AUC of 0.939 and 0.935, respectively. However, the predicted mortality had higher concordance with actual mortality in the modified ABSI model (1) compared to the original model and modified ABSI model (2): (...). Conclusion: It seems that although the overall accuracy of the original and modified ABSI models are similar, the latter has a much better mortality prediction in Iranian population. These results support the suggested revision for this famous prediction tool.