Improvement of Predictive Scores in Burn Medicine through Different Machine Learning Approaches.

The mortality of severely burned patients can be predicted by multiple scores which have been created over the last decades. As the treatment of burn injuries and intensive care management have improved immensely over the last years, former prediction scores seem to be losing accuracy in predicting survival. Therefore, various modifications of existing scores have been established and innovative scores have been introduced. In this study, we used data from the German Burn Registry and analyzed them regarding patient mortality using different methods of machine learning. We used Classification and Regression Trees (CARTs), random forests, XGBoost, and logistic regression regarding predictive features for patient mortality. Analyzing the data of 1401 patients via machine learning, the factors of full-thickness burns, patient's age, and total burned surface area could be identified as the most important features regarding the prediction of patient mortality following burn trauma. Although the different methods identified similar aspects, application of machine learning shows that more data are necessary for a valid analysis. In the future, the usage of machine learning can contribute to the development of an innovative and precise predictive score in burn medicine and even to further interpretations of relevant data regarding different forms of outcome from the German Burn registry.

The effect of self-inflicted burns on ABSI score prediction power: A four-year prospective multicenter study of the German Burn Registry.

BACKGROUND: Suicide attempted by self-inflicted burns are associated with lower survival rates compared to accident related burns. OBJECTIVE: We investigate the relation between self-inflicted burns (SIB) and survival rates and how this relation is moderated by variables used to predict survival rates in the ABSI score, a widely used measure. Additionally, we compare the predicted survival rates by the ABSI score to the actual rates in our sample for SIB and accident patients. METHODS: In this prospective multicenter study data from the German Burn Registry are statistically analyzed using two sided t-test and multivariate linear regression models. RESULTS: 5330 patients (214 with SIB) met our inclusion criteria. We find a 6.8 percentage points lower survival rate for patients with SIB when we control for patient condition with the five ABSI components as covariates. These higher mortality rates can be explained by the higher rate of therapy restrictions for patients with self-inflicted burns. Additionally, different ABSI modifications can improve the predictive power of the score. CONCLUSION: Patients with SIB have lower survival rates compared to accident patients. Recently proposed modifications of the ABSI score can improve the accuracy of survival rate prediction for SIB.