Evaluation of a National Sample of 16,671 Pediatric Burn Admissions: Identifying Predictors of Non-accidental Pediatric Burns

Background: Burn injuries remain one of the leading causes of injury and death in children. Studies have demonstrated a higher mortality for pediatric burns associated with non-accidental injury. Using data from a burn registry, our study aimed to discern potential factors associated with non-accidental burn injuries. Methods: We utilized the American Burn Association database from 2016 to 2018, which collects data from over one hundred burn centers across the United States, to evaluate a large pediatric burn population. Patients aged <= 14 years were analyzed. The population was then divided into suspected non-accidental versus accidental burn injuries. A multivariable logistic regression model was utilized to evaluate for predictors of burn injuries. Additional models were used to assess the relationship between suspected nonaccidental burn injury and mortality, intensive care unit (ICU) stay, and hospital length of stay. Results: 16,671 pediatric patients were included. Of those, 1228 (7.4%) patients suffered non-accidental burn injury. A majority of children who sustained non-accidental burn injury were younger, non-white, and sustained scald burns. The regression model demonstrated predictors for non-accidental burn injuries included younger age, Black race, chemical/corrosion burns, government insurance, and increased total body surface area (TBSA) burn. Overall mortality for the population was 0.5%. Conclusion: This evaluation of a national burn registry reveals differences in pediatric patients sustaining non-accidental burns compared to accidental burns. The findings in this study identify pediatric populations at risk for suspected non-accidental burn injuries which may assist in preparing the families for expectations after admission for a burn injury. Level of Evidence: III. (c) 2024 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.