A comparative analysis of the Trauma and Injury Severity Score and the Injury Severity Score in predicting high-value care outcomes in children with traumatic brain injury.

Abstract

OBJECTIVE Predicting high-value care outcomes is crucial in managing pediatric traumatic brain injuries (TBIs), where timely and accurate prognostication can significantly influence treatment decisions and resource allocation. This study aimed to enhance understanding of how well scoring systems such as the Trauma and Injury Severity Score (TRISS) can forecast high-value care outcomes. Furthermore, the authors compared the predictive power of TRISS with the routinely used Injury Severity Score (ISS). METHODS The authors performed a retrospective review of their institutional database from June 2016 to June 2023 to identify cases of TBI based on a modified Centers for Disease Control and Prevention framework. Prolonged length of stay (LOS) was defined as a hospital stay falling into the upper quartile of the overall cohort. Discharge to an inpatient rehabilitation facility, acute care hospital, or foster care or death was defined as a nonroutine discharge disposition. Emergency department (ED) transfer to the intensive care unit (ICU) or operating room (OR) was defined as a proxy for severity of injuries. Multivariate logistic regression models were used to explore the association between ISS, TRISS, and high-value care outcomes. The DeLong test was used to assess the differences between the areas under the receiver operating characteristic curve (AUROCs). RESULTS This study included 2705 patients with a mean age ± SD of 7.28 ± 5.46 years (63% male). In the overall cohort, 28% experienced prolonged LOS, 7% had a nonroutine discharge disposition from the hospital, and 23% were transferred to the ICU/OR from the ED. In multivariate regression models, both TRISS and ISS were correlated with higher odds of prolonged LOS, nonroutine discharge disposition, and transfer to the ICU/OR from the ED (all p < 0.001). TRISS had a significantly greater AUROC than ISS for nonroutine discharge disposition (0.883 vs 0.849, p < 0.001) and transfer to the ICU/OR (0.898 vs 0.887, p = 0.045), but this result was not significant for prolonged LOS (0.873 vs 0.880, p = 0.140). CONCLUSIONS TRISS and ISS are effective tools for predicting high-value care outcomes in pediatric TBI. Utilizing these resources can assist healthcare providers in making informed, risk-adjusted predictions.