Post Doctoral Research Fellow, Neurosurgeon Johns Hopkins University
Introduction: Traumatic brain injury (TBI) is a significant public health issue requiring extensive medical resources. Accurate, individualized risk assessments for extended length of stay (LOS), non-routine discharge, ICU or OR transfer, and direct ED discharge are essential for optimizing patient care. This study aimed to develop a tool to aid clinicians, social workers, and researchers in risk stratification for pediatric TBI patients.
Methods: A retrospective review of an institutional database was conducted to identify pediatric TBI cases using ICD-10 codes based on the CDC framework. Injury Severity Scores (ISS) and Social Deprivation Index (SDI) were documented, with higher scores indicating more severe injuries and greater deprivation. Multivariate regression with backward elimination was used to create the most parsimonious model, assessed for discrimination and calibration by AUC and Spiegelhalter’s z-test. An online calculator was developed using R’s Shiny package.
Results: In a cohort of 2,954 TBI cases (mean age 7.05 years), 28.4% experienced extended LOS, 8.3% had non-routine discharge, 23.4% were transferred to ICU/OR, and 52.3% had direct ED discharge. The ISS had a mean ± SD of 8.11 ± 7.23, and the SDI had a mean ± SD of 54.63 ± 30.25. In multivariate logistic regression, higher ISS scores were correlated with increased odds of extended LOS, non-routine discharge, higher ED transfer to ICU/OR, and lower direct ED discharge (all p < 0.001). Higher SDI scores were associated with higher odds of non-routine discharge (p = 0.030). All models demonstrated good predictive ability, achieving AUCs of 0.89, 0.87, 0.89, and 0.88, respectively. Spiegelhalter z-tests indicated adequate fit (p > 0.05). The models were used to develop an open-access online calculator at https://jhpedsnsgy.shinyapps.io/JHPedsNSGY/.
Conclusion : By integrating readily accessible data in the ED, these predictive models and the online calculator can help clinicians deliver personalized risk assessments for pediatric TBI patients.
.jpg)
