Authors: Oliver Tang; James Yoon, BS; Mohamed Labib, MD; Michael Lawton, MD (Providence, RI)

Introduction: Previous research has documented better outcomes at high-volume centers for adult traumatic brain injury patients. However, the volume-outcome relationship for pediatric neurotrauma patients remains poorly characterized. Methods: We identified all pediatric admissions for traumatic intracranial hemorrhage in the National Inpatient Sample database from 2002 to 2014. We classified hospitals within each year by pediatric neurotrauma volume as high-volume (top 20%) or low-volume (bottom 80%). Our outcome variables were routine discharge, length of stay, and inpatient complications. We performed survey-weighted univariate and multivariate analyses to assess national outcomes and adjust for the following confounding variables: patient age, sex, race, insurance, household income, risk of mortality and severity of illness scores, weekend admission, transfer status, hospital ownership, location and teaching status, region, number of presenting hemorrhage subtypes (epidural, subdural, subarachnoid, intraparenchymal), external ventricular drain placement, and performance of decompressive craniectomy or craniotomy. We maintained statistical significance at P <0.05. Results: There were 97,344 total admissions for pediatric traumatic intracranial hemorrhage. The majority of admissions (n=63,655, or 65%) were treated at high-volume hospitals. Patients at high-volume hospitals were more likely than low-volume hospital patients to have higher risk of mortality and severity of illness scores, present with multiple hemorrhage subtypes, and undergo surgical intervention over medical management (all P <0.01). Following multivariate adjustment, admission at a high-volume hospital was associated with increased likelihood of routine discharge (odds ratio=1.77, P <0.001). Length of stay was not associated with volume. Finally, pediatric traumatic intracranial hemorrhage admissions at high-volume centers exhibited decreased odds of developing overall complications (odds ratio=0.77, P =0.007) and respiratory system failure (odds ratio=0.72, P =0.003). Conclusion: Although high-volume hospitals had more critical pediatric neurotrauma patients, these institutions had more frequent routine discharges and lower complication rates than low-volume centers. These findings suggest an increased need to discuss centralization of neurosurgical care and development of transfer networks.