The pediatric Brain Injury Guidelines: a retrospective clinical validation study.

Abstract

Objective: Pediatric traumatic brain injury (TBI) represents a significant public health concern and source of resource utilization. The aim of this study was to establish the ability of the previously published pediatric Brain Injury Guidelines (pBIG) to identify patients with traumatic intracranial hemorrhage (ICH) who might not require routine repeat neuroimaging, neurosurgical consultation, or hospital admission in a large level I and level II trauma cohort.

Methods: Pediatric patients who presented with traumatic ICH between 2018 and 2022 at the included institutions were retrospectively reviewed and sorted into pBIG categories using clinical and radiographic criteria. Nonaccidental trauma was excluded. Repeat neuroimaging and results, neurosurgical intervention, length of stay (LOS), and 30-day mortality and re-presentation to healthcare were collected as outcomes.

Results: A cohort of 955 patients (median age 7.0 years, with 64.5% of patients being male) were included. Overall, 9.7% of patients had pBIG 1 injuries, 30.0% had pBIG 2 injuries, and 60.2% had pBIG 3 injuries. A total of 368 (38.5%) of patients underwent repeat neuroimaging, of whom 144 (39.1%) showed progression of hemorrhage. Neurosurgical intervention was performed in 129 (13.5%) patients, with 127 (98.4%) of them meeting pBIG 3 criteria on arrival. The two remaining patients met pBIG 2 criteria on arrival and then progressed to meet pBIG 3 criteria within 24 hours. Patient meeting pBIG 3 criteria were significantly more likely to have progression on repeat imaging, require neurosurgical intervention, and experience 30-day mortality (p < 0.001). Within the pBIG 3 cohort, there was not a significant relationship between progression on repeat imaging and the need for intervention (p = 0.61). Post hoc pairwise testing of individual radiographic pBIG groupings revealed pBIG 3 criteria for all categories except subarachnoid hemorrhage (SAH) to be predictive of need for neurosurgical intervention (p < 0.05).

Conclusions: Algorithmic management of mild TBI is beneficial to patient care. With zero and near-zero rates of neurosurgical intervention and mortality in patients with pBIG 1 and pBIG 2 injuries, respectively, the pBIG are valid in stratifying a larger and broader population of pediatric TBI patients. In contrast to other pBIG 3-defined compartment ICHs, "scattered" SAH does not correlate with need for neurosurgical intervention. However, these guidelines have the ability to safely improve care and decrease unnecessary resource utilization without negatively affecting patient outcomes. Utilization of guidelines of any sort are not intended to supersede clinical judgment. Prospective studies are needed.