OBJECT Traumatic head injury (THI) is a highly prevalent condition in the United States, and concern regarding excess radiation-related cancer mortality has placed focus on limiting the use of CT in the evaluation of pediatric patients with THI. Given the success of rapid-acquisition MRI in the evaluation of ventriculoperitoneal shunt malfunction in pediatric patient populations, this study sought to evaluate the sensitivity of MRI in the setting of acute THI. METHODS Medical records of 574 pediatric admissions for THI to a Level 1 trauma center over a 10-year period were retrospectively reviewed to identify patients who underwent both CT and MRI examinations of the head within a 5-day period. Thirty-five patients were found, and diagnostic images were available for 30 patients. De-identified images were reviewed by a neuroradiologist for presence of any injury, intracranial hemorrhage, diffuse axonal injury (DAI), and skull fracture. Radiology reports were used to calculate interrater reliability scores. Baseline demographics and concordance analysis was performed with Stata version 13. RESULTS The mean age of the 30-patient cohort was 8.5 ± 6.7 years, and 63.3% were male. The mean Injury Severity Score was 13.7 ± 9.2, and the mean Glasgow Coma Scale score was 9 ± 5.7. Radiology reports noted 150 abnormal findings. CT scanning missed findings in 12 patients; the missed findings included DAI (n = 5), subarachnoid hemorrhage (n = 6), small subdural hematomas (n = 6), cerebral contusions (n = 3), and an encephalocele. The CT scan was negative in 3 patients whose subsequent MRI revealed findings. MRI missed findings in 13 patients; missed findings included skull fracture (n = 5), small subdural hematomas (n = 4), cerebral contusions (n = 3), subarachnoid hemorrhage (n = 3), and DAI (n = 1). MRI was negative in 1 patient whose preceding CT scan was read as positive for injury. Although MRI more frequently reported intracranial findings than CT scanning, there was no statistically significant difference between CT and MRI in the detection of any intracranial injury (p = 0.63), DAI (p = 0.22), or intracranial hemorrhage (p = 0.25). CT scanning tended to more frequently identify skull fractures than MRI (p = 0.06). CONCLUSIONS MRI may be as sensitive as CT scanning in the detection of THI, DAI, and intracranial hemorrhage, but missed skull fractures in 5 of 13 patients. MRI may be a useful alternative to CT scanning in select stable patients with mild THI who warrant neuroimaging by clinical decision rules.

OBJECTIVEThe authors reviewed the transfer requests for isolated pediatric traumatic brain injuries (TBIs) at a Level I/II facility with the goal of identifying clinical and radiographic traits associated with potentially avoidable transfers that could be safely managed in a non–tertiary care setting.METHODSThe authors conducted a retrospective study of patients < 18 years of age classified as having TBI and transferred to their Level I tertiary care center over a 12-year period. The primary outcome of interest was identifying potentially avoidable transfers, defined as transfers of patients not requiring any neurosurgical intervention and discharged 1 hospital day after admission.RESULTSOverall, 70.8% of pediatric patients with isolated TBI did not require neurosurgical intervention or monitoring, indicating an avoidable transfer. Potentially avoidable transfers were associated with outside hospital imaging that was negative (86%) or showed isolated, nondisplaced skull fractures (86%) compared to patients with cranial pathology (53.8%, p < 0.001) as well as age ≤ 6 years (81% [negative imaging/isolated, nondisplaced fractures] vs 54% [positive cranial pathology], p < 0.001). The presence of headaches, nonfocal deficits, and loss of consciousness were associated with necessary transfer (p < 0.05). Patients with potentially avoidable transfers underwent frequent repeat CT studies (19.1%) and admissions to the pediatric intensive care unit (55.9%) but at a lower rate than those whose transfers were necessary (p < 0.001). Neurosurgical interventions occurred in 11% of patients with cranial pathology, which accounted for 17.9% of necessary transfers and 5.2% of all transfers.CONCLUSIONSIn the authors’ region, potentially up to 70% of interfacility transfers for pediatric brain trauma in the absence of other systemic injuries warranting surgical intervention may not require neurosurgical intervention and could be managed locally. No patients transferred with isolated, nondisplaced skull fractures or negative CT scans required neurosurgical intervention, and 86% were discharged the day after admission. In contrast, 11% of patients with CT scans indicative of cranial pathology required neurosurgical intervention. Age > 6 years, loss of consciousness, and nonfocal deficits were associated with a greater likelihood of needing a transfer. Further studies are required to clarify which patients can be managed at local institutions, but referring centers should practice overcaution given the potential risks.