The epidemiology of vasospasm in children with moderate-to-severe traumatic brain injury.

OBJECTIVE: To gain a description of the prevalence and time course of vasospasm in children suffering moderate-to-severe traumatic brain injury.
DESIGN: A prospective, observational study was performed. Children with a diagnosis of traumatic brain injury, a Glasgow Coma Score less than or equal to 12, and abnormal head imaging were enrolled. Transcranial Doppler ultrasound was performed to identify and follow vasospasm. Diagnostic criteria included flow velocity elevation more than or equal to 2 sd above age and gender normal values for the middle cerebral and basilar arteries. Additional criteria required for vasospasm diagnosis in the middle cerebral artery was a ratio of flow in the middle cerebral artery to extracranial internal carotid artery more than or equal to 3.
INTERVENTIONS: None.
MEASUREMENTS AND MAIN RESULTS: Sixty-nine children were included. The prevalence of middle cerebral artery vasospasm in children with moderate traumatic brain injury (Glasgow Coma Score, 9-12) was 8.5% and was 33.5% in those with severe traumatic brain injury (Glasgow Coma Score, ≤ 8). The prevalence of basilar artery vasospasm in children with moderate traumatic brain injury was 3% and with severe traumatic brain injury was 21%. Mean time to onset of vasospasm was 4 days (± 2 d) in the middle cerebral arteries and 5 days (± 2.5 d) in the basilar artery. Mean duration of vasospasm in the middle cerebral artery was 2 days (± 2 d) and 1.5 days (± 1 d) in the basilar artery. Children in whom vasospasm developed were more likely to have been involved in motor vehicle accidents, had higher Injury Severity Scores, had fever at admission, and had lower Glasgow Coma Score scores. Good neurologic outcome (Glasgow Outcome Score Extended Pediatric version of ≥ 4) at 1 month from injury was seen in 76% of those with moderate traumatic brain injury without vasospasm and in 40% of those with vasospasm. In those with severe traumatic brain injury, good neurologic outcome was seen in 29% of those children without vasospasm and in 15% of those with vasospasm.
CONCLUSIONS: Vasospasm occurs in a sizeable number of children with moderate and severe traumatic brain injury. Children in whom vasospasm developed were more likely to have been involved in a motor vehicle accident, had higher Injury Severity Scores, had fever at admission, and had lower Glasgow Coma scores than in those whom vasospasm did not develop. Based on these findings, we recommend aggressive screening for posttraumatic vasospasm in these patients. Future studies should establish the relationship between vasospasm and long-term functional outcomes and should also evaluate potential preventative or therapeutic options for vasospasm in these children.