Vasospasm Surveillance by a Simplified Transcranial Doppler Protocol in Traumatic Brain Injury

Abstract

To detect post-traumatic vasospasm in patients with traumatic brain injury (TBI), we implemented a simplified transcranial Doppler (TCD) surveillance protocol in a neurointensive care setting. In this study, we evaluate the yield of this protocol. Adult patients with TBI admitted to the neurointensive care unit were examined with TCD by 2 intensive care nurses trained in TCD examinations. Flow velocities of the middle cerebral arteries were recorded. TCD suspected vasospasm was defined as the mean flow velocity >120 cm/s, and when detected, the protocol recommended a supplementary computed tomography angiography. The rate of detection of TCD suspected vasospasm and the subsequent rate of radiological diagnosis of vasospasm were recorded. In multivariate logistic regression analysis, we evaluated age, initial Glasgow Coma Scale, craniotomy, and decompressive craniectomy as potential predictors of developing increased TCD velocity. A total of 84 patients with TBI with a median initial Glasgow Coma Scale score of 6 were examined by TCD. TCD suspected vasospasm was found in the middle cerebral arteries of 18% of examined patients. Two-thirds of patients with TCD suspected vasospasm were investigated with a subsequent computed tomography angiography, and 80% of these patients received a radiological diagnosis of vasospasm. In logistic regression analysis, decompressive craniectomy was significantly associated with increased risk of developing TCD suspected vasospasm (odds ratio: 11.57, 95% confidence interval: 2.59-51.73, P= 0.001). The implementation of a simplified TCD surveillance protocol in a neurointensive care setting yielded an 18% detection rate of TCD suspected vasospasm. In our cohort of patients with TBI, decompressive craniectomy was associated with increased risk of developing TCD suspected vasospasm.