Selective computed tomographic angiography in traumatic subarachnoid hemorrhage: a pilot study

Computed tomographic angiography (CTA) tends to be overused in patients with traumatic subarachnoid hemorrhage (tSAH) to rule out intracranial aneurysmal disease. We hypothesized that there are two exclusive subsets of patients with tSAH that maybe at increased risk for aneurysm and thus should undergo CTA, those "found down" with an unknown mechanism of injury and those with "central subarachnoid hemorrhage" (CSH, in the subarachnoid cisterns and Sylvian fissures). This pilot study was performed to provide more information on the validity of our hypothesis.

A retrospective analysis was performed on trauma patients with tSAH who underwent CTA of the brain. Patients presented to a level I trauma center from January 2008-December 2012. Our principal outcome was the diagnosis of an intracranial aneurysm. Student t-test, chi-squared test, Mann-Whitney U test, and binary logistic regression were used for statistical analysis, with significance set at alpha = 0.05.

Of 617 total patients with tSAH, 186 patients underwent CTA. Majority of patients were male (64%), with median age of 56 y. Median Glasgow coma scale on presentation was 15, and the median injury severity score was 16. Thirteen patients (6.99%) had an aneurysm on the follow-up CTA. Of those, 8 of 13 (61.5%) were felt to have presented with a ruptured aneurysm. Among those, 5 of 8 (62.5%) sustained a fall and 3 of 8 (37.5%) resulted from a motor vehicle crash. Among the 14 patients (7.5%) "found down", none had an aneurysm. All eight patients with a ruptured aneurysm (100%) had CSH, whereas none of the five patients with unruptured aneurysm had CSH. On multivariate analysis, suprasellar cistern hemorrhage was the most predictive noncontrast computed tomographic finding with regard to aneurysm presence (odds ratio, 4.78; 95% confidence interval, 1.33-17.1). Patients with an aneurysmal disease had a significantly higher mean arterial pressure on presentation (median, 115 mm Hg) than those without an aneurysm (median, 96 mm Hg, P < 0.05). Of the eight ruptured aneurysms, six underwent neurosurgical clipping or coiling, one underwent a ventriculostomy, and one underwent a craniotomy for evacuation of hemorrhage.

These preliminary data support a more selective approach to screening CTAs in patients with tSAH. CTA should be used in those patients with CSH regardless of mechanism of injury. A more restrictive approach should be used in patients with only peripheral subarachnoid hemorrhage.