Study objective: Emergency physicians use noncontrast cranial computed tomographic (CT) imaging of headache patients to identify subarachnoid hemorrhage caused by aneurysms or arteriovenous malformations. Given sufficiently high sensitivity, CT imaging could be used as a definitive diagnostic study in these patients. The purpose of this study is to determine the sensitivity of noncontrast cranial CT in detecting all spontaneous subarachnoid hemorrhages and those caused by aneurysm or arteriovenous malformation.
Methods: This was a retrospective review performed at an urban tertiary academic emergency department (ED). Using a combination of noncontrast cranial CT radiology coding, lumbar puncture results, International Classification of Diseases, Ninth Revision discharge diagnosis, and medical record review, we identified all patients who presented to a tertiary care academic ED from August 1, 2001, to December 31, 2004, with spontaneous subarachnoid hemorrhage. We determined whether patients were diagnosed by cranial CT or lumbar puncture, the presence of headache and level of consciousness at ED presentation, and whether or not they had an aneurysm or arteriovenous malformation.
Results: We identified 149 patients who were diagnosed with spontaneous subarachnoid hemorrhage during the study period. Noncontrast cranial CT scan diagnosed 139 patients, and 10 were diagnosed with lumbar puncture. This yielded an overall CT scan sensitivity of 93% (95% confidence interval [CI] 88% to 97%). Of the 149 with subarachnoid hemorrhage, 117 (79%) had aneurysm or arteriovenous malformation; cranial CT scan demonstrated subarachnoid hemorrhage in 110 of the 117, for a sensitivity of 94% (95% CI 88% to 98%). For the 67 patients presenting with headache and normal mental status who had a subarachnoid hemorrhage and vascular lesions (either aneurysm or arteriovenous malformation), the sensitivity of cranial CT scan was 91% (95% CI 82% to 97%).
Conclusion: Noncontrast CT imaging exhibits inadequate sensitivity to serve as a sole diagnostic modality in detecting spontaneous subarachnoid hemorrhage caused by aneurysm or arteriovenous malformation.