Determining the sensitivity of computed tomography scanning in early detection of subarachnoid hemorrhage

This study aims to determine the sensitivity of modern computed tomography (CT) scanners in detecting subarachnoid hemorrhage (SAH) and to determine whether there is a continued need for lumbar puncture to exclude the diagnosis.

This retrospective study was conducted from January 2000 to December 2005. The study population consisted of all patients referred on suspicion of SAH or with verified SAH. All medical records were reviewed together with results from CT scan, angiography, and cerebral spinal fluid analysis. Clinical history, examination findings, and time from onset of symptoms until CT scan (days) were recorded. Patients with a negative CT scan had a lumbar puncture done.

A total of 499 patients were included. In 203 patients the diagnosis was excluded by a negative CT scan and negative lumbar puncture. Two hundred ninety-six patients were found to have a SAH. The diagnosis in 295 of these patients was based on a positive CT scan. In a single patient, on day 6, the diagnosis was based on a positive lumbar puncture. From day 1 to day 5, CT scanning was found to have a sensitivity of 100%. Overall, CT scanning had a sensitivity of 99.7 % (95% confidence interval: 98.1-99.99%).

CT scanning is excellent for diagnosing SAH. We demonstrate that in the first days after ictus a negative CT scan is sufficient to exclude SAH. Data do not allow for any specific cutoff point to be made. We suggest leaving out lumbar puncture in the first 3 days after ictus if the results of the CT scan are negative.