ABSTRACT
Study objective: Current clinical practice assumes a negative computed tomography (CT) head sean result and a negative lumbar puncture result together are adequate to rule out subarachnoid hemorrhage in patients with acute headache. Our objective is to determine the sensitivity of a negative CT result combined with a negative lumbar puncture result to exelude subarachnoid hemorrhage. Methods: This prospective cohort study was conducted at 2 tertiary care emergency departments (EDs) during 3 years. We enrolled all patients who were older than 15 years, had a nontraumatic acute headache and normal neurologic examination result, and who had a CT head sean and a lumbar puncture ifthe CT result was negative (ie, no blood in the subarachnoid space). Patients were followed up with a structured telephone questionnaire 6 to 36 months añer their ED visit and electronic hospital records review to ensure no missed subarachnoid hemorrhage. We calculated sensitivity, specificity, and likelihood ratios of the strategy of CT and then lumbar puncture for subarachnoid hemorrhage. Results: Five hundred ninety-two patients were enrolled, including 61 with subarachnoid hemorrhage. The mean patient age was 43.6 years, with 59.1 percent female patients. Aneases of subarachnoid hemorrhage were identified on initial CT or lumbar puncture. One patient without subarachnoid hemorrhage was subsequently diagnosed with cerebral aneurysm, requiring surgery. The strategy classified patients with subarachnoid hemorrhage with sensitivity, specifícity, and positive and negative likelihood ratios (with 95 percent confidence intervals fCIsJ) of 100 percent (95 percent CI94 percent to 100 percent), 67 percent (95 percent CI 63 percent to 71 percent), 3.03 (95 percent CI 2.69 to 3.53), and 0. For diagnosis of subarachnoid hemorrhage or aneurysm, these were 98 percent (95 percent CI91 percent to 100 percent), 67 percent (95 percent CI63 percent to 71 percent), 2.98...