Study objective: This study investigated the hypothesis that modern computed tomographic (CT) imaging is sufficient to exclude subarachnoid hemorrhage (SAH) in patients with severe headache.
Methods: All 38,730 adult patients who presented to Hermann Hospital in Houston, Texas, during a 16-month period were prospectively screened to detect those with "the worst headache of my life." Two neuroradiologists blinded to the study hypothesis interpreted the CT scans. Patients with negative scans underwent comprehensive cerebrospinal fluid (CSF) analysis including cell count in first and last tubes, visual and spectrophotometric detection of xanthochromia, and CSF D-dimer assay.
Results: A chief complaint of headache was elicited in 455 patients, and 107 of these had "worst headache" and were enrolled in the study. CT-confirmed SAH was found in 18 of the 107 (17%). Only 2 patients (2.5%, 95% confidence interval, .3% to 8.8%) had SAH detected by CSF analysis among those with negative CT imaging result. CSF spectrophotometric detection was the most sensitive test for blood. Three patients with less than 6 red blood cells in tube 1 had positive spectrophotometric results, but in all 3, tube 4 was negative on spectrophotometric analysis, suggesting a high false-positive rate.
Conclusion: Modern CT imaging is sufficient to exclude 97.5% of SAH in patients presenting to the ED with "worst headache" symptoms.