The timing of computed tomography in acute stroke: a practice audit.

OBJECTIVE

To compare the patterns of practice and diagnostic yields for early and late computed tomography (CT) in patients with acute stroke.

PATIENTS AND METHODS

Among the 191 consecutive patients with acute stroke admitted to a university teaching hospital from Sept. 1, 1989, to Sept. 1, 1993, charts were available for review for 185. In addition to the timing and results of CT in these patients, data were collected on the demographic features of the patient group and the presence of any cardiovascular risk factors or other features suggestive of a nonstroke cause for the neurologic deficit.

RESULTS

Of the 185 patients for whom charts were available, 177 (95.7%) underwent CT. In 107 (60.4%) of the cases, scanning was performed early, within 24 hours of the onset of neurologic deficit, and only 25 (23.4%) of these scans demonstrated a lesion. Of the 70 initial scans obtained more than 24 hours after the deficit onset (delayed scanning), 41 (58%) revealed a diagnostic lesion (relative probability of finding a lesion with early vs delayed scanning, 0.40; 95% confidence interval 0.27 to 0.59; p < 0.0001, Fisher's exact test). Of the 107 patients who underwent early scanning, 45 (42.0%) underwent repeat CT, and previously unrecognized lesions were seen in 28 (62%) of these. Scanning was repeated in only 17 (24%) of the 70 patients who underwent delayed initial scanning (p = 0.02), and previously unrecognized lesions were seen in only 4 of these (24%). All lesions documented on repeat scans after initially nondiagnostic scanning were ischemic. Patients presenting with features thought to justify early CT were no more likely to undergo this intervention than those without such features (68% and 56% respectively, p = 0.11).

CONCLUSIONS

Initial CT was often carried out within 24 hours of the onset of deficit in patients with suspected acute stroke at this institution, CT was a low-yield investigation, and scanning was frequently repeated. Although the literature suggests that clinical features can distinguish that subset of patients who should undergo urgent neuroimaging, prospective studies are needed to establish the indications for early CT in patients presenting with acute neurologic deficit.