Role of arteriography in the selection of patients for carotid endarterectomy.

Background: Duplex ultrasonography is used increasingly to select patients for carotid endarterectomy. This study aims to clarify whether arteriography is still required.

Methods: A total of 272 patients in whom duplex imaging indicated significant disease of the internal carotid artery underwent arteriography and the results were compared. Regarding the arteriogram as the 'gold standard', three aspects were considered: the accuracy of duplex ultrasonography in detecting significant stenosis, the ability of duplex imaging to identify patients who should have arteriography and whether the arteriogram provided important additional information that might influence a decision to operate.

Results: Three patients (1 per cent) suffered a permanent neurological deficit as a result of arteriography. Of the 272 patients with significant stenosis identified by duplex ultrasonography, 241 (89 per cent) were confirmed as significant on arteriography. Duplex imaging was not able to predict accurately which arteriograms would provide useful additional information (sensitivity 59 per cent, specificity 65 per cent), whereas 89 arteriograms (33 per cent) contained information that might have influenced subsequent management.

Conclusion: Duplex imaging is unable to detail the anatomy or determine the extent of carotid disease and may therefore lead to an inappropriate decision to perform carotid endarterectomy. Duplex imaging alone is not adequate before carotid endarterectomy.