Purpose: Findings from the Asymptomatic Carotid Atherosclerosis Study (ACAS) indicate that carotid endarterectomy can be beneficial in symptom-free patients with 60% to 99% carotid artery stenosis. However, patients in ACAS who underwent contrast angiography (CA) before carotid endarterectomy were exposed to an additional 1.2% risk of stroke.
Methods: We used the methods of decision analysis to assess whether the overall 5-year stroke risk in symptom-free patients with suspected carotid artery disease can be reduced by preoperative imaging with magnetic resonance angiography (MRA) or duplex ultrasonography (DU). We compared four strategies for the preoperative evaluation of carotid artery stenosis in symptom-free patients: 1) CA alone, 2) MRA alone, 3) DU alone, and 4) MRA and DU with CA when the results of these tests disagree. Accuracies of MRA and DU were estimated from 81 patients exposed to all three procedures; stroke risks for patients with 60% to 99% carotid artery stenosis were obtained from ACAS.
Results: For predicting 60% to 99% carotid stenoses, sensitivity and specificity for noninvasive tests, optimized to reduce morbidity, were as follows: DU (0.96, 0.66), MRA (1.00, 0.76), DU/MRA (1.00, 0.86; 26% would require CA). The 5-year stroke risk of these four strategies in order of decreasing benefit was MRA, 6.17%; MRA/DU, 6.34%; DU, 6.35%; and CA, 7.12%. In sensitivity analyses, noninvasive tests were advantageous even if the stroke rate with CA diminished to 0.4%, or if the sensitivity and specificity of noninvasive tests fell to 70%.
Conclusion: The preoperative use of noninvasive tests resulted in a lower 5-year stroke risk compared with CA in symptom-free patients with suspected carotid artery stenosis.