Purpose: The Asymptomatic Carotid Atherosclerosis Study (ACAS) showed that carotid endarterectomy reduces stroke risk in symptom-free patients with 60% or greater internal carotid artery (ICA) stenosis. This will surely lead to the performance of an increased number of screening duplex examinations. Assuming that positive study results will lead to arteriography or endarterectomy and keeping in mind the modest benefit for prophylactic endarterectomy demonstrated by ACAS (absolute risk reduction for ipsilateral stroke of 5.8% at 5 years), duplex criteria for 60% or greater ICA stenosis must have high positive predictive values (PPV). Determining criteria for 60% or greater stenosis, which emphasized high accuracy and PPV, forms the basis for this study.
Methods: Stenoses detected by angiography in 352 ICAs were blindly compared with those detected by duplex scanning. Duplex criteria were determined for highest overall accuracy in detection of 60% or greater ICA stenosis and for 95% or greater PPV.
Results: Maximal accuracy for detection of 60% or greater stenosis was 90%. This was achieved by the combination of a peak systolic velocity of 260 cm/sec or greater and an end diastolic velocity of 70 cm/sec or greater (sensitivity 84%, specificity 94%, PPV 92%). The 95% PPV for 60% or greater stenosis results from combining peak systolic velocity of 290 cm/sec or greater and end diastolic velocity of 80 cm/sec or greater.
Conclusions: With use of these criteria duplex scanning accurately detects with high PPVs the threshold level of ICA stenosis defined in ACAS as receiving stroke reduction benefit from prophylactic carotid endarterectomy. These criteria should be useful for carotid artery screening and minimizing unneeded intervention.