Optimal Peak Systolic Velocity Thresholds for Predicting Internal Carotid Artery Stenosis Greater than or Equal to 50%, 60%, 70%, and 80%

Background: We aimed to determine the optimal peak systolic velocity (PSV) thresholds for predicting >= 50%, >= 60%, >= 70%, and >= 80% internal carotid artery (ICA) stenosis. Methods: Patients who underwent both carotid ultrasonography and cerebral angiography during hospitalization were consecutively and retrospectively enrolled. The degree of ICA stenosis was calculated using the North American Symptomatic Carotid Endarterectomy Trial method on cerebral angiography. The optimal PSV thresholds for predicting >= 50%, >= 60%, >= 70%, and >= 80% ICA stenosis were determined using receiver operating characteristic curves and the Youden index. Results: A total of 127 ICAs were analyzed. The optimal PSV thresholds for predicting >= 50%, >= 60%, >= 70%, and >= 80% ICA stenosis were 130 cm/s (sensitivity, 95%; specificity, 85%; positive predictive value [PPV], 75%; negative predictive value [NPV], 97%; overall accuracy, 88%), 160 cm/s (sensitivity, 91%; specificity, 94%; PPV, 83%; NPV, 97%; overall accuracy, 93%), 200 cm/s (sensitivity, 96%; specificity, 95%; PPV, 83%; NPV, 99%; overall accuracy, 95%), and 270 cm/s (sensitivity, 89%; specificity, 94%; PPV, 74%; NPV, 98%; overall accuracy, 94%), respectively. Conclusions: The optimal PSV thresholds for predicting >= 50%, >= 60%, >= 70%, and >= 80% ICA stenosis were 130 cm/s, 160 cm/s, 200 cm/s, and 270 cm/s, respectively. All of them had high diagnostic accuracies.