NESVS16. Carotid Duplex Velocity Criteria Recommended by the Society of Radiologists in Ultrasound and Endorsed by the Intersocietal Accreditation Commission Lack Predictive Ability for Identifying High-Grade Carotid Artery Stenosis

Abstract

Carotid duplex ultrasound is the first-line imaging modality for characterizing degree of carotid artery stenosis. The Intersocietal Accreditation Commission, in published guideline documents, has endorsed use of the Society of Radiologists in Ultrasound (SRU) criteria to characterize ≥70% stenosis: peak systolic velocity (PSV) ≥230 cm/s. We sought to perform a validation of the SRU criteria using computed tomography (CT) angiography as a “gold standard” imaging modality and to perform a sensitivity analysis to determine optimal velocity criteria for identifying ≥80% stenosis. We queried all carotid duplex ultrasound examinations performed at our institution between 2008 and 2017. Patients with ≥70% carotid stenosis, based on previous criteria, and further narrowed to those with minimum PSV ≥300 cm/s or end-diastolic velocity (EDV) ≥100 cm/s were identified. Of these patients, those who also had CT angiography of the neck within 1 year formed the study cohort. Patients who underwent carotid revascularization between the two imaging dates were excluded. Degree of stenosis, as reported from the CT angiogram, was considered the true degree of stenosis. Receiver operating characteristic curves were generated to evaluate the SRU criteria and to identify the optimal discrimination threshold for high-grade carotid stenosis. Of 37,204 carotid duplex ultrasound examinations, 3478 arteries met criteria for ≥70% stenosis. Of these, 344 patients had CT angiography within 1 year of the carotid duplex ultrasound examination (mean time between studies, 55 days; standard deviation, 6.5), and 240 (69.8%) were consistent with ≥80% carotid stenosis. The predictive ability of the SRU criteria to identify ≥70% stenosis was poor, with an area under the curve of 0.51. A sensitivity analysis to identify ≥80% stenosis demonstrated the optimal discrimination threshold to be PSV ≥450 cm/s or EDV ≥120 cm/s, with an area under the curve of 0.66 (Fig). In this validation study, the SRU criteria, endorsed by the Intersocietal Accreditation Commission, to identify ≥70% carotid stenosis have no predictive value. For detection of ≥80% stenosis, the optimal criteria are PSV ≥450 cm/s and EDV ≥120 cm/s. This study demonstrates the critical importance of carotid duplex ultrasound examination validation.