Measuring carotid stenosis. Time for a reappraisal.

BACKGROUND AND PURPOSE: Data from recent multicenter carotid endarterectomy trials have questioned the validity and reliability of Doppler ultrasound in the assessment of carotid stenosis. METHODS: We prospectively analyzed 45 patients undergoing carotid angiography to compare the North American Symptomatic Carotid Endarterectomy Trial (NASCET) and European Carotid Surgery Trial (ECST) methods of measuring carotid stenosis with those of direct visualization ("eyeballing") and duplex ultrasound. Linear NASCET and ECST measurements were also converted into area using the pi r2 function and termed "squared NASCET" (N2) and "squared ECST" (E2). In 15 of 45 patients undergoing carotid endarterectomy, the carotid plaque was removed intact, sectioned, and photographed for computer measurement of cross-sectional area. Comparison of this "gold standard" was then made to each method of measurement. RESULTS: Comparison between duplex and the various angiographic measurement techniques revealed significant differences between NASCET and duplex (P < .0001), ECST and duplex (P < .01), and E2 and duplex (P < .01) but not between N2, eyeballing, and carotid duplex methods. Even the NASCET and ECST methods themselves differed significantly (P < .006). When comparison was made with computerized planimetric measurements of the carotid plaque, there were significant differences for both NASCET (P < .0007) and ECST (P < .007). Correlation was demonstrated only between planimetry and N2, E2, and duplex. CONCLUSIONS: NASCET and ECST angiographic methods of measurement consistently underestimate the "true" anatomic stenosis. As such, they represent only "indexes" of carotid stenosis severity. Duplex provides a more accurate measurement of carotid stenosis.