Prognostic value and reproducibility of measurements of carotid stenosis. A comparison of three methods on 1001 angiograms. European Carotid Surgery Trialists' Collaborative Group.

BACKGROUND AND PURPOSE The use of three methods of measuring carotid stenosis, which produce different values on the same angiograms, has caused confusion and reduced the generalizability of the results of research. If the results of future studies are to be properly applied to clinical practice, and if noninvasive methods of imaging are to be properly validated against angiography, a single, standard method of measurement of stenosis on angiograms must be adopted. This standard method should be selected on the bases of its ability to predict risk of ipsilateral carotid distribution ischemic stroke and its reproducibility. METHODS The method of measurement of carotid stenosis used in the European Carotid Surgery Trial (ECST), that used in the North American Symptomatic Carotid Endarterectomy Trial (NASCET), and a method based on the measurement of the common carotid (CC) lumen diameter were studied. Their use in the prediction of ipsilateral carotid distribution ischemic stroke was assessed in 1001 consecutively selected patients randomly assigned to medical treatment in the ECST. Carotid stenosis was measured by two observers working independently, using all three methods, on the angiographic view that showed the most severe stenosis of the symptomatic carotid bifurcation. Interobserver agreement was determined, and 50 angiograms were remeasured to determine intraobserver agreement. RESULTS There was little difference in the ability of the three methods to predict ipsilateral carotid distribution ischemic stroke. The CC method was consistently the most reproducible of the three, particularly for stenosis in the clinically important range of 50% to 90%. CONCLUSIONS The CC method of measurement should be adopted as the standard method of measuring the degree of carotid stenosis on angiograms.