Effect of coronary stenosis severity on variability of quantitative arteriography, and implications for interventional trials

Abstract Quantitative coronary arteriography is now routinely used in studies assessing arterial remodelling in response to interventions such as lipid lowering and percutaneous transluminal coronary angioplasty. Although this methodology provides both relative and absolute measurements that can be evaluated statistically as continuous variables, it is, however, often desirable or necessary to establish categorical responses to the interventions based on the variability inherent in the quantitative methods (i.e., it is often necessary to state whether individual patients, as opposed to entire groups, have “restenosed,” “progressed,” “regressed,” and so forth, and these categorical designations are based on critical limits of variability). 1,2 These limits are generally based on an analysis of a rather diverse range of lesion severity, and the question arises as to whether the variability of measuring changes in lesion or segment morphology is affected by the initial severity of the lesion being studied. It is conceivable that variability may be greater at 1 end than at the other end of the spectrum of lesion severity. If this is true, then the use of 1 critical value to designate whether a subject has responded in a certain way would be inappropriate, and this designation would be better achieved by using criteria that more directly reflect the measurement variability of lesions of a specified, initial severity. This report examines whether the variability of measuring morphologic parameters of stenoses is affected by the baseline severity of the lesion.