Algorithms to measure carotid intima-media thickness in trials: a comparison of reproducibility, rate of progression and treatment effect.

摘要

BACKGROUND: Current ultrasound protocols to measure carotid intima-media thickness (CIMT) in trials differ considerably. The best CIMT protocol would be one that combines high reproducibility, a large and precise estimate of the rate of CIMT progression and a large and precise estimate of the treatment effect. We performed a post-hoc analysis to determine the best algorithm for determining CIMT using data from the METEOR study, a randomized double-blind, placebo-controlled study of the effect of rosuvastatin on CIMT progression in 984 low coronary heart disease risk individuals with increased CIMT. METHODS: CIMT information was collected from two walls (near and far wall), three segments (common carotid, bifurcation and internal carotid artery), five different angles (for the right carotid artery - 60, 90, 120, 150, and 180 degrees on the Meijer's carotid arc; for the left - 300, 270, 240, 210, and 180 degrees) of two sides (left and right carotid artery), resulting in possibly (2 x 3 x 5 x 2 =) 60 measurements. On the basis of combinations of these measurements, we built 66 different ultrasound protocols to estimate a CIMT for each individual (22 protocols for mean common CIMT, 44 protocols for mean maximum CIMT). For each protocol we assessed reproducibility [intraclass correlation (ICC), mean difference of duplicate scans], 2-year progression rate in the placebo group with its corresponding standard error and treatment effect (difference in CIMT progression between rosuvastatin and placebo) and its corresponding standard error. RESULTS: Data of duplicate ultrasound examinations at baseline and end of study were available for 688 participants (70% of 984). The ICC based on duplicate baseline examinations ranged from 0.81 to 0.95. CIMT progression rates in the placebo group ranged from 0.0046 to 0.0177 mm/year, with SE ranging from 0.00134 to 0.00337. Treatment effects ranged from 0.0141 to 0.0388 mm/year. The protocols with highest reproducibility, highest CIMT progression/precision ratio and highest treatment effect/precision ratio were those measuring both near and far wall for at least two angles. CONCLUSION: Ultrasound protocols that include CIMT measurements at multiple angles of both near and far wall give the best balance between reproducibility, rate of CIMT progression, treatment effect and their associated precision in this low-risk population with subclinical atherosclerosis.