Reproducibility of coronary plaque detection and characterization using low radiation dose coronary computed tomographic angiography in patients with intermediate likelihood of coronary artery disease (ReSCAN study)

Abstract

The purpose of this study is to evaluate the interscan, interobserver and intraobserver agreement for coronary plaque detection, and characterization using low radiation dose high-pitch spiral acquisition coronary CT angiography (CTA). Two experienced observers independently evaluated coronary CTA datasets from 50 consecutive patients undergoing two 128-slice dual source CT scans within 12 days. Mean (±SD) estimated radiation exposure was 1.5 ± 0.2 mSv per scan. Observers recorded the presence and characterization of coronary plaques as non-calcified or calcified. A "segment involvement score" (SIS) was computed by summing the numbers of segments with any coronary plaque per patient. Reproducibility was assessed using kappa (κ) statistics, paired t test and Bland-Altman analyses. Interscan, interobserver, and intraobserver agreement (κ-values) for detection of any or calcified plaques were 83-94% (κ-values 0.57-0.85), and 67-84% (0.31-0.67) for non-calcified plaques on a patient level. No significant difference was observed in mean interscan or interobserver SIS. Mean (95% CI) intraobserver SIS difference was -0.88 (-1.25; -0.51), P < 0.001, with limits of agreement from -4.7 to 2.9. Low radiation dose high-pitch coronary CTA permits detection of any or calcified plaques with high interscan, interobserver, intraobserver agreement. However, variability for the detection of non-calcified plaque is substantial.