Serial Electron Beam CT Measurements of Coronary Artery Calcium: Has Your Patient's Calcium Score Actually Changed?

Abstract

The objective of our study was to develop a model for determining the smallest statistically significant change in the coronary artery calcium score (CAC) between serial measurements in a given subject. We assembled a convenience sample of 2,217 pairs of repeated electron beam CT coronary calcium scans acquired in quick succession. Each scan consisted of forty 100-msec, 3-mm sections obtained at 60% of the ECG R-R interval. A single observer quantified calcium in each scan independent of knowledge of calcium quantity in the repeated scan. We then modeled a relationship between the variation of the differences between repeated measurements of calcium and the magnitude of the calcium score and formulated 95% repeatability coefficient equations for the Agatston and volumetric CAC score. The equations allow determining the smallest statistically significant interval change in the calcium score between two serial measurements in a given subject. In a subject with measurable CAC at baseline, the smallest statistically significant interval change is +/- (4.930 x square root of baseline Agatston CAC score) or +/- (3.445 x square root of baseline volumetric CAC score). In a subject with no measurable CAC at baseline, a follow-up CAC score exceeding 11.6 Agatston units or 9.5 mm3 qualifies for statistically significant progression. The results were similar in men and women. By examining repeatability of quantitative electron beam CT measurements of coronary calcium as a function of the magnitude of the calcium score, we developed a model to determine the smallest statistically significant change between serial measurements in a given subject.