BACKGROUND: Coronary artery calcification (CAC) is a specific feature of atherosclerotic cardiovascular disease. Continued improvements in computed tomography (CT) technology may complicate CAC progression assessment in longitudinal studies that change scanner protocols. Herein, we assessed agreement in CAC measures derived from consecutive non-ECG-gated, chest CT scans performed on the same visit and differing only in radiation dose. METHODS: Agreement in quantitative CAC scores and CAC score categories were determined from a subset of COPDGene participants (N = 115; age: 66 (9) years; 50% male) who received full- (mean effective radiation dose of 6.5 (1.0) mSv) and reduced- (1.5 (0.7) mSv) dose non-ECG-gated CT chest scans at maximal inspiration during their Phase 2 COPDGene study visit. RESULTS: = 0.96, SE = 0.01, 95% CI: 0.95-0.98). With the full-dose protocol as reference, the reduced-dose protocol had high sensitivity and specificity for determining CAC presence, 94.5% (95%CI: 87.6%-98.2%) and 100% (95% CI: 85.8%-100%), respectively. A Bland-Altman analysis revealed a non-significant higher CAC score with the full-dose protocol: mean bias between protocols of 26 with 95% limits of agreement of -185 to 237. CONCLUSION: CAC scores from non-ECG-gated, chest CT scans are robust to changes in radiation dose at the group level; however, when CAC burden increases, the variability of the CAC score differences increases. GOV IDENTIFIER: NCT00608764.
Moore et al. (Fri,) studied this question.