Probabilistic Model for Prediction of Angiographically Defined Obstructive Coronary Artery Disease Using Electron Beam Computed Tomography Calcium Score Strata

Abstract

Background —Electron beam CT (EBCT) is an accurate, noninvasive method to detect and quantify coronary artery calcification, a marker of coronary artery disease (CAD). This investigation examined the accuracy of EBCT to detect obstructive CAD (≥50% stenosis) and determined the optimal strata for quantity of coronary artery calcification to facilitate clinical decision-making. Methods and Results —Clinical research patients (n=213) were examined with coronary angiography and EBCT (angiography group), and 765 research participants were examined with only EBCT (nonangiography group). Of the angiography group, 53% had obstructive CAD. After adjustment for verification bias, the estimated sensitivity and specificity for calcium score ≥1 were 97.0% and 72.4%, respectively. Likelihood ratios for strata of calcium score associated with obstructive CAD were calculated in each sex and 2 age groups. Among those ≥50 years old, the same 4 strata of EBCT calcium scores were identified in each sex; likelihood ratios ranged from 0.03 (calcium score 0) to 12.85 (calcium score ≥200). The same 3 strata of EBCT calcium scores were identified in each sex among those <50 years old; likelihood ratios ranged from 0.13 (calcium score 0) to 190 (calcium score ≥100). Conclusions —A calcium score ≥200 among those ≥50 years old and calcium score ≥100 among those <50 years old provided strong evidence that patients of either sex had obstructive CAD. A calcium score of 0 provided strong evidence that patients ≥50 years old did not have obstructive CAD.