Comparing predictive risk to actual presence of coronary atherosclerosis on coronary computed tomography angiography: a retrospective cohort analysis

Abstract Background There is limited data supporting the ability of cardiovascular risk scores (CVRS) to accurately predict asymptomatic coronary artery disease (CAD). Purpose To examine the predictive accuracy of CVRS in detecting presence and extent of atherosclerosis, determined by coronary computed tomography angiography (CCTA). Methods Asymptomatic individuals without known CAD, undergoing a screening CCTA with available CVRS data, were examined retrospectively. Natural language processing of the CCTA report data extracted the coronary artery calcium score (CACS) and the extent and severity of coronary plaque (low, moderate, or extensive, using a modified CAD-RADS 2.0 classification). Normal was defined as both zero plaque and zero CACS. CVRS was categorized as high (>15%), moderate (10-15%), low (1-9%) and "zero" (