Advanced electrocardiography predicts computed tomography coronary artery calcium score

BACKGROUNDCoronary artery calcification (CAC) has been linked to an increased risk of cardiovascular events. Its detection in asymptomatic individuals is valuable for reclassifying cardiac risk and informing management strategies. We hypothesised that an advanced electrocardiography (A-ECG) score derived from the standard 12-lead ECG predicts CAC with good diagnostic accuracy compared to computed tomography (CT). METHODSThis retrospective study included patients that had undergone a 12-lead ECG and CT CAC scoring demonstrating either the absence (n=265) or presence (n=255) of any CAC. Multivariable elastic net logistic regression was used to generate an A-ECG score validated by nested resampling. RESULTSAn A-ECG score for detecting CAC comprised of age, sex, and four ECG measures encompassing the duration of the Q wave in lead I, vectorcardiographic measures derived from the 12-lead ECG related to the spatial direction of the QRS complex (two measures) and the magnitude of the ST segment (one measure). Nested resampling estimated performance for predicting the presence of any CAC with an area under the receiver operating characteristics curve [95% confidence interval] of 0.78 [0.77-0.79], sensitivity 73 [72-75]%, specificity 66 [65-68]%, positive predictive value 70 [68-71]%, negative predictive value 71 [69-72]%, positive likelihood ratio 2.3 [2.1-2.4], and inverse negative likelihood ratio 2.6 [2.4-2.7]. CONCLUSIONSThe standard 12-lead ECG analysed by A-ECG analysis can predict the presence of CAC with good diagnostic performance. A-ECG may hold clinical utility as a low-cost and widely available initial screening test for the presence of CAC and cardiac risk prediction.