Explainable advanced electrocardiography has a high negative predictive value for ruling out significant coronary artery disease on cardiovascular computed tomography

BACKGROUNDAdvanced electrocardiography (A-ECG) has been used to improve the diagnostic performance of the ECG in a number of cardiac disease states. We hypothesised that A-ECG can improve the diagnostic assessment of intermediate risk chest pain by optimising an A-ECG score for significant coronary artery disease (CAD) by cardiovascular computed tomography (CCT). METHODSParticipants attending an outpatient rapid access chest pain clinic underwent a 12-lead ECG and CCT. Significant CAD was defined as luminal stenosis >50%. Multivariable logistic regression was performed using measures from the conventional ECG, derived vectorcardiography, and singular value decomposition measures of waveform complexity. RESULTSOf included patients (n=171, 60% male, age 59{+/-}13 years), 37 (22%) had >50% stenosis in at least one coronary artery, with single, double, or triple vessel disease in 38%, 38%, and 24%, respectively. A four parameter A-ECG score to detect significant CAD had an area under the receiver operating characteristic curve [95% confidence interval] of 0.87 [0.78-0.94], sensitivity 89 [69-97]%, specificity 82 [68-94]%, positive predictive value 55 [43-78]%, negative predictive value 96 [92-99]%, positive likelihood ratio 4.6 [2.9-13.1] and inverse negative likelihood ratio 6.4 [2.9-27.2]. CONCLUSIONA-ECG can rule out significant CAD on CCT with a high negative predictive value and overall good diagnostic performance. This supports the use of A-ECG to screen patients in a chest pain clinic setting who would benefit from further testing or not.