Comparison of exercise echocardiography and the Duke treadmill score for risk stratification in patients with known or suspected coronary artery disease and normal resting electrocardiogram

Background Although exercise echocardiography (EE) is not clearly indicated in patients with normal electrocardiogram (ECG) as the first evaluation, there is a lack of data regarding its superiority over the Duke score for prognosis. We investigate whether EE has incremental value over the Duke score for predicting outcome in patients with normal ECG. Methods One thousand six hundred forty-seven patients with interpretable ECG referred for EE were followed up for 2.5 ± 1.4 years. There were 58 hard events (myocardial infarction or cardiovascular death). Results There were 38 events in 735 patients with abnormal EE versus 20 events in 912 with normal EE (P < .0001). The Duke score, resting wall motion score index, and ischemia were independently associated to events (incremental P value of EE = .03). The Duke score allowed stratification of patients with abnormal EE (P = .001) or ischemia (P = .01) into different risk categories but did not stratify patients without these characteristics. Exercise echocardiography variables stratified patients with the low Duke score (left anterior descending artery territory P = .04, left anterior descending artery ischemia P = .03) and with the intermediate Duke score (abnormal EE P = .005, necrosis P = .0009, ischemia P = .004, resting ejection fraction P < .00001, resting wall motion score index P < .00001, peak ejection fraction P < .00001, peak wall motion score index P < .0001, number of territories P = .002, left anterior descending artery territory P = .001, and left anterior descending artery ischemia P = .002) but did not with the high Duke score. Conclusions Exercise echocardiography has incremental value over clinical variables, the Duke score, and resting echocardiography for the prediction of hard cardiovascular events in patients with normal resting ECG.