Exercise echocardiography after coronary artery bypass surgery: correlation with coronary angiography
H Kafka,
AJ Leach,
and
GM Fitzgibbon
Cardio-Pulmonary Unit, National Defence Medical Centre, Ottawa, Ontario, Canada.
OBJECTIVES. Our aim was to assess, in patients after coronary artery bypass surgery, how well exercise echocardiography predicts the presence of vascular compromise on angiography. BACKGROUND. Because late graft failure frequently occurs after bypass surgery, a reliable noninvasive technique is needed to identify those patients who would benefit from angiographic study. METHODS. In 182 patients, a total of 213 symptom-limited treadmill exercise electrocardiograms (ECGs) and exercise echocardiograms were performed in association with coronary and bypass angiography 2 weeks to 21 years after bypass surgery. RESULTS. There were more inconclusive exercise ECGs (28%) than exercise echocardiograms (9%). The positive predictive value was 85% for the exercise echocardiogram versus 62% for the exercise ECG; the corresponding negative predictive values were 81% versus 52%. The accuracy of the exercise echocardiogram was linked to the degree of underlying vascular compromise. After excluding cases with nondiagnostic results, due to either submaximal stress or poor image quality, the exercise echocardiogram detected 46 of the 60 cases with vascular compromise in one region (sensitivity 77%) and 47 of the 49 cases with compromise in two or three regions (sensitivity 96%). Similarly, an abnormal exercise echocardiogram had a positive predictive value of 71% for vascular compromise in one region and 98% for compromise in two or three regions. Most false negative exercise echocardiographic results were associated with posterolateral single-region vascular compromise on angiography. CONCLUSIONS. This study confirms a high positive and negative predictive value of exercise echocardiography in the detection of vascular compromise in patients after bypass surgery. It is clearly superior to exercise electrocardiography in predicting which patients will have angiographically significant graft or arterial lesions, and it can be used to obtain a better selection of patients for angiographic study.
