The effect of baseline electrocardiographic abnormalities on the diagnostic accuracy of exercise-induced ST segment changes.

Although exercise-induced ST segment depression is thought to be unreliable marker of myocardial ischemia in the presence of resting electrocardiographic changes, this conclusion is based on limited and disparate data from studies often lacking acceptable measures of ischemia. To determine the diagnostic accuracy of the ST segment response in a blinded prospective protocol, we compared ST deviation to thallium201 SPECT scintigraphy in 95 patients during exercise. Diagnostic accuracy was poor in the 95 patients with resting abnormalities: left bundle branch block (LBBB) = 70%, complete right bundle branch block (cRBBB) = 75%, incomplete right bundle branch block (incRBBB) = 79%, intraventricular conduction delay (IVCD) = 44%, left ventricular hypertrophy (LVH) = 59%, digitalis = 53%, compared with a diagnostic accuracy of 90% in 29 patients without resting changes. There were 20 false negative and 17 false positive ST segment responses. The extent and direction of resting ST deviation varied substantially and had no influence on diagnostic accuracy. The extent of change in ST deviation with exercise required for a positive response did not alter diagnostic accuracy: -1.0 mm = 61%, -1.5 mm = 63%, and -2.0 = 61%. While the location of regional ischemia did not influence the accuracy of ST segment analysis, a QRS duration less than 120 msec did improve diagnostic accuracy. Our data confirm that ST segment analysis with exercise testing is not reliable in patients with resting electrocardiographic abnormalities and demonstrates that accuracy is not improved by adjusting for either resting or exercise-induced ST segment changes or for location of the ischemic region.