Postoperative electrocardiographic (ECG) changes are frequently present after insertion of implantable cardioverter-defibrillators (ICD) and may mimic perioperative myocardial infarction (MI). The purpose of this study was to assess the incidence and clinical significance of postoperative ECG changes in relation to clinical, laboratory, and implantation data. In 25 (16%) of 156 patients undergoing ICD implantation, significant ECG changes (> or = 50% reduction in R-wave amplitude in > or = 3 leads or new Q waves in > or = 2 leads) were present 1 to 3 days after the operation and persisted at hospital discharge in 12 (8%). Presence of thoracotomy, the total number of induced ventricular fibrillation episodes, and the number of defibrillation shocks required during defibrillation threshold (DFT) testing correlated with postoperative ECG changes. Other factors associated with a significant R-wave loss in the lateral precordial leads included left-sided pleural effusion, lung infiltrates or atelectasis, and large defibrillator patch electrodes over the left ventricle or the lateral chest wall. Myocardial necrosis documented by elevated cardiac enzymes occurred in 6 (5%) of 151 patients without significant ECG changes and in 3 (12%) with (p value not significant). However, postoperative ECG changes associated with elevated enzymes were indistinguishable from changes unrelated to necrosis. Therefore the sensitivity and specificity of the surface ECG for detection of MI after ICD placement is poor. Multiple factors such as thoracotomy, myocardial injury from DFT testing, electric insulation, or shielding of the heart may contribute to the development of electrocardiographic pseudo-infarct patterns.