Objectives: To assess the prevalence of abnormal electrocardiograms (ECG) recorded routinely in patients undergoing elective surgery as well as the prevalence of ECG abnormalities that could not be predicted by the patient's medical history. To determine the influence of ECG results on patient care. To identify patient groups at high risk for abnormal ECG.
Patients and methods: Retrospective study of 2,146 patients scheduled for non cardio-thoracic surgery requiring anesthesia. Two physicians reviewed the preoperative ECG reports in consultation with a cardiologist. The reports were considered normal, abnormal but expected or abnormal and unexpected based on agreement or not with the patient's medical history.
Results: The 161 (7.4%) patients for whom data was missing were excluded. Preoperative ECGs were performed in 1,898 (95.6%) cases. Abnormal ECG results were reported for 351 (18.5%), most often for men, patients over 45 years of age, those classified ASA III-V, those with cardiovascular disease, kidney failure and diabetes mellitus. Two hundred seventy unexpected abnormalities were found in 234 (12.3%) patients. Of these, 117 (43.3%) were considered major. The unexpected abnormalities motivated changes of attitude toward 4 (1.7%) patients, but new treatment was established for only 2 (0.8%) of them. The prevalence of unexpected abnormalities was higher in men, patients over 45 years of age and those classified as ASA III-V.
Conclusions: The prevalence of unexpected preoperative ECG abnormalities among non cardio-thoracic surgical patients is high but the influence of such results is minimal. The sensitivity and specificity of preoperative ECG for detecting electrocardiographic abnormalities are low. We suggest that performance of ECG before surgery is useful in patients over 45 years of age or in those with cardiovascular disease, chronic kidney failure, diabetes mellitus, physical state ASA III-V and those for whom it has not been possible to obtain an adequate medical history or perform a complete physical.