The diagnosis of ischemic cardiomyopathy is not well established. Our objective is to determine predictive variables of coronary disease in unselected patients with ventricular dysfunction. This study is a retrospective cohort study of consecutive patients with left ventricular dysfunction and no known history of ischemic heart disease. We analyse the demographic variables, clinical data, electrocardiogram, and echocardiogram that are associated with the presence of coronary stenosis on coronary angiography. A total of 536 patients with left ventricular dysfunction were studied, with 135 (25.2%) of them having significant coronary lesions. In the multivariate logistic regression analysis, age ≤ 50 years, female gender, and the presence of atrial fibrillation on the electrocardiogram (ECG) were predictors of the absence of coronary lesions. Diabetes, hypercholesterolemia, the existence of Q waves in the ECG, and segmental alterations in contractility in the echocardiogram were predictors of coronary heart disease (C-Statistics 0.771, 95% CI 0.727 to 0.814). The information obtained from the clinical history, the ECG, and the echocardiogram of patients with ventricular dysfunction allows us to select subjects in whom coronary angiography has shown poor performance in diagnosing coronary disease.
Keywords: coronariography; diagnosis; prognostication; ventricular dysfunction.