Variables predictive of survival in patients with coronary disease. Selection by univariate and multivariate analyses from the clinical, electrocardiographic, exercise, arteriographic, and quantitative angiographic evaluations.

A progression of univariate followed by multivariate analyses was applied to 46 variables selected from the clinical examination, exercise test, coronary arteriography, and quantitative angiographic assessment of left ventricular function in patients with coronary disease to determine those variables most predictive of survival. For the 733 medically treated patients, the final Cox's regression analysis showed that the left ventricular ejection fraction was most predictive of survival, followed by age, number of vessels with stenosis(es) greater than or equal to 70%, and ventricular arrhythmia on the resting electrocardiogram. For the 1870 surgically treated patients, ventricular arrhythmia on the resting electrocardiogram was most predictive of survival followed by ejection fraction, heart murmur, left main coronary artery stenosis greater than or equal to 50%, and use of diuretic agents.