Prognostic assessment of patients with coronary artery disease by exercise testing.

To assess whether exercise testing could help predict cardiac mortality, we analyzed 14 exercise and 10 clinical variables in 292 patients treated medically, who underwent treadmill exercise testing and cardiac catheterization and were followed annually for a mean of 2.5 years. None of the individual variables could accurately predict subsequent cardiac mortality with predictive values ranging from 6% to 44%. Combinations of variables were then analyzed in the subset of 113 patients with multivessel coronary disease. A high-risk subgroup (n = 59) consisting of patients with either severe exercise ischemia (greater than or equal to 2 mm ST depression lasting greater than or equal to 5 minutes involving greater than or equal to 3 leads) or left ventricular dysfunction (treadmill time less than or equal to 3 minutes, S3 gallop, or cardiac enlargement) had a mortality of 20%; this was significantly greater (p less than 0.01) than a low-risk subgroup (n = 54) with neither severe exercise ischemia nor left ventricular dysfunction whose mortality was 2%. We conclude that combining clinical and exercise variables to distinguish high- and low-risk subgroups of patients with similar coronary anatomy is useful in predicting cardiac mortality.