Variables predictive of survival in patients with coronary disease
Seattle Heart Watch is a cooperative, community-wide registry that closely observes patients with known or suspected coronary heart disease with the intent of identifying those at risk for sudden cardiac death. It is the purpose of this report to define the variables from the clinical, electrocardiographic, exercise, arteriographic, and quantitative angiographic evaluations that are univariately and multivariately predictive of cardiac death. This analysis is based on 1836 and 730 patients who initially received medical treatment for coronary disease, all of whom were entered into the Seattle Heart Watch at the time of coronary arteriography between 1969 and 1974. Average follow-up for this analysis is 42 months (range, 30 to 78 months). The surgical cohort is unique in that it represents most (94%, 1870/1985) patients undergoing direct myocardial revascularization in Seattle during this time. The medical cohort is an approximate 40% sample of patients undergoing arteriography during this 5-year period and initially treated medically. Follow-up is by mailed questionnaire at 6-month intervals. In the medical cohort, 70 of the 85 deaths were of cardiac causes (82%); in the surgical cohort, 151 of the 173 deaths (87%) were of cardiac causes.
Baseline characterization of these patients is complete and includes more than 300 data items from the history, physical examination, resting electrocardiogram, chest roentgenogram, exercise electrocardiogram, and cardiac catheterization. Left ventricular function was assessed by measuring left ventricular volumes and ejection fraction from the left ventricular angiogram. The goal of our analyses was to select from this large number of variables . . .