• The hypothesis being tested is that a strategy of initial surgical revascularization is superior to percutaneous intervention in preventing death or myocardial infarction in diabetics with severe ischemic heart disease. [ Time Frame: Date of Death and non-fatal MI ] [ Designated as safety issue: No ]
  

• To include the costs associated with each treatment strategy [ Time Frame: Date of Death and non-fatal MI ] [ Designated as safety issue: No ]
  
• Patient health status following each type of revascularization [ Time Frame: Date of Death and non-fatal MI ] [ Designated as safety issue: No ]
  
• To develop risk models for predicting procedure-related outcomes in diabetic patients undergoing revascularizationTo develop risk models for predicting procedure-related outcomes in diabetic patients undergoing revascularization [ Time Frame: Date of Death and non-fatal MI ] [ Designated as safety issue: No ]
  
• To examine angiographic myocardial jeopardy as a predictor of adverse outcomes following revascularization [ Time Frame: Date of Death and non-fatal MI ] [ Designated as safety issue: No ]
  

Diabetic patients (HbA1c greater than 6.0% if treated or greater than 6.9% untreated) with severe ischemic heart disease and clinical indications for coronary revascularization will be screened and, if agreeable, entered into the study. Qualified patients unwilling to participate in the randomized trial will be approached for participation in a Registry that will collect computerized health data. The study consists of a 4-year accrual period. The median follow-up will be 3.7 years with a maximum of six years and a minimum of two years. Six sites will begin recruitment in the first year followed by nine additional sites for the remaining three years. The study is powered to detect a 40% relative difference between the groups in time to death or MI with a two-sided, type I error of 5% and a power of 90%.