Clinical characteristics associated with poor long-term survival among patients with diabetes mellitus undergoing saphenous vein graft interventions

Abstract

Limited data exist on the long-term outcomes among diabetic patients undergoing saphenous vein graft (SVG) interventions. Thus, the baseline clinical factors associated with long-term adverse events in these patients are less known. Accordingly, we analyzed 1,160 consecutive patients (37.7% with diabetes) undergoing SVG interventions from the Duke Cardiovascular Disease Database (1990-2003). Cox proportional hazards modeling was used to identify predictors of long-term death in diabetic patients. The most significant model predictors were then used to construct a decision tree providing unadjusted Kaplan-Meier survival estimates at a median follow-up of 4 years. At median follow-up of 4 years, death (33.3% vs 18.1%, P < .0001; unadjusted hazard ratio 1.98, 95% CI 1.64-2.38) and death or myocardial infarction (49.6% vs 32.9%, unadjusted hazard ratio 1.71, 95% CI 1.462.00) were significantly higher in patients with diabetes mellitus compared with those without it. In patients with diabetes undergoing SVG interventions, a simple clinical decision algorithm, based on the most significant model predictors, demonstrated that 88% of patients without heart rate >80 beat/min, congestive heart failure, renal insufficiency, or hypertension survived after SVG intervention at median follow-up of 4 years. In contrast, none of the few patients with all these 4 factors survived at follow-up (100% mortality). Compared with patients without diabetes, diabetic patients undergoing SVG intervention have significantly worse long-term outcomes with one third dying at median follow-up of 4 years. We provide a simple decision tool that allows stepwise risk-stratification using baseline factors in diabetic patients undergoing SVG interventions and identify 4 risk factors associated with extremely poor long-term survival in this cohort.