Yale University School of Medicine

Diabetics have higher mortality after myocardial infarction (MI), yet little is known regarding the impact of quality of care on long-term survival in older post-MI diabetics. Using data from the Cooperative Cardiovascular Project (CCP), a national cohort of 234,769 Medicare patients aged 65 or older hospitalized with confirmed AMI between 1994-95, we assessed differences in 10-year mortality outcomes between diabetics and non-diabetics using Cox proportional regression. To account for quality of care, a composite measure among ideal candidates was constructed and entered into the final model, adjusting for use of aspirin and beta-blocker on admission/discharge, angiotensin-converting enzyme inhibitors at discharge, reperfusion within six hours of admission, and smoking counseling at discharge. We also assessed the relationship between insulin use, sulfonylureas/biguanides and statin therapy, and long-term mortality within the diabetic cohort. The final study sample included 203,658 cases: 32 percent were diabetics. Compared to non-diabetics, diabetics were younger (75 vs. 76, p<0.001), female (53 percent vs. 47 percent, p<0.001), had more comorbidities, and were unlikely to receive evidence-based care (59 percent vs. 64 percent, p<0.001). The unadjusted HR for mortality among diabetics vs. nondiabetics was 1.38 (95 percent CI: 1.37-1.40). After adjusting for demographics, past medical history, procedures during hospitalization, medications on admission/discharge, and quality of care, the HR was 1.29 (95 percent CI: 1.27-1.31). Among diabetics, those on insulin or oral hypoglycemic therapy during the initial hospitalization for AMI had the highest risk of mortality during the last seven years, after adjustment for demographics, clinical characteristics, and quality of care (HR insulin=1.30, 95 percent CI: 1.25-1.35; HR oral hypoglycemics=1.11, 95 percent CI: 1.08-1.15), whereas those on statin therapy were not at increased risk (HR statin=0.95, 95 percent CI: 0.90-1.02). As compared to non-diabetics, older diabetics had a 29 percent increase in mortality even after adjusting for demographics, clinical variables during hospitalization, and quality of care (HR=1.29, 95 percent CI: 1.27-1.31). Additionally, within the diabetic cohort, the risk of long-term mortality was highest among those on insulin or oral hypoglycemic therapy during initial hospitalization for AMI. Our study demonstrates that neither patient characteristics nor quality of care fully account for the poor outcomes among diabetics suggesting that metabolic risk factors associated with diabetes ultimately require therapies beyond those currently recommended for post-MI patients.