Clinical and nonclinical correlates of racial and ethnic differences in recommendation patterns for coronary revascularization

BACKGROUND: We sought to determine whether gender or racial differences exist in recommendations for coronary revascularization in a multiracial patient population undergoing their first coronary angiography at an academic institution from 1990-1993 for the evaluation of coronary artery disease (CAD). HYPOTHESIS: For patients with clinically significant CAD, no racial differences exist in the recommendation to revascularization following coronary angiography. METHODS: The main outcome measure was a recommendation for coronary revascularization such as percutaneous transluminal coronary angioplasty (PTCA) or coronary artery bypass graft (CABG) for patients with clinically significant CAD (n = 590). The primary multiple logistic regression analysis focused on only those patients with angiographically severe disease, defined as triple-vessel or left main CAD (n = 180). Race was trichotomized into Hispanic, black, and white to ascertain whether any differential effects of race/ethnicity existed while controlling for age, gender, ejection fraction, angina, diabetes, hypertension, and peripheral vascular disease. A medical record review for all patients with severe CAD, who were given a recommendation for medical therapy, was conducted to ascertain whether previously unmeasured clinical factors or nonclinical factors may have precluded a PTCA/CABG recommendation. RESULTS: Hispanics with severe disease were significantly less likely than whites to be given a recommendation for PTCA/CABG following angiography [odds ratio (OR) = 0.39; 95% confidence interval (CI) (0.17, 0.92)]. Blacks were 67% as likely as whites to be given such a recommendation [OR = 0.67; 95% CI (0.17, 2.71)]. Medical records, reviewed for 35 of 40 of these patients given a recommendation for medical therapy, revealed that 6 patients eventually had PTCA/CABG within 6 months due to precipitating ischemic events; 9 had such severe or diffuse disease that revascularization did not appear to be an alternative, and 2 patients opted for medical therapy. CONCLUSIONS: Racial differences were manifested in the recommendations made following angiography and may be explained by previously unmeasured clinical as well as nonclinical factors.