This study aimed to assess the use of revascularization procedures among patients with cardiovascular disease in the Washington, DC metropolitan area. This areas was chosen because the relative proportions of black patients and white patients evaluated for coronary artery disease (CAD) are more evenly distributed there than in other areas of the country.

All male patients admitted to a Veterans Administration hospital who had undergone a coronary angiography procedure during the period from November 1986 to November 1992 and were of either black or white race were included in the study. A total of 1,460 patients were included; the population consisted of almost equal numbers of blacks and whites (726 versus 734). Blacks typically had a higher prevalence of hypertension and diabetes, but a lower prevalence of myocardial infarction, anginal syndromes, and dyslipidemias. The percentage of black patients without significant CAD was almost twice that of the white patients (37% versus 19% had no vessels involved). Of those with CAD, however, the black and white groups had similar proportions of patients with 3-vessel and left main disease (35% versus 37%). Blacks were less likely to be classified by study clinicians as either "should revascularize" (24% of blacks versus 38% of whites) or "possibly revascularize" (11% of blacks versus 16% of whites).

Only about 60% of those who needed revascularization received it: 7% of blacks and 9% of whites received PTCA, and 6% of blacks and 11% of whites received CABG. Ethnicity was not a significant independent predictor for revascularization in multivariate analyses that controlled for clinical characteristics. When the analysis was confined to patients with significant CAD, ethnicity was again not significantly associated with revascularization. Additionally, the rates of revascularization did not differ by race even when stratified by the number of diseased vessels.

However, black patients had significantly decreased survival beyond 2 to 3 years after cardiac catheterization (p<0.05). The mean survival was 3,101 days for black patients and 3,346 days for whites. Risk of death was associated with advancing age, ventricular function, and clinically significant CHF. CABG (but not PTCA) was protective. The OR for mortality for white versus black race in a multivariate analysis that included all of these variables was 0.57 (95% confidence interval=0.43 to 0.76). These results were similar when the sample was limited to patients with significant CAD.

The authors conclude, "This study has confirmed that (when anatomic and functional factors are assessed) there is not statistically significant evidence of an ethnic or racial bias at work to explain the overall difference in the revascularization rates of black veterans and white veterans." With regard to the discrepancy in findings between this study and some previous reports, the authors note that previous studies frequently had either relatively small numbers of black patients or a lack of precise anatomic data stratifying the degree of coronary disease. The typical practice of substituting morbid events such as myocardial infarction and anginal syndromes for anatomic data might be misguided, as these events have no predictive value regarding the extent of abnormal coronary anatomy. (Although the authors of this study acknowledge that "why such morbid events do not prompt the routine performance of coronary angiography is difficult to fathom without assuming some influence of provider bias or patient preferences.") Furthermore, previous studies based on administrative datasets may not have completely controlled for readmission rates (due to privacy issues in these types of datasets). However, the authors note that it is not possible to state with authority that the large black population of this medical center makes these findings atypical in comparison to other VA facilities.

With regard to the higher long-term mortality for blacks, the authors state that this "is very likely reflective of a frequently cited, society-wide phenomenon and its poorly defined and confusing explanatory variables." They conclude that, "it is not possible from the data presented in this study to assess the reasons that led to higher long-term mortality rates among black patients. It can be speculated that possible repeat procedures or even known risk factors that are more prevalent among blacks (e.g., hypertension, diabetes, kidney failure) may have played a role."