The purpose of this study was to use a large national database to evaluate whether racial differences in coronary artery bypass graft surgery (CABG) rates varied depending on the geographic area, the need for care, and the supply of physicians. Data for this study were derived from 1986 MEDPAR files (for the number of coronary artery bypass graft surgeries and myocardial infarctions), the Bureau of Health Professions area resource file, and county population estimates from the 1985 Bureau of the Census. The population of interest was whites and blacks over 65 years of age (Medicare eligible).

There was great geographic variation in the rate of CABG across the United States. States in the northeast generally had low rates, and states in the northwest and south generally had the highest rates. One quarter of the states had (age and sex adjusted) rates of 31.0 per 10,000 or higher and one quarter had rates of 23.2 per 10,000 or lower. The racial disparity in rates also varied greatly by geographic location. Overall, the national age- and sex-adjusted rate for whites was 27.1 per 10,000 in comparison to 7.6 per 10,000 for blacks. Thus nationally, the ratio of white to black CABG rate was 3. However, for the southeastern states, the ratio was greater than 6, and in other areas it was much lower. Among men, the national CABG rate was 40.4 for whites and 9.3 for blacks; among women, the national CABG rate was 16.2 for whites and 6.4 for blacks. Regional differences in rates of CABG for blacks were limited to non-SMSA (rural) regions.

In multivariate analysis, neither the number of cardiologists nor the number of admissions for myocardial infarction was associated with CABG rates for either racial group. However, both location in the Southeast and the density of thoracic surgeons were significantly associated with CABG rates for whites (but not for blacks; separate analyses were conducted for blacks and whites). As a proposed explanation for an association between surgery rates and physician supply, the authors suggest that physicians in areas of high concentration may maintain their income by encouraging patients to have more medical services. This may explain the variation in racial disparities across states.

The authors consider other possible explanations for the racial differences in rates of CABG. They rule out the possibility that whites have a greater need for surgery because the racial difference in hospitalization for myocardial infarction was smaller than the racial difference in CABG. Additionally, there was no evident geographic variation in myocardial infarction rates. (Note that supporting data were not presented. Rates were presented by state, but formal analyses of potential geographic variation were not presented.) Other explanations presented include poverty among blacks, reluctance among blacks to have surgery, physician prejudice, and lack of Medicare eligibility/co-pay insurance.

This study is limited by its inability to control for individual factors that might explain geographic variation in rates, but, nonetheless, it suggests that racial disparities exist in rates of CABG that are not explained by patient need. The authors conclude that “the discrepancy in the CABG rates between blacks and whites and the variations in this discrepancy with sex, region, and availability in physician suggest that social and cultural factors may interfere with access to care even if a national health insurance policy is implemented.”