Ayanian JZ, Udvarhelyi IS, Gatsonis CA, Pashos CL, Epstein AM.
Racial differences in the use of revascularization procedures after coronary angiography.
JAMA 1993;269(20):2642-6.

This study examined rates of coronary revascularization procedures after angiography by race in a national cohort and then compared these rates across hospital types to determine whether racial differences in the process of care vary by site of care. Data for Medicare enrollees were derived using abstracted discharge data in the Medicare Provider Analysis and Review (MEDPAR) file provided by the Health Care Financing Administration (HCFA). A random sample of enrollees was selected, and, for each enrollee, a longitudinal record of hospitalization for coronary angiography and revascularization procedures during 1987 and 1988 was created. Linkages to Medicaid and HMO files were also made. From these records, all black or white enrollees from 65 to 74 years of age who received coronary angiography during 1987 and had a principal diagnosis indicating coronary heart disease were included in the study.

The authors tested three specific hypotheses: (1) teaching and public hospitals might have lesser racial differences in the use of revascularization because of their strong commitment to caring for minority patients, (2) rural hospitals might have greater racial differences than urban hospitals because of their more limited facilities for cardiac procedures, and (3) blacks might be more likely to receive angiography in hospitals that do not perform revascularization procedures, and these hospitals might have greater racial differences in subsequent procedure use because additional effort is required to refer the patient for further care.

The study sample was comprised of 96% whites and 4% blacks. Blacks accounted for 2.3% of patients receiving angiography in the West, 3.3% in the Midwest, 3.2% in the Northeast, and 5.5% in the South. There were several demographic and clinical differences between blacks and whites. A larger proportion of blacks than whites was female and eligible for Medicaid. Blacks had a principal diagnosis of myocardial infarction more frequently (22.5% for blacks versus 19.0% for whites) and more often had secondary diagnoses of congestive heart failure, diabetes mellitus, and chronic renal failure. Whites had a principal diagnosis of chronic ischemia (44.5% for whites versus 37.4% for blacks) more frequently and a secondary diagnosis of chronic obstructive lung disease more frequently. Blacks were more likely to be discharged from a public hospital (15.1% of blacks versus 9.6% of whites) and a teaching hospital (71.0% of blacks versus 66.1% of whites). Whites were more likely to be discharged from an urban/suburban hospital (92.7% of whites versus 90.2% of blacks) and a hospital where revascularization procedures are available (84.2% of whites versus 77.7% of blacks).

The odds of receiving a revascularization procedure within 90 days after angiography for whites relative to blacks (after adjusting for age, sex, region, Medicaid eligibility, principle diagnosis, secondary diagnoses, and hospital characteristics) was 1.78 (95% confidence interval=1.56 to 2.03). In similar analyses, the adjusted odds of receiving percutaneous transluminal coronary angioplasty (PCTA) specifically was 1.56 (95% confidence interval = 1.29 to 1.89) and the adjusted odds of receiving coronary artery bypass surgery (CABG) specifically was 1.56 (95% confidence interval = 1.34 to 1.80). The odds ratios for race did not differ substantially by any of the hospital characteristics (ownership, teaching status, location, and availability of revascularization facilities). The odds ratio for race was statistically significant in all but rural hospitals (OR=1.63; 95% confidence interval=0.93 to 2.86) after adjustment for potential confounders.

The authors conclude “that the odds of proceeding to a revascularization procedure within 90 days after coronary angiography were substantially higher for whites than blacks with coronary heart disease after controlling for region, Medicaid availability, comorbid diagnosis, and hospital characteristics. The consistent racial differences in use of revascularization procedures across all types of hospitals indicate that important disparities are widespread, and that the institutional characteristics (analyzed in this study) are not significant mediators of these differences. These findings also suggest that unequal access to specialty care is not the primary explanation for racial differences in rates of coronary revascularization procedures, because almost all patients in this (study) population were likely to have been evaluated by a cardiologist at the time of angiography”. They emphasize that “these findings do not reflect impaired access to cardiologists who perform angiography or hospitals that perform revascularization procedures.”

The authors also considered whether differences observed in this study might be due to racial differences in severity of coronary health disease. This study showed that adjustment for principal and secondary diagnoses yielded little change in the race effect on receiving a revascularization procedure (a minor reduction in the adjusted odds ratio). Some other studies that used more detailed clinical data support this finding, while some do not. In particular, the CASS study (Maynard, et al., 1986) reported a lower rate of coronary artery disease at angiography among blacks than whites. The authors here note that a subsequent analysis of the CASS study, which focused on patients with coronary artery disease (rather than with angiography) documented racial differences in procedure use after controlling for severity. Thus, this subsequent analysis raises two other possible explanations for the findings of racial differences in revascularization use in the present study as well as in the CASS study: physician bias and patient preference. However, the authors note that the data used in the present analyses were insufficient to assess whether “differing rates of revascularization procedures represent a racial bias among physicians or whether race is a proxy for other cultural and socioeconomic factors that effect physician or patient behavior.” They also add that “these interracial differences may reflect overuse in whites or underuse in blacks.”