Do race-specific models explain disparities in treatments after acute myocardial infarction?

Background Racial differences in healthcare are well known, although some have challenged previous research where risk-adjustment assumed covariates affect whites and blacks equally. If incorrect, this assumption may misestimate disparities. We sought to determine whether clinical factors affect treatment decisions for blacks and whites equally. Methods We used data from the Cardiovascular Cooperative Project for 130 709 white and 8286 black patients admitted with an acute myocardial infarction. We examined the rates of receipt of 6 treatments using conventional common-effects models, where covariates affect whites and blacks equally, and race-specific models, where the effect of each covariate can vary by race. Results The common-effects models showed that blacks were less likely to receive 5 of the 6 treatments (odds ratios 0.64-1.10). The race-specific models displayed nearly identical treatment disparities (odds ratios 0.65-1.07). We found no interaction effect, which systematically suggested the presence of race-specific effects. Conclusions Race-specific models yield nearly identical estimates of racial disparities to those obtained from conventional models. This suggests that clinical variables, such as hypertension or diabetes, seem to affect treatment decisions equally for whites and blacks. Previously described racial disparities in care are unlikely to be an artifact of misspecified models.