Patient-provider racial/ethnic concordance and adherence to cancer screening recommendations

Background: Racial/ethnic concordance between patients and providers has been suggested as an indicator of shared cultural, language, and social characteristics and has been hypothesized to be associated with better health care quality. Whether such concordance is associated with receipt of cancer screening services is, however, particularly unclear. Objective: We examined patient-provider pairs, stratified by provider race/ethnicity, among respondents of the Medical Expenditure Panel Survey, administered during years 2003 to 2010, in order to test the hypothesis that racially/ethnically concordant pairs are more likely than discordant pairs to receive breast, cervical or colorectal cancer screenings. Measures: Receipt of cancer screening tests (breast, cervical and colorectal). Odds ratios were adjusted for highest education level achieved, percentage of Federal Poverty Level, type of insurance coverage, marital status, self-rated physical and mental health, number of chronic conditions and the year in which cancer screening was reported. Results: Adjusted odds ratios indicated that, relative to racially/ethnically discordant patient-provider pairs, white patients seeing white providers were significantly less likely to receive breast (OR=0.81) and/or cervical cancer (OR=0.69) screening, black patients seeing black providers were more likely to receive breast (OR=2.2) and/or cervical cancer (OR=2.1) screening and Hispanic patients seeing Hispanic providers were less likely to receive colorectal cancer screening (OR=0.62). Conclusions: Lack of screening among Hispanic patients was expected; that this was significantly only among patients seeing concordant providers was not surprising. White providers appear to be making a conscious attempt to increase screening among their non-white patients, perhaps due to awareness of disparities. Among patients of black providers, the increase in the relative odds of breast and cervical cancer screening services appears to be due to a combination of implicit patient and provider biases though explicit biases cannot be ruled out.