Conjoint analysis to measure physician discrimination toward African Americans

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): The National Research Council defines discrimination as "differential treatment on the basis of race that disadvantages a racial group." Decades of research have found that well-intentioned physicians discriminate by providing unequal healthcare recommendations to minorities versus whites. Physicians' discrimination, though often unintentional, can play a significant role in creating and perpetuating disparities. The National Institute of Health (NIH) (PA-11-164) recognizes the critical need to empirically measure racial discrimination in healthcare in order to reduce health disparities. Despite the serious public health repercussions of physicians' discrimination, no gold standard assessment method has yet been developed to measure such discrimination. To date, the assessment methods that have been used to measure physicians' discrimination (e.g., self-report) have serious limitations (e.g., tainted by social desirability effects). Given the limitations in the previously studied assessment methods, PA-11-164 calls for research to "improve the measurement of racial/ethnic discrimination in healthcare delivery" in order to move toward eliminating the unequal treatment of minorities in our healthcare system. In direct response to this call for research, the proposed R03 aims to test an innovative assessment method, conjoint analysis, to examine the role of patients' race on physicians' treatment recommendations. Conjoint analysis is an assessment method that asks participants to respond to a series of computer-generated vignettes that vary on multiple factors (e.g., hypothetical patients' race, income, and health status). Conjoint analysis has the unique ability to overcome the limitations in the previously used assessment methods because: 1) it can detect socially undesirability effects (e.g., discrimination); and, 2) it has been found to be an excellent predictor of real worl behavior. For the proposed project, conjoint analysis will be used to detect physicians' discrimination against African Americans in the delivery of colorectal cancer (CRC) screening recommendations (i.e., colonoscopy, FOBT, FIT). Physicians' differing CRC screening recommendations can have direct consequences on CRC disparities. Aim 1 is to examine whether conjoint analysis can detect the role of patients' race on physicians' recommendations for CRC screenings. Aim 2 is to explore whether physicians' beliefs about patients' health behaviors (as determined by conjoint analysis) will mediate the relationship between patients' race and physicians' recommendations. Aim 3 is to explore whether physician characteristics (e.g., gender, race) moderate the relationship between patients' race and physicians' CRC screening recommendations. To achieve these aims, we will enroll 732 primary care physicians to participate in the conjoint study. The results of the R03 will contribute to the growing literatre on physicians' discrimination toward African Americans in recommending colorectal cancer screenings.