Diversifying the Racial and Ethnic Composition of the Physician Workforce FREE

This position paper from the American College of Physicians states that a diverse workforce of health professionals is an important part of eliminating disparities among racial and ethnic groups in the United States (1). An implicit assumption underlying this position is that increasing the number of ethnic minority providers will not only reduce health care disparities but also improve the health of minorities. Arguably, the most definitive and direct evidence for this assumption would be a clinical trial that randomly assigned patients to minority and majority physicians and followed them longitudinally while assessing changes in objective measures of health status. This trial does not exist and is extremely unlikely to be performed because many patients would object to random assignment to physicians. Choosing a physician is a very personal matter that most people who have a choice do not leave to chance.

Since direct evidence does not exist, one could try to construct a chain of logic by using observational data from separate studies. This chain might lead inexorably to the conclusion that a diverse workforce would improve health. How solid is this chain of logic? Ample evidence suggests that minority physicians are more likely to return to communities from which they came and that minority physicians are more likely to treat patients with lower socioeconomic status (2 - 7). While evidence has shown that more health care (and possibly physician care) does not always lead to better health status in relatively well-insured populations (8 - 9), providing health care to a geographic area where none to little existed may influence health status.

Furthermore, concordance (patients and their physicians sharing similar characteristics, such as race or ethnicity, language, and sex) is associated with better patient-reported outcomes. In race-concordant visits, patients are more satisfied with their care and feel that they are more involved in decision making about their care (10 - 11). Language concordance also has these effects and results in improvements in self-reported health status (12 - 13). Unfortunately, we do not clearly understand why concordance leads to better patient-reported outcomes. If we did, we could teach all physicians how to achieve these outcomes whether or not they and their patients were of the same race or ethnicity. Differences in communication patterns in race-concordant and race-discordant relationships do not explain why patients in race-concordant relationships rate their care more highly (11).

In arguing to expand the number of minority clinicians, is it necessary to claim that concordance improves health outcomes? Suppose that patients' health status was the same in race-concordant and race-discordant relationships. Would not patients choose to enter a relationship that brought greater satisfaction and participation in decision making? Some patients might even prefer greater satisfaction and participation in health care to slightly better health status, although this is unproven.

Furthermore, projections from the 2000 U.S. Census indicate that the U.S. population will grow more diverse from 2000 to 2050 (18% to 27.9% ethnic minorities and 12.6% to 24.4% Hispanics) (14). Many believe that the social and economic future of the United States depend on how well the workforce reflects the population it serves. They argue that diversity can bring value not only through better services but also through better workforce functioning. Other segments of our global society recognize this value. So might the health care industry. Therefore, although the direct evidence that increasing the diversity of the physician workforce improves health status is not ironclad, we believe that the College's position paper takes a reasonable stance in supporting actions to diversify the health professional workforce.