Impact of Physician Race on Patient Decision-Making and Ratings of Physicians: a Randomized Experiment Using Video Vignettes

BackgroundStudies suggest that black patients have better interactions, on average, with physicians of their own race. Whether this reflects greater "cultural competence" in race-concordant relationships, or other effects of race unrelated to physician behavior, is unclear. It is also unclear if physician race influences patient decision-making.ObjectiveTo determine whether physician race affects patients' ratings of physicians and decision-making, independent of physician behavior.DesignRandomized study using standardized video vignettes.ParticipantsPrimary care patients with coronary risk factors or disease.InterventionsEach participant viewed one of 16 vignettes depicting a physician reviewing cardiac catheterization results and recommending coronary artery bypass graft (CABG) surgery. Vignettes varied only in terms of physicians' race, gender, age, and communication style (high vs. low patient-centeredness).Main MeasuresParticipants rated the video physician's communication, interpersonal style, competence, trustworthiness, likability, and overall performance (0-4 Likert scales). They also rated the necessity of CABG (0-5 scale) and whether they would undergo CABG or obtain a second opinion if they were the video patient (0-3 scales).Key ResultsParticipants included 107 black and 131 white patients (72% participation rate). Black participants viewing a black (vs. white) video physician gave higher ratings on all physician attributes (e.g., overall rating 3.22 vs. 2.34, p<0.001) and were more likely to perceive CABG as necessary (4.05 vs. 3.72, p=0.03) and say they would undergo CABG if they were the video patient (2.43 vs. 2.09, p=0.004). Patient-centered communication style reduced, but did not eliminate, the impact of race concordance. Physician race was not associated with any outcomes among white patients.ConclusionsBlack patients viewed the doctor in a scripted vignette more positively, and were more receptive to the same recommendation, communicated in the same way, with a black vs. white physician. Patient-centered communication reduced but did not eliminate the effect of physician race.