Sex and Gender
Background: Women and underrepresented minority (URM) physicians confront significant workplace discrimination. We investigated whether women and URM physicians face systematic discrimination in patient ratings compared to white male physicians.
Methods: We conducted randomized patient analog experiments using a clinical vignette of an emergency department visit for abdominal pain, on two nationally representative samples of online respondents using validated crowdsourcing platforms (Amazon MTurk and Lucid). The intervention included 40 physician images from the Chicago Face Database, normed to minimize differences in appearance. Race (black or white) and gender of physician images were assigned in a 2x2 factorial design. Doubt in the physician diagnosis was introduced with a conflicting diagnosis from an Online Symptom Checker. Primary outcomes were patient confidence and satisfaction. Measures were collected using validated survey instruments. Data analyses included linear regression and Bayesian Additive Regression Trees (BART) to assess treatment heterogeneity. We estimated a design with 1,600 subjects had a minimum detectable effect of 0.20 standard units with 0.80 power at alpha = 0.05.
Results: Between 01/18-04/18, 3,308 participants were recruited (1,619 MTurk; 1,622 Lucid), 67 withdrew and 3,241 analyzed. Respondents had mean age 46 years (±16), were 75% white, 11% black, and 52% female. Of the 87% with health insurance, 59% was private, 36% Medicare/Medicaid and 5% other. We observed no statistically significant difference in composite confidence and satisfaction on a 100 point scale across treatment dyads: white male (WM) 67 (95%CI 65-68), white females (WF) 68 (95%CI 67-69), black male (BM) 68 (95% CI 67-69), black females (BF) 68 (95%CI 67 -70). BART revealed little evidence of effect heterogeneity for interactions among participant’s race, sex, measures of racial prejudice or sexism. These results were not explained by perception of warmth, competence, fairness of charge or willingness to sue.
Conclusions: Systematic discrimination from patients may be less of a barrier to a diverse physician workforce than other sources of workplace discrimination.