Model of the indirect effect of patient race on participant willingness to
prescribe PrEP. Unstandardized coefficients and standard errors [B (SE)] are
included for all paths. Participant sociodemographic and medical training
characteristics (gender, race, sexual orientation, age, social class, current
year of medical school, and past clinical experience with HIV-positive
patients); other clinical judgments (predicted patient adherence, perceived
patient risk of HIV infection without PrEP, and risk reduction associated with
PrEP); and racial bias (perceived importance of patient request and general
feelings toward Black versus White patients) were statistically controlled in
the analysis