Physician Characteristics Associated With Bias in Contraceptive Recommendations [ID: 1349236]

Abstract

INTRODUCTION: We sought to determine the effect of obstetrician–gynecologist characteristics on contraceptive recommendations for patients of various racial and socioeconomic backgrounds. METHODS: U.S. obstetrician–gynecologists were administered an online survey depicting photographic scenarios of reproductive-aged women of differing races (White, Black, Latina, Asian) and described socioeconomic status (SES) (high versus low), with all other factors identical, and asked to provide contraceptive recommendations. We used logistic regression to analyze recommendation differences based on physician and patient characteristics. Institutional review board approval was obtained. RESULTS: One hundred thirty-five total respondents were included: 55% attending physicians, 45% trainees (residents/fellows); 73% White, 27% non-White. Among all patients: Non-White physicians had significantly lower odds of recommending tubal (odds ratio [OR] 0.493), vaginal ring (OR 0.401), and patch (OR 0.513) compared to White physicians. Non-White physicians had significantly higher odds of recommending any long-acting reversible contraception (LARC) (OR 1.846) and condoms (OR 1.438) compared to White physicians. Trainees had significantly lower odds of recommending tubal (OR 0.632), condoms (OR 0.595), pills (OR 0.428), vaginal ring (OR 0.326), and patch (OR 0.536) compared to attendings. Trainees had significantly higher odds of recommending any LARC (OR 18.374) and medroxyprogesterone injection (OR 1.797) compared to attendings. All physicians had a significantly lower odds of recommending any self-administered contraceptive method to Black high-SES patients compared to White high-SES patients (OR 0.418). Other variations were noted but did not reach statistical significance. CONCLUSION: In our study, contraceptive recommendations differed by race and training level of the recommending physician and by patients' perceived race and SES. This suggests that physician bias contributes to contraceptive recommendations.