Responding to Bias: Equipping Residents With Tools to Address Microaggressions

Abstract

Introduction: Resident physicians frequently experience bias at work, with patients and families often being the source. Women and other trainees underrepresented in medicine are disproportionately impacted by these negative experiences, and experiencing bias contributes to resident physician burnout. Unfortunately, many resident physicians feel inadequately prepared to respond to bias.

Methods: We developed a 45-minute, peer-led, case-based workshop that equipped trainees with tools to respond to patient-expressed bias. Our toolkit centered on resident physicians by including an assessment of the trainee's emotional well-being, a team-based response, and an embedded debrief. The toolkit provided resident physicians with possible responses to bias directed towards themselves (bias-towards-self) or bias directed towards others (bias-towards-others). Surveys were administered pre- and postworkshop to assess change in participants' comfort in responding to patient-expressed bias.

Results: Thirty-seven residents completed both surveys. The workshop significantly increased comfort in responding to bias-towards-self (p < .001; 95% CI, 1.00-1.50) and bias-towards-others (p < .001; 95% CI, 1.00-1.50).

Discussion: We improved resident physicians' comfort responding to patient bias-towards-self and bias-towards-others through a toolkit and workshop designed specifically for trainees. The toolkit centers the resident physician perspective, incorporates clinical context, and embeds a debrief. Our novel approach situates the toolkit's teaching in a highly scalable, case-based workshop.

Keywords: Anti-racism; Bias; Case-Based Learning; Communication Skills; Diversity; Equity; Gender Identity; Gender Issues in Medicine; Inclusion; Microaggression; Well-Being/Mental Health.

Figure 1.. Word cloud generated during the workshop from participant responses to the prompt “What names other than ‘doctor’ have patients called you?”

Figure 2.. A: Overall summary data for comfort responding to bias-towards-self and bias-towards-others pre- and postintervention presented as first quartile, median, and third quartile, as scored on a 7-point Likert scale (1 = very uncomfortable, 7 = very comfortable). Significance testing with Wilcoxon signed rank test was significant for both bias-towards-self (p < .001; 95% CI, 1.00–1.50) and bias-towards-others (p < .001; 95% CI, 1.00–1.50). The bars demonstrate the ranges of the first and fourth quartiles. The shaded areas are the second and third quartiles, with the horizontal bar being the median (not visible on all boxplots due to tighter clustering of scores, hence the Xs as mean markers). The asterisks (***) indicate level of significance between pre- and postintervention scores (<.001). The open circle is an outlier point. B: Participant comfort responding to bias-towards-self represented with individual Likert scores pre- and postintervention. C: Participant comfort responding to bias-towards-others represented with individual Likert scores pre- and postintervention.