Respect for diverse identities—including race, ethnicity, gender identity, sexual orientation, cultural and religious background, and age—is paramount in the teaching hospital. Failing to respond to implicit or explicit bias reinforces a longstanding “hidden curriculum” for trainees of all disciplines and perpetuates a culture of hostility toward those from minoritized backgrounds [1]. Patients and families are frequently cited as the individuals making biased comments in the teaching hospital setting [1,2,3,4,5,6] and experiences of mistreatment are correlated with physician burnout [7, 8]. Unfortunately, faculty response is often felt to be insufficient or absent [1, 4].

Strategies for responding to biased comments from patients and families have been taught via webinars [9], scenario-based workshops [10,11,12,13,14], and reenactment [15]. None of these, however, asks participants to practice responding in real time, and only one [10] specifically confronts the challenge of addressing incidents of bias from patients with mental health conditions, whose mental status may be compromised due to mania, psychosis, delirium, or intoxication. Building on the “ERASE” framework [16] described below, we created an innovative, experiential workshop for psychiatric educators using roleplays tailored to situations that routinely occur on psychiatry teaching services, providing an opportunity for learners to actively practice responding to different types of bias. To our knowledge, there are no other roleplay-based modules specifically adapted for educators within psychiatry that teach skills in responding to biased comments from patients and families, and no similar modules adapted for a virtual platform. Here, we describe the workshop, present qualitative and quantitative data on how it can help participants build confidence in responding to bias incidents, and provide recommendations for psychiatric educators.

Workshop Learning Objectives and Outline

By the end of this workshop, participants will understand the impact of bias incidents on trainees, institutional culture, and patient care; recognize the different types of bias that might occur in a clinical setting; and practice specific strategies to respond to bias incidents in a professional manner, in real time and after the event. All members of the multidisciplinary team are invited to participate, including attending physicians, residents, nurse educators, social workers, and psychologists. The content of this workshop builds upon the “ERASE” framework (expect mistreatment will happen, recognize when mistreatment occurs, address the situation in real time, support the trainee after the event, and establish a positive culture) for responding to trainee mistreatment by patients [16].

To encourage adult learning and to maximize workshop time, the workshop begins with a flipped classroom approach. Prior to participating in the workshop, participants watch a short video introducing specific strategies for responding to different types of bias, including overt derogatory language (explicit bias incidents) and microaggressions (brief, everyday exchanges that send denigrating messages to an individual on the basis of personal characteristics or group membership [17]) from patients and families. Participants receive a content handout summarizing key points from the video.

Our group has moderated hybrid and virtual workshops 60–90 min in length for faculty and multidisciplinary staff, with separate sessions for residents. Since the content of the workshop may bring up difficult emotions for participants, each workshop begins with establishing a safe space, including asking participants to turn on their cameras (if virtual) and to keep the session private. Workshop moderators acknowledge the challenges of discussing bias incidents and that roleplays might use uncomfortable language. They invite participants and moderators to acknowledge their own imperfections and discomfort and to adopt a growth mindset, while setting the expectation that all are there to learn. For trainee workshop sessions, moderators provide reassurance that in addition to teaching these skills to trainees in supervisory roles, they are delivering the workshop to faculty and staff within the department. Moderators encourage trainees to expect a culture change in which attending physicians and other team members also intervene when bias incidents occur.

A workshop moderator gives verbal instructions for roleplay exercises, while another moderator creates breakout rooms with 1–2 moderators and 3–4 participants in each group. Examples of roleplays are given in Table 1. The instructions include the following: first, roleplays involve several roles, including a patient, a trainee, and a supervisory clinician (e.g., attending); second, group members (in their assigned roles) read a short script portraying a bias incident; and then third, continue to improvise how they might respond in real time. A moderator then distributes roleplay materials and participants separate into their breakout rooms.

Table 1 Roleplay examples
Full size table

Within breakout rooms, moderators assign roles to each participant. If there are insufficient participants in a breakout room, the moderator plays the role of the patient. Moderators then invite participants to read aloud the roleplay dialog and continue improvising “off script” to respond to the bias incident. Moderators pause the roleplay for discussion when the participants have had sufficient time to improvise responding to the incident, and then facilitate a discussion of the roleplay, including emphasizing any strategies used by participants. Participants are encouraged to repeat the same roleplay with participants taking on different speaking roles prior to moving on to the next roleplay, with each participant having at least one opportunity to engage in a roleplay during the small group session.

Once participants return to the large group, one or two moderators then facilitate a large-group discussion. Each small group is invited to share questions, reflections, and ideas. Moderators are encouraged to contribute to the discussion. To conclude the workshop, moderators review the learning objectives and provide participants with a list of resources for further reading.

Workshop Experience: Outcomes

We conducted 16 hybrid sessions of the workshop, including two at other institutions. Due to the COVID-19 pandemic, in 2020–2021, we transformed the 60–90-min workshops to a virtual format via a popular videoconferencing platform; 10 workshops were held during staff meetings for multidisciplinary members of each residency teaching service, including the consultation-liaison service, inpatient units, emergency psychiatry service, and outpatient clinics. Four separate sessions were conducted for residents. Workshops included 5–30 participants per workshop and required 1–2 small group moderators for every 3–4 participants. This project was undertaken as a quality improvement initiative at our institution and was granted a waiver by the Institutional Review Board.

To assess the effectiveness of the workshop, we invited participants to fill out anonymous pre- and post-workshop surveys based on the learning objectives of the workshop. Responses were linked by a unique identifier created by participants and unknown to the authors. Survey responses were scored on a 5-point Likert scale (1 = not at all confident to 5 = extremely confident). The primary outcome for our analysis was the question: “How confident are you in your ability to use specific strategies to respond to biased comments or actions against you, your colleagues, and/or your trainees in real time?” We chose this as the primary outcome because lack of confidence in the ability to respond effectively is a major barrier to action; this was further supported by qualitative feedback from participants.

We analyzed participants’ responses in Excel using paired two-sample t-tests. For this analysis, we used responses that included both pre- and post-survey Likert scale responses during the first “round” of workshop rollout (all initial sessions). We also analyzed free text responses from participants for qualitative themes.

During the period of collection, 136 participants completed the survey, with 116 participants completing the pre-survey and 92 completing the post-survey. Seventy-two participants completed both pre- and post-session surveys, and this subset of responses was used for quantitative comparison. Free-response questions were added after the first few workshops and all participant responses were included in the qualitative analysis.

For our primary outcome, participants’ mean confidence in their ability to respond in real time improved from 2.75 to 3.79 (p < 0.0001). Secondary outcomes (ability to describe the impact of bias, ability to recognize bias, and ability to respond after the event) also demonstrated significant improvement (Table 2).

Table 2 Comparison of pre- and post-didactic survey results (n = 72)
Full size table

To assess whether participants who completed only one survey (pre- or post-didactic) differed from those who completed both, we used two-sample t-tests to compare ratings from the primary data set to those who completed only one survey. Of these comparisons, the groups differed significantly only in their pre-workshop ability to describe the impact of bias, with those who completed both surveys expressing greater confidence (3.35 vs. 3.05; p = 0.04). No other comparisons were statistically significant. Notably, as an educational intervention we did not track total participant numbers, which may impact the validity of our statistical analyses. We also acknowledge that our data were based on the workshop roleplays and therefore do not assess real-world use of the skills.

Numerous themes emerged in both the pre- and post-didactic surveys. In the pre-workshop survey, participants were asked to identify factors that prevent them from intervening during bias incidents. Common responses included a perceived lack of skill at recognizing bias; lack of confidence in ability to respond to bias; concerns about the patient escalating (including to physical violence); lack of time to respond; concerns about damaging relationships with patients; and not wanting to “undercut” the targeted individual. Several respondents mentioned concerns about experiencing “retribution” or being targeted themselves due to their own personal characteristics.

When asked what they found most valuable about the session, numerous respondents identified the roleplay format as being more useful than didactic content. Several noted the importance of creating a “safe space” for practicing the skills and discussing issues of bias. Other themes included learning and practicing the skills and deepening their understanding of bias incidents through the lens of roleplay.

When asked about ways to improve the workshop, most cited insufficient time and/or wanting additional sessions for more practice. Regarding ways in which they planned to change their practice after the workshop, participants stated that they planned to be more proactive in identifying and addressing bias and more attentive to supporting trainees during and after bias incidents.

We also asked participants after the workshop to identify ongoing barriers. Fewer participants cited barriers after the session, though some wished for more practice and others continued to note time and mental energy as potential barriers. One participant who expressed concerns about retaliation from program leadership in the pre-session survey continued to express these concerns after the workshop. Level of comfort with intervening remained a concern for many participants, though several expressed confidence that they could intervene more often with practice.

Lessons Learned: Teaching Points for Educators in Psychiatry

Based on our experiences as facilitators of this workshop, we have identified common questions from participants and developed suggested responses.

Should the Response to Incidents of Bias Change if a Patient Is Experiencing Mania, Psychosis, Delirium, or Intoxication?

We teach that safety comes first, including ending the interview, stepping away, and allowing security to intervene if necessary. Setting limits is encouraged; racism, homophobia, sexism, and other types of implicit or explicit bias are not part of the DSM-5 criteria of any psychiatric disorder [18]. We recommend that bias incidents are always addressed, even if mental status is compromised, because doing so actively demonstrates a commitment to a safe and supportive learning environment for trainees; work environment for staff; and milieu environment for other patients. Moreover, it is our clinical experience that patients may say or do uncharacteristic things while experiencing an acute mood or psychotic episode; if they are permitted to continue, they may later regret their actions [19]. We find that setting limits may help patients realize that their behavior is inappropriate and may prevent further harmful behavior [3].

What Should We Do if a Patient Refuses to Work with a Provider Because of a Personal Characteristic?

We encourage participants to become familiar with their institution’s policies for when a patient makes this request and highlight our own institution’s policy. The Association of American Medical Colleges encourages institutions to develop policies if they are not already in place [20]. Patients who are acutely unstable cannot simply be discharged. However, it may be useful to redirect the patient toward focusing on the chief complaint rather than the physical characteristics of their ideal treatment team.

How and When Should Debriefing Take Place After a Bias Incident?

Bias incidents should always be debriefed. We recommend avoiding assumptions about a trainee’s response to bias, checking in with trainees after the event, and considering practicing or modeling vulnerability. Our content handout includes suggestions for debriefing bias incidents. All trainees should be empowered to speak up and collaborate on a proactive plan—offered initially by the teacher—for when bias incidents occur. Supporting trainees includes creating a safe learning environment, cultivating cultural humility, and becoming familiar with institutional offices and procedures for escalating response to bias incidents.

Conclusion

Developing skills to address bias directed toward members of the healthcare team is necessary to creating a safe learning environment, especially within psychiatry. Practicing these skills with an experiential workshop is an effective method to gain experience and build confidence. We believe that utilizing roleplays is a particularly effective format for teaching this material; our results show a statistically significant improvement in participant confidence in ability to respond in the moment. This workshop can be provided in routine staff meeting time frames in multiple settings. We encourage all training programs to consider adding experiential training such as this workshop as an in-service to required faculty curricula, and our materials are available upon request. As the ERASE framework [16] establishes, being prepared to recognize episodes of mistreatment, addressing the situation in real time, and supporting learners in the moment are foundational skills for creating a positive, inclusive workplace culture and a learning environment where all are valued.