Words Matter: An Antibias Workshop for Health Care Professionals to Reduce Stigmatizing Language

Introduction Biased language influences health care providers' perceptions of patients, impacts their clinical care, and prevents vulnerable populations from seeking treatment. Training clinicians to systematically replace biased verbal and written language is an essential step to providing equitable care. Methods We designed and implemented an interactive workshop to teach health care professionals a framework to identify and replace stigmatizing language in clinical practice. The workshop included a reflective exercise, role-play, brief didactic session, and case-based discussion. We developed the program for a broad target audience of providers and initially delivered it at three academic conferences. We used descriptive statistics to analyze Likert-style items on course evaluations and identified themes in open-text responses. Results A total of 66 participants completed course evaluations; most believed the workshop met its objectives (4.8 out of 5.0) and strongly agreed that they would apply skills learned (4.8). Participants planned to incorporate reflection into their verbal and written language. Potential barriers to applying course content included perceived difficulty in changing entrenched practice habits, burnout, and fatigue. Suggestions for improvement included more time for group discussions and strategies to teach skills to colleagues. Discussion Participants found the course material highly engaging and relevant to their clinical practice. Learners left the workshop feeling motivated to engage in more mindful word choice and to share key concepts with their colleagues.


Introduction
There is increasing evidence that biased spoken and written language influences perceptions of patients and negatively affects treatment plans. [1][2][3][4][5] Goddu and colleagues' 2018 study exploring the impact of biased documentation found that medical students and residents randomized to read a clinical vignette using biased language to describe a patient with sickle cell disease were more likely to have a negative perception of the patient and less likely to recommend opioid analgesicss. 1 A similar pattern was found among mental health professionals who read a clinical vignette of a patient referred to as a substance abuser versus substance user; participants who read descriptions of a substance abuser were more likely to make character judgments and recommend punitive measures. 2 Biased language in the electronic health record may negatively influence future provider perceptions and patient care as well.
Patients with stigmatizing conditions such as diabetes, obesity, substance use disorder, and chronic pain are affected by the language of their health care providers. [6][7][8][9][10][11] Biased language can worsen feelings of shame, decreasing patients' motivations to complete their treatment plans or engage in treatment at all. 9,12 Therefore, biased language, likely due to assumptions made from the environment and throughout training, plays a central role in perpetuating health care disparities. As medical records become increasingly accessible to patients and their families, the importance of unbiased documentation that supports patient agency is magnified.
MedEdPORTAL has several examples of workshops promoting antioppressive language use. For example, Mayfield and colleagues' LGBTQ-inclusive sexual history taking curriculum teaches the importance of using appropriate terms and genderneutral language. 13 Stagno, Crapanzano, and Schwart's workshop for mental health audiences highlights the importance of using person-first language and removing judgmental words such as denies or claims. 14 However, there are no antibias workshops that provide a framework for identifying verbal and written bias in other clinical contexts.
Here, we describe the development of a language-based, antibias workshop and provide the materials necessary to present this training program. The course specifically addresses stigmatizing language in both written and verbal communication, including health records and clinical presentations, using a framework that reinforces antibias and antioppressive skills. The target audience is broad, including any provider, trainee, or staff member who works with patients in a clinical environment. We describe the power of stigmatizing language in medicine and situations in which clinicians often use biased language, as well as providing learners with tools to replace biased language.

Methods
This workshop was designed by participants and facilitators at the Stanford School of Medicine Leadership Education in Advancing Diversity (LEAD) program, including residents, fellows, program administrators, and teaching faculty. 15 Several members of this team had prior experience developing and facilitating antibias trainings. Three investigators were clinician educators with experience in workshop design and faculty development; three investigators were LEAD program instructors with content expertise in equity and inclusion; one trainee had prior experience developing and presenting an antibias workshop. Team members completed an extensive literature review of stigmatizing language in medicine to identify workshop content; search terms included the specific conditions of obesity, addiction, and chronic pain. We selected Sukhera and Watling's framework for integrating implicit bias recognition into health professions education 16 as our conceptual model, including these key recommendations: (1) create a safe space for learning, (2) emphasize how biased language impacts patient outcomes, (3) increase self-awareness of personal biases, (4) discuss strategies to overcome these patterns, and (5) enhance awareness of influence of implicit biases on others.
Our target audience was all health professionals caring for patients and documenting in the electronic medical record, including novice learners. There were no prerequisites for learners. We piloted the workshop with a diverse group of learners at three different professional conferences: the Stanford Medicine Second Annual Diversity and Inclusion Forum, 17 the Academy for Professionalism in Heath Care (APHC) Seventh Annual Meeting, 18 and the Academic Pediatric Association (APA) Region 9/10 Annual Meeting. 19 The LEAD program provided our workshop facilitators with trainings that included reducing unconscious bias and microaggressions and addressing privilege and allyship. We recommend that facilitators have experience with these topics. Audience participation was voluntary, and there were no incentives.
The workshop was designed to be 90 minutes, and course materials were inexpensive to duplicate. The facilitator's guide (Appendix A) provided a detailed agenda, and the PowerPoint presentation (Appendix B) contained detailed presenter notes below each slide. Key features of the workshop included the deliberate creation of a brave space, 20 a skit demonstrating biased language in a clinical presentation (Appendix C), a short didactic to define important terms, introduction of the Mindful Language Toolkit (Appendix D), group discussions, and opportunities to practice learned skills (Appendix E). Clinical cases can be adapted to the clinical discipline of the audience. However, we believe that the cases represent a wide variety of biased language commonly encountered in any discipline or position in health care.
All workshop attendees completed course evaluations (Appendix F) consisting of four 5-point Likert-style items (1 = strongly disagree, 5 = strongly agree), which were analyzed using descriptive statistics. Additionally, participants were asked four short-answer questions that were analyzed using the thematic content approach, which included data familiarization, identifying themes, developing and applying a coding scheme, and organizing codes and themes. 21 One investigator (Ria Pal) developed the coding framework, and two other investigators (Julia Raney and Kim Hoang) reviewed the coding scheme and made adjustments as necessary. This project was deemed exempt by the Stanford School of Medicine Institutional Review Board (Protocol #55660).

Results
We collected course evaluations from 66 participants ( Table 1).
Most participants found the sharing of personal stories, display of stigmatizing language on word clouds, and case-based examples to be particularly helpful activities (Table 2). Participants also enjoyed learning from group discussions with their colleagues and identified the Mindful Language Toolkit as a course highlight. Several participants offered feedback to better structure the workshop, including more time for small-group discussion. Many participants planned to teach lessons learned in the workshop to their colleagues and trainees, and some participants recommended adding dedicated workshop time to learn teaching and implementation strategies. Additionally, some participants suggested including more subtle and nuanced examples of bias and patient perspectives on the topic. Participants planned to use the Mindful Language Toolkit, workshop handouts, or reflective exercises to practice language substitutions discussed in the presentation. However, participants also identified several barriers to change, including the rigid social structure in medicine, cultural norms, entrenched practice habits, normalization of stigmatizing language, and assumptions that colleagues or supervisors would lack interest in the topic. Time, fatigue, and burnout were also cited barriers. Interestingly, participants were concerned about staying caught up with appropriate terminology and being unaware of antibias language choices.

Discussion
Based on our experience piloting this course, we believe it fills an important gap by providing health professionals with a practical "Incorporate teaching on biases and language into divisional or residency curriculum." "Compassionately challenge colleagues on their language." Use Mindful Language Toolkit provided "Replicate code of conduct card for colleagues." "Use 'states' instead of claims." Barriers to application of lessons learned Entrenched habits and/or lack of interest from colleagues or superiors "Difficult to change culture." "Long held habits." Lack of social capital and skills needed to effect change "I'm beginning third year rotations so my social capital is very low to influence care." "How to provide constructive feedback to others without making them feel bad." Time, fatigue, burnout "Burnout and how it leads to lack of empathy." "Complacency." Staying caught up with appropriate unbiased terms "Using the correct terms because language is always changing."

Most valuable aspects of workshop
Interactivity (sharing personal stories, word clouds, cased-based group work, large-group discussion) "The examples shared by colleagues and the experiences they have faced." "The discussion between providers of different backgrounds." "Variety of teaching methods, including high tech approaches." Mindful Language Toolkit "Pocket card and [toolkit] handout [helpful to prevent] forgetting." Facilitation of self-reflection "Self-assessment of biases." "Made me think critically about the language I use." Improvements to workshop Incorporate discussion of how to teach and address with colleagues "As an educator, I would appreciate specific tips on how to teach this or address this when I see others falling into this pattern." "How can we change the culture? People in attendance at this workshop are interested. How do you engage colleagues that may not be as aware and/or interested?" Provide more subtle bias and evidence-based examples "More space for small groups to discuss nuances." Provide patient perspectives "Hear from patients about their experience." More discussion time for small groups "Perhaps more space for small groups to discuss nuances." framework to identify and replace biased language. Stigmatizing language is normalized in medicine, and providers require concrete skills to make meaningful changes to their long-standing language use. Most of our participants learned new actionable skills in the workshop that they intended to incorporate in their practices, such as avoiding labeling patients, reframing situations, and reflecting before documenting. Prior implicit bias research has shown that teaching intentional strategies, such as those in our workshop, is key to reducing biased behaviors following an intervention. 22 Our workshop offers antibias skills and strategies in the Mindful Language Toolkit and provides activities that allow participants to practice these techniques in a safe environment. However, the Mindful Language Toolkit should not stand alone without interactive teaching and facilitated discussion.
Participants described important barriers to the implementation of antibiased language in their workplaces. This feedback emphasizes the need to engage key stakeholders, such as unit supervisors and hospital administrators, who can facilitate normative practice changes. In response to this identified barrier, we lengthened the final group discussion to include strategies for building institutional support.
Participants offered several suggestions for how to overcome cultural barriers to practice change. For example, they recommended including discussion of how to educate colleagues about stigmatizing language. This topic is often included in implicit bias trainings 23,24 but was not initially covered in our workshop. However, given that social pressure is one of the most commonly cited barriers to addressing implicit bias, we modified our workshop to address this topic.
Lessons learned from piloting the workshop include the importance of providing ample time for practice of new skills and discussion with peer learners, as well as the broad application of the material to learners from different stages of training and discipline. The Mindful Language Toolkit is a critical feature of the training that summarizes key content and implementation strategies. We will extend the length of the workshop to better address potential barriers to change and to offer teaching techniques to participants interested in disseminating content to their colleagues and trainees.

Limitations
The setting of the workshop likely introduces some selection bias, as participants who attended may have been motivated to reduce disparities. Course evaluations focused on participants' self-reported reactions to the workshop, and we are unable to comment on knowledge or behavioral changes. One way to evaluate these in future studies would be to query electronic medical records for biased words such as alcoholic or frequent flyer following a training. In addition, course evaluations did not assess the impact of the interprofessional learning environment. Anecdotally, the presenters remember many interdisciplinary small groups and rich interprofessional discussions, so we believe the content translated well to our broad audience. Lastly, there was no presession assessment for pre/post comparisons, and we did not record the total number of participants, which would have allowed for a calculation of course evaluation response rate. We recommend that future course evaluations include presession surveys, the professions of the learners, and their experiences with the interprofessional education.

Conclusion
Learners in our workshop on stigmatizing language in health care were highly engaged and found the material clinically relevant. Participant feedback suggests that the workshop provided actionable skills for replacing biased language in clinical presentations and electronic health records. Barriers to practice change are important to address in future iterations of the workshop.