As medical students at Columbia University Vagelos College of Physicians & Surgeons (VP&S), we (L.J.B. and C.T.) recognized implicit bias in the curriculum and wanted to take action to address it.1 Implicit bias refers to unconscious but learned beliefs that lead to negative associations regarding a person or group based on irrelevant characteristics of that person or group.2 Attending to this issue was particularly important to us given that mounting empirical studies and reports, such as the Institute of Medicine’s report Unequal Treatment, argue that implicit biases can exacerbate health disparities.3–5
Early in our training, we found that the curriculum often used White bodies as examples of healthy or normal, inadvertently implying that the bodies of people of color are abnormal or unhealthy. For example, in a lecture about the anatomy of the oral cavity intended to help students differentiate between healthy and inflamed gingiva, students were shown an image of pink-colored gingiva and taught that “healthy gums are pink.” This message was troubling to us because we know that healthy gums are not always pink; they can have purplish or brownish hues due to differential melanin deposition, particularly in populations of color. By emphasizing that healthy gums are pink, the lecturer missed the opportunity to describe how inflamed or otherwise unhealthy gums might appear in a variety of patients with different pigmentation. It soon became apparent that while this incomplete education has very explicit and harmful effects, the educator often communicated these biases unintentionally. In fact, this unconscious bias is merely the perpetuation of the lessons learned over many years and decades of medical education. We noticed that such implicit bias in various forms is common not only in the curriculum but also in medical literature more broadly, including in materials from other institutions, case presentations, textbooks, and standardized test questions.6–8
A growing body of literature suggests that these narrow depictions of “normal” may be largely the result of implicit biases to which we, including our teachers and writers of textbooks or educational materials, are all prone.9 In an effort to make positive change, in early 2016, we conveyed to medical school leadership our thoughts on the ways that implicit bias relating to race and ethnicity was affecting our curriculum. Administrators responded with interest and invited us to present our concerns to the school’s curriculum governing committees over the next several months. The faculty was open to our feedback and recognized implicit bias in the curriculum as an important issue that merited further exploration.
Development and Implementation of a Task Force and Guidelines
In response to our concerns, that summer, the VP&S vice dean for education convened a Task Force for Promoting a Bias-Free Curriculum composed of students, faculty, administrators, and staff to address issues of harmful racial and ethnic bias in our curriculum. We, along with coauthor D.G., were members of the task force, with D.G. serving as task force chair. Early on, the task force recognized a need to broaden its attention beyond a focus on race and ethnicity because it was evident that other harmful biases, such as in the areas of sexual orientation and gender identity, were also present in the curriculum. Over the following year, the task force adopted a 2-pronged approach to address such biases in the curriculum: (1) the development and dissemination of guidelines for educators to reduce bias in the curriculum and (2) the creation of a feedback portal to allow for ongoing discussion and adaptation of curricular materials.
The task force created guidelines for educators to consider when developing and implementing the curriculum in both classroom and clinical settings. It adopted a spirit of “co-creation” to draw from the collective insights of the community and create shared ownership of issues and solutions, while acknowledging the task force members’ own implicit biases. To that end, during guideline development, it actively sought input from additional students and faculty outside of the task force. The guidelines to promote a bias-free curriculum are brief, clear, and nonjudgmental, and they describe examples of bias along with possible solutions. Further, the guidelines (see Table 1) and reporting portal represent a collective effort by many different members of the VP&S community to address bias in our medical school curriculum.
The task force also recommended adapting the current learning environment reporting and review process to elicit feedback from students regarding their experience of bias or lack of inclusion in educational experiences. At VP&S, students now report bias in the curriculum in 2 ways. First, the VP&S Office of Education added the Bias Reporting Portal to the Learning Environment Reporting Portal, an anonymous survey used by this office to gather reports about the learning environment and medical student mistreatment. Students use this portal to report observed bias in real time. Additionally, students are oriented to the portal via email and in a first-year lecture on culture and bias in medicine. The VP&S Office of Education may flag incidents of bias from reports to either the Bias Reporting Portal or the Learning Environment Reporting Portal. Second, an automated process screens anonymous course evaluation comments for reports of bias using a search of comments for more than 100 keywords associated with mistreatment, poor learning environment, or bias. Education specialists in the Center for Education Research and Evaluation (A.S.S. and S.C.Q.) review flagged comments from all of these sources, and comments eliciting concern are further reviewed by the senior associate dean of student affairs (J.A.). Those comments are reviewed in a meeting of education deans and student leaders every 6 weeks (see Figure 1).
Bias reports from the portals and course evaluations are sent to the senior associate dean for curricular affairs, who contacts the faculty member (e.g., lecturer, course director) associated with the comments to share and debrief the reports as well as to develop a follow-up plan.
The curriculum committee approved the guidelines drafted by the task force and distributed them to students and educators in August 2017. In addition, the bias monitoring process was integrated into the school’s procedures to monitor the learning environment.
To begin to evaluate the impact of these interventions, we used our guidelines to help identify student reports of bias from the learning environment portal and in-course evaluations from the year preceding the introduction of the guidelines and the year following (see Table 2). From August 2016 to July 2018, there were a total of 245 student comments related to learning environment from course evaluations, the learning environment portal, and the bias portal. Of these 245 comments, 106 related to bias, including 26 comments reporting multiple instances of bias. The number of comments reporting bias increased from 42 in the year before the intervention to 64 in the year after the intervention. While we do see a rise in the number of reports related to bias, it is important to note that because this evaluation was done over 2 different years, the student cohorts represented are not identical due to different matriculation and graduation years. Therefore, further quantitative analyses including additional years would be necessary to draw definitive conclusions about this aspect of the impact of the intervention.
In addition to this increase in reports, substantive changes have been made to preclerkship and clerkship curricula. For example, after student reports and discussion with the lecturer, the aforementioned lecture on gingiva now states that “healthy gums range from pink to brown” with several representative example images. Moreover, a lecture on osteoporosis was identified because it discussed osteoporosis primarily in terms of the archetypal elderly White female patient without discussing other groups that might be affected or providing an explanation for their absence from the lecture. Following the report and discussion with the faculty member, the subsequent iteration of the lecture acknowledged the skew in the racial and ethnic makeup of the preponderance of subjects in the major research studies. In addition to the formal reports from the reporting tool, there have also been changes made from informal conversations following the introduction of the guidelines. In one example, coauthor D.G. reports that he made edits to his lectures after one such conversation with a student. While his slides had been the same for many years, the issues of bias had not been brought up before the introduction of the guidelines and bias portal.
We have also monitored the discussions between the senior associate dean of curricular affairs and the faculty member who is associated with the bias-flagged comment. The primary goal of these discussions is to reduce implicit bias in the curriculum by raising awareness in our community using real and recent examples reported online. We found that these discussions are sometimes difficult for a number of reasons: (1) the reporter and the reportee may not agree about the presence of bias, (2) the reporter and the reportee may have different thresholds for considering bias to be important, and (3) the reportee may feel accused of being explicitly biased. The discussions are intentionally nonpunitive and designed to be constructive, both for the reportee and reporter as individuals and also in the effort to increase open communication and awareness about bias.
Lessons Learned and Next Steps
Identifying and addressing implicit bias requires broad institutional support and a dedicated team of students, faculty, and staff. In our institution, the student effort was crucial, in particular L.J.B. and C.T., who elevated the issue into the general institutional conversation as well as all of the students who have reported using the portal. In the first year since the implementation of these resources, the rise in reports associated with bias may be due to a combination of factors, including: (1) heightened awareness of bias by students, (2) student willingness to report bias given additional institutional attention to this matter, and (3) enhanced surveillance for bias via the portals and screening of course evaluation comments. While this structure continues to rely on student participation, faculty and administrators at all levels of the institution must also continue to engage in the efforts to promote a bias-free and inclusive learning environment.
We hope that our efforts will be useful to others across a variety of educational environments and institutions. While our institution already had deans and faculty who were very accessible to the student body, a curricular governance structure that incorporates students and administrators on all committees, and a robust system for evaluating the learning environment, every medical school has its own infrastructure that can lend itself to such an intervention. In our experience, the critical first steps were describing the problem and creating a space to discuss potential solutions. Implicit bias cannot be addressed by a single person or even a single innovation. It requires the investment and dedication of individuals at all levels and in all parts of an educational community. Such a concerted effort to address bias goes a long way toward making medical education more comprehensive, inclusive, and helping all of us grapple with the challenges of training physicians to be good doctors to all of their patients.
Our goal is to better prepare trainees to care for the wide variety of patients they will encounter in their careers by creating a process for identifying and addressing bias in the curriculum and establishing best practices for bias-free and inclusive teaching. We recognize that such a goal cannot be fully realized all at once, if indeed, it can be achieved at all. As such, future directions of our work will require periodically reviewing and updating the guidelines to ensure that they identify comprehensive, inclusive, and relevant examples of bias, with practical suggestions for bias-free teaching. Additionally, we will use the guidelines in future curriculum review and innovation efforts. As evidenced by the importance of the students’ direction in this initiative, all future efforts must ensure that they amplify all voices, regardless of societal and institutional rank.
While this paper describes the perspective of one medical school, we developed the guidelines with the hope that they would be a useful tool for addressing bias not only at VP&S but also at other institutions. Furthermore, medical licensing bodies, textbook publishers, and board and licensing examination review companies may want to consider adopting the guidelines and implementing their own mechanisms for identifying and remedying bias as part of their quality improvement efforts. Bringing implicit bias in our curricula into the light is difficult work, but these kinds of efforts offer us great hope in promoting diversity, equity, and belonging in the health professions.
The authors would like to thank Lisa Mellman, MD, Boyd Richards, PhD, and Florante Garcia for their work over the years to make this project possible.
1. Gowda D, Benoit LJ, Travis CI. A common purpose: Reducing bias in the curriculum. Acad Med. 2017; 92:274–275
2. Green AR, Carney DR, Pallin DJ, et al. Implicit bias among physicians and its prediction of thrombolysis decisions for black and white patients. J Gen Intern Med. 2007; 22:1231–1238
3. Institute of Medicine. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. 2003, Washington, DC: National Academies Press
4. Phillips RS, Hamel MB, Teno JM, et al. Race, resource use, and survival in seriously ill hospitalized adults. The SUPPORT investigators. J Gen Intern Med. 1996; 11:387–396
5. FitzGerald C, Hurst S. Implicit bias in healthcare professionals: A systematic review. BMC Med Ethics. 2017; 18:19
6. Tsai J, Ucik L, Baldwin N, Hasslinger C, George P. Race matters? Examining and rethinking race portrayal in preclinical medical education. Acad Med. 2016; 91:916–920
7. Metzl JM, Hansen H. Structural competency: Theorizing a new medical engagement with stigma and inequality. Soc Sci Med. 2014; 103:126–133
8. Acquaviva KD, Mintz M. Perspective: Are we teaching racial profiling? The dangers of subjective determinations of race and ethnicity in case presentations. Acad Med. 2010; 85:702–705
9. Krishnan A, Rabinowitz M, Ziminsky A, Scott SM, Chretien KC. Addressing race, culture, and structural inequality in medical education: A guide for revising teaching cases. Acad Med. 2019; 94:550–555