Most medical schools teach students how to communicate effectively with a diverse patient population. However, diversity is usually interpreted as a range of ages, ethnicities, and physical and mental challenges rather than diversity of patient sexual orientation. Accrediting organizations such as the Accreditation Council for Graduate Medical Education (ACGME) (1) and the Liaison Committee on Medical Education (LCME) (2) recognize the need for medical schools and medical students to understand and be sensitive to diversity of sexual orientation. When students are not taught how to discuss sexual orientation openly with their patients, they run the risk of taking the long path to a correct diagnosis, or missing it altogether. They may also miss opportunities for effective patient education on wellness and disease prevention. In short, if medical students do not understand a patient’s sexual orientation, they are missing an essential component of what makes each patient unique.

The National Coalition for Lesbian, Gay, Bisexual, and Transgender Health reports that lesbian, gay, bisexual, and transgender (LGBT) patients lack equal access to quality health care and face many barriers when attempting to access health care. Consequently, LGBT patients are more likely to neglect their health (3). Many medical students are aware of this gap in their education and see the value of adding more content on sexual health because they feel “undertrained” in this area (4).

This gap in medical education is long-standing. In 1992, a study in Academic Medicine (5) explored faculty perceptions of how homosexuality was taught in medical school. This first exploration of LGBT health care content in the curriculum surveyed directors of psychiatry clerkships. The study found that homosexuality was taught in a marginalized manner (i.e., a curricular approach that could potentially trivialize the importance of LGBT health care). The authors concluded that this content should be wholly integrated throughout the curriculum to improve knowledge and comfort level.

A 1994 study (6) reported that the majority of psychiatric residency programs surveyed did include gay and lesbian content in their curriculum, but program directors varied on how this should best be done. Another 1994 study by Schatz et al. (7) found that 88% of physicians and medical students surveyed had heard colleagues disparage homosexual patients, and 67% had witnessed denial or reduction of care.

This study investigated the amount of instructional time in the undergraduate medical curriculum dedicated to teaching LGBT health care, analyzed current curriculum in terms of the estimated national average, and identified faculty perceptions of barriers to including LGBT content in the curriculum so that decisions about improving the LGBT curriculum would be evidence-based.

Because the Association of American Medical Colleges (AAMC) graduation questionnaire findings revealed that 29% of University of Louisville graduates felt that the amount of time devoted to human sexuality was inadequate (compared to 19% nationally) (8), we hypothesized that the number of instructional hours in the undergraduate medical curriculum focused on LGBT patient care would be less than the national average, that responding faculty would identify barriers to including LGBT content in the undergraduate medical curriculum, and that there would be a negative correlation between the number of barriers faculty members identified and the number of hours in their courses or clerkships addressing LGBT health care.

This study was unfunded and deemed exempt by the institutional review board; the informed consent process consisted of a brief letter of explanation to potential participants providing a description of the study rather than a signed release. Course and clerkship directors were invited to complete the questionnaire during a regularly scheduled meeting. Both clinical and basic science courses were included because our curriculum committee encourages the integration of basic science and clinical application. The questionnaire included seven quantitative and two qualitative questions. Fourteen directors representing 19 required courses/clerkships completed the survey. Four course directors representing five required courses (all in the basic sciences) chose not to participate.

Because the population was small, the data were not adequate to justify seeking a Cronbach alpha coefficient. Instead, the face validity of the instrument was based on a review by an external content expert in LGBT health care education (Mark Townsend, M.D., of Louisiana State University in New Orleans) and University of Louisville’s Medical Education Research and Evaluation Unit. The “barriers” listed in the questionnaire were derived from the literature and anecdotal knowledge of faculty reaction to including LGBT content in their courses. Data were not anonymous because faculty assignments to specific courses and clerkships were common knowledge, but all data were treated as confidential.

We used SPSS version 17 to analyze quantitative data. Descriptive statistics were calculated for all replies. The correlation between “actual hours” and “number of barriers identified” was calculated using the Spearman correlation coefficient (H₀ rejected at p<0.05). Qualitative data from the open-ended questions were analyzed using a variation of Glaser and Strauss’ 1964 Constant Comparison method. Survey results indicate a marked difference between the number of LGBT content hours in basic science and interdisciplinary courses and the number of hours that these course directors see as “ideal” (Table 1).

Directors representing six of the courses and clerkships (32% of respondents) indicated that they did not see any significant barriers to including LGBT content in their courses and clerkships. Eleven directors felt that a lack of instructional time was a barrier, eight cited a lack of relevance to course content, and seven identified a lack of professional development on how to teach the topic. Two course directors cited a lack of personal comfort with the topic, and two identified a lack of validated content on LGBT topics as potential barriers (Figure 1).

A Spearman’s rank correlation coefficient revealed a significant negative relationship between the number of instructional hours devoted to LGBT topics in courses and clerkships and the number of barriers identified by respondents (r_s=−0.47, p=0.047); directors who perceived more barriers to teaching LGBT topics devoted fewer hours to those topics in their curricula.

The two qualitative questions on the survey were “If you address LGBT issues directly or indirectly in your course or clerkship, please describe” and “Where do you think topics related to LGBT health care should be taught in the medical school curriculum?” When asked where LGBT content should be included in the curriculum, 13 of the 14 respondents replied. Four replies were general (e.g., “as needed” or “throughout the curriculum”), but several others were much more specific. For example, one director specified, “outpatient and inpatient primary care, especially pediatrics outpatient with focus on adolescents, and in small group instruction.”

The University of Louisville course and clerkship directors’ estimate of 6 hours dedicated to teaching LGBT content was almost twice the 3.5 hours estimated in the American Medical Student Association Plus One Initiative report (9). The estimated 6 hours, however, were certainly less than the 11 hours proposed by course directors as “ideal.” Both the University of Louisville and national estimates are self-reported and therefore a reasonable comparison. However, a more precise measure of the national average may result in a study now in progress from Stanford University School of Medicine (10).

It was not surprising to find the barrier most often cited was “lack of instructional time” because this barrier is often mentioned when any new content is proposed for the already-full medical student curriculum. The two barriers cited next most often, “lack of relevance to my course content” and “lack of professional development on how to teach this topic,” are good news for medical educators because it is much easier to address extrinsic issues such as “lack of professional development” than intrinsic issues such as “lack of comfort with the topic” with professional development.

Limitations of this study include the self-reported estimates of curricular hours and the small population from a single institution. This study could be replicated at other institutions with greater emphasis on why faculty members do not perceive LGBT instruction as relevant to their courses, especially in the basic sciences, and which strategies are most effective for integrating LGBT health care content into the curriculum. Offering faculty development on how to include LGBT content and providing validated content should decrease the perception of “barriers” and increase the number of hours dedicated to teaching LGBT health care topics.

The ultimate goal is to increase medical student knowledge and sensitivity for communicating with a diverse patient population. Identifying barriers to this goal is a small but important first step in a long process of curricular change.

N=14 faculty representing 19 courses and clerkships

At the time of submission, the authors reported no competing interests.