Participants

In Taiwan, most medical schools are undergraduate entry level, with years one to four as the pre-medical and pre-clinical stages and years five to six as the clinical stage. All medical graduates attend the compulsory two-year postgraduate (PGY) general medicine training program. Undergraduate medical students/trainees were recruited in two phases. In the first phase, a survey on cross-cultural care competence involving diverse cultural group (Table 1) was administered to 1545 medical students from one public and one private medical school in southern Taiwan and 1120 (72.5%) responses were received.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Diverse groups included in cross-cultural care competence survey.

https://doi.org/10.1371/journal.pone.0270862.t001

Initial results from the survey were then analyzed to serve as baseline for qualitative data collection. The researchers recruited from each school a random sample of students from the cohorts of students that completed the survey and conducted at least one focus group (FG) and one interview for each year of study, involving a total of 89 students who participated in 14 FGs (with four to six participants in each session) and 19 individual interviews. Foreign students studying abroad in Taiwan were excluded. For FGs, cohort representatives first posted recruitment messages on class bulletins, then students were recruited through convenience and snowball sampling. Interview participants were recruited through purposive sampling via email or phone. Demographic characteristics of participants in FGs and interviews are shown in Table 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Demographics of student participants from two medical schools in Taiwan: 2015–2017.

https://doi.org/10.1371/journal.pone.0270862.t002

Design

Qualitative data were triangulated to obtain in-depth descriptions of participants’ perceptions and experiences. Researchers conducted most interviews and FGs on medical school campuses or meeting rooms of universities’ affiliated hospitals, with one interview conducted at a quiet café close to campus. Structured questions were prepared (See Box 1 for English translation of the questions, which were given in Mandarin), covering general attitudes to students’ and trainees’ preparedness and competence in caring for diverse groups, including LGBT individuals. Interviews and FGs lasted approximately 40–60 minutes each and were audio-recorded and transcribed verbatim. The same semi-structured questions, which were created based on findings from quantitative data and validated by a panel involving project researchers and two clinicians, were used for both interviews and FGs. We are aware that FGs and interviews are different methods of data collection, however, we believe these differences do not warrant the need to conduct separate sets of analyses as the phenomenon identified by participants in both groups were similar. Thus, these data were triangulated and interpreted together.

Ethics approval for this study was obtained from the National Cheng Kung University Governance Framework for Human Research Ethics (Approval No.: NCKU HREC-E-106-291-2/ NCKU HREC-E-104-035-2). Participants in focus groups and interviews provided written consents at the beginning of the sessions.

Analysis

Two members of the research team reviewed and discussed each transcript to reach consensus on differences in coding. Inductive thematic analysis was conducted. Transcripts were reviewed, coded manually, and processed using Nvivo 10. Researchers independently reviewed transcribed data and identified recurring topics. From these topics, codes for texts that expressed similar contents were established and students’ quotes were sorted according to these codes. Codes were collated into various themes that were then reviewed, defined, and labeled. Only data relevant to LGBT care provision is extracted and discussed. Similar codes were clustered to identify emerging themes and sub-themes. Recurring contents were grouped into four main themes and nine categories. The established themes and categories were translated to English and compared to the original data to ensure they reflect the original contents.

Attitude and perceptions

Bias and stereotypes

Respondents observed that throughout training, they had heard about or experienced certain degree of stereotypes and biases in relation to LGBT individuals. However, they also recognized that biases and stereotypes might affect their reflexive thoughts and judgements.

Preparedness and skillfulness

Curriculum

Bias formed in medical education

Bias against LGBT individuals could be formed in the process of medical training and cultural stereotyping is a problem for doctors [42]. Implicit biases are also formed through the combination of cultural socialization and repeated experiences [43], and these biases might be difficult to unlearn once they become habitual [44]. Medical students and trainees from younger generations were aware that biases and stereotypes embedded in training might have negative influences on their judgements. For instance, indicative of students’ awareness, stigmatization of HIV repeatedly emerged. A Taiwanese study also indicated some teachers may even express openly to students their hostilities toward LGBT individuals [45]. However, students also justified that having unconscious or conscious associations of bias with illness in their training environment is helpful. It is possible students may have unconsciously integrated biases and stereotypes related to LGBT individuals as part of their medical knowledge or training, rather than differentiating between the breadth of their medical knowledge and professional skills.

Examples most often mentioned by students were related to HIV, AIDS, and other sexually transmitted diseases. However, LGBT individuals are also at higher risks for certain cancers, experiences of discrimination and trauma within the healthcare system, mental health disorders, risk behaviors, and other health disparities [46,47]. Since respondents hardly mentioned these other issues, it is possible students were more influenced by common LGBT stereotypes than they were by deeper understanding of the epidemiologic patterns of disease and risk factors. This points to the need for additional education on caring for LGBT patients.

It is also evident that as students advanced through different stages of medical program, associations formed could be increasingly instilled into reflexive clinical reasoning processes, which might lead to the institutionalization of bias in medical culture [48]. This implies that the norm for medical training is to make a quick assumption of the most likely diagnosis based on clues of a patient’s social background. However, the mindset of using this pattern of reasoning in medical practice may pose the risk of or even harden the stigmatization of diseases. This might result in medical practitioners becoming less sensitive to psycho-social concerns of patient-centered care, as well as possibilities for misdiagnosis or making assumptions that contribute to medical errors [49]. Medical practitioners’ bias toward LGBT individuals might also further perpetuate physical and mental health disparities experienced by LGBT patients [50]. As forming associations between LGBT patients and diseases is not unique to Taiwan’s medical students/trainees [49], medical education programs globally should consider the contexts in which these associations are formed and the impact they have on LGBT patients’ health care to minimize these associations’ negative influences on LGBT patients’ access to adequate care.

Self-perceived preparedness versus actual competence

Despite some mixed responses, most students nevertheless expressed strong confidence in their preparedness to care for LGBT patients and in their belief that they held non-discriminatory attitudes toward LGBT individuals as they objectively viewed scientific evidence as the basis for treating all patients. This mirrors similar findings from previous research [42], among Harvard medical students [51] and among residents in a national survey [52] who felt well-prepared, but were less confident in their specific skills of caring for LGBT patients. The findings support the idea that social changes in Taiwan have had a significant impact on the students’ attitudes toward LGBT patients. Exposures to social movements and LGBT individuals have evidently helped to increase understanding of and familiarity with these groups. Whereas students considered themselves as being prepared to care for LGBT patients, very few of them could actually articulate in detail the nature of their competence. They tended to equate non-discriminatory attitude as being sufficiently prepared, while their skills were shown to be inadequate when it came to real interactions with LGBT patients. It also did not occur to most students that biopsychosocial knowledge (such as traumatization by discrimination) and communication skills are equally essential for developing mutual trust in physician-patient relationships [53] and thus providing appropriate care. It is evident that students were not sure where to draw the line between asking questions that were “too sensitive” and those that were “necessary” because they might affect provision of effective care [54,55]. The incongruence in perceived and actual preparedness should be considered across different layers of Taiwan’s social environment and medical education curriculum.

Curriculum gap

Results suggest medical education in Taiwan has not evolved along with nor responded sufficiently to social changes and diverse patient needs. Despite legal advances in LGBT rights and requirement for medical education programs to include training to eliminate gender biases, LGBT related issues have not been well addressed in the curricular content of Taiwanese medical institutions. Other countries are also dealing with similar curricular gaps [29,32,35,56,57]. It is apparent that the curricular content lacks consistent and systematic design across formal, informal, and hidden curriculum when it comes to LGBT related issues, yet students’ views of factors that contribute to their preparedness level appear to be tied to medical education curriculum. There are still gaps in formal curriculum regarding the training stage in which relevant courses are taught. Although a few courses in the earlier stages of medical education were mentioned, it was apparent that the aforementioned criteria stated by TMAC [40] and LCME [41] still need to be more systematically and fully integrated in Taiwan’s medical education curriculum.

Additionally, teachers’ cultural and gender sensitivity could play a part in the hidden curriculum, and conscious or unconscious biases and stereotypes embedded in teaching remain a challenge to change in the educational climate [45,58–60]. Further faculty development will be necessary to address these issues, to identify barriers to inclusion of LGBT care competency, and to design and implement effective curricula [56].

Some students believed they would ‘naturally’ and ‘quickly’ acquire the skills once they reach clinical stage, but in an earlier study of clinical students in a psychiatry rotation, students still have limited understanding of how to engage with transgender patients despite having reached the clinical setting [38]. The intended learning outcomes of relevant training at the clinical stage appears to be unclear and may be largely affected by clinicians and their role modeling behavior, and the hidden curriculum in clinical environment.

Teachers as role models

Although teachers are important role models, messages that emerged about their influence are paradoxical. Students were aware of biases and stereotypes that exist in teaching; however, observing clinical teachers’ communication skills and behaviors was seen as an effective way of learning. If clinical teachers associated certain health habits with particular cultural groups, students might reflexively or subconsciously include these behaviors in their understanding of diseases. Clinical observation is important, but whether students have adequate role modeling is not assured. Teachers’ cultural competence warrants caution and requires further faculty training, particularly if they were seen as role models in students’ development of professionalism [60–62]. Since current health care providers in Taiwan were trained when LGBT issues were less openly discussed, it follows that the informal learning environments both on campus and in clinical affiliates may be lacking in quality mentoring [60–62]. Similar concerns were expressed in western countries such as the US as many clinical educators also expressed receiving little training on LGBTQ+ health themselves or ways to teach these contents [63].

The degree to which the curricular content has served to eliminate health disparities among LGBT patients appears to be insufficient. Medical educational program directors and educators need to identify and address the gap between expected competence, students’ self-assessed competence, and the curricula throughout the various stages of students’ medical education.

Limitations

This study has several limitations in terms of representativeness. We conducted qualitative research in only two medical schools, which may limit the generalizability of findings. It is possible that students who held different views or had greater knowledge about LGBT patients did not participate. However, there is reason to be confident that the views expressed by respondents are typical of those held by students of different Taiwan medical schools. First, the diversity of students in this study may be limited by uniform selection criteria adopted by medical schools across Taiwan. Nevertheless, uniform selection criteria also make these perspectives more generalizable to other medical students. Thus, despite the different curriculum design and duration, students of both schools had similar perceptions. Participants’ actual knowledge of specific medical services that LGBT patients might need was not assessed. Since this study was cross-sectional, cohorts of students were not followed as they advanced in medical training. Thus, the results may be confounded by variations in training stages. Participants might have also withheld their own opinions out of fear of repercussion or disapproval from peers. These concerns will require further consideration in future studies.