Medical students' views on the value of trigger warnings in education: A qualitative study

Abstract Background Trigger warnings—advance notification of content so recipients may prepare for ensuing distress—feature in discussions in higher education. Students' expectations for warnings in some circumstances are recognised, and some educators and institutions have adopted use. Medical education necessitates engagement with potentially distressing topics. Little is known about medical students' expectations regarding warnings in education. Methods All students from a 4‐year graduate‐entry UK medical degree programme were contacted via digital message outlining study details and were openly sampled. Qualitative methodology was chosen to explore participant expectations, experiences and meanings derived from experiences. Students participated in semi‐structured interviews exploring perspectives on functions, benefits and drawbacks of trigger warnings in classroom‐based medical education. We analysed interview transcripts using thematic analysis. Results Thirteen semi‐structured, qualitative interviews were undertaken. Themes in the following areas were identified: (1) students' experiences influence understanding of trauma and trigger warnings, (2) warnings as mediators of learning experiences, (3) professional responsibilities in learning, (4) exposure to content, (5) professional ethos in medical education and (6) how to issue trigger warnings. Students recognised the term ‘trigger warning’, and that warnings are an accommodation for those affected by trauma. Students' conceptualisation of warnings was influenced by personal experiences and peer interactions both within and outside education. Students expressed both support and concerns about use of warnings and their ability to influence learning, assuming of responsibility and professional development. Discussion Diverse student opinions regarding warnings were identified. Most students suggested that warnings be used prior to topics concerning recognised traumas. Incremental exposure to distressing content was recommended. Students should be supported in managing own vulnerabilities and needs, while also experiencing sufficient formative exposure to develop resilience. Greater understanding of trauma prevalence and impacts and underpinnings of warnings amongst students and educators are recommended to optimise education environments and professional development.

students and educators are recommended to optimise education environments and professional development.

| INTRODUCTION
Trigger warnings-prior notification allowing recipients to prepare for or avoid sensitive content and ensuing distress-are widely encountered in communications. 1,2 These advisories are considered to have originated online as an accommodation for survivors of sexual violence or other trauma and who may experience symptoms of posttraumatic stress disorder (PTSD). 3,4 Although associated terminology may have changed, the practice and construct predate their online use. 3 Their use has been widely adopted in relation to diverse settings and topics. 2,5 Discussion continues about their role in education, with evidence of some students expecting warnings in relation to distressing topics. 6,7 In some cases, educators and institutions have shared these sentiments, voicing support and citing rationale for adoption of warnings in practice or policy, 8,9 including desire to curate inclusive learning environments. 5,10,11 Support has not been unanimous, with opposition to the construct, underpinning principles and use of warnings in education noted. 10 Concerns expressed include promotion of avoidance, 12 hypersensitisation of recipients 1,13,14 and censorship effects. 12,15 Despite routine use and relevance to classroom settings, current literature regarding trigger warnings is largely derived from opinion pieces based on individual or few author perspectives and rigorous academic evaluations or empirical evidence regarding trigger warnings remain lacking. 16 As an accommodation for affected individuals, trigger warnings and associated discussions have relevance to clinical education contexts where discussions of trauma, suffering and inequalities are integral and commonplace. 17,18 Graduating professionals need to regularly encounter and manage these subjects, while maintaining own well-being. Medical students' perspectives in this area remain relatively unexplored. Inquiry may provide insights into students' experiences of distressing content, how best to prepare graduates for managing distressing experiences and whether warnings may have a role. Without student consultation, educators risk maintaining inconsistent approaches, outwith a framework for practice, culminating in suboptimal learning environments.
Sensitive or trauma-related subjects in medical education may hold personal relevance for students. [17][18][19] Experiences in medical education themselves have the potential to traumatise, irrespective of personal histories. 20,21 Although currently there is limited literature regarding secondary traumatisation in medical students, 20 there is substantial evidence regarding depression and burnout amongst medical students, [22][23][24] entities that are more prevalent amongst medical student than general and other student populations. 22,23,25 These issues are further compounded by reported stigmatisation of medial students experiencing mental illness. 26,27 Incorporation of trigger warnings could promote accessibility by enabling reasonable accommodations for students with trauma histories or mental health difficulties 10,11 and signal that well-being is valued in the organisation and profession, as previously described by medical educators. 28 Diversity of medical student populations is increasing internationally, in response to measures to ensure representation of the served patient populations. 29,30 Significant increases in numbers of students admitted from educationally and socially disadvantaged backgrounds as well as groups from ethnic minorities and students experiencing disability 31 are noted-all groups noted to experience higher incidence of adversity. 17 Consideration of trauma-informed approaches to distressing content, including use of warnings, appears increasingly justified and necessary in this context. 17,32 Given established impacts of emotion 33 and PTSD 3 on learning, efforts to identify evidence for efficacy of trigger warnings in general education literature have explored impacts on arousal and distress. One experimental study identified that trigger warnings' had 'trivial' effects on participants' ratings of negative material and distress symptoms, 14 while acknowledging that warnings may have other effects not assessed by their study. Bellett et al.'s larger replication study overturned their original finding that TWs affect some domains of resilience, leading them to conclude that warnings are 'inert'. 13 They acknowledged that educators may use warnings for other reasons not explored in their study. Recruitment from non-traumatised populations limits generalisability of these findings. Results of each of these studies 13,14 were also limited by reliance on participant self-reporting of symptoms.
Evidence regarding trigger warnings in medical education is limited. Our previous semi-structured interview study exploring the views and practice of medical educators identified that educators regularly employed warnings in classrooms settings. 28 They cited various rationales beyond mitigating hyperarousal as well as a number of concerns relating to use of warnings. A single study of medical student perspectives, a subsection of a larger survey, exploring students' views suggested warnings may have a role in teaching about trauma but did not establish clear consensus regarding support for warnings. 34 Survey methodology, however, prohibited depth of discussion.
Despite clear relevance of trauma-related content to medical student populations and experiences, and impetus to consider appropriateness of warnings in managing related impacts in education settings, medical student perspectives remain underexplored. This current interview study aimed to explore medical students' experiences and perspectives regarding the role of trigger warnings in classroom-based medical education. Noting that experiences of traumatising content are personal to the individual, 17 we wished to broadly explore students' perspectives and constructs of trigger warnings, including both within and outwith education experiences. As trigger warnings may have pedagogical function beyond preventing hyperarousal for individuals identifying as affected by trauma, we wished to explore perspectives of students identifying with varying personal experiences of adversity.
We formulated the following research questions: Do medical students perceive value in the use of trigger warnings? What are medical students' perspectives on the function, benefits and drawbacks of trigger warnings in classroom-based medical education?

| METHODS
In this study of medical students' perspectives and expectations regarding trigger warnings, we wished to explore social behaviours and experiences, meanings derived from experiences and factors underlying expectations of an educational phenomenon; thus, a qualitative methodology was used. 35 Individual semi-structured interviews facilitated deeper discussion of more complex questions allowing participants to share detailed accounts of experiences, interpretations and perspectives. 36 The study was approved by University of Warwick Biomedical Research Ethics Committee.

| Participants
Students on a 4-year medical degree programme (MBChB) were recruited. Eligible participants needed to have completed at least 3 months of the programme, ensuring adequate experience of classroom-based teaching, including lectures, case-based learning and small group sessions. At the time of recruitment, the most junior students were nearing completion of first year, so were eligible to participate. Additionally, senior students could offer reflections on the appropriateness of early teaching as preparation for clinical practice.
Unlike previous studies in this area, we did not limit participation to individuals who identified as not having experienced past trauma. 13 Personal trauma history may influence perspectives and value participants assign to warnings. However, we also recognised that trigger warnings may be viewed as serving wider pedagogical function, including development of understanding and empathy towards trauma amongst non-affected individuals, as suggested by previous studies 10,28,34 or, conversely, impeding learning experiences; thus, all students on the 4-year programme were openly sampled, capturing diverse perspectives and their evolution through programme progression.
Students were contacted via a digital message and provided with information outlining study details. The participant and recruitment information highlighted that responses would be anonymized and decisions regarding participation had no bearing on academic progression. We stated that we did not intend to directly explore distressing personal experiences, but that participants may refer to previous experiences in responses. Participants were informed of steps that would be taken if at any stage they needed support including discontinuing interviews and signposting to appropriate services. After addressing any questions, participants provided written consent to participate.

| Context
Our programme is atypical in UK medical education as a graduateentry programme, compared with standard school-leaver entry.
Degree holders from any academic background are admitted. These criteria widen participation in medical education by traditionally underrepresented groups and in relation to student sociodemographic profile. Students are older, come from more varied backgrounds and have greater life experience. As an accelerated programme, ours is intensive and academically demanding; thus, student well-being is emphasised in curricular development and delivery. Guidelines for practice in relation to warnings have been developed, based on a student feedback and disseminated amongst educators, and discipline teams have in some cases determined best approaches in their subject area. However, there is not currently an overarching institutional or departmental mandated policy on use of trigger warnings.

| Data collection: Interviews
We developed interview questions aligned with our overarching research questions. As this study builds on a previous study of educators' views in this programme setting, 28 these findings, in addition to existing literature and researcher discussion, influenced focus. Nonetheless, we wished to identify novel, unanticipated areas of priority for student participants. A semi-structured interview approach with open-ended questions was adopted to enable discussion of participants' experiences, with clarification and probing to qualify responses.
Interview guide (Appendix S1) provides detail of questions.
As the study was conducted during the COVID-19 pandemic and In the previous study of this subject with medical educators, the term trigger warning was not used initially, to accurately explore educators' use of warnings. As students may have experienced personally impactful content and, due to perceived power differential between students and faculty, it was ethically imperative to highlight intended discussion of trigger warnings and related circumstances during recruitment. Further, it was anticipated that students would readily recognise the term as it had been noted in student feedback. Students had an opportunity to clarify views on what constituted a trigger warning.
The interview guide was structured to initially explore participants' experiences of trigger warnings in day-to-day life and then in classroom-based education. This initial, general approach sought to establish general familiarity with the subject in a less personalised way, before the possibly more contentious subject of warnings in education. Semi-structured interviews allowed an approach that was often more fluid, enabling participants to freely discuss experiences and areas of greater priority to them. After initial interviews, a further question exploring previously unanticipated areas was added.

| Data analysis
Thematic analysis was used to identify, analyse, organise, describe and report themes within the dataset. Here, a theme is a notable feature of the data relating to the research questions and adds meaning. 37 Thematic analysis was chosen to assess participants' perspectives and identify similarities and differences in responses and unanticipated perspectives, creating a rich account of the data. 38 HN transcribed verbatim audio recordings. We both read and reread transcripts for immersion and familiarisation. Analysis of initial interviews commenced as further interviews progressed. Any striking features and patterns noted in the data during collection, transcription and analysis were recorded and discussed. These were then incorporated in identifying preliminary codes. Further codes were identified, developing a full coding framework. HN iteratively coded all transcripts using Nvivo12. LR read all interview transcripts to triangulate and establish agreement. As new codes were identified, these were applied to previously coded transcripts. We reached consensus on suggested codes and how these were assigned to data. All data assigned a particular code were collated. The complete codebook was reviewed, searching for relationships and patterns, thereby identifying themes inductively. Diagrams were used to organise themes and subthemes. Proposed themes were compared with the dataset, confirming key findings had been reported. Titles of themes were then revised, ensuring appropriateness. Detailed notes were maintained throughout analysis, beginning in the transcription phase. Features and patterns in the dataset and codes and development and hierarchies of themes were documented in a reflexive diary, providing an audit trail.

| Reflexivity, positionality
We both have leadership roles with oversight of student feedback and programme enhancement, including issues of accessibility and duty of care. LR has a senior educational leadership position and is known to students. HN has a quality leadership role and is directly known to fewer students. While involved in curriculum development, we are not substantively involved in programme delivery. Noting that roles may impact students' willingness to share experiences, HN conducted all interviews.
Regular researcher meetings occurred throughout the study, discussing reflexive notes and observations. Discrepancies in analysis, interpretation and reporting were considered, including regarding coding and theme titles. We maintained awareness of potential bias during analysis and actively sought evidence of contrary views, ensuring that our individual perspectives did not disproportionately influence interpretation or reporting.

| RESULTS
We conducted 13 semi-structured qualitative interviews with students (six males and seven females). Data collection occurred between June and October 2021. Average interview duration was 65 min (range 47-101 min). Participants from various backgrounds and from each of the four programme years participated (Table 1), sharing diverse perspectives and experiences and providing data relevant to the research questions and unanticipated areas. Noting the quality of dialogue, variety of participant perspectives and insights shared, we identified that we had appropriate data to address the research questions after completion of 13 interviews. 39 In addressing research questions to explore student perceptions of value, function, benefits and drawbacks of trigger warnings, six thematic areas were identified, shown in Table 2. Quotes are presented by participant number and year of study. Support for warnings varied; individuals who identified personal need for warnings were exclusively supportive of use, whereas individuals who did not identify own need fell into two groups of supporters or opponents (fully or in part).

Participant number
Year of study Gender There was also recognition that beyond specific topic areas, trauma was not uncommon and that this may be associated with increasing self-awareness and emotion recognition.
My generation are known as the snowflake generation.

| Responsibilities in learning
Participants frequently discussed educators' and students' professional responsibilities in the educational process, which included respective responsibilities to provide and engage with content that could be distressing. There were a range of views in relation to expectations of individuals and where the balance of responsibility lay within that relationship.
A small number of participants appeared to view the duty of care and accountability for student well-being to be predominately the remit of the medical school, with students being recipients of educa- There was value in students learning to tolerate negative emotional reactions that would be relevant to experiences such as professional failure or poor patient outcomes.
Others similarly discussed the need for students' self-awareness and self-management of vulnerabilities. However, they did not view warnings as being antagonistic to students assuming responsibility and developing self-awareness. Where warnings were provided as overview statements of content, these fulfilled the function of informing recipients, prompting them to reflect and determine their readiness or need for support. A gradual approach to withdrawal of warnings could scaffold students' self-awareness and self-reliance, allowing development and assumption of responsibility over time and enabling self-identification and action planning regarding sensitivities.
There's a certain element of autonomy and self- The classroom was primarily associated with learning and students identified their personal learning as being of greatest priority. It was therefore expected to be a safe, supportive environment. Where exposure to distressing content featured, warnings could allow students to prepare themselves.
Within an education setting … you are in an environment which is supposed to be nurturing and supporting you as a student. When you are in placement … it's a professional capacity … When you are sitting in a classroom, you are expecting to … be able to think … reflect

| How to provide warnings
Participants discussed the principle of providing advance notification and how warnings were given. Some students who expressed limited support for warnings and underlying principles conceded that use may be appropriate in limited circumstances related to recognised traumas.
Wide variation was noted regarding the most appropriate way to do this and the primary rationale. Several participants expressed disapproval for the term due to connotations associated with trigger warnings: the expanded use to encompass content of widely varying severity and as a tool that could permit avoidance or stifle debate.
Others, noting that warnings presupposed or could precipitate harmful responses, proposed alternative terminology such as content statements, providing objective information. This measure, coupled with signposts to supports, delegated responsibility to students as partners in the education process to consider own needs.
Where someone just says, "we are going to be discussing this today" … so it does not really come as much of a surprise. "If you find this content upsetting then please feel free to step out of the room and rejoin when you feel comfortable". I think that's a bit more sensitive to people … giving them information and ground rules, as opposed to just announcing, "Trigger warning!", because it's … a thing that needs to be done. (P1, year 2) Others expressed expectation for directive warnings in some circumstances, demonstrating educators' compassionate acknowledgment regarding the impact of content on students. Some participants did not feel that regular forewarnings were a justifiable expectation, preferring to be informed during induction regarding the breath and nature of content that would be encountered and how to access support, if required.

| DISCUSSION
This study sought to explore medical students' experiences, perspectives and priorities regarding use and value of trigger warnings in classroom-based education. In contrast to our previous study with educators, all participants readily recognised the term from contexts including media, therapeutic and educational settings. In keeping with previous findings, arguments both in favour of and against use of warnings were presented. 3,28,34,40 Opponents expressed concerns that use of warnings made inferences regarding students' limited coping abilities. Others cited drawbacks similar to those identified by educators, 28 including disrupting flow of discussion and hypersensitising recipients. Warnings could be primarily used as defence against anticipated complaints. Although warnings were a concession required by a minority, they were imposed on all students. Some suggested that enrolment to medical education implied readiness to encounter routine content and that the need for warnings was overstated, implying trauma was not a concern for this population.
More participants recognised a role for warnings, due to personal need or as reasonable concessions for others. Trauma-related content was recognised as causing adverse emotional reactions, impeding learning, which could be ameliorated by advance warnings. The role of warnings appeared to have been debated amongst some students, leading to occasional discordance and polarisation in views.
A key aspect of contention surrounding warnings has been whether they promote avoidance of distressing content or facilitate engagement, 5  Participants discussed responsibility for learning and managing individual sensitivities in education settings. Students, as developing professionals, recognised the realities of medical education and the requirement for self-awareness in managing their own needs. 42 Duty to acknowledge and take responsibility for this was noted. Elsewhere, evidence was noted of students' identifying these responsibilities as lying with educators, suggesting an external locus of control that may be at odds with readiness for self-directed learning. 43,44 Warnings could thereby contribute to disempowerment. Others highlighted that educators typically controlled session content, further contributing to the power differential between educator and student and placing a responsibility on educators to consider recipients' needs. 5 In undertaking a journey to professionalisation, students commenced in a novice role. Models of self-directed learning highlight that learners' progress through different stages of self-directedness and that selfdirectedness traits can be acquired, nurtured and developed. 45 Professional preparedness could be acquired through incremental, scaffolded approaches, where warnings accommodated individual needs. It has previously been positioned that future professional responsibilities should not be prioritised over students' own current needs 28 and self-care should be enshrined in professionalisation. 46,47 Participants who identified a personal need for warnings desired compassion and acknowledgement from educators and that classrooms be preserved as nurturing environments, as indicated by differential expectations in classrooms and clinical environments. Where practice conveyed educators' consideration towards students, this demonstrated acceptance of students' needs and circumstances, a suggestion also proposed by educators. 28 Professional identities are influenced by the culture of learning environments and processes of socialisation. [48][49][50] Caring attributes, expected in future professionals, should be upheld and role-modelled in these contexts. 49 Participants explained how educators' treatment of trauma-related content was indicative of organisational attitudes and professional norms. These experiences impacted sense of belonging and identity and relationships with peers and the organisation, factors noted as inherent to professional identity formation. 49 These hidden curriculum experiences echoed those of students from previously underrepresented backgrounds. 51 Intersections of the experiences of traditionally underrepresented students, including minority group experiences, power hierarchies and social inequalities, with trauma provide further impetus for consideration of trauma-informed approaches, 17 including content warnings, for increasingly diverse student cohorts.
Participants shared perspectives regarding personal growth arising from exposure to challenge and stress. An optimal degree of stress-'eustress'-can be motivational and performance enhancing. 52,53 Trauma, a related but distinct experience, characterised as more severe and with persisting adverse effects, 54 and positioned further along this stress spectrum, was also discussed. Stress and trauma are opportune areas to explore in medical education, an experience widely acknowledged as intensive and having potential to harm practitioner well-being. 20,21 Trauma-informed medical education advocates for integration of trauma-informed approaches in curricular development, delivery and learning environments. 17,19 This includes teaching about science underpinning trauma and its effects, acknowledgement of potential impacts of trauma-related content on students, and accommodation for this through use of content overviews and advance warnings, thereby promoting understanding amongst students and educators, 17 a recommendation shared by some participants. Greater evidence-based understanding of stress and trauma, as both affordances and hazards, may be achieved through use of warnings, promoting empathy and avoiding marginalisation of either supporters or opponents.
Considerable variation in views was noted regarding the primary intended purpose of warnings and the best way to provide these.  55 -impacts academic outcomes when employed by students. 56 Regulation by cognitive or arousal reappraisal, which aims to change the type of stress response, encourages individuals to reconceptualize stress as a coping tool. 57 This technique has been explored in both therapeutic contexts, 58 resulting in decreased PTSD symptom severity, 59 and academic contexts, 41 showing effectiveness in improving student outcomes. 60 Compared with other regulation strategies, specifically suppression, cognitive reappraisal was associated with lesser symptom severity in PTSD 58 and lower levels of academic burnout. 56

| LIMITATIONS
The study was conducted at a single UK graduate-entry medical school that may limit generalisability of findings to school-leaver entry populations. However, typical, younger school-leaver entrants may not have had sufficient experience to develop professional maturity and resilience, meaning issues of vicarious traumatisation are pertinent. Furthermore, diversity in graduate-entry populations, with students from traditionally underrepresented backgrounds, enhances representativeness, meaning that a diverse population was sampled.
Experiences of adversity and trauma may be more likely in this population, allowing these participants to provide richer, more nuanced insights.
Unlike previous studies regarding trigger warnings, we did not limit participation to individuals identifying as having no past history of trauma. 13 Arising from ethical considerations, we did not explicitly inquire about individual trauma histories but facilitated discussion of such experiences when volunteered by participants. Absence of trauma history categorisation of participants may be considered a limitation. However, noting suggested broader pedagogical functions or drawbacks, 10,28 and pervasiveness of trauma in medical education and practice, 17,21 we recognised the relevance of trigger warnings to all students.
Students with more severe traumatic histories may have been reluctant to participate, despite assurances in recruitment information regarding confidentiality and that past traumatic experiences would not be explicitly explored by the researcher. Further, both researchers have education leadership roles at the study setting and were aware these roles could lead to student reticence to participate or discuss experiences. However, we captured a variety of participant perspectives in relation to the research questions, including reflections on both trauma and resilience required of clinicians. Participants also shared experiences of trigger warnings in both therapeutic and educational contexts. These two points provided assurance of an adequate sample. Expanding the study to additional settings and increasing the sample size would capture further perspectives and enhance generalisability of findings.

ACKNOWLEDGMENTS
We wish to thank the students at WMS MB ChB who shared their time and ideas in participating in this study.
We also wish to thank the reviewers for their constructive comments and suggestions.

CONFLICTS OF INTEREST
We declare that we have no competing interests.

AUTHOR CONTRIBUTIONS
HN conceived and developed the idea for the project, interviewed participants and analysed data and identified themes. HN drafted the early versions of the manuscript and made subsequent critical revisions for important intellectual content. LR developed the idea for the project, analysed data and identified themes. LR reviewed the early versions of the manuscript and made substantial contributions to the content and direction of the manuscript. Both authors approve the final version and agree to be accountable for all aspects of the work including questions related to the accuracy or integrity of the work.

ETHICS STATEMENT
This study was reviewed and approved by University of Warwick Bio-