Impact statement
Beyond documenting stigma and unmet need, this study identifies three service processes that participants associated with the usability of formal mental healthcare. Interviews with transgender women in Ho Chi Minh City linked engagement to whether a service could be trusted without sacrificing privacy, whether clinicians kept the presenting mental health concern central rather than treating gender as a curiosity and whether assessment explained confidentiality, sensitive questions and possible consequences clearly enough for fuller disclosure. Mental health authorities, service managers, clinicians and trans-led organizations can act by making credentials, fees, confidentiality and referral steps verifiable; funding accountable community navigation pathways and treating ordinary respect, clinically relevant inquiry and plain-language assessment as core elements of care quality. These changes may improve care for other patients whose social position makes unfamiliar services risky to enter.
Background
Expanding mental health services does not ensure that people can recognize, approach or remain in care. In global mental health, treatment gaps reflect workforce and financing constraints, but also stigma, opaque referral routes, poor service legibility and encounters that make disclosure risky (Patel et al., Reference Patel, Saxena, Lund, Thornicroft, Baingana, Bolton, Chisholm, Collins, Cooper, Eaton, Herrman, Herzallah, Huang, Jordans, Kleinman, Medina-Mora, Morgan, Niaz, Omigbodun, Prince, Rahman, Saraceno, Sarkar, De Silva, Singh, Stein, Sunkel and Unutzer2018). Levesque et al.’s access framework locates access at the interface between service characteristics and people’s abilities to perceive, seek, reach, pay for and engage with care, while candidacy describes eligibility as negotiated rather than simply possessed (Dixon-Woods et al., Reference Dixon-Woods, Cavers, Agarwal, Annandale, Arthur, Harvey, Hsu, Katbamna, Olsen, Smith, Riley and Sutton2006; Levesque et al., Reference Levesque, Harris and Russell2013). Both perspectives direct attention to the threshold between a service being present and being usable.
Transgender and gender-diverse populations experience disproportionate depression, psychological distress, self-harm and suicidality, patterned by structural stigma, violence, administrative exclusion and material insecurity rather than by gender diversity itself (White Hughto et al., Reference White Hughto, Reisner and Pachankis2015; Reisner et al., Reference Reisner, Poteat, Keatley, Cabral, Mothopeng, Dunham, Holland, Max and Baral2016). Gender minority stress and structural-stigma perspectives connect such distal conditions to anticipated rejection, vigilance and concealment; access and candidacy frameworks show how they are negotiated at service thresholds (Dixon-Woods et al., Reference Dixon-Woods, Cavers, Agarwal, Annandale, Arthur, Harvey, Hsu, Katbamna, Olsen, Smith, Riley and Sutton2006; Levesque et al., Reference Levesque, Harris and Russell2013; White Hughto et al., Reference White Hughto, Reisner and Pachankis2015). Together, these perspectives locate entry and disclosure in person–system interaction rather than individual readiness. Healthcare can reproduce exclusion through misgendering, invasive or irrelevant questioning, poor provider knowledge and uncertainty about competent care (Cicero et al., Reference Cicero, Reisner, Silva, Merwin and Humphreys2019; Rees et al., Reference Rees, Crowe and Harris2021; Panchal et al., Reference Panchal, Piper, Whitmore and Davies2022; Holt et al., Reference Holt, Hope, Mocarski and Woodruff2023). Evidence remains thinner on how an unfamiliar mental health service becomes credible enough to try and safe enough for meaningful disclosure.
The Vietnamese context combines policy movement with institutional gaps. Article 37 of the 2015 Civil Code provides for civil-status change after gender reassignment, but the implementing Gender Affirmation Law remained pending as of 2026, leaving no routine legal recognition pathway (National Assembly of the Socialist Republic of Vietnam, 2015; United Nations Population Fund, 2026). Studies document limited provider preparedness, discrimination, violence and restricted psychosocial support (Do and Nguyen, Reference Do and Nguyen2020; Nguyen et al., Reference Nguyen, Cu, Nguyen, Nguyen and Cu2025). A 2026 survey of 347 transgender and nonbinary adults found 62.5% with high or very high psychological distress; discrimination and alienation-related internalized transphobia were associated with greater distress, while resilience was protective (Nguyen et al., Reference Nguyen, Tran, Tran, Nguyen and Thai2026). Community-led and HIV-linked primary-care models offer transgender-competent services, including some mental health screening and counseling (Doan et al., Reference Doan, Vu, Nguyen, Green, Phan, Janamnuaysook, Vu, Le, Do, Tran, Ngo, Basu, Tran and Humeau2022), but do not create a legible citywide pathway outside HIV networks. Professional care remains concentrated in psychiatric and urban outpatient services, with a small, uneven psychosocial workforce and substantial out-of-pocket payment (Le et al., Reference Le, Hahn, Tran, Mavituna and Ta2024).
These gaps are historically produced. After Đổi Mới, ‘social evils’ and HIV discourses linked gender and sexual diversity to pathology, sex work and disorder. Donor-supported HIV programs gave community groups space to build health services. Activists translated health information, trained providers, organized safer services and sought legal recognition through negotiated service delivery. This expanded support, but project funding, registration limits, and incomplete legal recognition constrained continuity (Mai, Reference Mai2025). Community organizations alone thus cannot repair the conditions that make formal care hard to verify.
When formal services are unavailable, distrusted, or unaffordable, people may turn to relatives, friends, transgender peers, community organizations, online groups, religious actors or traditional remedies (Nguyen et al., Reference Nguyen, Cu, Nguyen, Nguyen and Cu2025). These sources can provide recognition, practical aid and referral knowledge, but access depends on social and material resources and cannot guarantee confidential, sustained clinical care. This analysis treats community mediation as a route toward or away from formal mental healthcare, not an adequate substitute for an accountable system.
Population counts are unreliable because transgender women have often been subsumed within broader surveillance categories. A community-informed needs assessment estimated 2,000–3,000 transgender women in Ho Chi Minh City; later respondent-driven studies recruited 205 and 456 without providing a census (APCOM, 2014; Colby et al., Reference Colby, Nguyen, Le, Toan, Thien, Huyen, TDA, Friedman and Stall2016; Vi et al., Reference Vi, Long, Hong, Anh, Ngoc, Tam, Lai, Hanh, Phuong, Trang and Dung2020). The city concentrates psychiatric, private counseling, HIV and community-based services, making it an analytically important setting where care remains fragmented, unevenly affordable and difficult to navigate (Trang et al., Reference Trang, Ly, Lam, Brown, To, Sullivan, Worthman, Giang and Jovanovic2021; Le et al., Reference Le, Hahn, Tran, Mavituna and Ta2024).
Across Southeast Asia, these forces combine differently rather than forming a linear hierarchy of progress. Thailand’s public visibility and specialized private sector coexist with cost and healthcare discrimination; Indonesian peer-led organizations navigate religious and political hostility and uneven public provision; and Philippine mental-health reforms coexist with fragmented, substantially self-financed transgender care. Vietnam’s comparative strength lies in community-led and HIV-linked services, not a legible general mental-health pathway (Fisk and Byrne, Reference Fisk and Byrne2020; Asia Pacific Transgender Network and Youth LEAD, 2023; Alibudbud, Reference Alibudbud2024). Across these settings, trans-led groups build peer referral, provider training, service directories and policy advocacy; what they can sustain depends on public authority, markets, donor priorities social norms and civic space. Community infrastructure can make services navigable but cannot substitute for public coverage, accountability or legal protection.
Against this background, the study asked: what conditions did transgender women with depressive symptoms associate with approaching, remaining in and disclosing within formal mental healthcare in Ho Chi Minh City? The question was open rather than organized around the eventual themes. The analysis sought plausible service mechanisms that could inform equitable mental healthcare in Vietnam and other settings where formal care exists but remains difficult to claim.
Methods
Design and analytic orientation
This article reports a focused analysis from a broader two-phase focused ethnography of depressive distress among transgender women in Ho Chi Minh City. The broader study included 30 in-person interviews and observation before strict COVID-19 restrictions, followed by digital ethnography and 10 remote interviews with a separate phase-specific sample during Directive 16 in 2021. The analysis centers on all 40 interview transcripts, supplemented by observational and digital-ethnographic materials. It covers formal-service engagement from pre-entry through consultation and assessment, leaving informal care practices, embodiment and community crisis response outside its analytic scope.
An interpretivist framework suited the aim of understanding how participants made sense of actual and anticipated encounters, not adjudicating one objective account of each clinic. A contextualist orientation connected accounts to service organization, cost, stigma, documentation and community knowledge without treating them as direct reports of external facts. Reflexive thematic analysis generated patterned meaning while recognizing the researcher’s interpretive role (Braun and Clarke, Reference Braun and Clarke2006, Reference Braun and Clarke2021). Availability means that formal care exists; accessibility concerns person-service fit; entry is movement from locating a service to first contact; and engagement is participation before and during care. Usability denotes the practical ability to identify, approach, remain in and communicate within care. Legitimacy denotes professional credibility and accountability. Disclosure-supportive assessment describes conditions enabling fuller accounts without moralizing silence as dishonesty.
Reporting followed the Consolidated Criteria for Reporting Qualitative Research (COREQ) (Tong et al., Reference Tong, Sainsbury and Craig2007). A completed COREQ checklist is provided as Supplementary Material 1.
Setting, participants and recruitment
Participants were eligible if they self-identified as transgender women, were aged 18 years or older, had lived in Ho Chi Minh City for at least six months, scored 16 or higher on the Vietnamese 20-item Center for Epidemiologic Studies Depression Scale (CES-D) and reported avoiding, declining or discontinuing professional mental healthcare for current distress. The CES-D was used solely for screening and sampling; it did not establish a diagnosis or imply that psychiatric treatment was necessarily indicated (Radloff, Reference Radloff1977). In a validation study of 400 Vietnamese HIV outpatients in Ho Chi Minh City, the scale demonstrated good internal consistency (Cronbach’s alpha 0.81) and an area under the receiver operating characteristic curve of 0.88. At the cutoff of 16, sensitivity was 79.8% and specificity was 83.0% (Thai et al., Reference Thai, Jones, Harris and Heard2016). The implications of validation in an HIV outpatient sample rather than specifically among transgender women are addressed in the Limitations. The Vietnamese CES-D instrument used to determine eligibility is provided as Supplementary Material 2.
Recruitment used purposive, criterion-based, snowball and peer-referral approaches in collaboration with G3VN, a transgender women-led organization. Of 65 people screened, 21 scored below the CES-D threshold and four eligible people declined because of time constraints, leaving 40 participants. Sampling sought variation in age, duration of residence, education, livelihood, housing, transition-related intervention, religion and service-contact history. Recruitment routes reflected access to a marginalized population, not membership in organized community groups.
Fifteen participants reported some lifetime contact with formal mental healthcare and 25 had not entered formal services. Reported contacts included primary care for insomnia (n = 8), psychiatric consultation in general medical or primary-care settings (n = 6), mental health hotlines (n = 5) and private counseling (n = 5); categories overlap. Accounts without direct contact informed analysis of searching, attempted entry, anticipated risk and community verification, but not claims about events inside consultations. Findings on provider conduct and assessment were grounded primarily in the 15 direct-contact accounts. Table 1 reports characteristics relevant to engagement. Sexual orientation was not collected as a separate sociodemographic variable.
Selected participant characteristics relevant to formal service engagement (N = 40)
Table 1. Long description
Among 40 participants, 20 (50.0%) were aged 18–30 years, 10 (25.0%) were 31–40, 7 (17.5%) were 41–50 and 3 (7.5%) were 51 years or older. Education was: 6 (15.0%) with no formal schooling, 7 (17.5%) primary school, 15 (37.5%) lower secondary, 7 (17.5%) upper secondary and 5 (12.5%) tertiary education. Employment was: 5 (12.5%) contracted, 33 (82.5%) informal or non-contracted and 2 (5.0%) not currently employed. Housing was: 30 (75.0%) in rented accommodation, 6 (15.0%) living with family and 4 (10.0%) experiencing housing instability or homelessness. Fifteen participants (37.5%) reported any lifetime formal mental healthcare contact and 25 (62.5%) reported none. Informal or non-contracted work included sex work, nightclub promotion, funeral or event performance, lottery-cabaret performance, nail or makeup work and street vending; participants could report more than one activity.
* Among participants classified as being in informal or non-contracted work, reported income-generating activities included sex work, nightclub promotion, funeral or event performance, lottery-cabaret performance, nail or makeup work and street vending. Participants could report more than one activity.
Data collection and analysis
The semi-structured topic guide covered recognition and naming of distress, knowledge of professional mental healthcare, attempts to seek help, direct service experiences, provider interaction, assessment, reasons for refusing or discontinuing care and everyday responses to distress. It was refined with the five-member lived-experience advisory board and did not prespecify the three themes. A four-person research team, comprising one physician and three master’s-trained researchers, conducted interviews after preparation in qualitative interviewing and trans-sensitive research. No non-participants were present. Thirty interviews took place in private participant-selected or community-organization locations, and 10 were conducted remotely during lockdown. Interviews lasted 40 to 120 min, were audio-recorded with consent and transcribed verbatim in Vietnamese; observational field notes and digital-ethnographic materials contextualized the interview analysis. Participants received VND 200,000 for their time.
Earlier in the broader study, the author and two other researchers independently coded the same purposively selected 10 Vietnamese-language transcripts. They compared readings to surface interpretive differences and jointly revised a preliminary coding structure; disagreements were discussed rather than adjudicated by vote or a reliability statistic. The structure served as a sensitizing resource, not a fixed codebook. For this article, the author alone defined the focused question, coded all 40 transcripts in Vietnamese using ATLAS.ti 8.4.5 and developed, named and interpreted the final themes. The other two preliminary coders did not participate in article-specific theme development or interpretation.
The service-engagement domain was selected after familiarization because participants repeatedly connected non-use to service organization and conduct, not only distress management. The focused question was defined before the themes. Across all 40 transcripts, the author indexed passages concerning discovery, evaluation, approach, use or avoidance of professional services, including accounts that complicated the interpretation. Codes were revised, combined, discarded or added as required by the full dataset. Subsequent coding compared service verification, referral, cost and procedural fit, naming and address, the relevance of gender-related inquiry, confidentiality, assessment language and withholding. Analytic memos traced relationships among codes, and whole transcripts were revisited to preserve context.
Themes were developed recursively across extracts, whole transcripts, memos and the study question. Contrasting accounts in which similar conditions inhibited or enabled engagement refined boundaries. Provisional themes were examined against direct-contact, pre-entry and discordant cases. Final themes were assessed for coherence, distinction and capacity to account for meaningful variation. This judged interpretive adequacy, not saturation. Broader-study research-team and advisory-board discussions supplied alternative readings considered during the focused analysis; neither group co-developed this article’s final themes.
Reflexivity, translation and ethics
The author is a cisgender Vietnamese man with long-standing professional relationships with transgender community organizations; the four-person interview team comprised two cisgender men and two cisgender women. Familiarity could support trust and contextual understanding but also make expected narratives seem self-evident. Interviewers therefore used open prompts, avoided assuming that gender explained every experience, kept reflexive field notes and debriefed difficult interviews. During analysis, the author returned to full transcripts and relevant observational and digital-ethnographic materials, compared divergent accounts and used the lived-experience advisory board as a critical interlocutor rather than treating community consultation as confirmation. The four interviewers, including the author, had no prior personal relationships with any participants. Before providing consent, participants were informed that the study examined depressive distress and help-seeking among transgender women and of the interviewers’ roles in the research.
Analysis was completed in Vietnamese. After themes had been developed, the bilingual author translated selected quotations using conceptual and interactional equivalence rather than word-for-word substitution. No professional translation service was used. The study’s translation protocol required two bilingual team members to check each English extract against the Vietnamese transcript, with attention to forms of address, tone and colloquial expressions of distress; differences were resolved through discussion. Quotations received only minor readability edits. Participants later received a summary of the broader findings, and none requested corrections.
Separate ethical approvals were granted by the London School of Hygiene and Tropical Medicine Research Ethics Committee (25561) and the Hanoi University of Public Health Institutional Ethical Review Board (130/2021/YTCC-HD3). All participants gave written or audio-recorded verbal informed consent according to preference. Study information was read aloud when a participant could not read. Referral information for available mental health support was offered, and a safety protocol was available for acute distress.
Results
Participants’ accounts generated three connected, but not compulsory or uniformly ordered, service processes: establishing a credible route into care, maintaining clinical relevance in the consultation and making assessment safe and intelligible enough for fuller disclosure. Each theme is organized from conditions that inhibited engagement to those that enabled it. The processes are interpretive associations, not independently measured service attributes or causal effects.
From service visibility to a credible route into care
Finding a clinic name or telephone number did not make it approachable. Participants described sparse information about credentials, fees, confidentiality, appointment steps and what would happen on arrival. When these questions remained unanswered, searching could stop short of contact.
I found one clinic and saved the phone number, but I didn’t call. I asked a few chị em first. One said she didn’t know that place, another said it looked expensive, and someone else said to be careful because they ask many strange things and it doesn’t seem useful. After that I just left it there. I thought maybe later, but I never went. (38-year-old, freelance performer and sex worker)
Participants distinguished public visibility from trustworthy visibility. Prominent psychiatric signage could expose the reason for attending, yet an unobtrusive clinic with little verifiable information could appear informal or unsafe. Privacy and professional accountability therefore had to be established together.
If a place has ‘Psychiatry’ written big outside, we already feel exposed before even opening the door. But if it is too hidden, then we ask each other, is it legal, is the doctor real, what happens if something goes wrong? Quiet is not enough. We still need to know the place is legitimate. (29-year-old, freelance singer and sex worker)
Social fit formed another part of this judgment. Highly polished private clinics signaled cost and unfamiliar class expectations. For women living with informal work, debt, unstable housing or limited literacy, entry raised the prospect of being unable to pay, navigate forms, speak in the expected way or appear to belong.
Those clinics look like places where you already need to know how to speak, how to behave, how to pay. When I look at them, I feel like I am going into someone else’s world. Before the doctor says anything, I already feel I don’t belong there. (44-year-old, freelance funeral performer)
Transgender peers and community organizations sometimes converted this uncertain visibility into a credible route. Participants used chị em, literally “sisters,” for other transgender women whose first-hand knowledge could connect professional credibility with discretion, realistic cost information and accompaniment. Their role was not merely encouragement; it was practical verification where formal information was incomplete.
I looked online first. There were names, photos and addresses, but I still didn’t know if I could really go there. I didn’t know if the doctor was properly licensed, if they would charge me for things I didn’t understand, or if people would stare when I arrived. I went only because one older sister said she had been there before and offered to go with me the first time. (34-year-old, lottery-cabaret performer)
A usable entry pathway therefore joined verifiable legitimacy, privacy, affordability and social claimability. Community verification operated as access infrastructure when the formal system did not communicate these conditions reliably.
Community advice did not replace trust in professional care. It created a chain of accountability in which peer experience made formal claims about credentials, privacy and cost testable. Without that chain, public information could remain advertising rather than evidence; so, adding a service to a directory might improve visibility without improving entry.
From trans-related scrutiny to a clinically relevant consultation
Open hostility was not the only way a consultation failed. Some providers presented themselves as knowledgeable about transgender health but expressed that knowledge through curiosity about bodies, surgery, hormones or sex before addressing the reason for attendance.
The doctor said she understood transgender people, so at first I felt more comfortable. But then she asked a lot about my body, hormones, surgery and sex. I kept waiting for her to ask about my sleep and why I was scared all the time. I came because of that, but the conversation kept going to other things. (36-year-old, nail technician and makeup artist)
Participants did not want gender ignored. They wanted its clinical relevance established. They judged competence through ordinary conduct: being greeted without spectacle, being asked what name or form of address to use and having sleep, panic, sadness or daily functioning treated as the presenting concern. Trans-specific knowledge became useful when it supported that work rather than displacing it.
At the better places, they don’t start by making me explain what kind of person I am. They greet me, ask why I came, ask about my sleep, ask what has been hardest. If they don’t know how to call me, they ask, or they use neutral words first. Then I feel I am there as a patient, not as something strange for them to look at. (33-year-old, freelance performer and sex worker)
The opening minutes carried disproportionate weight because they indicated whether the clinician could hold gender and distress in proportion. A steady greeting and initial focus on the complaint did not require elaborate affirming language; they showed that the participant had entered a clinical relationship rather than an encounter organized around difference.
The first few minutes tell me a lot. If the doctor starts normally, like asking what name to use and why I came, I can breathe a bit. But if they start by looking at my body or asking what I am, I already know I won’t say much. (21-year-old, freelance makeup artist and singer)
Participants’ standard of professionalism was thus specific rather than generic: respectful address, attention to the stated mental health problem and gender-related questions only when clinically relevant or invited. A provider could possess trans-related facts and still appear incompetent if those facts pulled the encounter away from care.
This theme also contained a diagnostic tension. Gendered stigma, transition-related healthcare, family conflict and violence could be central to distress, but treating gender as automatically explanatory narrowed the assessment before the participant’s problem had been heard. Clinical relevance was therefore negotiated through timing, explanation and responsiveness, not decided by avoiding or routinely prioritizing gender.
From high-stakes questioning to disclosure-supportive assessment
Assessment was not experienced as a neutral technical step. Participants entered with stories circulating among chị em about impatient clinicians, intrusive questions, family disclosure and consequences that were difficult to predict. They therefore evaluated not only what to say but how much could be said safely.
Before anyone even goes, chị em have already told you what to expect. They say the questions are strange, the doctors get impatient, and if you answer carelessly it can become a whole problem. One sister did not even know how her family found out she had gone to the clinic. So when it was my turn, I wasn’t thinking, now I will tell the truth. I was thinking, how do I answer without making things worse? (44-year-old, freelance performer and sex worker)
Fixed questions about suicidality, self-harm, worthlessness or whether life felt worth living were especially difficult when their purpose and consequences were not explained. Participants distinguished wanting unbearable pressure to stop, wanting to disappear and wanting to die, but could not always fit these experiences into the response expected by the clinician.
The doctor asked whether I felt life was not worth living and whether I wanted to die. I remember thinking those are not the same thing, but I didn’t know how they wanted me to understand it. Sometimes I wanted the pressure to stop. Sometimes I wanted to disappear for a while. Is that the same as wanting to die? I asked what would happen if I answered yes, and nobody really explained. After that, I chose the safer answer, not necessarily the truer one. (28-year-old, unemployed)
Ambiguity about language combined with uncertainty about confidentiality, family involvement and escalation. Under these conditions, a less-accurate answer could be a rational form of self-protection. The clinical problem was not an absence of distress but an assessment process that failed to establish shared meaning and predictable boundaries.
Attempts to clarify could worsen the problem when treated as ignorance, delay or resistance. Participants then shifted from describing distress to ending the encounter with minimal risk.
When I said I didn’t fully understand one question, the doctor tapped the paper and said, ‘Just answer first. We don’t have all day.’ After that I stopped thinking about what I really felt. I just thought, fine, tell him what will let this finish quickly. (22-year-old, nightclub promoter)
Fewer accounts showed the enabling contrast. Clinicians made assessment workable by explaining why sensitive questions were asked, restating unfamiliar language, protecting privacy and allowing uncertainty rather than demanding an immediate categorical answer.
There was one counselor who did it differently. She said, ‘Some of these questions sound blunt, but I ask them because I need to understand how heavy things have become for you, not because I’m judging you.’ When I didn’t understand a phrase, she said it another way, like how bad has it felt, what have your nights been like, have you been afraid you might do something. That was the first time I answered properly. (24-year-old, lottery singer and sex worker)
Participants described fuller disclosure when assessment functioned as a clinical conversation rather than a demand for decontextualized answers. Figure 1 integrates the three processes and their associated engagement states. Table 2 links each process to responsible actors, implementation responses and supporting participant evidence.
Service processes associated with mental healthcare usability. Participants described cross-cutting structural and service conditions alongside three connected service moments and paired engagement states. The processes were not compulsory or uniformly ordered and could recur. Lines show analytic associations in participants’ accounts, not estimated causal effects.
Figure 1. Long description
A top box lists cross-cutting conditions described by participants: stigma, weak service legibility, financial insecurity and uncertain confidentiality. Plain connector lines lead to three service-moment boxes. Before entry, participants associated verifiable legitimacy, privacy, cost, procedures, social fit and community verification with a credible route. During consultation, they associated respectful, clinically relevant conduct with continuation. During assessment, they associated clear language, explained confidentiality and consequences, and respectful clarification with fuller disclosure. The three boxes are paired, respectively, with accounts of entry or non-entry, continuation or discontinuation, and fuller or minimal disclosure. The connectors denote analytic associations, not causal effects.
Evidence-linked service and implementation responses
Table 2. Long description
The table links eight service findings to responsible actors, implementation responses and participant evidence. Visibility alone did not establish trust, so services should make accreditation, fees, confidentiality, procedures and complaint routes clear. Privacy and legitimacy could conflict, requiring discretion without making services unverifiable. Community verification was associated with entry; funded referral and accompaniment pathways were recommended as implementation responses. Cost, literacy and social fit shaped whether care felt claimable. Clinical competence required respectful, relevant conduct that kept the mental health concern central. Participants associated clearer explanations of confidentiality, possible escalation and sensitive questions with fuller disclosure. Standard clinical categories did not always fit lived distress, especially around suicidality. Plain language and respectful clarification were therefore important for valid assessment.
Across the themes, enabling conditions did not erase material constraint or guarantee continued treatment. Participants described them as reducing avoidable uncertainty at successive points of contact. A credible route was associated with considering attendance, a relevant consultation with continued participation and explained assessment with fuller accounts. These service moments could recur when participants changed providers, returned after discontinuation or encountered a new assessment task.
Discussion
Principal contribution
Participants described formal mental healthcare as more usable when three linked conditions were present: a credible route into care, a consultation that kept the mental health concern central and an assessment whose meaning and consequences were clear enough for fuller disclosure. Vietnamese studies document distress, discrimination and constrained psychosocial support; this analysis specifies plausible service mechanisms through which those inequities were experienced at entry, continuation and disclosure (Nguyen et al., Reference Nguyen, Cu, Nguyen, Nguyen and Cu2025, Reference Nguyen, Tran, Tran, Nguyen and Thai2026). Here, mechanism denotes a recurrent interpretive connection in participants’ accounts, not an estimated causal effect.
The themes were not predetermined. They emerged by comparing direct-contact, attempted entry and non-entry accounts. Participants described visibility as compromising privacy or improving access, trans-related knowledge as reassuring or redirecting and sensitive assessment as identifying risk or accompanying strategically safer answers. These conditional tensions, rather than a universal sequence, are the article’s analytic contributions.
Credible entry and community-mediated access
Access frameworks describe approachability and care-seeking ability; participants’ accounts show these capacities were negotiated collectively rather than individually (Dixon-Woods et al., Reference Dixon-Woods, Cavers, Agarwal, Annandale, Arthur, Harvey, Hsu, Katbamna, Olsen, Smith, Riley and Sutton2006; Levesque et al., Reference Levesque, Harris and Russell2013). Peers supplied missing evidence about credentials, cost, discretion, reception and likely treatment. In Vietnam, this intermediary role extends community organizations’ work of translating information and building services within constrained civic and donor environments (Mai, Reference Mai2025). This grounds circuitous help-seeking in fragmented mental healthcare, where much transgender-competent provision remains linked to HIV programs (Doan et al., Reference Doan, Vu, Nguyen, Green, Phan, Janamnuaysook, Vu, Le, Do, Tran, Ngo, Basu, Tran and Humeau2022; Holt et al., Reference Holt, Hope, Mocarski and Woodruff2023).
Trans-led organizations should be resourced as co-design and navigation partners without carrying unpaid clinical risk. Authorities can publish accredited referral options; services can make fees, confidentiality, appointment steps and complaint routes visible; and funded community navigators can test those promises in practice (Asia Pacific Transgender Network and Youth LEAD, 2023).
Clinically relevant transgender-inclusive care
Participants’ accounts sharpen the meaning of transgender-inclusive mental healthcare. They did not ask clinicians to avoid gender, but to establish why a gender-related question mattered and to keep the presenting concern in view. This accords with reviews emphasizing respectful communication, individualized care and avoidance of assumptions and with research on gender-related medical misattribution and invasive questioning (Panchal et al., Reference Panchal, Piper, Whitmore and Davies2022; Wall et al., Reference Wall, Patev and Benotsch2023). The finding adds a practical criterion: trans-competence is demonstrated through disciplined clinical relevance, not the volume of trans-specific questions.
Training should rehearse observable conduct rather than stop at attitude change: using the patient’s name and form of address, beginning with the stated concern, explaining identity-related questions and redirecting curiosity that displaces care. Professional associations, training institutions and supervisors can incorporate these behaviors into teaching, observed practice and quality review. The same discipline can benefit patients whose social position is too readily treated as the explanation for every symptom.
Disclosure as a condition of clinical validity
Assessment linked equity to clinical validity. Participants sometimes gave answers that are least likely to trigger exposure or loss of control, not answers that best described their distress. Misunderstanding or withholding should not automatically be read as low literacy, resistance or lack of need. It can indicate that the service has not explained confidentiality, the purpose of the question or what escalation might follow.
Qualitative evidence from mental health service users similarly favors assessment that is collaborative, humane and protective of agency (Patel et al., Reference Patel, Kennedy, Carr, Gillard, Harris and Sweeney2025). Clinics can operationalize this by stating confidentiality and its limits before sensitive questions, explaining why suicide and self-harm questions are asked, inviting narrative clarification and describing possible next steps before demanding a categorical response. These practices do not weaken risk assessment; instead, they may improve the conditions under which clinically useful information is offered.
Regional interpretation
The three mechanisms are transferable as questions, not claims of identical operation. Cost, uncertain pathways, provider inexperience and discrimination recur in Indonesia, Thailand and the Philippines, but patients navigate different institutional configurations (Fisk and Byrne, Reference Fisk and Byrne2020; Asia Pacific Transgender Network and Youth LEAD, 2023). Thailand’s visible private sector may improve recognition while preserving cost barriers; Indonesian peer networks may carry more navigation work amid uneven public provision and political-religious hostility; and Philippine reforms coexist with fragmented, substantially self-financed transgender care. Vietnam’s distinctive configuration is a strong community- and HIV-linked provision alongside an unclear general mental-health route. Comparison therefore directs attention to how public authority, market provision, legal protection and community infrastructure combine at each service threshold.
Policy and implementation implications
Table 2 distributes action because training alone cannot resolve these barriers. Authorities can publish information and confidentiality standards and fund accountable navigation; clinics can audit first contact, fees, naming, privacy and complaints, professional bodies can assess clinically relevant communication and resourced trans-led organizations can co-design and monitor pathways. Vietnam’s integrated community-led clinics offer a feasible starting point, while regional evidence shows that cost, pathway complexity and discrimination can persist despite nominal provision (Doan et al., Reference Doan, Vu, Nguyen, Green, Phan, Janamnuaysook, Vu, Le, Do, Tran, Ngo, Basu, Tran and Humeau2022; Alibudbud, Reference Alibudbud2024; Reback et al., Reference Reback, Lin, Larkins, Vi, Diep, Lai and Dung2024).
Implementation should be evaluated at each identified breakdown. Indicators include complete public information, community referrals reaching first appointments, time and cost to contact, patient-reported relevance and respect, understanding of confidentiality limits and continuation after assessment. Measures should be disaggregated by employment and housing insecurity and interpreted with qualitative feedback because attendance alone cannot show that care was safe or useful. Community partners should share authority and resources for indicator design, navigation and monitoring.
Legal gender recognition cannot substitute for service improvement, but it shapes names, documents and privacy. Progress on Vietnam’s Gender Affirmation Law and anti-discrimination protections could reduce routine exposure if linked to mental health-service standards (UNDP and UNAIDS, 2021; United Nations Population Fund, 2026). Transparent navigation, respectful address, relevant questioning and explained assessment may also assist people navigating poverty, disability, migration or stigmatized mental health conditions.
Transferability and limitations
This single-city study purposively selected transgender women with CES-D scores of at least 16 who had avoided, declined or discontinued professional care. It does not represent transgender women across Vietnam, gender-diverse people, rural settings or people satisfied with formal services. Community and peer recruitment may have underrepresented those disconnected from social networks. Sexual orientation was not collected separately, limiting examination of intersecting identities. The CES-D was validated among Vietnamese HIV outpatients rather than transgender women and used only as a screening heuristic. Transferability rests on mechanisms and contextual detail, not statistical prevalence.
Although observational and digital-ethnographic materials supplemented interviews, the study did not observe consultations or interview clinicians. It therefore cannot establish causal relationships among service conditions, engagement and disclosure, determine provider intent or verify individual encounters. Only 15 participants reported direct formal service contact; so, consultation and assessment are narrower than pre-entry claims. Translation may have reduced nuance despite bilingual checking. Interviews occurred in 2021; services, law and referral pathways have since changed. The findings identify interpretive associations and plausible mechanisms, not causal effects or a current provider’s performance assessment.
Conclusion
For participants in this study, accounts of approaching, continuing and disclosing in formal mental healthcare were associated with a credible route, a clinically relevant consultation and explained assessment. Service expansion alone is unlikely to close global mental health inequities. Verifiable pathways, accountable community partnerships, disciplined provider conduct and disclosure-supportive assessment are concrete components of inclusive care and may assist other patients who must manage exposure, misrecognition or loss of control when seeking help.
Open peer review
To view the open peer review materials for this article, please visit http://doi.org/10.1017/gmh.2026.10331.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/gmh.2026.10331.
Acknowledgments
The author thanks Tom Shakespeare, Shaffa Hameed and Ritsuko Kakuma for academic guidance; G3VN and the lived-experience advisory board for contributions to study design, recruitment, safety and interpretation; the research team for interviewing, preliminary coding, broader-study interpretive discussion and bilingual quotation checks; and, above all, the participants.
AI statement
No AI-generated or AI-assisted text, analysis, interpretation, references or other substantive content was incorporated into this manuscript. The author independently developed, wrote and revised the manuscript and takes full responsibility for its content. The graphical abstract and Figure 1 were created manually by the author in Canva using standard, non-generative design tools; no AI-generated images or visual elements were included in the final version.
Author contribution
Conceptualization, Methodology, Investigation, Formal analysis, Funding acquisition, Project administration, Writing – original draft, and Writing – review and editing. The author approved the submitted version and accepts accountability for the work.
Competing interests
The author declares no conflicts of interest.
Data availability statement
De-identified supporting data may be available from the corresponding author on reasonable request, subject to ethical approval, participant confidentiality and institutional permission. Full transcripts, observational field notes and digital-ethnographic materials are not publicly available because their sensitive, contextual detail could permit re-identification or compromise participant safety.
Ethics and approvals
The study received separate ethics approvals from the London School of Hygiene and Tropical Medicine Research Ethics Committee (25561) and the Hanoi University of Public Health Institutional Ethical Review Board (130/2021/YTCC-HD3). All participants provided written or audio-recorded verbal informed consent.
Funding statement
This work received partial research support from the 2021 London School of Hygiene & Tropical Medicine Johnson & Johnson Global Mental Health Award and the 2021 King’s College London Global Mental Health Award. The funders had no role in study design, data collection, analysis, interpretation, manuscript preparation or the decision to submit.
Inclusivity statement
A five-member lived-experience advisory board of transgender women advised on recruitment, terminology, safety procedures, interpretation and the framing of implications. G3VN, a transgender women-led organization, supported community entry and recruitment. Participants chose interview locations and consent format and received a summary of the broader findings. Community input challenged deficit framings and shifted attention from individual reluctance toward the conditions under which services become usable. The advisory board did not take final authorship or decision-making responsibility, consistent with members’ preference not to be publicly identified.
Open access
