LGBT+ partner bereavement and appraisal of the Acceptance-Disclosure Model of LGBT+ bereavement: A qualitative interview study

Background: Support from social networks is vital after the death of a partner. Lesbian, gay, bisexual and/or transgender (LGBT+) people can face disenfranchisement and isolation in bereavement. The Acceptance-Disclosure Model (of LGBT+ bereavement) posits that experiences are shaped by the extent to which individuals feel able to disclose their bereavement to others, and whether that loss is acknowledged appropriately. Aim: To explore LGBT+ specific experiences of partner bereavement; determine decision-making processes regarding disclosure of relationships/identities; and appraise the Acceptance-Disclosure Model using primary qualitative data. Design: Exploratory in-depth qualitative interview study positioned within a social constructivist paradigm. Data were analysed using inductive and deductive reflexive thematic analysis. Setting/participants: 21 LGBT+ people from across England bereaved of their civil partner/spouse. Results: Participants described LGBT+ specific stressors in bereavement: lack of recognition of their loss; inappropriate questioning; unwanted disclosure of gender history; and fears of discrimination when accessing support. Disclosure of LGBT+ identities varied across social networks. Some participants described hiding their identities and bereavement to preserve relationships, and challenging intersections between LGBT+ identities and other aspects of culture or self. These findings provide primary evidence to support the Acceptance-Disclosure Model. Conclusions: LGBT+ people face additional stressors in bereavement. Not all LGBT+ people want to talk directly about their relationships/identities. Sensitive exploration of support needs, aligned with preferences around disclosure of identities, can help foster trust. Five recommendations for inclusive practice are presented. Further research should consider whether the Acceptance-Disclosure Model has utility to explain bereavement experiences for other isolated or disenfranchised groups.


Background
The impact of bereavement is felt most heavily by those closest to the deceased, 1,2 with primary caregivers at risk of poor bereavement outcomes. 3 However, bereaved partners infrequently access health care 4 despite increased odds of worsening or new physical illness 5 and mortality. 6 Access to support from social networks is vital to avoid adverse bereavement outcomes including psychosocial morbidity and prolonged grief symptoms, 7 and to enable adaptation after the loss. 8 Lesbian, gay, bisexual and transgender (LGBT+) people constitute minority groups with specific healthcare needs, 9 including greater all-cause mortality 10 ; increased risk of common mental disorders, substance misuse and suicidality [11][12][13] ; more health risk behaviours 14,15 ; and therefore increased risk of life-limiting illnesses. [16][17][18] With higher rates of mental health conditions, LGBT+ people may have worse bereavement outcomes, particularly where compounded by increased isolation reducing essential social support. Indeed recent US research which mapped social networks for older LGBT adults found that those with the most restricted social networks had poorer mental health. 19 Similarly, in bereavement, LGBT adults who reported good social support, described more positive coping behaviours, 20 highlighting the importance of informal support in bereavement.
Despite protection under UK law (Equality Act (2010), Health and Social Care Act (2012)), experiences of discrimination for LGBT+ people persist, including at the end of life, 17 and into bereavement. 21,22 Sensitive communication is therefore vital to build trusting relationships between LGBT+ people and their clinicians. 23,24 Incorrect assumptions about relationships lead to disengagement and loss of trust, resulting in reluctance to access healthcare. 25,26 Lack of acknowledgement of the depth of the relationship, or experiences of discrimination around the bereavement period, can have devastating implications for LGBT+ bereaved partners. Such disenfranchisement reduces their access to vital social support afforded during grieving. 27 A recent systematic review describing the breadth of additional barriers and stressors for LGBT+ partners in bereavement 21 informed the development of the Acceptance-Disclosure Model of LGBT+ bereavement (see Figure 1). This posits that the experience of bereaved LGBT+ people is shaped by the extent to which individuals feel able to disclose the nature of their relationship to others, and the degree to which that relationship is accepted. The model describes four domains of bereavement based on the four dimensions; spoken, unspoken, accepted, not accepted.
The domain within the model that an individual occupies is theorised to influence their bereavement experience What this paper adds?
• • Additional barriers and stressors in bereavement can limit access to essential support for LGBT+ people.
• • LGBT+ people demonstrate agency in their decisions regarding disclosure of identities and relationships. The extent of disclosure often varies across their social networks. • • Challenges were described due to intersections between LGBT+ identities and other aspects of culture or identity.
Some individuals chose to hide their relationship and bereavement in order to retain important relationships and avoid disenfranchisement. • • Five recommendations for inclusive care of LGBT+ people facing bereavement derived from the data are presented.

Implications for practice, theory or policy
• • Direct questions about relationships and identities can feel confrontational for LGBT+ people due to fears and previous experiences of discrimination. • • Sensitive communication, and respecting autonomy regarding disclosure, can support person-centred care.
• • Our findings support the Acceptance-Disclosure Model to explain experiences of LGBT+ bereavement. This model could also inform assessment and support strategies for other potentially isolated or disenfranchised groups. Figure 1. Acceptance-disclosure model of LGBT+ bereavement experiences. 21 and their ability to access support. If for example an individual occupies the domain of 'overt acceptance', their relationship is known and accepted by those around them, and therefore they are recognised as facing bereavement and will be able to access support. In contrast an individual occupying the domain of 'invisibility', has not felt able to share the nature of their relationship with those around them, and their grief will also remain invisible. Therefore their access to the grieving role, and the support afforded by that, would be limited. The domain of 'unspoken acceptance' represents those individuals who, for whatever reason, prefer not to directly explain the nature of their relationship, however they would assume that those around them recognise the depth of the relationship (also called tacit acknowledgement 28 ). The risk for these individuals is that the relationship may be misunderstood and therefore they may not receive adequate support. Finally, those in the domain of 'overt exclusion' have felt able to share the nature of their relationship, however this has been met with rejection and exclusion from those around them, adding markedly to the trauma of their loss. Primary quantitative data provides preliminary support for this model, as loneliness, social support and caregiver burden are associated with the intensity of grief, and psychological symptoms are greater for bereaved same-gender partners. 29 The present study aimed to (1) investigate LGBT+ specific experiences of partner bereavement; (2) to examine decision-making processes in relation to disclosure of relationships and identities; and (3) appraise the Acceptance-Disclosure Model of LGBT+ bereavement using primary qualitative data.

Design
Our research team comprises mental health and palliative care clinicians, qualitative methodologists, social scientists, psychologists, health services researchers, researchers experienced in LGBT+ health research, and LGBT+ community members. This exploratory qualitative interview study is positioned within a social constructivist paradigm, which suggests that learning and knowledge develop from an individual's interactions with their culture and society. This qualitative study was nested in a larger population-based study (ACCESSCare-B) incorporating a national cross-sectional survey of bereaved samegender and different-gender partners 29 6-10 months post bereavement.

Participants and settings
Individuals who had registered the death of a civil partner or spouse in England and Wales were invited to participate in the ACCESSCare-B survey. 29 All survey participants were asked whether they would be willing to be contacted for a subsequent one-to-one qualitative interview. A comprehensive demographic variables section was included within the ACCESSCare-B survey. This enabled the research team to purposively sample individuals who had consented to be contacted for interview and had reported the death of a same-gender partner, by the following criteria (of both participant and decedent), to maximise the diversity of the sample and increase transferability of the findings: gender, age, sexual orientation, gender history and ethnicity.

Recruitment
Participants were contacted by the study researcher (LT), who explained the interviews in more detail. Participants were offered a choice of interview modality (in person or telephone) and location (their own home, a university building, another setting). All participants gave written informed consent prior to the interview. No participants were known to the study team prior to recruitment. No incentives or payments were provided to participants in this study, however all participants were given a leaflet on bereavement and a signposting resource for bereavement support services.

Interview data collection
All interviews were conducted by the study postdoctoral researcher (LT), a psychologist new to qualitative interviewing, but with training in person-centred counselling and experience as a helpline volunteer. They were supervised by the study lead (KB), a qualitative methodologist. The interview topic guide was devised by KB, LT and RH and refined through discussion with the study research team including LGBT+ community members (see Supplemental Materials for a copy of the interview topic guide). Each interview commenced with demographic questions about the participant and their deceased partner, to allow individuals to describe these important elements of self in their own terms without the restrictions afforded by a survey. They then explored the illness experience, and the participant's involvement in their partner's care and care planning. Subsequent questions explored preferences regarding disclosing relationships, sexual orientation and/or gender identity to others in their social networks and to health care professionals. Questions then explored the partner's death, funeral and sources of support both pre-and post-bereavement. The interviews closed with an opportunity to share recommendations for practice. After each interview a reflective diary was completed by the interviewer (LT), and discussed with the study lead (KB) and other members of the research team. The diary included emergent themes related to the aims of the study, commonalities and differences in experiences described compared to preceding interviews, and any environmental factors that may have influenced the conduct of the interview. These emergent themes were presented, discussed and revised with the whole steering group, including LGBT+ community members. An iterative process of discussion of the reflective diary, alongside the study aims, informed the decision to stop recruitment, and move to analysis. Recruitment continued until saturation was indicated and no new themes related to the study aims were being identified.

Analysis
Interviews were audio recorded and transcribed verbatim. Analysis was conducted in two phases supported by NVivo qualitative data analysis software. First, an inductive reflexive thematic analysis 30 was conducted to explore bereavement experiences and sources of support. Analysis followed six stages: familiarisation, generating initial codes, generating themes, reviewing potential themes, defining and naming themes and reporting. Analysis was led by KB, but interpretation was collaborative and iterative throughout the stages of the analysis, drawing on the diverse perspectives and experiences of those within the research group. Descriptions of themes and coded extracts were shared with the research team, including LGBT+ community members, to comment on and revise interpretations. A secondary deductive reflexive thematic analysis was conducted to appraise the extent to which the data supported the Acceptance-Disclosure Model of LGBT+ bereavement. 21 This stage of the analysis included a return to the initial codes, and generation of themes within the four domains of the model.
The present study is reported in line with the COREQ (consolidated criteria for reporting qualitative research) guidance. 31 Labels for verbatim quotes state gender modality only where this was shared during the interview. All names were replaced with culturally appropriate pseudonyms.

Participants
Twenty-one individuals were interviewed (October 2018-September 2019) who were bereaved of a samegender civil partner (n = 17) or spouse (n = 4). Nine identified as women, eleven as men and one as non-binary.
Although not explicitly asked about gender modality (how a person's gender identity stands in relation to their gender assigned at birth 32 ), two participants shared that they were transgender, and three participants shared that their deceased partners were transgender. The remaining participants did not share information about their own or their partner's gender modality during the interviews. Most participants described their sexual orientation as gay, lesbian, bisexual, or same-sex attracted (n = 18), one as heterosexual, two preferred not to say and one did not feel that the label was important. Six of the participants were from minority ethnic communities. The median age of participants was 57 (range 37-85), and median relationship duration was 19 years (range 7-61 years). Participants were recruited from across England, and the median interview duration was 79 min (range 38-194) ( Table 1).

Findings
Participants' bereavement experiences are presented below. Part 1 focuses on descriptions of the emotional and social response to bereavement, which we conceptualised as falling into two categories; those that accorded with universal experiences of partner loss (as derived from other published literature), and those that appeared to be specific to LGBT+ groups. The second part presents the appraisal of the Acceptance-Disclosure Model of LGBT+ bereavement. For those who spent much time alone, some talked of a need for this as self-preservation, while others described a process of evolution, as they learnt that time alone was what they needed to grieve, regain control and heal. There were also challenges for some individuals at the intersection between their LGBT+ identities and other identities, such as age and ethnicity. Some who had experienced bereavement at a relatively young age described intrusive questioning, or individuals minimising their loss. Questions tended to relate to finding a new partner; a question they felt would not be asked of heterosexual bereaved partners. There were also specific challenges described relating to LGBT+ identities themselves. For individuals for whom their deceased partner had been their first same-gender partner, the bereavement led them to question themselves and their identity despite not feeling ready for another relationship. Specific challenges were also shared in relation to gender identity. Participants described universally respectful and caring interactions with funeral providers, however these could be undermined by bureaucratic requirements to use the individual's birth name on official documentation. There were also LGBT+ specific considerations in relation to sources of support. Participants described fragile relationships with family members, but also how their lesser likelihood of having biological children resulted in many LGBT+ people having less support. Others saw official environments and services (bound by equality legislation in the UK) as affording protection and giving them the confidence to share their identity. This was in contrast to negative experiences in public spaces. In the second part of our analysis we applied a more deductive analytic approach to appraise the extent to which the data supported the four domains of the Acceptance-Disclosure Model of LGBT+ bereavement 21 .
We focussed on data that captured participants' views, experiences and preferences related to disclosure of their relationships and identities and the reasons behind these decisions. We were interested in whether they felt able to disclose the nature of their relationship to those around them (spoken vs unspoken), relating this to the degree to which that relationship felt recognised (accepted vs not accepted). Participants described experiences which support the four domains of the model. However, they also highlighted a degree of autonomy and agency in these experiences, which had not emerged so strongly from the initial modelling of the literature. Generally, the data demonstrated varying degrees of agency in decisions to disclose, ranging from actively choosing their level of disclosure, through to this being elicited passively. As this agency has been demonstrated in relation to the position an individual occupies in the model, we have described this in terms of active positioning and passive positioning. Often participants occupied more than one domain concurrently, disclosing their relationship openly in some settings, while adopting a more unspoken or invisible position in others. The findings below are structured around the four domains of the model. a. Overt acceptance. Participants described achieving overt acceptance both actively and passively.
Active positioning: The positive impact of overt acceptance was confirmed by the interview findings, with individuals describing how feeling able to talk about their bereavement enabled support from those around them. Some talked about using their legal status (e.g. 'This is my civil partner') to achieve that, and to ensure they were involved in decision making.
'I mean the reason we did our civil partnership was really for the legality of it that would in law, make me next of kin. That was the driving force for it, not because we wanted to shout it from the rooftops [. . .] At the end of the day we're just people and they accepted that we were in a relationship so we came as a duo and a package and that was the end of it''. Caroline, gay woman in her 60s Passive positioning: Others in this domain did not show such a degree of agency, but instead effected disclosure more passively. This might be in response to questions about their relationship from the clinical team, and would not have been disclosed otherwise.
'So she put it in the way 'Are you together?' and that's a fine piece of words. That would be a fine medical approach of asking. When people ask you 'Are you married?' or 'How long have you been together?', when people were genuinely interested, that was a positive experience'. Aoife, bisexual woman in her 40s Asking about relationship status was advocated by participants to ensure relationships were recognised, and shift responsibility away from the patient and partner to disclose this themselves.
'I don't know, if they asked her or if she told them, but definitely, they should have a conversation about it when the patient first goes in. Ask them, ask them who the visitors are gonna be, who they are, what their relationship is and then they know'. Lorraine, same-sex attracted, non-binary person in their 50s b. Unspoken acceptance. Participants described achieving unspoken acceptance both actively and passively.
Active positioning: Despite being in a legally recognised relationship, many participants described a preference for unspoken acceptance. For some participants this active decision was framed positively. They talked about LGBT+ relationships being normalised, thus removing the need to disclose anything. However, for others this decision was framed more negatively and related to fears or experiences of discrimination, discomfort talking about sexual orientation, or sometimes simply a preference for privacy. Others described the decision more neutrally, or in terms of necessity. They described a need for the patient rather than the couple to be the focus of care, with any disclosure about their relationship not being a priority. Passive Positioning: There were also examples of passive adoption of unspoken acceptance. For some this was due to a preference for not talking about the nature of their relationship, and therefore allowing their relationship to be presumed, and not initiating discussions themselves. Others however talked about this more positively in relation to social support. Supportive behaviours from those around them were more important than explicit acknowledgement of the nature of the relationship itself. However, on some occasions the decision not to be explicit about the relationship resulted in incorrect assumptions about the relationship, which added an additional layer of embarrassment and potential offence. c. Invisibility. Participants provided examples of their relationship not being accepted or spoken about, but only described this invisibility in the context of this being intentional (i.e. achieved actively).
Active Positioning: Despite the implications for impaired access to support, several participants chose to maintain invisibility within certain social spheres. Reasons for this included historical experiences of the need for caution around disclosure of relationships, a preference not to talk about relationships, and fears of discrimination. Some participants talked about not feeling comfortable to be disclose their relationships at work, and therefore their employers and colleagues being unaware of their bereavement. Participants provided examples of their relationship not being accepted when spoken about, but only described this overt exclusion in the context of this being non-intentional (i.e. achieved passively). In these situations, the participant had disclosed their relationship, however the response from the individual forced them into a position of overt exclusion, thus passively achieved.
Passive Positioning: Participants described experiences of overt exclusion following disclosure due to bureaucracy (such as the requirement to use birth name of deceased transgender partners). However, they also described overt exclusion enacted by individuals aware of but refusing to recognise the nature of their relationship.

Main findings
LGBT+ people experience devastation at the loss of a life partner, just as captured by the literature for loss of spousal partners. 2 This is exacerbated by additional barriers and stressors specific to their sexual orientation and gender history. Participants described lack of recognition of the depth of their loss, inappropriate questioning, disclosure of gender history against their wishes, and fears of discrimination or rejection when accessing bereavement support. There were also challenges related to the intersection between their LGBT+ identities and other aspects of their culture and identity. Our findings in relation to disclosure of relationship status provide some primary evidence to support the Acceptance-Disclosure model of LGBT+ bereavement. 21 However, for one domain, overt exclusion, we found examples only of passive rather than active positioning, and this finding helps in the interpretation of the model. We posit that it is the action of exclusion by another individual that forces the LGBT+ person into the position of overt exclusion, thus removing their source of agency.

What this study adds
These findings build on previous theories on the concepts of disenfranchised grief 27,33 and loss to provide new perspectives. Rather than disenfranchisement being binary in nature, individuals talked about a continuum of support and disenfranchisement within their social networks. For example, one individual might describe their relationship and bereavement being known about and acknowledged by friends, not openly discussed with neighbours, and unknown or unacknowledged by biological family and employers. This demonstrates the dynamic nature of disclosure preferences, and the importance of understanding these in order to best support individuals pre-and post-bereavement. Of particular note in relation to the Acceptance-Disclosure model of LGBT+ bereavement was the role of agency and choice in decisions regarding disclosure of identities, and the decision for some LGBT+ individuals to actively choose invisibility to retain relationships within certain social groups. This theory serves as a valuable tool for professionals caring for LGBT+ individuals peri-bereavement. However, it has relevance throughout the lifecourse for many LGBT+ people and their significant others, where decisions regarding disclosure of identities are ever present, due to the legacy of a lifetime of exclusion and discrimination. 34 It helps explain the decisions individuals make about disclosure of their relationships and identities, the role agency plays in this process, and the implications of these on their access to bereavement support. As such, it may also have utility for individuals from other minoritised and marginalised groups whose relationships flout cultural or societal norms, and whose social support resources may be reduced.
These findings also challenge the current discourse and policy drive to mandate sexual orientation and gender identity monitoring. 35,36 Participants were universal in their agreement of the importance of understanding who matters to the patient, and who is next of kin. However, many individuals did not feel their specific LGBT+ identity was relevant or found questions about the nature of their relationship confrontational. This view was held by both older and younger participants. The arc of legal and social change (from criminalisation and discrimination, through legalisation, to acceptance and protection) seems to have led some individuals back to a preference for unspoken acceptance. Older participants who had experienced relationships prior to legalisation of same-gender relationships were habituated into not talking about their relationships, whereas younger individuals described experiences of LGBT+ relationships and identities being normalised, obviating the need to talk about them. This finding highlights the important role of implementation research for LGBT+ monitoring, and the contribution required by both services and individuals to enable inclusive practice that is person centred. 23,24 Service level indicators of inclusivity, and sensitive inclusive communication are key. Through using neutral language, avoiding assumptions, and being aware of non-verbal communication and the environment in which discussions are being held, professionals can offer individuals the choice and opportunity to share information about relationships and identities in line with their preferences. 24 This in turn enables LGBT+ people to feel supported, included, safe and confident to access health and social care when they need it. 37 Drawing on the findings of the present study five recommendations for inclusive care of LGBT+ people facing bereavement are presented (see Table 2).

Strengths and limitations
This study purposively sampled individuals who had responded to a national population-based post-bereavement survey. This enabled recruitment of a more diverse sample of participants than is common in LGBT+ research. Our data included experiences of two transgender participants, three people bereaved of a transgender partner, and individuals from minority ethnic and religious groups. The diversity of this sample will enable greater transferability of the findings. One limitation of our sampling was that all individuals were bereaved of a civil partner or spouse. It could be argued that these individuals are more likely to be well supported in bereavement, because of their legally recognised relationship. Further research including those for whom the experience of invisibility predominates, or who experience internalised stigma, would increase transferability and resonance to wider groups, and give further understanding as to how agency (active and passive positioning) contributes to the experience and disclosure preferences for these individuals.

Conclusions
LGBT+ bereaved people face additional barriers and stressors in bereavement due to legal and financial issues, and sub-optimal support due to fears regarding disclosure of their relationship and bereavement within certain groups. Not all LGBT+ people want to talk directly about their identity or the nature of their relationship. Sensitive communication is required to build trust, and to enable discussions peri-bereavement to ensure support is in place. Particular consideration is needed for those who may be isolated or disenfranchised in their grief due to intersections related to their culture or ethnicity, and for people experiencing damaging posthumous disclosures of gender history. The legal requirements currently negate transgender inclusive care. Further work is needed to ensure bereavement support is culturally competent, and to consider how the Acceptance-Disclosure Model could inform support strategies for other potentially isolated or disenfranchised groups.

Author contributions
KB & RH conceived the study; LT conducted the interviews; KB led the analysis with RH, LT, SM, and DB; KB led the interpretation and manuscript drafting with contributions from RH, LT, DB, SM, AP, AR, KJ, ED, PC, RR, DY, PY, WG, KA. All authors* approved the final version of the manuscript before submission.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work was supported by the Marie Curie Research Grants Scheme grant reference MCRGS-07-16-45.

Ethical approval
Ethical approval was provided by King's College London Research Ethics Committee (HR-17/18-5668).

Data management and sharing
The full transcripts are not available for sharing to ensure anonymity of participants.

1.
Avoid making assumptions about individuals and relationships, because incorrect assumptions can be damaging, particularly in bereavement.

2.
Use the relationship label(s) (e.g. partner) and pronouns (e.g. she/her) that individuals themselves use, and ask if you do not know. 3. Explore sources of personal and social support by asking: Who in their lives is important to them? Who is aware of their bereavement, and the nature of that loss? This information will help you to understand the true nature of support available from family, friends, social networks and employers, and also ensure you do not inadvertently disclose their relationship to others against their wishes. 4. Be cognisant and respectful of the hesitancy and fears individuals may have around accessing formal bereavement support.

5.
Facilitate connections with local bereavement support services, and make the first introduction if required, to allay fears and enable individuals to access the support they need.