‘So isolation comes in,discrimination and you find many people dying quietly without any family support’: Accessing palliative care for key populations – an in-depth qualitative study

Study design

Cross-sectional in-depth qualitative interview and focus group study.

Recruitment

Using the above definitions, key populations were defined as adult (aged at least 18 years) sex workers and sexual and gender minorities (LGBTI people and other sexually non-conforming people). The selection was focused on those socially excluded populations whose behaviours led to higher incidence of infectious and non-communicable life-limiting illness. ¹³ As the socio-political context precluded a conventional recruitment strategy via health centres, these legally vulnerable and hard to reach populations were accessed through non-governmental organisations that support sex workers and LGBTI people. The researcher met with all representatives to introduce the study and the sampling frame. They approached eligible individuals about the study to avoid any potential coercion from direct contact with the researcher at this stage. Purposive sampling criteria including sexual orientation, gender identity, age and sex worker status, guided recruitment. Representatives were not offered training to recruit for this study but were familiar with the process and all succeeded in recruiting requisite numbers. Data regarding refusal to participate were not collected by support agencies, therefore it is not possible to know whether there were any patterns to non-participation.

Due to non-response from the Ministry of Health, government health professionals were not included in the sample. Health and palliative care providers and representatives of support and programme agencies were purposively sampled across palliative care, sexual health and support agencies. The sample was drawn from those living in two urban areas as well as smaller centres near border posts, mines and truck stops known to attract high-risk populations in Zimbabwe.

Ethical considerations

Ethical approval was obtained from the Medical Research Council of Zimbabwe on 24 September 2014 (Ref: MRCZ/A/1881). Recruitment was undertaken with clear safety protocols to minimise risk to participants and researchers. A distress protocol was developed in case any interviewee became distressed, fatigued or unable to continue for other reasons. A representative from a relevant support organisation was available to address any adverse events. Interviews were conducted in places considered safe by our partner organisations. These included their own secure premises, or at alternative spaces considered low risk with no link to key populations. Before commencing the interview/focus group discussion and having received an explanation of the study, each participant signed consent forms to participate and to audio record. Confidentiality was ensured throughout the analysis and reporting process by using pseudonyms and removing identifying information.

Data collection

A focus group discussion for sex workers and another for LGBTI people, each lasting approximately 1.5 h, was conducted in the four centres (n = 8 focus group discussions in total) using a semi-structured topic guide. The proximal towns of Beitbridge and Masvingo were combined as a single centre focus group discussion. Individual interviews of approximately 1 h were conducted with a small sample of participants from the focus group discussion, who self-identified as having a life-limiting illness. A further 12 interviews were conducted with health professionals: five palliative care nurses, three doctors, two support organisation administrators, one programme coordinator and one outreach worker across the sites. Data collection was undertaken from November 2014 to January 2015.

The topic guides were developed with reference to the study objectives and review of the literature on LGBTI people and sex worker health. Two pilot focus group discussions helped refine phrasing and sub-headings (data not included). The broad areas of enquiry were illness history and experiences of accessing health and palliative care, exploration of sexual identity in clinical consultations, communication and sexual identity, involvement of partner/significant others and support structures during serious illness and after death. Interviews were conducted and audio-recorded primarily in English with questions and topic guides translated into the vernacular (Shona or Ndebele) by the research assistant as required. The lead researcher (J.H.) is Zimbabwean female, fluent in English and conversational in Shona; the research assistant (S.C.) is Zimbabwean female, and fluent in all three local languages (English, Shona and Ndebele). It was not possible to return translated transcripts to participants for checking due to the challenges and sensitivities of re-approaching participants.

Analysis

Data were transcribed verbatim (with translation into English where required) by the research assistant and analysed using thematic analysis ²⁰ with five key stages: familiarisation, coding, theme development, defining themes and reporting. Analysis was led by J.H. The first stage of analysis was familiarisation in which the researcher(s) read, reread and annotated the transcripts alongside research questions. Transcripts were coded line-by-line within the sub-headings of the interview schedules; however, inductive coding was undertaken within each theme to identify emergent patterns and sub-themes, and across transcripts for higher level overarching themes. Subsequently, a coding frame was developed to demonstrate emergent patterns and themes. The coding frame and transcripts of three interviews/focus groups were reviewed by the research team. After discussion, themes were developed and the coding frame revised and applied to the full data set. During subsequent analysis, the themes were developed further and refined with particular attention paid to non-confirmatory cases, where emerging themes contradicted more common ideas. Additional themes not captured under the sub-headings were also noted during this stage of analysis and added to the thematic compilation. Themes were defined and finalised through discussion (J.H., K.B., R.H.), and all researchers agreed with the final analysis, interpretation and reporting.

Participants

A total of 72 individuals participated (see Table 1): from key populations (n = 60) recruited for focus group discussions, and from healthcare providers and support organisation representatives (n = 12) recruited for interviews. Focus group discussion participants who self-identified as living with a life-limiting illness were also interviewed (n = 1 sex worker in Mutare, n = 1 sex worker in Beitbridge and n = 1 lesbian woman in Masvingo).

OPEN IN VIEWER

Table 1.

Sample characteristics.

Key populations	60	Health professional	12
Site		Site
Harare	17	Harare	8
Bulawayo	16	Bulawayo	1
Mutare	15	Mutare	2
Beitbridge/Masvingo	12	Beitbridge/Masvingo	1
Age		Age
<35	32	<35	3
35 and over	27	35 and over	8
Info not provided	1	Info not provided	1
Gender		Gender
Male	18	Male	4
Female	42	Female	8
Ethnicity		Ethnicity
Shona	39	Shona	10
Ndebele	16	Ndebele	1
Other	5	Other	1
Sexual orientation/gender identity		Status
Lesbian	6	Palliative care nurse	5
Gay	15	Doctor	3
Bisexual	3	Director/administrator	2
Transgender	2	Coordinator	1
Intersex	1	Outreach worker	1
Sex worker	33

In all, 49 participants were of Shona ethnicity, 17 were of Ndebele and 6 were of ‘other’ ethnicity. Of the participants, 50 were female and 22 were male. Focus group discussion participants self-identified as lesbian (n = 6), gay (n = 15), bisexual (n = 3), transgender (n = 2), intersex (n = 1) and sex worker (n = 33).

Findings

Two dominant themes emerged from the data: (a) minimal understanding of, and negligible access to palliative care significantly increased the risk of painful, undignified deaths and (b) discriminatory beliefs and practices from healthcare providers, family members and the community had wide-ranging effects on care for those living with life-limiting illness and their wishes at the end of life.

Few key population participants had heard of palliative care. Understanding was incomplete and generally linked to anti-retroviral therapy adherence. Some participants had heard of a local hospice, but were unclear what was offered there:

Maybe it’s for people who are tired of pain and go there to die. Maybe get injected and die. (FGSWB3 Sex worker female, 21, focus group Bulawayo)

Healthcare providers had limited experience of providing palliative and end-of-life care for key populations:

I haven’t had terminally ill patients. Most of the patients who are presenting are still at a manageable stage. (KIIH8 doctor male, 30, interview Harare)

Despite HIV being a common reason for palliative care referrals in Zimbabwe, exploration of sexual risk behaviours, particularly among key populations who are known to be at risk, was uncommon. Health professional assumptions of heterosexuality and monogamy enabled LGBTI people and sex workers to remain ‘undercover’ and access health and palliative care along with mainstream populations:

Our social history just goes to, ‘are you married?’ If yes then we take it that they have one sexual partner. (KIIH7 palliative care nurse, female, 33, interview Harare)

Hiding sexual risk behaviours, however, resulted in untreated symptoms. Together with inaccurate self-diagnosis, this was likely to increase transmission of infectious diseases, non-adherence to treatment, risk of co-morbidities and mortality:

[When} they die within a week you wouldn’t have done a lot of work because they were a late referral … because they would be hiding … and you wouldn’t have achieved much. (KIIH1 palliative care nurse female, 59, interview Harare)

Pain assessment and management were lacking in personal testimonies and anecdotal accounts of others who had died in their communities. Participants had access to only basic pain medication regardless of severity. Most were unaware of medications for severe pain and described scenarios of friends dying in pain, partly due to their resistance to seek healthcare.

The six healthcare providers associated with palliative care organisations were conversant with palliative care principles, but some acknowledged personal difficulties in applying the approach to key populations. Attitudes to key populations were strongly determined by religious beliefs and personal experience:

… we tend to intertwine [unfinished personal business] with what the patient brings and at the end of the day it will impact negatively on the type of treatment that we give a patient. (KIIM2 palliative care nurse female, 42, interview Mutare)

Healthcare providers reported less experience with, and more prejudice against, sexual minorities than sex workers, a population they were familiar with in their personal lives:

The commercial sex workers I don’t think I will have much fuss with them but the gays and lesbians. I think I am not used to them. (KIIM2 palliative care nurse female, 42, interview Mutare)

That familiarity also jeopardised care for seriously ill sex workers:

… they are corrupting and taking our husbands … they are looting from our husbands and we are now running short of finances because of these commercial sex workers … if you are a female nurse and a sex worker comes as a patient and sometimes your husband, it happens doesn’t sleep at home and sleeps elsewhere or books a room for him and a sex worker then you won’t be good to that lady … and you would rather say let her die because she is corrupting my husband and disturbing my marriage. (KIIBB1 coordinator male, 48, Beitbridge)

Participants offered reasons for delayed presentations at health facilities such as lack of finance and anticipated discriminatory behaviours from health workers. When patients did seek care, judgmental and inconsistent medical care led to patients defaulting on treatments, with consequent incremental effects on co-morbidities and mortality:

They get questions like, ‘why are you doing this? Have you read the bible?’ Or the nurses would scream ‘look at that homosexual’, ‘is this a woman?’ and you get those giggles. And people feel like next time I won’t go. They default on treatment. They do that. Sometimes people get stressed and default. (KIIH3 Director male, 42, interview Harare)

Some participants chose to visit private doctors who they felt were professional and confidential. Others sought advice and medicines from pharmacists. Many relied on colleagues and friends. None reported relationships with healthcare providers that supported their specific needs. Untreated symptoms, uncontrolled illness, lack of health monitoring and social support led to painful, lonely and undignified deaths:

I had a friend of mine … she had a cancer on her leg and she went to the Apostolic church. So they used some herbs to heal the wound. So they used to take some [local grain] and put it in there saying that there were worms in the wound that needed to eat the food they put rather than eat her flesh … the whole room would be filled with a bad smell. So people started to run away from her until they stopped giving her food, then one day when we went to pay her a visit we found no one there, flies everywhere and the moment we opened the door we discovered that she had passed away. (KIIBB2 Lesbian, 40, interview Masvingo)

Stigma from family members and the wider community further compromised support for the ill and dying. Discrimination in this environment was described as extreme to the extent that lives were in danger:

They can be killed or burnt alive. Because people regard them as animals and not as human beings. (KIIM2 palliative care nurse female, 42, interview Mutare)

Patient care appeared contingent upon attitudes of family and community towards key populations, regardless of the severity of illness, even at the end of life. Some participants preferred to die in hospital as they could not count on family support:

So isolation comes in, discrimination and you find many people dying quietly without any family support … They will say you were doing your gay things on your own that’s why now it’s payback time for your sins … People will look at you and say it’s now punishment for not having reproduced, getting married and not following the family tradition and going against God. (KIIH3 Director male, 42, interview Harare)

Spiritual support was also compromised by discriminatory views of some religious leaders:

… when the pastor comes … and you say this is my wife and he is a man the pastor will definitely freak out. ‘This is a taboo, an abomination. What are you doing? No wonder you are in this situation’. (KIIH6 palliative care nurse female, 48, interview Harare)

Participants believed final wishes at the end of life would not be respected because of their ‘bad behaviour’:

I would want my partner to stay with some of my things, but if you are staying with your partner and you die, your family will just come and take away your things and just say you were a friend and this property belonged to my daughter. (FGLGB3 lesbian 42, focus group Bulawayo)

Strengths and limitations of this study

This was the first study undertaken in low- and middle-income countries to describe palliative care experiences for sexual and gender minorities and contributes to the growing body of research on sex workers. It is likely that palliative care issues experienced by gender diverse minorities were under-described as only two transgender and one intersex persons were included in the sample; therefore data saturation is unlikely to have been achieved. In this socio-politically punitive environment, their recruitment, however, was a positive achievement and addressed the absence of this demographic in other research. ²² The sexual orientation of sex workers was not recorded, therefore, we may have missed the opportunity to explore experiences for those with both LGBTI and sex worker status. Diagnosis information was not included when recruiting key population participants, beyond a request for participants with life-limiting illness to volunteer for individual interview. Data regarding refusal to participate were not requested from the support agencies who all successfully recruited requisite numbers, therefore, it is not possible to know whether there were any patterns to non-participation. Information surrounding serious illness, dying and death was often based on witnessing deaths of colleagues and friends. Permission was sought from the Ministry of Health and Child Care to interview government health workers, but non-response in the time available excluded them from this study. Recruitment of 12 other health workers to discuss care for key populations was challenging given the socio-political situation. Translated transcripts were not returned to participants for checking, due to the challenges and sensitivities of re-approaching individuals.