The burden of EoL care
In most of the articles, EoL care was reported to be a significant financial, physical, or social burden for carers. Providing appropriate food, medicines and water financially impoverishes carers and increases the burden of care [
17‐
27]. Funeral costs were also mentioned as a large economic burden [
27,
28]. Time spent on caring meant less time for food [
17] and livelihood production [
24], and carers sometimes sacrifice their own resources for the patient [
23].
The physical consequences of caring for someone with EoL care needs were mentioned in several studies [
18,
22,
24,
25,
28‐
30]: carers were exhausted and developed pains after moving patients and carrying out strenuous household tasks. Several studies also described how HIV positive caregivers' health was compromised through caring for someone else [
20,
21,
25]. The studies likewise provided evidence of the psychological impacts of caring; the emotional demands of caring resulted in loneliness, depression, and isolation [
18,
23,
24,
28,
30‐
33]. Factors that increase the stress of caring include the stigma of the patient's illness, parental carers needing to come to terms with their child's HIV status, patients being abusive and unpleasant to carers, and a lack of knowledge regarding the illness or best care practices [
17,
23,
25,
30,
31]. Grant et al. reported communication difficulties in families due to a fear of discussing care and illness, as this would increase their vulnerability to "evil forces" [
19].
The burden of care is also often compounded by gender and age related issues. Many authors discussed the material and emotional burden of care that women bear for their children, orphaned children, and relatives [
23,
25,
26,
28,
30,
31,
34‐
37]. Moreover, Matukula et al. argued that funds to pay for care should be directed specifically to women [
21]. In some instances, younger female carers face physical and sexual abuse from family members [
28]. Thomas' study however questioned the assumption that the burden of care falls on women [
24]. In regard to age, elderly carers in Togo [
22] and South Africa [
27] faced additional challenges as a consequence of role reversal: elderly parents had to care for their adult children, and to cater for the needs of their grandchildren [
27]. The health of older carers can suffer as a result of their care-giving, especially in Uganda where children seek parental care only when they are very sick and require intensive care [
28,
30]. The burden of care borne by older carers also impacts upon the living arrangements and family relationships [
38].
Low levels of support from other community members results in a social burden of EoL care for carers [
20,
25,
26,
31]. Identified causes of this lack of social support included caregivers' fears of community members' reactions towards people with HIV/AIDS, and carers withdrawing from social activities due to the demands of caring [
17,
21,
23,
25,
28,
30]. Additionally, the nature of the illness can cause social and personal disruption [
24]. Although carers had previously sought help from traditional healers, due to the costs and lack of confidence in their services, this was reported to be becoming less common in Botswana [
39].
Several studies demonstrated how in households the care, illnesses, and deaths of family members create a complex burden of care [
20,
23,
28,
30,
36]. Furthermore, a patient's role within the household may change due to the inability to reciprocate care or financial support and place strain on personal relationships [
22,
24]. Given its importance in care provision, two studies argued that approaches and policies towards EoL care should be focused at the household level [
20,
21].
Although the negative impacts of the care burden are commonly reported, Akintola [
40] detailed some of the rewards that carers receive in exchange for their assistance. The author argued that although the rewards do not outweigh the burden of care for volunteers caring for people living with AIDS, rewards include improving their own health, enabling them "achieve self-growth", acquiring new skills and gaining recognition from the community.
The burden of care for health professionals providing care for HIV/AIDS patients is also highlighted. Staff in South Africa and Zambia felt emotional stress, had fears of infection, and were unable to cope with the loss of patients [
32,
41‐
45]. Their burden of care was worsened by lack of hospital resources, such as beds, staff or medicines [
25]. Community carers faced difficulties, such as not being welcomed by patients' families for fear that their presence would stigmatise the family [
25] and undertook tasks that were not their responsibility such as arranging funerals [
45].
Training, support needs and tools
The improvement of training and support for informal and professional carers appears as a recurrent recommendation to lessen the burden of care. Identified areas of training include: basic hygiene, clinical aspects of care, symptom management (including fear of opioid addiction), prevention of patient to carer infection, care for children, counselling for patients and families, home-based care, information provision, and psychiatric and bereavement care [
17,
18,
23,
25,
28,
30,
33,
34,
36,
42,
44‐
47]. Mtalane et al. recommended training for professional carers regarding the emotional, spiritual and cultural needs of the dying patient [
48]. In contrast, Uys (2003) found that professional carers were comfortable meeting patients' spiritual needs [
48,
49]. Two studies that report the evaluation of an education strategy in rural Uganda [
50,
51] recognised the core role of education for service development.
Selman et al. [
47] identified that the information needs of patients receiving EoL care and their carers regarding symptoms, symptom management and disease cause, progression and treatment were commonly not met. The poor information supplied by healthcare personnel working in EoL care services in South African and Uganda led to high levels of distress amongst patients and their carers. The authors therefore recommended generalist EoL care training for all clinical staff, which should include communication and basic counselling skills. Information provision should also be individually tailored, proactive, open, documented and for both the patient and carer [
47].
Caregivers' psychological burden is the focus of identified unmet support needs. Counselling services, respite care, stress management courses and support meetings for carers need to be developed or improved [
17,
18,
28,
42‐
44,
49]. Furthermore, EoL care service managers need guidance in the management of HIV positive staff, especially in regard to dealing with workplace stigma [
43].
Nurses in Lesotho, who were interviewed about their knowledge of and attitudes towards EoL care and the WHO Integrated Management of Adolescent and Adult Illness (IMAI) guidelines, demonstrated limited knowledge of the presence of the palliative care booklet in the IMAI guidelines [
52]. In spite of their lack of familiarity with this tool, the nurses identified the need for "holistic" EoL care for people living with HIV/AIDS.
Place of care
Eleven papers focused on place of care in advanced illness. Hospice care initiatives are still a recent development and an evaluation of the hospice referral process was reported to promote improved access in South Africa [
53]. One study in a cancer palliative care setting addressed the related issue of information behaviour, pointing to its complexity and the need for further investigation [
54].
Four studies were undertaken in the context of home-based AIDS projects in South Africa [
25,
42,
45,
49], and one study on an intervention providing home care to a population with EoL care needs in Kenya [
55]. Several studies targeted home-based care, with family members as the sole providers of care, giving insight into patients' preferences for place of care and the conditions needed to supply home-based care [
22,
26,
33,
34,
37,
39,
56‐
58].
Uys et al. [
49] performed a multi-method study in seven sites in South Africa working with the integrated community-based home care model and found a relationship between dying at home and "good death". From which the authors concluded that this is the model of choice for patients with AIDS and asserted that home-based care is not a second best option for developing countries but has the capacity to improve the illness and dying experience. The study highlighted the need for outreach programmes and poverty relief as patients often sought hospital care in order to obtain food rather than treatment.
In Hunter's 2005 study [
25], patients and informal carers expressed a need for professional care because they encountered problems accessing clinics. Uys and De Saxe described the practices of community caregivers (CCGs) in a home-based AIDS care project [
42,
45]: CCGs were positive about their contribution [
45], however, CCGs received little salary and had low professional status which led to high levels of staff turnover [
42,
45]. The degree of community caregiver supervision also had an important effect on the success of the projects.
A mixed method study on the EoL care needs of HIV/AIDS patients in Rwanda found that the second most commonly perceived need after medical, psychosocial and financial care (77%) was home-based care (47%) [
57]. A later paper by the same authors in the same study site, focused on terminal AIDS patients' preferred place of care [
58]. Whilst the quantitative data showed that 67% of participants preferred hospital care, the qualitative findings attributed this to the lack of home care provision in Rwanda and to patients having lost family members in the 1994 genocide war who would have otherwise taken care of them.
Four studies [
22,
33,
34,
56] found that home care is not the evident choice at the EoL. Olenja's study of community attitudes towards home care for patients with AIDS in Kenya illustrated how home care was seen as unrealistic in a context of poverty, stigma and lack of knowledge of how to care for a person with AIDS [
34]. Two studies [
22,
33] from different regions reported similar findings that represent the carers' view: once a patient was discharged from hospital, carers found themselves unsupported and on their own. The capacity for care, traditionally available through the extended family structure, had changed permanently due to AIDS. Only the elderly directly related to the patient felt responsible to provide care for their children.
Murray et al. [
56] provided insight into how palliative care patients in Kenya experienced their stay in the hospital: patients felt alone and removed from their family, who were not always able to visit them due to the cost of transport.
Good death versus bad death and stigma
Two ethnographic studies focused on conceptions of death in Ghana [
59,
60]. In the 2002 study, Van der Geest explored older people's views about death in a rural town in Ghana [
59]. There was no fear of death. Rather, people spoke of death as a welcome visitor that brings peace and rest. Respondents maintained an agnostic stance towards the future. In Van der Geest's later study [
60], the accounts of the deceased relatives' last moments were not dominated by medical interventions and bad deaths were deaths that come too early, deaths that were believed to be punishments for sins. Those who died a bad death did not receive a proper funeral; a good funeral is part of a good death.
In rural Tanzania, using ethnographic methods Dilger [
29] identified a moral discourse around HIV/AIDS; understood as a bad disease, HIV/AIDS was brought by indecent behaviour of family members who migrated to the cities and this legitimised the poor quality care that in some cases people received. Moreover, HIV-infected migrants often returned to their rural families at the final stage of illness, placing increased burdens on their relatives.
Studies from Kenya and South Africa also revealed that the stigma surrounding HIV/AIDS was due to its association with immoral behaviour [
35,
61]. Another study from South Africa explained the stigma in terms of the 'terminality' of HIV/AIDS [
62]. People were aware of biomedical explanations for its cause but contested them with views based in local cosmologies and belief systems [
29,
35,
61]. In some instances, but not all [
38], the stigma of HIV/AIDS impacted on both the patient and the caregiver. In some cases stigma led to problems accessing local services, rejection and the breakdown of social support [
20,
22,
24,
28]. And in Botswana, there was a 'culture of silence', due to the stigma and denial of AIDS [
33].
In two studies from Kenya and Tanzania, death from HIV/AIDS was described as particularly disturbing because it supposed a 'permanent death' [
29,
61]: those who died from AIDS were denied the status of ancestors, and this threatened the continuity of the clan and community. Families attached great importance to burying the deceased in their village of origin, due to ancestral attachment of the deceased to their home. However, the shame linked to AIDS disrupted this cultural practice, leading to private funerals. The prospect of such a burial caused anxiety and affected people's dignity, impacting on their relatives and subsequent bereavement.
Culture-specific experiences and approaches to symptoms, illness, death and caring
Two studies addressed the culture-specific responses to symptoms and illness and the implications for its management [
46,
63]. Bor [
63] compared experiences of patients suffering from Kaposi sarcoma in the UK, USA, Zambia and Botswana, and found that the setting and culture mediate social responses to Kaposi sarcoma and disfigurement. Beck's ethnographic study in South Africa [
46] identified cultural variability of cancer as an illness, pain expectations, tolerance and expression, and treatment practices. Barriers to effective pain management were caused by the lack of standards in practice, knowledge and resources. Relations between patients and health workers and problems with communication extended to pain assessment and management [
46].
By focusing on how professionals think about and practice disclosure of a terminal cancer diagnosis in an African setting, Harris et al. [
64] showed the importance of cross-cultural differences in medical practice. Their study was conducted in one of the largest academic hospitals in Africa and could therefore contrast the views of both Tanzanian and expatriate physicians. Although Tanzanian physicians had been exposed to medical practices in foreign settings and were familiar with the values of Anglo-American bioethics they opted for a counselling approach derived from culture-specific discursive practices.