Patient-reported barriers and drivers of adherence to antiretrovirals in sub-Saharan Africa: a meta-ethnography


Corresponding AuthorSonja Merten, Swiss Institute of Tropical and Public Health, Socinstr. 57, Basel, Switzerland. E-Mail:


This meta-ethnography aims at providing a synthesis and an interpretation of the findings of recent social science research on the questions of retention in antiretroviral therapy (ART) programmes in sub-Saharan Africa (SSA). The literature reviewed comprises ethnographic studies of the barriers to adherence to ART in various cultural settings. The results show that the quality of services, treatment-related costs, as well as the need to maintain social support networks – which can be negatively affected by HIV-related stigma – are important barriers to adherence. In addition, they show how African concepts of personhood are incompatible with the way services are conceived and delivered, targeting the individual. In SSA, individuals must balance physical health with social integrity, which is sometimes achieved by referring to traditional medicine. The ability of local concepts of illness to address social relations in addition to health, together with a historically grounded distrust in Western medicine, explains why traditional medicine is still widely used as an alternative to ART.

What this review adds

  • • Weakening social networks in the absence of formal social security constrains ability of households to care for PLWH, increasing stigma
  • • High importance to maintain relationships pertinent to social security, leading to non-disclosure
  • • Traditional medicine may be perceived as manipulating relationships such as to prevent social exclusion, addressing the social and mental dimensions of health
  • • Exclusive knowledge claims of biomedicine are undermined by collective memories with coercive treatments and experiences with low quality services

Practical implications

  • • Create room to maintain own belief system in order to address the social and mental dimensions of health
  • • Importance of service provision that matches people’s socio-cultural pattern in which they feel comfortable
  • • Reconsider program approaches targeting the individual; instead focus more the community


Antiretroviral therapy (ART) has become available at low cost in many sub-Saharan African (SSA) countries (Egger et al. 2005; ART-LINC 2006). ART however requires lifelong treatment and consequent adherence. Although it has been shown that adherence in SSA can be comparable or even superior to that in Western countries (Mills et al. 2006a,b; Rosen et al. 2007), retention in programmes still poses problems (Weiser et al. 2003; Ferradini et al. 2006; Stringer et al. 2006; Chi et al. 2007; Carlucci et al. 2008; Krebs et al. 2008). Medication side effects and complexity of drug regimens, individual factors such as alcohol abuse, psychiatric co-morbidity or lack of social support are barriers in high- and low-income countries (Ammassari et al. 2002; Mills et al. 2006a,b; Nachega & Mills 2007; Selin et al. 2007). In sub-Saharan Africa, conditions of extreme poverty and livelihood insecurity add a cost dimension to adherence barriers (Feucht et al. 2007; Hardon et al. 2007; Yu et al. 2007). Non-disclosure because of fear of stigmatization compounds the problem (Mills et al. 2006a,b; Nachega & Mills 2007; Selin et al. 2007).

Although these barriers have been known for several years and recent research has highlighted a growing number of underlying reasons pertinent to treatment discontinuation, few attempts have been made to look at the way barriers relate to each other. This article presents an update and systematic review of findings from qualitative studies on ART adherence in SSA and discusses how these factors interact.


The method of meta-ethnography has proven to be an effective tool to synthesize qualitative studies to attain deeper understanding of complex health-related topics, such as adherence to medication for chronic conditions (Noblit & Hare 1988; Barroso & Powell-Cope 2000; Wallace et al. 2004; Pound et al. 2005; Garside et al. 2008). While a literature review summarizes the main findings of qualitative studies on a single theme, meta-ethnography aims at synthesis, which ‘involves the juxtaposition of studies and the connections between them, to develop a more sophisticated understanding (Harvey 2007). The focus is on interpretation instead of accumulation of information (Walsh & Downe 2005).

Search strategy and outcome

Medline, AIO, SSCI, CINAHL, IBSS, EMBASE and JSTOR databases were used for a global electronic search. Keywords were grouped around access to health care, treatment seeking, adherence or compliance, antiretroviral therapy, HIV and AIDS. Studies using a qualitative design and analysis method were included (Figure 1). Titles and abstracts were reviewed in duplicate by three researchers. Methods, regions and topics were coded. All studies that focused on ART access and adherence (n = 130) continued to the full-text review. Fourteen studies conducted in SSA passed quality appraisal (Wallace et al. 2004; Smith et al. 2005; Walsh & Downe 2005, 2006; Rolfe 2006) and were retained. After additional hand searches, 31 studies reporting findings between 2000–2008 for 12 countries (24 with free access to ART) were included in the synthesis (Table 1).

Figure 1.

 Flowchart of systematic literature search of March 31, 2009 (repeated on January 5, 2010, to include papers after March 31).

Table 1.   Study characteristics of included studies
 First author, yearCountry, Urban/RuralYear of researchARV available Free/ ChargeSample subgroupMethod of qualitative data collectionSample NAim
  1. **U = urban, R = rural.

  2. ***F = drugs available free of charge, C = drugs charged.

  3. ****FGDs = Focus Group Discussions.

  4. *****n.s. = not specified.

0Atuyambe 2008Uganda, Rn.a.n.a.Community members, policy makersFGD’s interviews1220To examine community perceptions and beliefs on whether enhanced access to ARVs increases risk behaviour
1Alliance 2004Zambia, U/R2002/03CCommunity members, HIV+ patients, health workersInterviews, FGDs, case studies, informal conversationsn.s.n.s.To learn about individual and community perceptions, knowledge and experiences of HIV/AIDS and related treatment
2Aspeling 2008South Africa, Un.s.F/CFemale HIV+ patientsInterviews11 To investigate adherence among HIV-infected black women attending an urban private health care facility
3Bila 2009Burkina Faso, U2003/05F/CPLWH and healthcare actorsInterviews, FGDs, participant observationn.s.n.s. × 6–12To better understand why women actually seem to have fewer difficulties than men in accessing HIV testing and care once they become HIV positive
4Byakika 2009Uganda, U2004/05FHIV+ patientsFGDs 3 × 6–4To examine whether having children in a household affected adherence to ART; the effect on adherence when all eligible HIVinfected members in a household are treated; and the factors associated with incomplete adherence in HIVinfected mothers
5Chinkonde 2009Malawi, U/R2005FHIV+ womenInterviews284 × 6–9To clarify why some women who were enrolled in a PMTCT programme did not fully participate in follow-up visits in the first six months after testing HIV+
6Cooper 2007South Africa, U2004/05FHIV+ patientsInterviews61 To examine HIV+ women and men’s reproductive intentions and factors influencing reproductive decision-making in Cape Town, South Africa
7Crane 2006Uganda, Un.s.CHIV+ patientsInterviews10 To better understand the social dynamics of the incentives and barriers to adherence to self-pay ART
8Ezekiel 2009Tanzania, R2006FIn-school and out-of-school youthFGDs 21 × 7–11To determine how ARVs therapy is understood and constituted in local discourses among young people
9Gilbert 2009South Africa, U2007n.s.Patients who attended the ARV clinicInterviews44 To provide insight into social complexities associated with the provision of ART in a resource-limited setting.
10Golooba 2007South Africa, R2002/03n.s.Community membersInterviews, participant observation27 To demonstrate the implications of health-seeking behaviour for the spread of infectious diseases
11Grant 2008Zambia, U2005FHIV+ patientsInterviews, FGDs401 × 12To gain patient perceptions on factors that facilitate and challenge access to VCT and ART access and adherence
12Gusdal 2009Ethiopia, Uganda, U/R2007F/CHIV+ patients on ART for 6 months or more, counsellors, care giversInterviews118 To explore the views and experiences of patients, providers, and peer counsellors on adherence to ART
13Hatchett 2004Malawi, R2000/01n.s.Family caregivers, HIV+ patientsInterviews46 To explore traditional and modern AIDS health seeking behaviour
14Irunde 2006Tanzania, U2005FHIV+ patients, health workers, key informantsInterviews, FGDs, observations1298 × n.s.To determine adherence levels, to identiy facilitators and barriers to ART adherence and to recommend possible interventions to improve adherence
15Kgatlwane 2006Botswana, U/R2005FPolicy-makers, HIV+ patients, health workers, community membersInterviews, FGDs, observations1778 × 8–10To determine adherence levels, to identiy facilitators and barriers to ART adherence and to recommend possible interventions to improve adherence
16Mitchell 2007South Africa, U/R2004F/CHIV/AIDS caregivers, HIV+ patients, health workers, community membersFGDs 7 × n.s. To understand perceptions relating to ART, thereby hoping to highlight issues that need to be addressed when the government’s ART rollout program expands in the future
17Mshana 2006Tanzania, R2003/04FHIV+ patients, community membersInterviews, FGDs1016 × 8–12To identify and mitigate barriers to seeking ART between the stages of testing and enrolling
18Murray 2009Zambia, U2006FHIV+ patientsInterviews80 To identify factors affecting HIV-infected women’s decision to accept or continue ART
19Nachega 2006South Africa, U2004FHIV+ patients, health workersInterviews, FGDs72 × 6To investigate patient and provider perceptions about characteristics of effective treatment support and to learn more about social and material barriers to HAART adherence
20Nakiyemba 2006Uganda, U/R2005F/CHIV+ patients, health workers, community membersInterviews, FGDs, observations706 × 8–10To identify factors that facilitate or constrain adherence to ART and to suggest possible interventions
21Nam 2008Botswana, Un.s.F/CHIV+ patientsInterviews32 To identify the psycho-social factors related to adherence behavior and to understand the inter-relationships between factors among paying and non-paying patients
22Nguyen 2007BurkinaFaso, Côte d’Ivoire, Mali, n.s***n.s.n.s.not specifiedEthnographic methods (n.s.)n.s.n.s.To explore the issue of adherence to ART in historical and anthropological context
23Olupot-Olupot 2008Uganda, n.a.2008FHIV+ patients from IDP camps, health workersInterviews, FGDs115 × 8To determine patient and health worker concerns regarding ARV adherence in a conflicted affected population
24Roura 2009Tanzania, R2006FHIV+ patients, health workersInterviews, participant observation53 To investigate the diversity and interplay of factors underlying attendance at ART clinics
25Sanjobo 2008Zambia, U/R2006FHIV+ patients, health workersInterviews, FGDs225 × 10To explore patients’ and health care professionals’ views on barriers and facilitators to ART adherence
26Schumaker 2008Zambia, U/R2004CHIV+ patients, health workersInterviews, FGDs, ‘pile-sort technique’312 × n.s.To examine patients’ and practitioners’ attitudes to the material characteristics of pharmaceuticals, ‘spoiling’ and rumours
27Smith 2007Nigeria, U2006FHIV+ patientsInterviews, case studies22 To understand marital and fertility aspirations and behaviours of ART patients and to evaluate the effects on sexual behaviour, disclosure and adherence
28Tuller 2009Uganda, U2007FHIV+ patientsInterviews41 To understand how transportation costs are perceived, experienced and addressed by HIV+ patients
29Unge 2008Kenya, Un.s.FHIV+ patientsInterviews26 To explore the reasons why HIV+ individuals in need of ART choose not to take treatment
30Ware 2009Nigeria, Tanzania, Uganda,U/Rn.s.FHIV+ patients, health workers, treatment partnersInterviews, cobservations250 To better understand exceptional adherence in Sub-Saharan Africa, with focuses on economic barriers
31Watt 2009Tanzania, U2006FHealth care providers, patientsInterviews42 To understand the dynamics of good adherence to ART among patients receiving free ART and HIV-related services from a clinic in Arusha, Tanzania

Synthesis process

Key themes and concepts were extracted and peer reviewed for inclusiveness. A distinction was made between first-order (original quotes) and second-order (author’s analysis) findings, and they were coded in duplicate. Discrepancy between codes was resolved by a third person. Second-order constructs pertinent to adherence were identified and cross-compared and are presented in the results section (Smith et al. 2005) (Table 2). Key themes found were consolidated into a line of argument (third-order analysis), which is presented in the synthesis/discussion section.

Table 2.   Translation Table
3rd order labels3rd order constructs2nd order constructsSummary definition (translation) of 2nd order constructsSource papers
Health systems: Evaluating healthcare experiencesFunding constraints on the national levelDirect costs of treatmentDrug costs, when it needs to be paid by patients, is as a barrier, which can explain the preference for cheaper alternative medicine02, 04, 07, 16, 17, 24
Sustainability of ART not guaranteedAnxiety about individual financial capacity to adhere to a lifelong treatment; fear that ART will not be available forever or that medication will be withhold for poor people01, 07, 12, 17, 20
Adequacy of healthcare services and communicationLow quality of service provisionLong waiting lines; lack of confidentiality and privacy; risk of forced disclosure caused by home based care or by attending certain clinics; drug stock outs; problems with clinic transfers; lack in health personnel; few health centre available, long distance to health centers01, 02, 05, 11, 14, 15, 16, 17, 20, 23, 24, 25, 26, 31
Historically and culturally shaped feminization of health careIn gender separated societies, female dominated health services can represent a barrier to men’s attendance; because of maternal health care (MRH), women have always been closer to health care system and may therefore accept more easily their positive status than men03, 15
Counseling and information influence beliefs and actionPeople may be highly supported by counselors; if individual counseling is poor and information unclear trust and understanding are compromised; difference in concepts of time renders the argument that ARVs prolong life difficult to understand; ART perceived as curative00, 01, 05, 08, 14, 15, 16, 19, 20, 21, 25, 29, 31
Historically and politically based distrust of the healthcare systemFeelings of being exploited and threatenedSuspicions pertinent to ART are based on associations with exploitation by wealthy nations and experiences with politicized national health policies, revealing commonly felt frustrations about political and economical inequality00, 13, 15, 17, 20, 22, 26
Compromised trust in science-based knowledge claimsART not always considered superior to alternative treatments, possibly because people have seen few positive examples of ART; more trust in TM providers than WM because of poor WM access and resources; people may hide TM treatment to WM health practitioners01, 02, 10, 13, 14, 16, 24, 25
Restoring social and spiritual health with TMHIV/AIDS related to witchcraftTM practitioners and faith healers are consulted in parallel to or instead of ART programs to address underlying spiritual and/or social origins of illness; Witchcraft may be considered to cause AIDS, taking away individual responsibility from the infected person and requiring treatment with TM; Magic or witchcraft rituals may protect a community against AIDS01, 10, 14, 16, 24
Traditional health beliefs and past experiences shape perceived efficacy of ARTDirect or indirect experiences affect the perceived lifesaving potential of ART; side effects can be interpreted as signs of strength or toxicity following TM beliefs; community fears, rumours, and preference for TM may change overtime, once people start seeing more persons with improved health through ART01, 12, 13, 14, 15, 16, 15, 20, 24, 26
Identity and agency: Reconciling HIV, and aspirationsReconfigured identity towards ‘living positive’Acceptance of diagnosis and integration in everyday lifeAcceptance of the diagnosis allows the positive integration of ART in everyday life; by experiencing spiritual sources of hope and faith, which support self-efficacy02, 21, 24, 29
Confessional technologies, conversion to a ‘new life’New perception of ‘self’ is a prerequisite for positive living; ‘remaking’ oneself through AIDS activism or Faith Based Organisation; changing lifestyle to lead a more moral life21, 22
Agency and perceived autonomy in coping with HIV and ARTSelf-efficacy and being in controlAdherence is positively influenced by a person’s perceived capacity of being in control over one’s own body and health; by one’s autonomy in prioritizing ART over other needs02, 11, 15, 18, 20, 21, 24, 28, 29
Depression and anxietyBeing tested HIV+ associated with despair, fear, loss; denial and lack of self-motivation, difficult disclosure and consequently adherence; seeing necessary lifestyle changes as impossible burden; lacking will to live with a chronic disease11, 15, 18, 21, 24
Understanding and controlling medication side effectsParticipants tolerate side effects when informed, otherwise it may be a barrier to adherence; willingness to tolerate side effects is influenced by past experiences with WM, affecting motivation to adhere
Increased hunger is a difficult side-effect to bear and affects motivation to adhere.
02, 09, 11, 14, 15, 20, 25, 26, 29, 31
Balancing life projects with ARTLife projects and social aspirationsLarger social expectations and responsibility to fulfill norms and expectations influence treatment behavior positively or negatively12, 27
Desire for marriage and procreationMarriage and procreation among the important life projects; child wish in order to maintain social status; ART enables procreation
Fear and rumours that ART leads to infertility, impotency and impossibility of sexual activity
06, 08, 09, 10, 12, 17, 19, 27, 31
Family responsibilityWanting to live to provide financial stability to the family; feeling responsibility to secure an income and take care for the children02, 03, 07, 11, 15, 20, 21, 31
Reconfiguring a desired body imageChanging health statusPhysical recovery can motivate adherence, or make people believe that they are better already and therefore push them to interrupt treatment; having a very low CD4 count can be a driver for adherence, or lead to drop-out07, 11, 15, 21, 24, 30, 31
Maintaining a desirable body imageRegaining life, strength and sexual power through ARV’s
Body image and attractiveness compromised by the disease OR treatment side effects (gain/loss of weight, changes in the quality of the skin), affecting adherence positively or negatively
04, 08
Social interaction: Negotiating entitlements and obligationsNature of relationship defines level of support/ obligationFamily support, support by partner/spouse Denial of supportSpouses and family members are the most important provider of emotional, practical and material support; difference in level of support comparing partner/spouse to relatives with children and parents providing more stable relationships than marriage/partnership.
Support may be denied by spouse or relatives due to the stigma associated with HIV and ART; family conflicts may result in lack of support
03, 04, 05, 07, 09, 11, 14, 15, 16, 19, 20, 21, 24, 31
Male power and intra-household decision-makingWomen are often not entitled to decide over the use of household resources. Men can deny access to ARV’s to their wives; women fear male rejection and end of marriage; men may also be abandoned by wife; female preference to disclose to female family, and not to sexual partner; women often blamed to ‘take the virus home’ leading to non-disclosure or violence; weighing the situation before disclosing in order to avoid a bad name; men feel threatened in their gender role as family providers01,02, 11, 15, 17, 18, 19, 20, 24, 27
Negotiating material support and reciprocity in wider social networkMaterial support and reciprocity LoansImportant role of informal social networks, neighbors, etc. for covering expenses that are not financed by relatives/spouse; provision of loans which must be paid back, usually with interest; having received support creates obligations on the part of the patients to become healthy in order to reciprocate and preserve one’s ‘social capital’, promoting adherence
Material support depends on the nature of the bonds before the illness, and on how the individual had contributed to the relationship
01, 02, 18, 30
Complementing social network with New ART related networksTreatment support groups, treatment supportersTreatment supporters/confidants, neighbours, or patient groups can have an important supporting role, especially where stigma is high (rejection of ‘self stigmatisation’, or felt stigma); they help to strengthen self-efficacy; the new forms of social relationships are different from family relationships01, 03, 19, 20, 21, 22, 24, 25, 30
Precarious livelihoods: Negotiating individual versus collective needsBalancing material and health needs and adjusting coping strategiesIndirect costsTransportation costs, availability, cost of opportunity, time off work generating loss of income, and laboratory tests costs01, 02, 07, 11, 12, 14, 15, 17, 20, 21, 23, 24, 28, 29, 30, 31
Coping by reducing costs or raising fundsPostponing treatment until severe health decline; alternating between one and two doses a day to stretch a prescription; sharing drugs among several infected family members; cutting on food to save for treatment; choosing who to treat within a family; loans borrowed from family and social networks, sometimes with interest; work to repay loans, or handouts not expected to be repaid; sharing transport costs between neighbours; selling assets01, 07, 14, 24, 28, 30
Competing needsLack of food, especially when hunger as side-effect appears; need for special ‘good’ food or special dietary supplements; no ART without food, resulting in reducing the dosage to once a day or skipping full days; wish to save and invest for future by purchasing land or housing; problems with paying school fees and books; on the other hand sometimes priority for treatment above everything else, including family needs01, 07, 12, 14, 15, 16, 17, 18, 19, 25, 21, 24, 25, 26, 28, 29, 30, 31
Social triageWithin a family or group, ARV’s will be allocated to the provider/the person considered the most important provider first; men have role of providers, therefore lack of ARVs for men is said to result in poverty for the whole family; sometimes only the man takes ARVs because of lack of money to buy for wife; strategic forms of social triage have been used in the past, prioritizing the most strategic person instead of most vulnerable person01, 07, 22
Morality and stigma; Coping with social exclusionMorality and fear of pollution create stigmaImmoral behaviorRejection and exclusion of a person by society and his/her own family; HIV can be seen as an agent of pollution brought on by immoral behavior; PLWH can be accused of spreading the disease; women sometimes fear being stigmatized as prostitute or being accused of adultery; self-stigmati zation related to immoral behavior, independent of enacted stigmatization01, 08, 14, 15, 16, 18, 21, 25
Physical signs of sickness as stigmataPeople with visible signs of sickness associated with HIV/AIDS may be avoided by others who fear contamination and pollution from those called ‘ dead already’04, 05, 09
Proving maternal responsibility decreases impact of stigmatizationMothers’ commitment for their children can contribute to restore a positive image after disclosure, which reduces AIDS stigma. This hasn’t been described for men03
Structural inequality and social exclusionThe disease of povertyHIV is seen as disease of poverty; loss of symbolic capital equals loss of social capital and thus endangers access to social support; hence, failure to disclose HIV status can be due to a need to avoid being considered poor01
Legitimacy to invest in death-prone individualExtra needs must be justified; short-term individual needs are balanced with long-term needs to sustain the family livelihoods; investments only legitimate if cure is expected (health status modifies readiness to invest in treatment), pertinent to poverty; fear of economic burden for family members; individual being accused of ‘not bringing money, only disease’, to their homes, thus increasing the overall economic burden01, 20, 25


Why (not) adhere: (Dis-) trust in the health sector

Motivational barriers to adhere – in contrast to the ability to adhere, which is mainly influenced by a cost dimension – relate to three main themes: the performance of health services, concepts of health and the body and a person’s aspirations or life projects.

Experiences with low-quality services and access constraints can undermine trust in health care services. Long waiting lines, lack of confidentiality and privacy, drug stock-outs and inconsistency of organizational systems within the clinic are common experiences (00, 01, 02, 05, 11, 14, 15, 16, 17, 20, 23, 24, 25, 26)1,2 Providers’ attitudes, the time they spend with patients (16)3 and low communication skills can jeopardize adherence to ART (00, 01, 08, 14, 16, 20, 25, 29).4 Programme requirements that imply disclosure, like joining a support group, accepting home visits, partner notification, or picking up drugs from special ARV outlets, may impede adherence when programme support cannot offset stigma (05, 15).5 Side effects, in particular, may be difficult to bear for a patient if they are not well explained and understood (02, 14, 15, 20, 26).6 Alternatively, the perceived consistency of a provider and the development of a confidential relationship facilitate treatment (02, 14, 20, 31).7

Historically based collective experiences beyond the quality of care affect the perception of services as well. Fears that treatment at low cost would not be available forever reflect prior experiences with international funding initiatives and global and local market forces and may evoke suspicions of unequal treatment (00, 10, 13, 17, 20, 26).8 In South Africa, medication side effects sometimes evoked fears of being poisoned, reflecting the country’s political history: ‘If you go to hospital, you will die; you will die. They should tell you that it is caused by witchcraft and send you home for the right treatment. If they don’t and instead give you an injection, you’re dead.’ (Unknown, South Africa) (Golooba-Mutebi & Tollman 2007).

In studies from Zambia and Malawi, too, a sense of not being in control is reflected in the suspicions of ‘medicine from whites’ (26) or that AIDS is a ‘government’s disease’ (13), assimilated by patients into a fear-based rationale that taking ART will bring ‘certain death’.

In addition to being distrusted, Western medicine does not address underlying spiritual and social dimensions of illness, which may partly explain why patients on ART commonly complement antiretrovirals with traditional medicine (02, 13, 14, 10).9 Even if the route of transmission in the individual is understood (sexual), an HIV infection may be seen as the ultimate consequence of a ‘bewitchment’ (01, 10, 13, 14, 16, 24).10

Positive motivations: Continuing life projects and perceived self-efficacy to succeed

Few motivations for adherence to ART were explicitly mentioned in the literature (31). Maintaining a positive identity and social role, such as having a family, and hoping to live as long as possible to provide for one’s children are a strong motivation for adherence (02, 06, 07, 08, 09, 11, 21, 31)11: ‘That what’s really inspired me to come around, because I have to think of my family now…I will still be living for them. That’s when I decided to come… back to medication.’ (36-year-old father, public sector, Botswana) (Nam et al. 2008).

But enacted or anticipated HIV-related stigma can constrain a person’s life plans – social aspirations and trajectories linked to identity. An HIV-positive woman narrates the reactions she experienced: ‘You scare people so much that they take you out of their thoughts. They think you are dead even before your time, and count you among those who died long ago. …. [My sister, not knowing my status,] said ‘I used to think you were infected. I had taken you out of all my plans’.’ (HIV-positive woman, Uganda) (Byakika-Tusiime et al. 2009).

If stigma is anticipated, disclosure is avoided to pursue future plans that involve others, like being able to marry (27).12 An intact image of the body (weight, smooth skin) must be maintained to avoid suspicions of HIV (04, 08).13 If weight loss or skin afflictions occur while on ART, medication may be stopped to avoid involuntary disclosure.

The motivation to adhere further depends on a person’s ability to cope with the personal crisis an HIV diagnosis may generate. The acceptance of an HIV diagnosis followed by developing a new, positive image of the ‘self’, reinforced by the participation in support groups, has been described as crucial for the successful management of ART (02, 21, 25, 31).14 Some people are going through a ‘conversion-like process’ or ‘rebirth’ to change their identity and lead a more ‘moral’ life (22, 31).15 Overall, the balancing of positive and negative factors at the individual and social level therefore needs to be positively skewed towards adherence (Figure 2). For example, feelings of inferiority, depression and social exclusion (11, 15, 18, 21, 24),16 must be offset by feeling in control (02, 11, 15, 18, 20, 21, 24, 28, 29, 30, 31)17 and obtaining social support through kin, treatment support groups, or faith-based organizations (01, 03, 04, 05, 07, 09, 11, 15, 14, 16, 19, 20, 21, 22, 24, 25, 27, 28, 30, 31).18

Figure 2.

 Third order construct of nested relationships of adherence.

How to adhere: a question of costs and competing livelihood needs

Adherence patterns not only reflect one’s willingness to take medication regularly but also the ability to do so (28).19 In settings characterized by resource scarcity, direct cost of drugs (02, 07, 16)20 or indirect treatment costs such as transport (02, 11, 12, 13, 14, 15, 17, 19, 20, 24, 25, 28, 30, 31)21 are important barriers to adherence. Rural patients who must travel great distances to the clinic are often unable to raise the money required for transport, especially when dispensaries require that patients pick up the medication themselves (19, 23)22: ‘I have many people in the village. They are dying because they don’t have money to transport themselves to the hospital. You need to have this money monthly.’ (ARV user, Uganda) (Kgatlwane et al. 2006).

Because of the chronic nature of the condition, adherence costs compete with short and long-term livelihood needs. The loss of opportunity to invest in assets relevant for future livelihood sustainability such as land (07, 28)23 and time off work resulting in a loss of income or risk of losing one’s job (02, 11, 17, 28)24 are weighted against the expected benefit of ART. When debts are incurred, interest is often high even if funds are borrowed from family and social networks (07, 24, 28, 30).25 Only a few studies mention contributions received by patients that were not expected to be repaid (30, 07), or the sharing of transport costs between neighbours (28),26 but they also note concomitant concern about the sustainability of such contributions (01,18)27: ‘This time there are about three people or four people who contributed so that I may buy the medication… because by myself I can’t afford it. I don’t know whether in the coming months I will be able to buy it.’ (Female ARV user, Uganda) (Crane et al. 2006).

Treatment costs also compete with basic needs such as food (01, 07, 12, 17, 18, 19, 24, 25, 28, 29, 30, 31).28 Cutting down on food to save for treatment is reported (07, 30),29 but other studies reveal dosage reduction or skipping in the absence of ‘good’ or ‘special’ food (16, 17, 25, 29)30: ‘I missed on Sunday because I could not find what to eat and yet I had to eat first before taking the tablets.’ (Female ARV user, Uganda) (Nakiyemba et al. 2006). The increased appetite ARVs may induce as a side effect adds to the problem (14, 20, 26, 30).31

Competing needs may open moral dilemmas, where health-related decision-making interacts with a person’s aims in life (25),32 such as being a responsible parent and provider (02, 07, 28, 30)33: ‘I wouldn’t think about it so much if I didn’t have to think about the word ‘transport.’… I think it is very important to come here for the medicine, but then I can’t refuse to buy books for my child so as to spare transport to come.’ (Female ARV user, Uganda) (Tuller et al. 2009). Another man said it more explicitly: ‘The compromising nature of the treatment is very difficult to adhere to. You have to sacrifice others for the sake of yourself.’ (Male ARV user, Uganda) (Nakiyemba et al. 2006). Decision-making and power-relations within households shape these decisions. Competing needs can have negative health consequences if priorities for the allocation of resources must be set within a family (07, 28).34

Mediating how and why: manipulating relationships and social dilemmas

Social networks play a central role in mobilizing material and personal resources for treatment, especially in the absence of formal social security and health insurance (01, 02, 03, 04, 05, 07, 09, 11, 17, 19, 21, 24, 28, 30, 31).35 Disclosure to a confident or a ‘treatment supporter’ is a prerequisite to mobilizing support to achieve optimal adherence, by helping to reorganize one’s daily activities to include treatment (02, 19, 24, 11)36 or by helping to mobilize the resources required (01, 02, 03, 04, 05, 07, 09, 11, 17, 19, 21, 24, 28, 30, 31).37 Yet disclosure has insecure consequences. Fears of being expelled from the family because of involuntary revelation may jeopardize the motivation to adhere (11, 18, 19, 25).38 As some economically dependent women put it, they ‘would rather die than lose the marriage’ (01, 11, 15, 18, 19, 22, 24, 30)39 or risk domestic violence (18).40 Men, on the other hand, expressed fear of not being cared for in case of sickness as a reason for non-disclosure: ‘If you don’t have a partner you are in trouble. The partner can take care of you when you are sick. The parents may say you only know them when you are sick …. They deny you and offer you negative support.’ (Male ARV user, Botswana) (Kgatlwane et al. 2006).

The fear of losing support by a spouse or close kin is substantiated by many accounts in the articles: ‘In my case, I stay with my father and mother. We are not on good talking terms because of this disease I have. My father told me that your friends bring money, but you have only brought this disease. So, even when there is food, they do not give me. Most of the time, I feel like I am confused, due to thinking too much. As a result I feel it is better to die because there is no one to care for me.’ (29-year-old, female ARV user, Zambia) (Sanjobo 2008).

Family members’ reactions are partly rooted in the fear of the economic burden implied by having to look after a sick person and the surviving children (01, 19, 25).41 Supporting ART adherence of a relative means investing valuable time, energy and financial resources into a single individual. This is only justifiable when there is an expectation of recovery. Investments thus create an obligation for the positive person to adhere to ART to become healthy and able to reciprocate investments (30).42

Synthesis and discussion: line of argument synthesis

Uncertainty of decision-making: evaluating expected outcomes

Experiences with ART are shaped by social interactions, which are often difficult to anticipate for an HIV-positive person. In other words, the effect of a possible determinant of adherence is modified by social processes, which are beyond individual control. For example, someone might want to disclose a positive HIV status to the family in order to get emotional and material support for treatment. Instead disclosure may incur stigma and rejection as described in southern Africa. Similarly, for example, home visits can be supporting adherence or may lead to involuntary disclosure and social exclusion. An individual must balance the expected outcomes of engaging in ART programmes with unknown collateral results making the prediction of desired outcomes difficult (Figure 2).

Socially delineated decision-making: poverty, power and stigma

Individual decision-making, however, is socially delineated and modulated by power-relations. Poverty, gender inequality and stigma are important modulators of individual adherence as they affect a person’s ability to access resources (Figure 2). The more powerful individuals define the entitlements and options of the weaker. They also set the moral norms and make compliance with social norms and expectations a requisite for the dependant (Parker & Aggleton 2003). HIV may be seen as a sign of antisocial or immoral behaviour, with a particular group, such as prostitutes, threatening the whole community, or as a disease of marginalized poor (Farmer 2004, 2006). People living in precarious conditions of poverty are likely to avoid any action that produces further marginalization and may thus prioritize their social network over investments in health. Non-disclosure and denial can therefore be understood as a long-term strategy to maintain a social network and support system (Figure 2). Particularly in SSA, where poverty is endemic and formal social security systems are absent, people have no other choice than to invest in social networks to be able to cope with hardships. As it has been reported that kinship networks have weakened over the last decades, people need to increase even more their efforts to maintain important relationships through constant investment and renegotiations (Benda-Beckmann 2004).

Negotiating personhood and social relationships with treatment

The juxtaposition of individual agency and social dependencies sheds a new light on culturally shaped attitudes towards ART, like the persistence of traditional medical and spiritual practices despite available science-based health information in the SSA context.

The use of the Western concept of the patient – an individual predominantly interested in restoring or maintaining physical health – tends to underestimate the importance of the social and spiritual dimensions of illness for health seeking (Figure 2). If social relationships must be reciprocal, revealing an illness with unclear outcome means imposing a burden on others. Disclosure may be understood as a (sometimes unwelcome) request for material involvement rather than an attempt to establish a trusting relationship for emotional support (Adams et al. 2005). In precarious conditions, a sick person may thus be perceived as a threat to the family’s material basis, endangering the group as a whole and sometimes leading to the expulsion of HIV-positive persons from their own homes (Alliance 2004). In societies where the self is experienced in terms of inherent connection with others (high interrelatedness), the fears an HIV-positive person can generate in others might be more expressed (Bongmba 2001).

Western medicine conflicts with these social bonds by targeting the individual alone. In traditional medicine, however, where diagnosis and treatment are primarily understood in social terms, a perceived threat can be addressed collectively. The suspicion of witchcraft or malevolent ancestral spirits, for example, can serve to examine and re-establish relations with others (Evans-Pritchard 1976; Allen 2007; Braude 2009). In the case of an HIV infection, it may help to relocate guilt from the individual to external forces, allowing the affected person to re-integrate into the community.

More generally, traditional medical concepts may enable an individual to negotiate illness beyond the individual body (Lock & Nichter 2002), in contrast to Western medicine. Especially where trust in the quality of Western medicine is historically as weak as in many African settings, accepting its knowledge claims as superior to traditional medicine is not self-evident. If a patient does not see any immediate benefit from the treatment, why invest to go back to the clinic? And if Western medicine is only partly accountable and effective, why not divide investments with traditional medicine which can at least address the underlying causes (Figure 3)?

Figure 3.

 Tuning adherence from an individual perspective.

Adding the ‘social’ to the ‘medication and moral career’

Decision-making related to ART is influenced by experiences with the health system, the ability to create a new, positive identity to include ART into one’s everyday life and the ability to negotiate positively one’s social relations, offsetting stigma. Social support is important for negotiating a ‘new’, positive identity as a PLWH and to counter stress-induced mental illness. From a biomedical perspective, the patient is seen as the recipient and performer of treatment regimens with the primary goal to restore or maintain physical health. The shortcomings of this perspective concern other dimensions of health: mental, spiritual and social well-being (Figure 2).

Similar to the findings of other meta-ethnographies on medication adherence in an industrialized context (Garside et al. 2008; Malpass et al. 2009), the results show that managing medication combines a ‘medication career’, including the practicalities of managing daily treatment (how to adhere), with a ‘moral career’ whereby meaning is created (why to adhere) (Malpass et al. 2009). In the context of ART in SSA, however, a ‘social career’ needs to be added as the third mediating dimension. Indeed, the social dimension is relevant for both willingness and ability to adhere to ART. Themes such as disclosure, discrimination, or life projects and their relation with medication adherence are substantially influenced by the social dimension.


In this article, the assumption of an autonomous individual maximizing physical health as first priority is refuted. Depending on levels of poverty, social support is a premise for long-term survival and health, making investments in social networks vital. Individuals must balance physical health with social integrity, sometimes by referring to traditional medicine. Yet, the mobilization of support has become more difficult because of weakening and overburdened social and family networks, turning a person in need of support into a threat to the group. Fear of exclusion, therefore, results in selective non-disclosure, which negatively affects adherence. Throughout the ‘medication’, ‘moral’ (Malpass et al. 2009) and social career, a person must evaluate expected outcomes on various levels and balance beneficial versus harmful effects of adhering to ART on their physical, mental and social well-being.


  • 1

    Please refer to the numbers in Table 1: ‘Study characteristics of included studies’.

  • 2

    Data were collected in the following countries: Botswana, Malawi, South Africa, Tanzania, Uganda, Zambia.

  • 3

    South Africa.

  • 4

    Kenya, South Africa, Tanzania, Uganda, Zambia.

  • 5

    Botswana, Malawi.

  • 6

    Botswana, Malawi, South Africa, Tanzania, Uganda, Zambia.

  • 7

    South Africa, Tanzania, Uganda.

  • 8

    Malawi, South Africa, Tanzania, Uganda, Zambia.

  • 9

    Malawi, South Africa, Tanzania.

  • 10

    Malawi, South Africa, Tanzania, Zambia.

  • 11

    Botswana, South Africa, Tanzania, Uganda, Zambia.

  • 12


  • 13

    Tanzania, Uganda.

  • 14

    Botswana, South Africa, Tanzania, Zambia.

  • 15

    Tanzania, West Africa.

  • 16

    Botswana, Tanzania, Zambia.

  • 17

    Botswana, Kenya, Nigeria, South Africa, Tanzania, Uganda, Zambia.

  • 18

    Botswana, Burkina Faso, Ivory Coast, Malawi, Mali, Nigeria, South Africa, Tanzania, Uganda, Zambia.

  • 19


  • 20

    South Africa, Uganda.

  • 21

    Botswana, Ethiopia, Malawi, Nigeria, South Africa, Tanzania, Uganda, Zambia.

  • 22

    South Africa, Uganda.

  • 23


  • 24

    South Africa, Tanzania, Uganda, Zambia.

  • 25

    Nigeria, Tanzania, Uganda.

  • 26


  • 27


  • 28

    Ethiopia, Kenya, South Africa, Tanzania, Uganda, Zambia.

  • 29

    Nigeria, Tanzania, Uganda.

  • 30

    Kenya, South Africa, Tanzania, Zambia.

  • 31

    Nigeria, Tanzania, Uganda, Zambia.

  • 32


  • 33

    Nigeria, South Africa, Tanzania, Uganda.

  • 34


  • 35

    Botswana, Burkina Faso, Malawi, South Africa, Tanzania, Uganda, Zambia.

  • 36

    South Africa, Zambia.

  • 37

    Botswana, Burkina Faso, Malawi, Nigeria, South Africa, Tanzania, Uganda, Zambia.

  • 38

    South Africa, Zambia.

  • 39

    Botswana, South Africa, Tanzania, Western Africa, Zambia.

  • 40


  • 41

    South Africa, Zambia.

  • 42

    Nigeria, Tanzania, Uganda.

Conflicts of interest

The authors have declared that they have no conflicts of interest.