Philippebourgois.net

AIDS and Behavior ( C 2006)DOI: 10.1007/s10461-006-9080-z The Price of Adherence: Qualitative Findings From HIV
Positive Individuals Purchasing Fixed-Dose Combination
Generic HIV Antiretroviral Therapy in Kampala, Uganda
J. T. Crane,A. Kawuma,J. H. Oyugi,J. T. Byakika,A. Moss,P. Bourgois,and D. R. Bangsberg
Contrary to early expectations, recent studies have shown near-perfect adherence to HIVantiretrovirals in sub-Saharan Africa We conducted qualitative interviews with patients pur-chasing low-cost, generic antiretroviral therapy to better understand the social dynamics un-derlying these findings. We found that concerns for family well-being motivate adherence,yet, the financial sacrifices necessary to secure therapy may paradoxically undermine familywelfare. We suggest that missed doses may be more due to a failure to access medicationrather than a failure to adhere to medications, and that structural rather than behavioral in-terventions may be most useful to insure optimal treatment response.
KEY WORDS: adherence; HIV; generic fixed-dose combination antiretroviral therapy; Africa; ethnog-
raphy.
INTRODUCTION
in several southern African countries (Laurentet al., Orrell, Bangsberg, Badri, and Wood, Early debates over public health responses Oyugi et al., including findings from to AIDS in sub-Saharan Africa assumed poor ad- our own group showing mean adherence rates of herence to HIV antiretroviral (ARV) medications 91-94% in an urban Ugandan cohort purchasing in impoverished countries (Stevens, Kaye, and self-administered, generic ARV therapy (Oyugi Corrah, Recent studies, however, have mea- et al., Given that adherence in the industrial- sured antiretroviral adherence rates of over 90% ized world-where treatment is often free-is only70%, these findings were surprising (Bangsberg andDeeks, 1Department of Anthropology, History and Social Medicine, Although sponsored programs providing free ARVs are slowly increasing, many individuals in Epidemiology and Prevention Interventions Center, Division ofInfectious Diseases, San Francisco General Hospital, UCSF, San Uganda directly purchase ARVs on a monthly or twice-monthly basis from local pharmacies and clin- 3Academic Alliance for AIDS Care and Prevention, Kampala, ics. Patients may opt to purchase drugs because free treatment is not available locally or because 4Makerere University School of Medicine, Kampala, Uganda.
5 they do not fit the eligibility criteria for free treat- Department of Epidemiology and Biostatistics, UCSF, SanFrancisco, California.
ment programs, which often require that patients are 6Positive Health Program, San Francisco General Hospital, treatment-na¨ive. Among the least expensive ARV therapies available in the country is a fixed-dose com- 7Correspondence should be directed to D. R. Bangsberg, Uni- bination of stavudine, lamivudine, and nevirapine versity of California at San Francisco, Epidemiology and Preven- manufactured by Cipla (Mumbai, India) under the tion Interventions Center, SFGH Building, 100 Room 301, POBox 1347, San Francisco, California 94143; e-mail: db@epi-center.
brand name Triomune R. Although inexpensive by Western standards (US$27 a month in 2003), the cost 1090-7165/06 C 2006 Springer Science+Business Media, Inc.
Crane, Kawuma, Oyugi, Byakika, Moss, Bourgois, and Bangsberg
of these medications approaches the median monthly income in Uganda (US$30). In an effort to better un-derstand the social dynamics of the incentives and barriers to adherence to self-pay antiretroviral ther- prospective cohort of 97 individuals. Participants in apy, we conducted in-depth qualitative interviews the qualitative sample were similar in age, sex, in- with a subset of Ugandan patients purchasing generic come, household size, employment, income, mean antiretroviral therapy followed in a prospective co- baseline log viral load, baseline CD4 cell count, and adherence compared to the remaining cohort(Table Participants were recruited from a longitudi- The Financial Burden of Antiretroviral Treatment
nal study measuring Triomune R adherence throughmonthly electronic medication monitor and unan- Consistent with the findings of several groups nounced pill count adherence assessments conducted indicating high levels of adherence (Laurent et al., at patients' homes (Oyugi et al., Individuals Orrell et al., Oyugi et al., partici- in the parent cohort were consecutively recruited pants rarely reported missing a dose of antiretroviral from pharmacies upon receipt of their first antiretro- medication during qualitative interviews. However, viral medication. All patients were treatment-na¨ive they described this excellent adherence as the prod- upon initiating therapy and were purchasing their uct of a constant battle to overcome the barrier of therapy themselves or with the assistance of an ex- drug cost. Participants routinely named the price of tended family network. In recruiting individuals for medication (rather than side effects, stigma, or in- qualitative interview, an effort was made to select a convenience) as the principal challenge to sustaining subsample reflecting the sociodemographic and clini- treatment, a finding consistent with those reported by cal characteristics of the larger cohort. Both English- speakers and non-English speakers were recruited.
who found that financial sacrifice is the most impor- Interviews not conducted in English were conducted tant barrier to sustained adherence to treatment.
in Luganda, the local language common to the Participants described purchasing and adhering to their antiretroviral regimen as a major life pri- Participants were recruited for qualitative in- ority. One single, working mother of two described terviews by the second author, who is also a home Triomune as the first thing she bought after being visitor for the longitudinal study and had estab- paid each month. She described the medicine as "the lished rapport and trust with participants. Interviews most important thing in my life right now." The were conducted with informed consent at partici-pants' homes by a pair of researchers, one Ugandanand one American. In all cases the Ugandan inter- Characteristics of the Qualitative Subsample and Re- viewer was one of two home visitors working with the longitudinal study, and therefore was known to the participant. Participants were asked to de- scribe how they initiated antiretroviral medication and the impact of HIV therapy on their health, their household finances, and their family. Interviews were coded according to theme, and analyzed using an inductive method (Miles and Huberman, Results reported here on adherence and medica- tion cost reflect "theoretical saturation," meaning these themes emerged consistently in the data and were redundant across interviews (Glaser and Straus, These results are presented in narrative form in order to highlight the complex social dynamics of medication access and adherence described by The Price of Adherence to Self-Pay Antiretroviral Therapy in Uganda
strong desire to adhere despite medication cost led In addition to contributing cash resources to- to rationing in the case of one participant who was ward the purchase of antiretrovirals, participants' unable to afford her next bottle of pills, and told a family members also suffered austerity measures in staff researcher that she was alternating between one order to support antiretroviral treatment. Sometimes dose a day and two doses daily in order to stretch a this involved cutting back on basic needs such as prescription because she did not want to go a whole What I know is that my people don't get enough to The financial sacrifices required to purchase make them satisfied. We are eight people and eat antiretrovirals encouraged participants to postpone 1 kg of maize flour. That is what I can afford. We therapy until they experienced a rapid decline in have one main meal daily. Yet a person is supposed their health or received a doctor's warning that to eat at least twice-that is, lunch and supper.
they would soon die without treatment. Their In another example, one participant's mother dramatic improvement, however, was often accom- sold a goat in order to help pay for her daughter's panied by new worries about the long-term sustain- second month of medication. Paying for HIV ther- ability of purchasing antiretrovirals. Three partici- apy also impeded participants' abilities to pay for pants described being unable to cover the full cost their children to attend school. A father of six, un- of the medication themselves, and so relied on as- employed due to illness, complained bitterly about sistance from family members and extended kin net- his inability to finance his children's education: works to purchase ARV medications. One woman,a widow with five children, purchased her pills us- At times I find things not working out. I fail to get ing money sent from an in-law working in Europe.
the money for school fees and they stay home with- Another woman was able to start therapy because out studying. Sometimes they miss a whole termwithout studying, or study for only 2 weeks during her grandmother paid for her medications, but was the beginning of the term. They are made to repeat forced to stop when her grandmother died.
classes but what can we do? Even if they repeat I still can't afford to pay for the whole year. They just worried about the sustainability of her family's fi- nancial assistance, who were also struggling to sup-port the eight children orphaned by the sudden (non-AIDS) death of her sister a month earlier: The Benefits and Burdens of High Adherence
This time there are about three people or four peo- When asked why they did not miss doses, many ple who contributed so that I may buy the medica- participants responded simply, "I want to live." tion, which means that it's a problem, because bymyself I can't afford it. I don't know whether in the While all participants in industrialized and develop- coming months I will be able to buy it. Because if this ing countries alike presumably take therapy to avoid is the second month, but three people contributed succumbing to AIDS, most Ugandans begin ther- for that 50 thousand shillings, you never know how apy at extremely advanced stages of disease (median CD4 68 in those we interviewed). Thus, it is possible In addition, many families had several HIV- that Ugandan participants correctly perceive them- infected members in need of treatment, further com- selves as close to death, and this proximity to death pounding the financial strains of therapy and forcing may drive near-perfect adherence. One participant families to choose whom they could afford to treat: described seeking out ARVs after doctors told herthat she needed to "look for ways of saving [her] I face problems, because sometimes the Triomune gets finished when I do not have the money, yet Ineed it. For example what I have right now is about Before I started taking Triomune I was not ok. I to get finished and I am already worried about where was very, very weak, I had diarrhea that had be- to get the money. I have no hope of where I will get come almost permanent, and fever every evening.
it . As you see, I have one pair of trousers and one My skin was very, very bad and I was scratching my- shirt but I cannot buy any other because any money self the whole night at times not sleeping. I had loss I get is used to buy Triomune . . It is only I who of weight, no appetite, vomiting. But I was lucky; use it, though it would be good if my wife used it.
I never got TB or pneumonia. When I went to the I do not know what you doctors say, but we cannot clinic, they tested my CD4 and they told me that it manage. She does not take it because I cannot buy was very low. They told me I had to look for ways for saving my life . Since I started taking Triomune Crane, Kawuma, Oyugi, Byakika, Moss, Bourgois, and Bangsberg
there is a lot of improvement, because now I feel their families, educating their children, saving very comfortable and safe, and I eat well, I sleep money, or investing in the future. One woman told well, no more diarrhea, and the skin is clearing. I am us, "Now I can't save anything. That money which I now becoming like a normal person, and I feel like I would save I now use to pay for the medicine. I have can now go back and work and lead a normal life.
to buy the medication because I want to be living." In addition, participants' desire to live was Thus, tragically, the very thing that motivates partic- strongly motivated by their obligations to kin, specif- ipants to adhere to ARVs-providing for their fami- ically by their concerns about the welfare of surviv- lies' future through the purchase of land or a house to ing family members. One father told us, "I have to pass on-is subverted by the financial imperative of take Triomune because I still want to live so that I paying for the medicine. In the words of one mother may take care of my children. This is only possible if I take this medicine." A mother of a young boy spokemore generally, saying, "Most people they have fam- I don't have any other plan. I only think about the ilies to look after, they have dependents and they medicine. I cannot say that I may buy a plot of landand build a house. I cannot.
have to live to look after those people. You haveto be healthy, you have to provide for children andthe dependents . So you have to keep strong andhealthy." DISCUSSION
Several participants hoped that ARV therapy could provide them with a few more years of life in Our qualitative inquiry suggests two observa- which they might be able to see their children into tions. First, participants described the duty to con- adolescence or adulthood, and provide them with a tinue caring and providing for their families as one degree of financial stability. A middle-aged man with of the major motivators for their high rates of ad- several financially dependent children and grandchil- herence. Yet, at the same time, the cost of medica- dren and an HIV-positive wife described his feelings: tion exacted a serious toll on family welfare and ba-sic needs. Thus, self-paid ARV treatment may force When I look at the family, where I am going to leave participants to undermine the very thing that moti- them, hm? I am not sure of their future. So I believeif I can still live some few years, I can push them and vates them to take the medication. This dilemma ex- emplifies what Whyte and colleagues have describedas the social (as opposed to pharmacological) activ- This man was the sole breadwinner for his fam- ity of HIV medications in Uganda (Whyte, Whyte, ily, and thus the possibility of his death implied broad repercussions. But even participants who were un- vent of programs offering free ARVs has significantly employed and depending on others for their sup- increased the availability of these drugs in Uganda port expressed a great desire to establish financial over the last 6 months, many such programs have el- security for their children before they died. One igibility criteria, which either exclude or deprioritize young woman who had been without work for a year participants who are not treatment-na¨ive. This means pleaded with the interviewers to let her know of any that participants who began buying medication job possibilities so that she might earn some money before free treatment became available, such as to build a house to leave for her children: those described here, may face barriers in attemptingto switch over to subsidized medication. Better ac- If I can get a job and work so that I can get somemoney to look after me and my children, at least I cess to free ARV treatment for both treatment-na¨ive could do for my children something in the future, and treatment-experienced family members would you never know . Because since now I am sick you avoid putting participants and their families in such never know when I might be down, when I might die.
But at least if you have got something where you get Secondly, in the rare instances when participants some money, you can build for them a small house.
did miss doses, it was most often due to an inability Providing for the welfare of their families, and to afford the pills. This suggests that missed doses in especially children, was a major motivator for partic- impoverished settings may be understood best as an ipants to continue therapy. Yet, at the same time, the issue of access, versus adherence to medication. The financial sacrifices required to sustain treatment of- distinction is more than semantic, because a problem ten precluded participants from adequately feeding with adherence suggests the need for intervention, The Price of Adherence to Self-Pay Antiretroviral Therapy in Uganda
whereas a problem of access suggests that efforts cation. Therefore, missed doses may be more due to would be best targeted toward providing a reliable a failure to access medication rather than a failure to supply of free treatment. Some have suggested adhere to medications and that structural rather than that the roll-out of antiretroviral treatment in sub- behavioral interventions may be most useful in insur- Saharan Africa should be modeled on the directly observed therapy (DOT) programs that have beenwidely used to ensure adherence to tuberculosis med-ication (Harries, Nyangulu, Hargreaves, Kaluwa, and ACKNOWLEDGMENTS
Salaniponi, One component of these programsis witnessed dosing (Farmer, Mukherjee, Gupta, The authors would like to thank Mary Kasango Tarter, and Kim, We argue that this type of for her assistance in data collection. This study was intervention is helpful in securing stable supply and supported by the NIMH, NIAAA, NIDA, Univer- distribution of medications, but the daily witnessed sity of California San Francisco-Gladstone Institute dosing component may be unnecessary in popula- of Virology and Immunology Center for AIDS Re- tions that are already highly motivated to adhere search, University of California, Berkeley Fogarty (Liechty and Bangsberg, More appropriately, International AIDS Training Program, Doris Duke resources should be allocated toward reliable drug Charitable Foundation, and University of California, supply, the distribution of free therapy, the training of medical providers to prescribe optimal therapy,and the clinical and laboratory infrastructure needed REFERENCES
to support the increasing numbers of participantson ARVs.
Bangsberg, D. R., and Deeks, S. G. (2002). Is average adherence There are a number of limitations to this study.
to HIV antiretroviral therapy enough? Journal of General In- The small sample size, absence of formal random se- ternal Medicine, 17(10), 812-813.
lection from the parent study, and urban location of Byakika-Tusiime, J., Oyugi, J. H., Tumwikirize, W. A., Katabira, E. T., Mugyenyi, P. N., and Bangsberg, D. R. (2005). Ad- the cohort limits the generalizability of our findings.
herence to HIV antiretroviral therapy in HIV+ Ugandan Participation in the larger longitudinal study may patients purchasing therapy. Int. J. STD. AIDS. 16(1), 38- have had a positive effect on participants' adherence Farmer, P., Leandre, F., Mukherjee, J., Gupta, R., Tarter, L., to medication, and possibly biased their responses and Kim, J. Y. (2001). Community-based treatment of ad- during the qualitative interview. However, the longi- vanced HIV disease: Introducing DOT-HAART (directly tudinal study found a very high correlation between observed therapy with highly active antiretroviral therapy).
Bulletin of the World Health Organization, 79(12), 1145- self-reported and objectively measured adherence rates in this cohort, and for this reason we feel that Glaser, B., and Straus, A. (1967). The discovery of grounded participants' descriptions of their adherence patterns are likely to be accurate. Participation in the longi- Harries, A. D., Nyangulu, D. S., Hargreaves, N. J., Kaluwa, O., and tudinal study also may have had a favorable impact Salaniponi, F. M. (2001). Preventing antiretroviral anarchy in on the qualitative study, as the rapport and trust pre- sub-Saharan Africa. Lancet, 358(9279), 410-414.
Laurent, C., Kouanfack, C., Koulla-Shiro, S., Nkoue, N., viously established with the Ugandan interviewers Bourgeois, A., Calmy, A., Lactuock, B., Nzeusseu, V., during monthly home visits is considered benefi- Mougnutou, R., Peytavin, G., Liegeois, F., Nerrienet, E., cial to the collection of accurate qualitative data.
Tardy, M., Peeters, M., Andrieux-Meyer, I., Zekeng, L.,Kazatchkine, M., Mpoudi-Ngole, E., Delaporte, E. (2004). Ef- Lastly, the willingness of participants to be vis- fectiveness and safety of a generic fixed-dose combination ited at home by a pair of interviewers may of nevirapine, stavudine, and lamivudine in HIV-1-infected have influenced participation and responses to adults in Cameroon: Open-label multicentre trial. Lancet,364(9428), 29-34.
Liechty, C. A., and Bangsberg, D. R. (2003). Doubts about DOT: In summary, Ugandans who purchase antiretro- antiretroviral therapy for resource-poor countries. AIDS, viral medications appear to be highly motivated to Miles, M., and Huberman, A. (1994). Qualitative Data Anal- adhere in order to sustain their health and enable ysis: An Expanded Sourcebook. Thousand Oaks, CA: them to rear their children. However, the high cost of therapy relative to their family incomes both period- Orrell, C., Bangsberg, D., Badri, M., and Wood, R. (2003).
Adherence is not a barrier to delivery of HIV antiretro- ically interrupts access to medications and competes viral therapy in resource-poor countries. AIDS, 17(9), with financing of necessary food, housing, and edu- Crane, Kawuma, Oyugi, Byakika, Moss, Bourgois, and Bangsberg
Oyugi, J. H., Byakika-Tusiime, J., Charlebois, E. D., Kityo, C., Stevens, W., Kaye, S., and Corrah, T. (2004). Antiretroviral ther- Mugerwa, R., Mugyenyi, P., et al. (2004). Multiple validated apy in Africa. Bmj, 328(7434), 280-282.
measures of adherence indicate high levels of adherence Whyte, S., Whyte, M., Meinert, L., and Kyaddondo, B. (2004).
to generic HIV antiretroviral therapy in a resource-limited Treating AIDS: Dilemmas of unequal access in Uganda. Jour- setting. J Acquired Immune Deficiency Syndromes, 36(5), nal of Social Aspects of HIV/AIDS Research Alliance, 1(1),

Source: http://www.philippebourgois.net/AIDS%20and%20Behavior%20Crane%20Adherence%20Uganda%202006.pdf