Objective: HIV counseling and testing (HCT) is a key intervention for HIV/AIDS control, and new strategies have been developed for expanding coverage in developing countries. We compared costs and outcomes of four HCT strategies in Uganda.
Design: A retrospective cohort of 84 323 individuals received HCT at one of four Ugandan HCT programs between June 2003 and September 2005. HCT strategies assessed were stand-alone HCT; hospital-based HCT; household-member HCT; and door-to-door HCT.
Methods: We collected data on client volume, demographics, prior testing and HIV diagnosis from project monitoring systems, and cost data from project accounts and personnel interviews. Strategies were compared in terms of costs and effectiveness at reaching key population groups.
Results: Household-member and door-to-door HCT strategies reached the largest proportion of previously untested individuals (>90% of all clients). Hospital-based HCT diagnosed the greatest proportion of HIV-infected individuals (27% prevalence), followed by stand-alone HCT (19%). Household-member HCT identified the highest percentage of discordant couples; however, this was a small fraction of total clients (<4%). Costs per client (2007 USD) were $19.26 for stand-alone HCT, $11.68 for hospital-based HCT, $13.85 for household-member HCT, and $8.29 for door-to-door-HCT.
Conclusion: All testing strategies had relatively low per client costs. Hospital-based HCT most readily identified HIV-infected individuals eligible for treatment, whereas home-based strategies more efficiently reached populations with low rates of prior testing and HIV-infected people with higher CD4 cell counts. Multiple HCT strategies with different costs and efficiencies can be used to meet the UNAIDS/WHO call for universal HCT access by 2010.
aUS Centers for Disease Control and Prevention (CDC), Global AIDS Program, National Center for HIV, Viral Hepatitis, STD, and TB Prevention, USA
bMacro International Inc, Atlanta, Georgia, USA
cHarvard University, Cambridge, Massachusetts, USA
dCDC-Uganda, Global AIDS Program, National Center for HIV, Viral Hepatitis, STD, and TB Prevention, CDC, and Uganda Virus Research Institute, Entebbe, Uganda
eMulago-Mbarara Teaching Hospitals' Joint AIDS Program (MJAP), Uganda
fMakerere University, School of Public Health, Uganda
gThe AIDS Information Centre (AIC), Uganda
hThe AIDS Support Organization (TASO), Kampala, Uganda
iCDC-Kenya, Coordinating Office for Global Health, CDC, Nairobi, Kenya.
Received 8 August, 2008
Revised 1 November, 2008
Accepted 11 November, 2008
Correspondence to Nick Menzies, Harvard University Health Policy Program, 14 Story Street, Cambridge, MA 02138, USA. Tel: +1 404 217 1076; fax: +1 617 496 2860; e-mail: firstname.lastname@example.org
HIV counseling and testing (HCT) is a key intervention for HIV/AIDS control in developing countries. HCT increases knowledge of HIV status, encourages safer sex, and is an entry point for HIV care and treatment services. Increasing HCT coverage can reduce HIV-associated denial, stigma, and discrimination, and mobilize communities to respond to the HIV epidemic . Uganda was the first African country to offer HCT services in 1990, yet population coverage is still low .
The traditional HCT strategy is client-initiated testing through free-standing clinics (stand-alone HCT) . This strategy has become relatively standardized as HCT programs have grown in number, size, and maturity. However, evolving epidemic dynamics and increased funding for HIV control have led to expanded service provision and targeting of population groups not reached by existing strategies. At the same time, the advent of HIV rapid test technology has simplified HCT implementation . New HCT strategies have been developed, which differ in their target groups and methods of accessing clients. These strategies include provider-initiated HCT offered to patients at health centers and hospitals, and mobile HCT offered in communities and homes. It is likely that these new strategies will differ in their costs and in their ability to contribute to HIV control objectives, such as expanding knowledge of status, identifying infected individuals to receive HIV treatment, and reducing HIV transmission.
Raising knowledge of status can be achieved through increasing the scale of HCT programs and offering services to traditionally underserved groups such as rural communities. Identifying infected individuals for treatment can be achieved through offering services to groups known to have high HIV prevalence. How HCT may reduce HIV transmission is more contentious. Whereas two randomized control trials of HCT [5,6] have reported strong preventive effects, other studies [7–9] have found little or no impact on risk behavior or HIV incidence, particularly in HIV-negative clients. Several meta-analyses [10–12] have found preventive effects to be strongest among HIV-positive clients and discordant couples, and a similar conclusion was reached by a recent meta-analysis of studies of HCT behavioral outcomes conducted in developing countries . Cohort studies following HIV-discordant couples in the Democratic Republic of Congo, Rwanda, and Zambia have found consistently strong beneficial effects of HCT on condom use and HIV incidence [14–17]. To the extent that HCT reduces HIV transmission, it is likely that prevention effects will be strongest when HCT is provided to infected individuals and discordant couples.
Policy makers must prioritize interventions that maximize impact while staying within budget constraints. Prior studies [18–21] have investigated HCT cost-effectiveness at stand-alone sites in developing countries, reporting per client costs from US$13 to US$36. Little data is available on the costs of HCT outside stand-alone sites, or how strategies differ in their success at reaching key population groups. We investigated the costs, outcomes, and cost-effectiveness of four different HCT strategies currently being used in Uganda.
HCT strategies were compared in terms of outcomes that reflect three HIV control objectives: expanding knowledge of HIV status, especially to previously untested clients; reducing HIV transmission; and identifying HIV-infected individuals to receive care and treatment services.
Population and setting
We collected data on four HCT projects in Uganda, each utilizing a different HCT strategy. Projects were selected from similar geographic areas to minimize differences in population characteristics and setting. Uganda has a mature generalized HIV epidemic predominantly driven by heterosexual sex, with an HIV prevalence of 7.5 and 5.0% in 15–49-year-old women and men, respectively. HCT coverage is low, and by 2007 only 25% of women and 21% of men reported ever having tested for HIV .
HIV counseling and testing strategies
All four HCT strategies follow similar procedures, with HIV testing provided in a single session using a serial HIV rapid test algorithm. Pretest and posttest counseling is provided, covering basic HIV/AIDS information, the testing process, risk-reduction strategies, the interpretation of positive or negative test results, partner communication and disclosure, and voluntary consent. Referral for HIV care and treatment is provided for clients diagnosed HIV-positive. Testing is free, voluntary, and private, and clients are encouraged to test with their partner (couples HCT).
Stand-alone HIV counseling and testing
Stand-alone HCT is the conventional HCT strategy , provided through free-standing centers. It is client-initiated, with attendance encouraged through promotional campaigns. Individuals attending the site are provided initial group counseling, with key messages reiterated and consent obtained in private sessions for couples or individuals. Clients diagnosed HIV-positive are referred to local health centers for follow-up.
Hospital-based HIV counseling and testing
Hospital-based HCT is a provider-initiated strategy, offered to all patients with unknown or undocumented HIV status attending hospitals and major health centers regardless of illness presentation, using an opt-out approach . Pretest counseling is provided in a group setting, with key messages reiterated and consent obtained in private sessions. HCT is provided by health providers with HCT-specific training, and counselor-assisted disclosure to sexual partners and family members is provided if requested by the patient. Hospital-based HCT has the potential to identify many HIV-infected individuals, as individuals attending health facilities are more likely to have HIV infection. Linkages to HIV care and treatment can be efficient if services are provided on-site, although treatment may be less effective if patients have advanced disease.
Door-to-door HIV counseling and testing
Door-to-door HCT is a home-based, provider-initiated strategy. Mobile teams offer HCT to clients at home, and community mobilizers ensure all households in a community are offered services . Door-to-door HCT is offered to all adults and to all children aged 0–14 years whose mother is HIV-infected, deceased, or of unknown HIV status. Initial group counseling is provided to family members and peers, followed by private pretest and posttest counseling for individuals and couples deciding to test. Through reducing barriers to testing, this strategy could expand HCT services to previously underserved groups and rural communities, and increase couples testing. HIV-infected individuals are referred to local health centers for follow-up.
Household-member HIV counseling and testing
Household-member HCT is similar to door-to-door HCT, with HCT offered to clients in their own homes. However, this strategy specifically targets the household members of individuals already identified as HIV-infected, reflecting research showing that household members are at higher HIV risk . Households are approached only after voluntary consent by the HIV-positive index client. This strategy is provided in conjunction with existing HIV treatment programs, and mobile teams travel to households of willing program participants, offering HCT to all adults, and to all children aged 0–14 years whose mother is HIV-infected, deceased, or of unknown HIV status. Individuals identified as HIV-infected are offered treatment through the treatment program serving the index patient.
Although relatively new, all three projects providing provider-initiated HCT have reported high levels of acceptance, with more than 84% of eligible individuals agreeing to participate when offered HCT [22–24].
Data collection and analysis
The present study collected de-identified data from project data systems and received ethical approval from the US Centers for Disease Control and Prevention. Projects were evaluated between June 2003 and September 2005, with each project assessed for an evaluation period of 6–12 months coinciding with a period of uninterrupted project operations. Data were collected on a total of 84 323 HCT clients. We adopted a programatic perspective, considering all economic costs incurred to provide HCT but excluding the potential cost-savings that might arise from earlier identification and treatment of HIV-infected individuals and from preventing new infections.
From project accounts and inventories and through interviews with project personnel, we collected data on HCT-specific costs and a portion of shared overheads determined via direct allocation . Input prices of major cost items (equipment, salaries, testing supplies) were standardized across strategies. Buildings and utilities costs were estimated from the equivalent building rental cost, and the costs of other investments (equipment, vehicles, training) were annuitized over the expected useful life with a 3% discount rate . Cost data were collected so as to include the costs of clients who declined services. All costs are reported as 2007 US dollars.
We extracted indicators on the client population and testing results from project data systems. Indicators included total client volume, general demographics (age group, gender), representation of key population groups (previously untested individuals, married or cohabiting couples), HIV diagnosis (including partner's diagnosis for couples testing), and session completion (percentage completing all intervention components). Data on the CD4 cell count distribution of HIV-infected HCT clients were collected for a separate sample of clients conducted by each project.
Cost-effectiveness analyses compared the crude cost per client as well as the costs of reaching key target groups. Average, rather than incremental, cost-effectiveness ratios are reported, as the strategies were judged to be independent programs (not mutually-exclusive) based on the different populations targeted by each strategy . Analyses were conducted using TreeAge Pro 2006 (Williamstown, Massachusetts, USA).
Client population characteristics
Client population characteristics are shown in Table 1. Whereas the age distributions for stand-alone HCT, hospital-based HCT, and door-to-door HCT were similar, the age distribution for household-member HCT was strongly skewed toward younger age groups, with 68% of clients less than 15 years of age. All strategies had relatively equal sex distributions except hospital-based HCT, where more female clients were seen (62% of all clients). All three nontraditional HCT strategies attracted proportionally more first-time clients than stand-alone HCT, especially both home-based strategies. Door-to-door HCT attracted the highest percentage of couples for testing. All strategies reported HCT completion rates above 98.5%.
HIV prevalence in HIV counseling and testing client groups
The HIV prevalence in HCT clients and subgroups is shown in Table 2. Both stand-alone HCT and hospital-based HCT reported high client HIV prevalence, at 19.1 and 27.2%, respectively. In contrast, the two home-based strategies reported relatively low client HIV prevalence. For household-member HCT, the low prevalence reflected a higher proportion of clients in the 5–14-year-old age group, which had a 2.4% HIV prevalence, compared with 17.0% prevalence in those more than 24-year-old. For door-to-door HCT, the low average HIV prevalence reflected a low HIV prevalence across all age groups. Across all strategies, the peak HIV prevalence was found in the 35–44-year-old age group. Women and first-time clients also exhibited higher HIV rates. A substantial percentage (>30%) of all HIV-infected clients had advanced immunosuppression (defined as CD4 cell count below 200 cells/μl) regardless of testing strategy, and almost one-quarter of HIV-infected clients identified by hospital-based HCT had a CD4 cell count below 50 cells/μl. The door-to-door strategy identified the highest proportion (69.3%) of HIV-infected clients with a CD4 cell count above 200 cells/μl.
Couples HIV counseling and testing
Although no strategy reported more than 25% of clients testing as part of a couple, many clients who tested as individuals were married or part of a cohabiting couple. There was wide variation in the percentage of clients who were married or part of a cohabiting couple who actually undertook couples HCT, ranging from 6.3 (hospital-based HCT) to 64.1% (household-member HCT). Married or cohabiting couples in hospital-based HCT and household-member HCT showed substantially higher rates of HIV discordance (Table 2), related to the high HIV prevalence in these populations, for example, hospitalized patients, or household members of people with HIV. The percentage of all clients who were part of a discordant couple ranged from 0.7% for hospital-based HCT to 3.1% for household-member HCT. It is notable that for all strategies, the infected partner was male in the majority of discordant couples.
Cost-effectiveness ratios were estimated for the crude cost per client and the costs for reaching key target groups (Table 3). Each of the three nontraditional strategies compared favorably with stand-alone HCT in terms of the crude cost per client, and door-to-door HCT was the least expensive strategy per client tested. Door-to-door HCT also appeared to be the most cost-effective in terms of reaching new clients, whereas hospital-based HCT appeared superior in terms of identifying HIV-infected clients.
The nontraditional HCT strategies we evaluated are increasingly being implemented to expand access to HIV diagnosis and prevention, care, and treatment services. Other studies have investigated the cost-effectiveness of stand-alone HCT in developing countries, but this study directly compared four different HCT strategies that a national HIV program might consider. Our results revealed clear differences between the four HCT strategies in terms of the populations served, the crude cost per client, and the relative cost-effectiveness in reaching key target groups.
Both home-based strategies were comparatively inexpensive and effective at reaching population groups with low rates of prior testing and appear to be the best strategies for addressing the first of three HIV control priorities identified earlier – expanding knowledge of HIV status in previously untested clients. In contrast, hospital-based HCT appears most effective at addressing other priorities. Hospital-based HCT identified a greater percentage of HIV-infected clients compared with other strategies, had a lower cost per HIV-infected client identified, and a high proportion had CD4 cell counts less than 50 cells/μl, and for these reasons may be the most efficient strategy for identifying HIV-infected individuals in immediate need of antiretroviral therapy. Although these individuals would be immediately eligible for treatment, research has shown poorer outcomes for individuals initiated on treatment with such advanced immunosupression. HIV-infected individuals would ideally be identified for treatment at an earlier stage of the disease, an apparent advantage of the home-based strategies. Providing HCT to HIV-positive individuals is also important for HIV prevention, and so hospital-based HCT appears a reasonable strategy for advancing both treatment and prevention goals. Stand-alone HCT also identified many infected individuals, and despite its higher cost per client was a more cost-effective strategy for identifying infected individuals than the two home-based strategies.
Household-member HCT reached many young clients aged 5–14 years with low HIV prevalence, increasing the cost per infected individual identified. In secondary analyses, we examined the potential benefits of modifying this strategy to target higher prevalence age groups (0–4-year olds and ≥20-year olds), and found that this modified household-member HCT strategy might achieve substantially improved cost-effectiveness results ($103 per HIV-positive individual identified) and provide couples counseling to many HIV-discordant couples (9% of all clients).
Couples counseling is a high-priority intervention and has strong preventive effects when received by HIV-discordant couples [14–17]. In a nationally representative survey in Uganda, 40% of all HIV-infected married persons had an HIV-negative spouse  and research suggests that a substantial proportion of all new infections occur within cohabiting couples [29,30]. In our analysis, the HCT strategies showed varying success at enrolling couples, and in all cases less than 50% of clients who were married or part of a cohabiting couple actually undertook testing with their partner. For all strategies, discordant couples represented less than 4% of all clients. Greater HIV-prevention effects might be possible if HCT projects are able to increase participation in couples HCT.
As this was an individual-level analysis, it does not take into account any potential population benefits of bringing HCT to scale, such as reducing stigma, mobilizing communities to respond to the epidemic, or community-wide reductions in HIV transmission. In addition, although this study focused on the programatic costs of HCT, it is likely that increasing HCT coverage would reduce future HIV treatment costs by averting new HIV infections. This may not have been true during the early years of the epidemic response (when treatment costs were determined by the limited availability of treatment slots), but becomes increasingly plausible as HIV treatment services are scaled-up.
This study highlights the need for ongoing empirical research as HCT services expand and diversify. In particular, research is needed to clarify the durability of HCT behavior change in resource-poor settings and to confirm the behavioral impact of HCT as implemented through provider-initiated and home-based strategies. Although there have been a number of studies investigating the effects of client-initiated HCT provided through stand-alone sites, it is not clear how these effects translate to provider-initiated HCT, in which services are provided in a variety of settings to clients who are not self-selected in the same way as client-initiated HCT.
The present research was conducted in Uganda, but many features of this setting – high HIV prevalence, widespread risky sexual behavior, and limited access to HCT and HIV care and treatment services – are common to many sub-Saharan African settings. In a generalized epidemic such as this, interventions need to be expanded beyond traditional risk groups and service provision strategies to achieve population-level impact. The results suggest that in addition to stand-alone HCT, expanding new hospital-based and home-based strategies may have the cost-effective ‘scalability’ needed to substantially increase access. Together these strategies provide new tools for countries aiming to meet the UNAIDS/WHO call for universal access to HCT by 2010 .
All authors contributed to the conception and design of the study. N.M. and B.A. oversaw data collection and conducted analyses. R.W., F.N., B.M., and A.C. assisted in gaining access to study data. R.B., J.M., J.M.B. provided technical oversight. All authors participated in manuscript writing. Thanks also to the following persons for assistance in conducting fieldwork: Muhamed Mulongo, Catherine Nabaggala, Francis Nahamya, Eriya Murana, Lucy Awor, Laban Waswa Bright, Michael Nankunda, Judith Asiimwe, Elioda Tumwesigye, Elly Muganzi, Blasio Matungo, Asa Owesiga, Godwill Wana, Edna Tindimwebwa, Susan Balyejjusa Mugumya, John Ssenkusu, Patrick Elyanu, Emmanuel Jawe, William Epalitai, Madina Apolot, and Alice Nakato.
The findings and conclusions in this study are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
1. WHO. The right to know: new approaches to HIV testing and counselling
. Geneva: WHO; 2003.
2. Uganda Bureau of Statistics (UBOS) and Macro International Inc. Uganda demographic and health survey 2006
. Calverton, Maryland, USA: UBOS and Macro International Inc; 2007.
3. Alwano-Edyegu M, Marum E. Knowledge is power: voluntary HIV counselling and testing in Uganda. Geneva: UNAIDS; 1999.
4. Downing RG, Otten RA, Marum E, Biryahwaho B, Alwano-Edyegu MG, Sempala SD, et al
. Optimizing the delivery of HIV counseling and testing services: the Uganda experience using rapid HIV antibody test algorithms. J Acquir Immune Defic Syndr Hum Retrovirol 1998; 18:384–388.
5. Kamb ML, Fishbein M, Douglas JM Jr, Rhodes F, Rogers J, Bolan G, et al
. Efficacy of risk-reduction counseling to prevent human immunodeficiency virus and sexually transmitted diseases: a randomized controlled trial. Project RESPECT Study Group. JAMA 1998; 280:1161–1167.
6. The VCT Efficacy Study Group. Efficacy of voluntary HIV-1 counselling and testing in individuals and couples in Kenya, Tanzania, and Trinidad: a randomised trial. The Voluntary HIV-1 Counseling and Testing Efficacy Study Group. Lancet
7. Corbett EL, Makamure B, Cheung YB, Dauya E, Matambo R, Bandason T, et al
. HIV incidence during a cluster-randomized trial of two strategies providing voluntary counselling and testing at the workplace, Zimbabwe. AIDS 2007; 21:483–489.
8. Matovu JK, Gray RH, Makumbi F, Wawer MJ, Serwadda D, Kigozi G, et al
. Voluntary HIV counseling and testing acceptance, sexual risk behavior and HIV incidence in Rakai, Uganda. AIDS 2005; 19:503–511.
9. Sherr L, Lopman B, Kakowa M, Dube S, Chawira G, Nyamukapa C, et al
. Voluntary counselling and testing: uptake, impact on sexual behaviour, and HIV incidence in a rural Zimbabwean cohort. AIDS 2007; 21:851–860.
10. Higgins DL, Galavotti C, O'Reilly KR, Schnell DJ, Moore M, Rugg DL, et al
. Evidence for the effects of HIV antibody counseling and testing on risk behaviors. JAMA 1991; 266:2419–2429.
11. Wolitski RJ, MacGowan RJ, Higgins DL, Jorgensen CM. The effects of HIV counseling and testing on risk-related practices and help-seeking behavior. AIDS Educ Prev 1997; 9(3 Suppl):52–67.
12. Weinhardt LS, Carey MP, Johnson BT, Bickham NL. Effects of HIV counseling and testing on sexual risk behavior: a meta-analytic review of published research, 1985–1997. Am J Public Health 1999; 89:1397–1405.
13. Denison JA, O'Reilly KR, Schmid GP, Kennedy CE, Sweat MD. HIV voluntary counselling and testing and behavioral risk reduction in developing countries: a meta-analysis, 1990–2005. AIDS Behav 2008; 12:363–373.
14. Allen S, Meinzen-Derr J, Kautzman M, Zulu I, Trask S, Fideli U, et al
. Sexual behavior of HIV discordant couples after HIV counseling and testing. AIDS 2003; 17:733–740.
15. Allen S, Serufilira A, Bogaerts J, Van de Perre P, Nsengumuremyi F, Lindan C, et al
. Confidential HIV testing and condom promotion in Africa. Impact on HIV and gonorrhea rates. JAMA 1992; 268:3338–3343.
16. Allen S, Tice J, Van de Perre P, Serufilira A, Hudes E, Nsengumuremyi F, et al
. Effect of serotesting with counselling on condom use and seroconversion among HIV discordant couples in Africa. BMJ 1992; 304:1605–1609.
17. Kamenga M, Ryder RW, Jingu M, Mbuyi N, Mbu L, Behets F, et al
. Evidence of marked sexual behavior change associated with low HIV-1 seroconversion in 149 married couples with discordant HIV-1 serostatus: experience at an HIV counselling center in Zaire. AIDS 1991; 5:61–67.
18. Aisu T, Raviglione MC, van Praag E, Eriki P, Narain JP, Barugahare L, et al
. Preventive chemotherapy for HIV-associated tuberculosis in Uganda: an operational assessment at a voluntary counselling and testing centre. AIDS 1995; 9:267–273.
19. Forsythe S, Arthur G, Ngatia G, Mutemi R, Odhiambo J, Gilks C. Assessing the cost and willingness to pay for voluntary HIV counselling and testing in Kenya. Health Policy Plan 2002; 17:187–195.
20. Sweat M, Gregorich S, Sangiwa G, Furlonge C, Balmer D, Kamenga C, et al
. Cost-effectiveness of voluntary HIV-1 counselling and testing in reducing sexual transmission of HIV-1 in Kenya and Tanzania. Lancet 2000; 356:113–121.
21. Thielman NM, Chu HY, Ostermann J, Itemba DK, Mgonja A, Mtweve S, et al
. Cost-effectiveness of free HIV voluntary counseling and testing through a community-based AIDS service organization in Northern Tanzania. Am J Public Health 2006; 96:114–119.
22. Wanyenze RK, Nawavvu C, Namale AS, Mayanja B, Bunnell R, Abang B, et al
. Acceptability of routine HIV counselling and testing, and HIV seroprevalence in Ugandan hospitals. Bull World Health Organ 2008; 86:302–309.
23. Nuwaha F, Muganzi E, Kasasa S, Achom M, Bunnell R, Kabatesi D, et al. District-wide, door-to-door, home-based HIV voluntary counseling and testing in rural Uganda
[abstract TUAC0101]. In: XVI International AIDS Conference
; 13–18 August 2006; Toronto, Canada.
24. Were WA, Mermin JH, Wamai N, Awor AC, Bechange S, Moss S, et al
. Undiagnosed HIV infection and couple HIV discordance among household members of HIV-infected people receiving antiretroviral therapy in Uganda. J Acquir Immune Defic Syndr 2006; 43:91–95.
25. Drummond MF, Sculpher MJ, Torrance GW, O'Brien BJ, Stoddart GL. Methods for the economic evaluation of healthcare programmes: third edition. Oxford: Oxford University Press; 2005.
26. Weinstein MC, Siegel JE, Gold MR, Kamlet MS, Russell LB. Recommendations of the panel on cost-effectiveness in health and medicine. JAMA 1996; 276:1253–1258.
27. Johannesson M, Weinstein MC. On the decision rules for cost-effectiveness analysis. J Health Econ 1993; 12:459–467.
28. Bunnell R, Opio A, Musinguzi J, Kirungi W, Ekwaru P, Mishra V, et al
. HIV transmission risk behavior among HIV-infected adults in Uganda: results of a nationally representative survey. AIDS 2008; 22:617–624.
29. Robinson NJ, Mulder D, Auvert B, Whitworth J, Hayes R. Type of partnership and heterosexual spread of HIV infection in rural Uganda: results from simulation modelling. Int J STD AIDS 1999; 10:718–725.
30. Bunnell RE, Nassozi J, Marum E, Mubangizi J, Malamba S, Dillon B, et al
. Living with discordance: knowledge, challenges, and prevention strategies of HIV-discordant couples in Uganda. AIDS Care 2005; 17:999–1012.
31. WHO. Towards universal access – scaling up priority HIV/AIDS interventions in the health sector: progress report
. Geneva: WHO; 2007.