An estimated US$ 7.5 billion will be needed to respond to HIV comprehensively in Asia by 2010 , with US$ 3.5 billion required for a more targeted approach . In the Asian context, 70–80% of the estimated needs relate to HIV prevention. It has been estimated that effective prevention can avert over 80% of new HIV infections in the most-at-risk populations [sex workers, injecting drug users (IDUs), men who have sex with men (MSM), and male clients of sex workers] . Yet there remains an important gap between the effectiveness of individual HIV prevention activities, and population coverage and impact.
The Asia Pacific region has an estimated 4.9 million people living with HIV, including 440 000 new cases and 300 000 deaths annually [1,3]. Over 95% of infections occur in eight countries: Cambodia, China, India, Indonesia, Myanmar, Nepal, Thailand, and Vietnam. In addition, within these countries, HIV is concentrated in the most-at-risk populations, wherein HIV prevalence is often over 5% . HIV prevalence of over 1% has been measured among antenatal clinic attendees (ANC) in Thailand, Cambodia, at least six states of India, and Myanmar . The majority of these countries show stabilizing ANC HIV prevalence, except Indonesia . However, HIV prevalence among most-at-risk populations remains significant and dynamic in many other countries in Asia .
Despite the gap in HIV prevention service coverage, Asia has been successful in reversing its three most important national epidemics: Thailand, Cambodia, and south India [5–11]. These countries suggest additional factors to incorporate in HIV prevention if it is to be successful at population level. They combined rapid, high level coverage of priority services in key groups, with intensive mobilization of most-at-risk populations to catalyze social change. It is important to identify these two modes of HIV prevention and not rely only on a service delivery mode, which can be delivered directly, modeled and attributed to a program. The delivery of services that protect individuals (type 1 prevention – individual service delivery) has largely been the focus of Universal Access. However, equally important is HIV prevention to mobilize those at risk directly at the population level (type 2 prevention – population social change). It is important to combine these two modes of HIV prevention to catalyze the impact of HIV prevention at the population level [12–15].
The aim of this study is to highlight the gap in coverage of individual prevention interventions for key most-at-risk populations. At the same time, it analyzes country situations where HIV has declined (Thailand, Cambodia, south India) to identify additional program components critical to catalyze a population level response. Finally, it assesses the barriers and challenges to more systematically combine these two critical elements to catalyze HIV prevention in Asia.
The review first assesses the latest data on the coverage of services by most-at-risk populations (sex workers, MSM, IDU, male clients of sex workers) and country for the period 2007–2009 drawing largely on data from United Nations General Assembly Special Session on AIDS (UNGASS) and Universal Access reporting to UNAIDS and WHO [3,16,17]. It also assesses the coverage of prevention of mother-to-child transmission (PMTCT) services. Modeling is used to assess at what levels of coverage these services may have a population level impact in reducing HIV incidence. The Asian Epidemic Model is used, whose details and parameters for Asia have been published elsewhere based on those used for the Commission on AIDS in Asia [3,18]. Given these findings, information on the unit costs and coverage is used to define a priority prevention package that is affordable and can lead to high population coverage. This draws directly on the costing analysis of the Commission on AIDS in Asia . The second part of the study analyzes national program information, epidemiological, and behavioral data from Thailand, Cambodia, and south India [4–10,12] to identify additional factors required to catalyze effective HIV prevention at the population level. Key components of these national programs are identified in addition to the delivery of services. Finally, the section reviews the literature to assess the barriers to combining high-level coverage of priority services with social mobilization of most-at-risk populations to catalyze effective HIV prevention in Asia.
HIV prevention coverage in Asia
There are two sides to HIV prevention in Asia. First, there are important gaps in the aggregate coverage of HIV prevention services in key populations (MSM, IDUs, sex workers, clients of sex workers) across the continent [1,3]. At the same time, where there has been a systematic, national approach to prevention, HIV prevalence has been reduced by up to 30–50%.
Table 1 shows the coverage of HIV prevention and treatment from the most recent country data collected for 2007–2009 [3,16,17]. This shows that less than one in two female sex workers (FSWs), one in five IDUs, and one in two MSM are being reached with HIV prevention services. In addition, coverage of PMTCT is still very low, at 32%.
The data from UNGASS and Universal Access reporting to UNAIDS and WHO showed the progress in coverage to 2009 and the variability by country [3,16,17]. Only in Nepal, parts of India, Myanmar, and Pakistan were more than 50% of IDUs reached with HIV prevention. Only in Mongolia, China, Nepal, Myanmar, and parts of India were more than 60% of MSM reached with prevention . Over 60% of FSWs were reached in China, Lao, and Mongolia. However, greater than 60% of FSWs reported use of condoms with their most recent client in all countries, except Pakistan and Iran .
Interestingly, clients of FSWs are often not well captured in these coverage estimates. They have not always been well targeted or measured (they often lie between the concentrated/generalized divide) as shown by the Universal Access estimates . It is estimated that at least 75 million men buy sex regularly in Asia (and up to 10 million women who sell sex) . The proportion of men who buy sex appears to be associated with the levels of HIV prevalence across Asian countries [3,16]. The percentage of men visiting sex workers is highly variable but substantial, from under 5% in the Philippines to 5–7% in Indonesia, Pakistan, and Malaysia, 7–10% in China, Vietnam, and Myanmar, 10–15% in India and Laos, and 15–25% in Cambodia and Thailand (all figures are for 2007, except Thailand for 1990). In addition, these behaviors can change significantly in response to social changes related to HIV and HIV prevention. For example, the number of men reported to be visiting sex workers halved in Thailand between 1990 and 1993, and among the military in Cambodia from 1997 to 1999 [9,11].
There is also a gap in financing and in targeting finance to the most effective interventions. The Commission on AIDS in Asia recommended a comprehensive package of 19 prevention, treatment, and impact mitigation activities, together with surveillance and management, at a total cost of US$ 6.4 billion per annum [3,19]. In 2007, only US$ 1.2 billion was available for an HIV response (19% of the estimated need). Important sources of funding include the Global Fund, World Bank, Gates, and bilaterals, including the UK Department for International Development (DFID), and in China, India, and Thailand domestic funding as discussed elsewhere .
Given the important gaps in funding and coverage, a recommended and costed package of prioritized interventions in Asia has been developed at a cost of US$ 3.1 billion . The package is based on estimates of the considerable variability in cost-effectiveness of prevention strategies. This package highlights prevention among sex workers and their clients (73% of the estimated infections averted) as well as among IDUs and MSM, and counseling and testing and PMTCT. The approach suggests that despite gaps in resources, a prioritized package of HIV prevention can be scaled to population level in many countries.
Based on this, a priority set of interventions, which can be applied systematically (to 80% of the populations in need), have been identified. The cost for prevention amounts to US$ 1.34 billion or 43% of the total annual resource need of US$ 3.1 billion (Table 2).
Finally, modeling suggests why a focused, HIV prevention strategy that rapidly achieves population coverage is important, even as a more comprehensive response is introduced. High priority interventions in Asia (among sex workers and their clients, IDUs, MSM) are effective individually and important at all levels of coverage. However, the population level effects in terms of reducing HIV incidence and prevalence require at least 60% coverage, as shown in Fig. 1. The coverage lines of 60 and 80% show the most important impacts in reducing HIV prevalence, while adult prevalence is not reduced with coverage levels of 50% and below. In fact, program research and evidence suggests that the reach and engagement of prevention needs to attain 80% coverage if the proportion of at-risk individuals who adopt protective behaviors is to reach 60% coverage [3,18]. Active surveillance and measurement are, therefore, critical components of systematic prevention to determine coverage, the scale and stage of the epidemic, as well as the most-at-risk population sizes, behaviors, and best means and places of engagement.
Additional components of population HIV prevention at the national level
The low coverage of individual prevention interventions in Asia is in contrast to the national successes in reducing HIV prevalence in the three most important epidemics, namely, in Thailand, Cambodia, and south India. Therefore, a review of these prevention campaigns may inform scale up in other countries and identify additional factors necessary to catalyze population level impact.
There is no doubt that improving the coverage of comprehensive interventions is critical and has an impact in key most-at-risk populations across the region. These have included programs among sex workers in India, China, Thailand, and Cambodia, among IDUs in China, Bangladesh, Malaysia, and Myanmar, and limited successes among MSM, for example, in Indonesia [20–23]. Community-based programs to prevent drug use have also shown results in China [24–26], but have not always been replicated. They were successful in mobilizing existing community resources with a focus on the family, school, community, and clinic, and involving community and local health leaders.
However, countries where HIV prevalence has declined at the population level suggest two additional characteristics. First, they identified a priority set of interventions, which were implemented rapidly to high population coverage in most-at-risk populations, as shown by the systematic condom use programs in Thailand and Cambodia. This may not have covered all comprehensive activities, for example, counseling and testing was only scaled out in Cambodia from 2000 to 2005. However, what was implemented was done to ensure population level coverage and engagement with all most-at-risk groups.
Second, the delivery of services to individuals (type 1 prevention – individual service delivery) was accompanied by activities that operated directly at the population level to catalyze social change (type 2 prevention – population social change). The intensity of direct engagement with networks is often underestimated in the service delivery models of HIV prevention. These included nongovernmental organization (NGO) outreach in factories and villages, national self-help groups of people with HIV, intensive prevention with army recruits, in private work places, and with taxi drivers in Thailand. Similarly, the intensity of social communication at the population level is not always appreciated or attempted in HIV prevention in other countries (see Table 3). In Thailand, AIDS messages were presented every hour on all the major 500 radio and seven TV stations (with 15-s slots). Interestingly, in all three countries, declining risk behaviors and infections [e.g., sexually transmitted infection (STI) and HIV prevalence trends in Thailand and southern India] preceded the high level coverage of discrete interventions as has been the case in other HIV prevention successes [9,11,14]. The much interpreted trends clearly show the onset and initial rapid decline in STI rates occurred early on when condom use with clients was low at 20–40%, less than the coverage that would explain these declines [9,11]. This suggests in these situations, social change in key populations may be catalyzed at early stages of the delivery of services and account for the initial 50% decline in risk behaviors. These population activities (mobilization of key networks and intensive social communication) directly impact HIV behaviors and prevalence and do not just act as a supporting environment for services. However, service delivery is a critical supportive component of social mobilization and in sustaining its impact. These aspects of successful HIV prevention are not always fully recognized, for example, in describing the Thailand ‘100% condom program’  or modeling based on the coverage of individual interventions .
Table 3 shows the importance of combining individual service delivery with population social change in programs that have led to national declines in HIV prevalence in Thailand and Cambodia.
These situations suggest the importance of a systematic response in terms of coverage of services and engagement of all most-at-risk populations. Important components in such programs are described below.
Disaggregation of the epidemic and sharing of results
Disease surveillance was not just used to monitor activities. It was an intervention in its own right, providing information on most-at-risk populations, evidence to base policy, and in dissemination to the population was directly important in catalyzing a population response. This included evidence of the scale of the epidemic among different groups and regions of the country, with sentinel HIV and behavioral surveillance established rapidly in Thailand and Cambodia. Such disaggregation was necessary to establish the basic parameters of the response, epidemic trends, size, coverage, and characteristics of vulnerable groups [28,29]. These approaches were also critical to initiate political momentum and financing, when presented to key decision-makers (the HIV trends among army recruits in both Thailand and Uganda played this role, at the level of the president and prime minister). This evidence can even initiate a population response when communicated transparently. In Thailand, the early behavioral surveys showing the large proportion of men who visited sex workers became an issue of wide public (and probably domestic) discussion [9,30]. In Cambodia, HIV and behavioral surveillance among 10 population groups every 12–18 months guided each stage of the response.
Prioritized, focused service delivery to 60% coverage
The rapid shift from individual services to population coverage required a defined, focused HIV prevention package. In Thailand, Cambodia, and southern India, this ensured that what was delivered was delivered at scale in specific populations, including sex workers, army recruits, and IDUs, and in different regions of the country. This involved prioritization absolutely by behaviors and at times geographically (for example, among army recruits in northern Thailand), and relative prioritization to the stage of the epidemic. In Thailand and Cambodia, high coverage among sex workers, clients of sex workers, and army recruits was established rapidly as a comprehensive response was built. In Thailand, a priority package of prevention was carefully defined, as well as the networks to ensure coverage in brothels, the army, work places, villages, and taxi drivers.
Mobilization of key populations for social change
In each of these situations, scale up involved systematic delivery of services and engagement with populations at risk, including establishing community and self-help groups. This required a strong implementation and leadership focus on key opinion leaders, nationally and at the community level. In Cambodia, the HIV/AIDS coordinating committee (HACC) brought together international and national NGOs and community-based organizations to ensure systematic engagement at community level. In Thailand, TV and radio media, together with targeted programs for sex workers, their clients, the army, and villages were central to the response. This was critical for the 60–80% coverage required for the effectiveness of individual services (type 1 prevention) and directly initiated the population response seen among sex workers, their clients, and other most-at-risk populations (type 2 prevention).
Program and disease surveillance to adapt the response
Strong program and disease surveillance generated the strategic information necessary to link services with behaviors and disease trends in all key populations. In Thailand, behavioral surveys were critical to raising the alarm early on, as were initial behavioral surveys in 1994 in Cambodia, together with the rapid establishment of sentinel HIV and behavioral surveillance. This ensures services are related to coverage and population trends. Second, approaches could be tested and adjusted as they were delivered and the epidemic matured. Important adjustments were made to address barriers to scale up, for example, violence as a key issue for sex workers in southern India and sex workers outside brothels in Thailand and Cambodia.
These stages are not prescriptive, but combined high coverage of priority services in key groups with intensive mobilization of most-at-risk populations to catalyze social change. They were guided by program and disease surveillance, which linked services to population characteristics, behaviors, and disease trends.
Challenges and barriers to scaling up HIV prevention
The contrast between modest coverage of interventions and remarkable national successes in Asia suggests greater attention is required to the detailed models and barriers of scale up.
The first set of barriers relates to the initial stages of an intense, focused strategy to achieve 60% coverage of all most-at-risk populations. This requires leadership, policy, and evidence-based resource allocation. There is a qualitative difference in the intensity of prevention programs that have succeeded and the commitment to mobilize a population level response. In Thailand, the government very rapidly developed a systematic response; in Cambodia and southern India, NGOs also contributed to initial policy and resource allocation decisions. Effective HIV prevention relies fundamentally on a decision at the national and community levels to implement HIV prevention seriously and at scale. Success has been characterized by high-level leadership (and at all levels into communities), sufficient resources, and wherever necessary careful decisions on prioritization. A supporting legal environment  and alignment of policy across sectors is important to ensure HIV prevention is consistent and systematic. However, in Cambodia, policy developed from programs rather than the latter waiting for a perfect policy framework.
There are also important barriers related to the mobilization of all most-at-risk populations. Given the importance of vulnerable populations, stigma, discrimination, and violence are critical factors to consider in implementation [3,32,33]. Among sex workers, the fear of violence seriously affects the acceptance of interventions and the risk behaviors of vulnerable populations. The Avahan program of HIV prevention in India has built in components to tackle violence in its community interventions . It is important to ensure HIV prevention goes beyond clinics into homes, bars, brothels, and the places where risk behavior occurs. Several programs, including those in Cambodia and in India among sex workers, carefully mapped out their interventions to places of risk. In many countries, punitive laws for MSM, sex workers, and IDUs severely limit critical engagement with these populations, together with the lack of supportive laws against discrimination .
Stigma puts strain on the human resources working in interventions with vulnerable groups . Human resources for HIV prevention are a major limitation in terms of training and ensuring the necessary support during deployment. There needs to be similar incentives and support for HIV prevention and treatment to ensure quality staff are deployed, supported, and retained [3,22,34].
Finally, a major barrier is sufficient attention to program and disease surveillance, linking services to population coverage, behaviors, and HIV trends. This evidence when shared widely can be an intervention in itself and is a critical guide to adapting programs. Evidence is critical for effective action at national, community and individual levels. The evidence for programming from extensive sentinel and behavioral surveillance in Thailand and Cambodia was central to implementation of the program and mobilizing political support. This included behavioral surveys, sentinel surveillance in most populations at risk and regions, and qualitative studies. Both Thailand and Cambodia very rapidly developed world-class surveillance systems to measure their projects and relate them to the epidemic.
Given the barriers to HIV prevention, different approaches and models for scale up will be necessary. It will necessarily be multipronged, but should systematically combine service delivery with mobilization of all most-at-risk populations.
There have been important limitations in the coverage of individual HIV prevention services in Asia. At the same time, there have been remarkable national successes where HIV prevention has combined high coverage of priority services with mobilization of a population response. The findings of this review suggest a number of important components: disaggregation of the epidemic and sharing of results; prioritized, focused service delivery to 60% coverage; mobilization of key populations for social change; and program and disease surveillance to adapt the response. It may be better to rapidly scale up priority HIV prevention in all most-at-risk populations as the first stage of a more comprehensive approach. Epidemiologic analyses and modeling suggest the need to reach 80% of all populations at risk to achieve a significant impact on HIV incidence in Asia.
There are many limitations to this analysis of prevention targeted to the diverse Asian epidemics. First, the study has focused on national responses and has not fully incorporated the many successes in subpopulations across Asia, among sex workers in several countries, IDUs in China and Bangladesh, and MSM in Indonesia [1,3]. When such populations are not engaged, for example, MSM in Thailand, serious epidemics can result [3,33]. Furthermore, there are lags in the data on HIV prevention coverage and recent increases in financing. Stigma and discrimination remain the most important barriers and often result in violence against the very populations we need to engage . In addition, human resources for HIV prevention and the need to combine community and national leadership require attention . There is bias in reporting the coverage of interventions, both underreporting and overreporting, particularly among most-at-risk populations and limited verification of national reporting. There is also a need to improve data on program descriptions, service delivery, and the population level social and behavior changes that accompany them.
Nevertheless, the HIV prevention successes in Thailand, Cambodia, and southern India show the importance of combining high-level coverage of priority services with intensive engagement of all most-at-risk populations to catalyze social change. Their responses also focused critically on clients of sex workers, who often fall between the concentrated/general population divide of HIV prevention. Combination HIV prevention should avoid implementation of an increasingly complex set of individual interventions, each with varying efficacy and which do not reach population coverage. Combination HIV prevention should combine two components, high coverage of priority services that protect individuals (type 1 prevention – individual service delivery) and intensive engagement with those at risk at population level (type 2 prevention – population social change). It is important to identify these two modes of individual and population level prevention, even if they are combined in many programs (clearly HIV testing and condom use involve behavior change). The approach focuses scale up on the need to address networks of people at risk very seriously and focus on the quality (and quantity) of engagement. In Thailand, Cambodia, and southern India, systematic HIV prevention facilitated the decline in risk indicators (STIs, behaviors, and even HIV prevalence) early. The population response preceded and exceeded the impact of service delivery.
Even with financial constraints, scaling priority HIV prevention to high population coverage among all most-at-risk populations is within the capability of most Asian countries. There is a need to build on the successes of HIV prevention in Asia, use resources in a prioritized way, and ensure systematic coverage and engagement in what is implemented. Attention to the networks for delivery and the quality of engagement is critical to support the social change, which can amplify HIV prevention. As with many public health interventions, prevention requires a shift in thinking from individual to population coverage to achieve significant impact. Catalyzing HIV prevention at population level will be essential to achieve universal access and support a sustainable response of HIV prevention, treatment, and care in Asia.
The article does not necessarily represent the views of the affiliate institutions. The authors acknowledge the contributions of Abu Abdul-Quader and the reviewers.
The views expressed in this paper are those of the authors and do not necessarily represent the position of the Global Fund to Fight AIDS, Tuberculosis and Malaria.
Conflicts of interest: None.
1. UNAIDS 2008 Regional Review of HIV and AIDS in Asia-Pacific Region; UNAIDS, Geneva.
2. Komatsu R, McLeod R, Sripong N, et al. Can we afford universal access for HIV in Asia in this time of an economic downturn? AIDS
3. Commission on AIDS in Asia. Redefining AIDS in Asia: crafting an effective response
. New Delhi: Oxford University Press; 2008.
4. UNAIDS/WHO. Epidemiological Fact Sheet on HIV and AIDS
: Indonesia. Geneva; 2008.
5. Nelson KE, Celentano DD, Eiumtrakol S, Hoover DR, Breyer C, Suprasert S, et al
. Changes in sexual behaviour and a decline in HIV infection among young men in Thailand. N Engl J Med 1996; 335:297–303.
6. Kumar R, Jha P, Arora P, Mony P, Bhatia P, Millson P, et al
. Trends in HIV-1 in young adults in south India from 2000 to 2004: a prevalence study. Lancet 2006; 367:1164–1172.
7. Steen R, Mogasale V, Wi T, Singh AK, Das A, Daly C, et al
. Pursuing scale and quality in STI interventions with sex workers: initial results from Avahan India AIDS Initiative. Sex Trans Infect 2006; 82:381–385.
8. Saphonn V, Parekh BS, Dobbs T, Mean CV, Bun LH, Ly SP, et al
. Trends of HIV-1 seroincidence among HIV-1 Sentinel Surveillance Groups in Cambodia, 1999–2002. JAIDS 2005; 39:587–592.
9. UNAIDS relationships of HIV and STD decline in Thailand to behavioural change. Best Practice Series, UNAIDS; Geneva; 1998.
10. Basu I, Jana S, Rotheram-Borus MJ, Swendeman D, Lee SJ, Newman P, Weiss R. HIV prevention among sex workers in India. JAIDS 2004; 36:845–852.
11. Morineau G, van Pelt M, Courdier F, Wolf R. The HIV pandemic: local and global implications.
In: Beck EJ, editor. Cambodia; 2006.
12. Guinness L, Kumaranayake L, Rajaraman B, Sankaranarayanan G, Vannela G, Raghupathi P, George A. Does scale matter? The costs of HIV prevention interventions for commercial sex workers in India. Bull World Health Organ 2005; 83:747–755.
13. Low-Beer D, Stoneburner R. AIDS communications through social networks: catalyst for behaviour changes in Uganda. Afr J AIDS Res 2004; 3:1–13.
14. Stoneburner R, Low-Beer R. Population-level HIV declines and behavioural risk avoidance in Uganda. Science 2004; 304:714–718.
15. Latkin CA, Knowlton AR. Social network approaches to HIV prevention: implications to community impact and sustainability. In: Trickett EJ and Pequegnat W, editors. Community interventions and AIDS: strategies for increasing community impact
. New York: Oxford University Press, 2005.
16. WHO/UNAIDS/UNICEF. Towards universal access: scaling up priority HIV/AIDS interventions in the health sector
. Geneva: WHO; 2010.
17. Stover J, Fahnestock M. Coverage of selected services for HIV/AIDS prevention, care and treatment in low- and middle-income countries in 2005
. Washington DC, Constella Futures, Policy Project; 2006.
18. Brown T, Peerapatanapokin W. The Asian Epidemic Model: a process model for exploring HIV policy and programme alternatives in Asia. Sex Transm Infect 2004; 80:19–24.
19. Economic and Social Commission for Asia and the Pacific, HIV and AIDS in Asia and the Pacific [discussion paper]; 2008.
20. Des Jarlais D. Structural interventions to reduce HIV transmission among injecting drug users. AIDS 2000; 14(Supplement 1):S41–S46.
21. Foss A, Watt CJ, Vickerman P, Azim T, Guinness L, Ahmed M, et al
. Could the CARE-SHAKTI intervention for injecting drug users be maintaining the low HIV prevalence in Dhaka, Bangladesh? Addiction 2007; 102:114–125.
22. Saidel TJ, Des Jarlais DC, Peerapatanapokin W, Dorabjee J, Singh S, Brown T. Potential impact of HIV among IDUs on heterosexual transmission in Asian settings: scenarios from the Asian epidemic model. Int J Drug Policy 2003; 14:63–74.
23. Pisani E, Girault P, Gultom M, Sukartini N, Kumalawati J, Jazan S, Donegan E. HIV, syphilis infection, and sexual pratices among transgenders, male sex workers, and other men who have sex with men in Jakarta, Indonesia. Sex Trans Infect 2004; 80:536–540.
24. Wu Z, Sullivan S, Wang Y, Rotheram-Bonus MJ, Detels R. Evolution of China's response to HIV/AIDS. Lancet 2007; 369:679–690.
25. Wu Z. Landmark government methadone maintenance program in mainland China
. 15th International AIDS Conference; Bangkok, Thailand; 11–16 July 2004.
26. Wu Z, Detels R, Zhang J, Li V, Li J. Community-based trial to prevent drug use among youths in Yunnan, China. Am J Public Health 2002; 92:1952–1957.
27. Rojanapithayakorn W, Hanenberg R. The 100 per cent condom programme in Thailand. AIDS 1996; 10:1–7.
28. UNAIDS/WHO. Guidelines for Second Generation HIV Surveillance
. Geneva: UNAIDS/WHO; 2000.
29. Chandrasekaran P, Dallabetta G, Loo V, Mills S, Saidel T, Adhikary R, et al. Evaluation design for large-scale HIV prevention programmes: the case of Avahan, the India AIDS initiative. AIDS
2008; 22 (Suppl 5)
30. Low-Beer D, Stoneburner R. Behaviour and communication change in reducing HIV: is Uganda unique. Afr J AIDS Res 2003; 2:9–21.
31. Gupta GR, Parkhurst JO, Ogden JA, Aggleton P, Mahal A. Structural approaches to HIV prevention. Lancet 2008; 372:764–775.
32. Azim T, Chowdhury EI, Reza M, Ahmed M, Uddin T, Khan R, et al. Vulnerability to HIV infection among sex worker and nonsex worker female injecting drug users in Dhaka, Bangladesh: evidence from the baseline survey of a cohort study. Harm Reduct J
33. Beyrer C, Jittiwutikarn J, Teokul W, Razak M, Suriyanon V, Srirak N, et al
. Drug use, increasing incarceration rates, and prison-associated HIV risks in Thailand. AIDS Behav 2003; 7:153–161.
34. Li L, Lin C, Wu Z, Wu S, Rotheram-Borus MJ, Detels R, Jia M. Stigmatization and shame: consequences of caring for HIV/AIDS patients in China. AIDS Care 2007; 19:258–263.