A review by the Reference Group to the United Nations on HIV and Injecting Drug Use in 2008 indicated that of the estimated 15.8 million (11.0–21.2 million) injecting drug users (IDUs) globally, 3.9 million (range 3.5–5.6 million), or 25%, live in south, east and south-east Asia . The prevalence of injection drug use in the 15–64-year-old population in selected countries is shown in Table 1 and ranges from 1.33% in Malaysia to just 0.02% in Cambodia . Due to the large size of these populations, even a low population prevalence of injection drug use represents a large number of IDUs (Table 1). The distribution of injection drug use is not homogenous, with eight countries accounting for approximately 85% (n = 3 306 473) of the estimated 3.9 million IDUs in Asia, whereas 11 countries account for 5% (n = 203 870).
Given the efficiency of HIV transmission through contaminated injecting equipment, injection drug use has been a significant factor in determining the course of HIV epidemics in several Asian countries. The prevalence of HIV typically rises fast among IDUs and either stabilizes at high levels (Myanmar and Thailand) or continues to increase (Indonesia) . An epidemic among IDUs may then ignite an epidemic among sex workers due to the interaction between injection drug use and sex work, including purchase and sale of sex by men and women who inject drugs .
This article will review the current situation and response to HIV in the context of injection drug use in Asia. Specifically, this review examines the current epidemiology of injection drug use and HIV among IDUs, the availability and coverage of selected recommended interventions, national responses to HIV epidemics, gaps in the national response, barriers to implementation, resource requirements for scaling-up prevention, care and treatment services, and recommendations for overcoming barriers to scaling-up services. The review builds on a situation update on HIV epidemics among IDUs in south-east Asia by Sharma et al. .
We conducted desk reviews of reports, documents and publications by governments, United Nations agencies, regional programmes, and nongovernmental organizations. The sources of data and information were retrieved from a repository of documents on injection drug use and HIV/AIDS compiled by the United Nations Office on Drugs and Crimes (UNODC). The review also drew from an electronic search of publications related to critical HIV prevention interventions, specifically needle and syringe programme (NSP) and opioid substitution therapy (OST), for IDUs using PubMed.
As part of universal access to HIV prevention, treatment and care for IDUs, the WHO, the United Nations Joint Programme on HIV/AIDS (UNAIDS), and UNODC have recommended a set of nine interventions . These include NSPs, OST and other drug dependence treatment, HIV testing and counselling (T&C), antiretroviral therapy (ART), prevention and treatment of sexually transmitted infections (STIs), condom programmes for IDUs and their sexual partners, targeted information, education and communication (IEC) for IDUs and their sexual partners, vaccination, diagnosis and treatment of viral hepatitis, and prevention, diagnosis and treatment of tuberculosis (TB). However, this particular review of programme coverage focused on NSPs and OST in view of the significant body of evidence on the effectiveness of these two interventions in preventing HIV infections .
Only low-income and middle-income countries, as defined by the World Bank, were included in the review . These included 19 countries: Afghanistan, Bangladesh, Bhutan, Cambodia, China, India, Indonesia, Lao People's Democratic Republic (PDR), Maldives, Malaysia, Mongolia, Myanmar, Nepal, Pakistan, Philippines, Sri Lanka, Timor Leste, Thailand, and Vietnam.
In several Asian countries, a large proportion of reported HIV infections have been associated with injection drug use. For example, in Bangladesh, IDUs accounted for nearly nine out of 10 HIV-positive cases in the 2006 serosurveillance round . In other countries, one-third to half of cumulative reported HIV cases have been traced back to injection drug use, including 70.6% in Malaysia  and 38.5% in China . In 2008, 57% of new infections in Malaysia  and 29.5% in China in the first five months of 2009  were associated with injection drug use.
Despite overall low rates of HIV, a high prevalence of HIV among IDUs has been reported in Indonesia (57.1% of women and 52.1% of men) , Thailand (38.7%) , Myanmar (36.3%) , Malaysia (22.1%) , Vietnam (18.4%) , China (9.3) , and Bangladesh (1.6% overall and 7.0% in Dhaka) , respectively. In a survey in Cambodia , 24.4% of IDUs were HIV-infected. Similarly, Afghanistan and the Philippines, which reported no HIV in this population during the 1990s, have since reported emerging epidemics, with prevalence of 7.1%  and 0.2% (2005), respectively . HIV prevalences reported in this section differ from those reported by Mathers et al. , in Table 1. Data in this section draw from UNGASS Country Progress Report submissions in March 2010.
Availability of and coverage of select HIV prevention interventions
Although WHO, UNODC, and UNAIDS recommended universal coverage of prevention, treatment, and care services for IDUs , these remain limited in scale despite the supportive policy frameworks and the high burden of HIV among IDUs in Asia. With the exception of Maldives, Thailand, and the Philippines, most countries have health-related or drug-related policy documents that make explicit supportive reference to harm reduction . Review of the national HIV responses by the Commission on AIDS in Asia indicated that some countries have made significant progress in implementing prevention, treatment, and care services focusing on IDUs during the last two decades . However, strategies, programmes, and implementation of the national responses have varied considerably in terms of quality and scale and, in many cases, harm reduction responses have lagged behind the epidemic or have been initiated too late to have the desired impact.
With the exception of Bhutan, Lao PDR, Maldives, Mongolia, Sri Lanka, and Timor Leste, NSPs are operational in the majority of countries reporting IDU in the region. Those countries without NSPs either report no injection drug use or low prevalence of injection drug use.
The number of NSP sites in Asia has increased from 238 in 2006  to 1902 as of the middle of 2010 (see Table 2). The increase in the number of NSP sites was primarily in four countries; Bangladesh, China, Indonesia, and Vietnam.
As seen in Table 2, the coverage of IDUs with NSP is limited in many countries. However, estimation of the size of the drug-using population is questionable, and monitoring data on NSP are often of poor quality, making calculation of NSP programme coverage a challenge in many countries.
Globally, treatment of opioid dependence with methadone or buprenorphine, is recommended as a highly effective intervention for reducing drug-related behaviours with a high risk of HIV transmission, although its impact on sex-related risk behaviours is lower . In Asia, OST is available in Bangladesh, Cambodia, China, Hong Kong, India, Indonesia, Maldives, Malaysia, Myanmar, Nepal, Thailand, and Vietnam. Geographic access within these countries, however, is limited. Whereas Pakistan has not started an OST programme, Cambodia and Bangladesh began to dispense methadone in the first OST clinic in the middle of 2010.
In countries where OST programmes are operational, the availability of OST sites has increased from 341 in 2006  to 1098 as of the middle of 2010 (see Table 3). The increase in the number of OST sites was most marked in China, followed by Indonesia and Malaysia. As seen in Table 3, the coverage of IDUs with OST was highest in China, with a cumulative total of 241 975, or 44.7%, of the 541 184 registered heroin injectors having benefited from OST . If the national IDU size estimate, 2 350 000 is used as the denominator, a cumulative total of 10.3% of the IDUs have accessed OST. Approximately 112 800 people are currently on treatment .
In Malaysia, 7455 IDUs were registered with government-operated OST clinics  and some 15 000 through private clinics . Coverage of IDUs with OST is below 10% in India and below 5% in Myanmar, Nepal, Thailand, and Vietnam. In Indonesia, approximately 1% of the estimated 219 130 men and women who inject drugs were accessing OST .
The large population size of IDUs, coupled with high HIV prevalence, translates into an estimated 735 000 IDUs living with HIV in Asia . Although countries in the region have made progress towards increasing access to ART in general, data on ART access by IDUs are scarce. Where data exist, they indicate that only a small proportion of IDUs in need of ART are receiving it. In Kuala Lumpur, only 2% of 315 persons receiving ART were IDUs, despite the majority of HIV infections being associated with injecting drug use .
Availability of size estimation and sentinel surveillance data for reporting and monitoring
Estimating the size of the IDU population is important for planning national prevention, treatment, and care programmes, as well as for monitoring intervention coverage and advocacy. However, estimates of the number of IDUs are unavailable in many countries. Some countries, such as Bangladesh , India , Indonesia , and China  have undertaken national size estimates of IDUs, whereas other countries, such as Vietnam, report on the number of registered drug users maintained by the national drug control authorities. Other countries are yet to make an estimate (Lao PDR, Maldives, Sri Lanka, and Timor Leste). In Cambodia, population estimates by different agencies using different methodologies have resulted in size estimates ranging from 600 to 10 000 .
In addition to population estimates, HIV surveillance systems aim to provide information that will inform the design of focused interventions in this population. Although a number of countries include IDUs in their sentinel surveillance system, some countries have yet to establish surveillance in this population.
Having become signatories to the resolution, ‘Declaration of Commitment’, adopted by the United Nations General Assembly , the heads of state and government and their representatives committed themselves to expand access to a wide range of prevention programmes, including sterile injecting equipment and other harm-reduction efforts . Some countries have been collecting and reporting on data on the recommended UNGASS core indicators, including levels of self-reported condom use, safe injecting, knowledge about where to have an HIV test, and receipt of condoms, needles, and syringes in the last 12 months. However, the indicators do not capture the reach of other interventions, such as OST. For example, prevention coverage for IDUs is highest in south and south-east Asia, at 62% , indicating that HIV prevention programme coverage has increased during recent years, but this fails to reflect coverage utilizing the recommended interventions.
The absence of a standard definition for what constitutes an IDU further impacts the quality and usefulness of population estimates and monitoring data. A review of the current definitions used by countries found that a total of 10 different definitions are currently in use . The use of different definitions directly impact estimates and consequently the denominators used for monitoring coverage, raising concerns about the meaningfulness of the coverage data.
Evidence on the effectiveness of interventions in Asia
Many countries in Asia initiated HIV prevention interventions, such as NSPs, only after injection drug use had been reported for a number of years or even decades and after HIV prevalence had already increased to over 10%, as was the case in Jakarta, Indonesia, where the prevalence had risen to 50%. In this context, evaluation of the effectiveness of NSPs in preventing HIV epidemics once the epidemic has already taken off is a challenge. Although, in theory, it is possible to evaluate the effectiveness of NSP in reducing HIV incidence once prevalence has reached high levels, the ability to evaluate the direct contribution is confounded by the parallel effect of other components of the prevention programme (e.g., peer education, voluntary counselling and testing (VCT), etc). Thus, it is difficult to tease out the impact of NSP, relative to other interventions in the comprehensive package, on reducing incidence. In addition to existence of other programmes, there are methodological challenges including the ability to recruit sufficient sample sizes to assess incidence and the use of appropriate comparison/control groups.
Few countries in Asia have evaluated NSPs . However, available evidence from Asian countries indicates many positive effects associated with NSPs. For example, in China, needle sharing among NSP attendees was 14.7% compared with 43.7% among those who did not participate . In Bangladesh, an evaluation of the NSP found a significant decline in reported levels of needle or syringe sharing, from 62% in 1997 to 18% in 2001 . In 2006, mathematical modelling used to estimate the impact of the NSP intervention on transmission among IDUs and their sexual partners predicted that NSPs may have reduced the incidence of HIV among IDUs by 90% in Bangladesh . A subsequent analysis confirmed that the intervention was cost-effective .
An evaluation of the pilot NSP in Malaysia documented a reduction in the percentage of IDUs passing their injecting equipment to others from 56% at baseline to 43% at follow-up and a reduction in being injected by a ‘street or port doctor’ from 42% at baseline to 33% at follow-up . Injecting services provided by street/port doctors have been acknowledged as a major factor in accelerating transmission among IDUs in Malaysia due to reuse of needles and syringes on more than one drug user. The evaluation found no evidence of any unintended negative consequences of the NSP such as increase in drug use, crime, or needle and syringe litter.
In south-east Asia evaluations of pilot OST programmes are available from China and Malaysia. In Malaysia, the national OST programme consists of both methadone and buprenorphine. The evaluation of the pilot methadone maintenance treatment (MMT) programme found a reduction in continued opiate use among patients from 45% at baseline to 10.7% at 6 months . Other positive impacts included an increased proportion of methadone patients with full-time employment, from 47.6% at baseline to 66.3% at 12 months follow-up and improved physical and psychological health and social relations measured by the WHO Quality of Life Assessment . Furthermore, Malaysia has documented a reduction in the number of drug users in detention centres, PusatSerenti, from 11000 to approximately 7000  a result of reduced relapse rates and lower levels of re-entry into the detention system due to the availability of community-based OST. The reduction in the number of inmates in detention centres is particularly important, as injection drug use cannot be ruled out inside the centres placing inmates at high risk of HIV infection.
In Malaysia, efficacy of buprenorphine, compared with naltrexone and a placebo, has also been studied and was found to be associated with greater time to first heroin use and time to heroin relapse . Compared with a placebo, buprenorphine was associated with a maximum number of consecutive days of abstinence. The evaluation also found a significant reduction in HIV risk behaviours compared with baseline . In India, OST using sublingual buprenorphine has been found to be acceptable by drug users [54,55]. Buprenorphine is also used for OST, on a smaller scale, in Indonesia, Nepal, and Thailand .
In China, an evaluation of the effectiveness of the first eight MMT clinics found a significant reduction in injection drug use reported by patients (from 69.1 to 8.8%), an increase in employment (from 22.9 to 40.6%), and a reduction in criminal behaviours (from 20.7 to 3.8%) at 12-month follow-up . They also reported a reduction in the incidence of HIV, which was 0% among IDUs who participated in MMT, compared with 8.8% in Xinjiang, 4.0% in Yunnan, and 3.1% among drug users in Guangxi who did not participate in the methadone programme . Studies also examined retention among patients. A prospective cohort study of 1003 MMT participants found a 56.2% retention rate at 14 months . Multivariate analysis identified higher methadone doses and intention to remain in treatment for life as predictors of retention. Reported challenges to service provision for staff include a lack of resources, professional training, and institutional support, along with heavy workloads and low incomes .
Although it is not a low-income or middle-income country, Hong Kong provides the longest-running example of MMT in Asia. It dates back to the early 1970s, when a general practitioner opened two methadone clinics . Available data indicate that the proportion of injectors has been consistently lower among methadone clinic attendees compared with those in drug treatment and rehabilitation centres . Also, lower needle sharing was reported among methadone clinic attendees compared with ‘street addicts’ or drug users in drug treatment and rehabilitation centres and those who were not in treatment  (despite the absence of NSPs in Hong Kong). From 1984 to June 2009, injection drug use accounted for 5.9% (n = 314) of the total 5320 reported HIV/AIDS cases .
Challenges in expanding access to prevention, treatment, and care interventions
Although a number of HIV prevention services have been implemented in most countries, significant challenges still remain in expanding access to services by IDUs. These include legal and policy barriers, including the legal status of NSPs and controlled substances. The national responses to illicit drugs in Asian countries are based on a ‘zero-tolerance’ approach with a vision to achieve ‘drug-free’ societies. National drug control polices are grounded in criminalization of trafficking, production, and possession of illicit drugs. These behaviours are subject to harsh legal penalties, including the death penalty. Although the majority of countries have national HIV or harm reduction policies and strategies, national drug control policies are not always aligned with public health strategies and policies. In 2008, 16 countries in south, south-east, and east Asia reported laws, regulations, or policies that present obstacles to introducing and scaling-up comprehensive HIV programmes for IDUs .
Access to OST medicines is a case in point. Even though the WHO has added methadone onto the list of Essential Drugs , it remains a classified substance in several countries. This has resulted in complex regulations that affect availability and access , whether for the purpose of OST or for clinical pain management. Even in south Asian countries where many governments have committed to introduce OST as part of a comprehensive package of interventions, its introduction or scale-up has been hampered by complex approval and regulatory procedures related to procurement of controlled substances.
Similarly, although most national strategies include an NSP component, the provision, distribution, or dispensing of needles and syringes to IDUs may result in penalties for inciting, aiding, or abetting an offence . Therefore, current laws frequently place NSP service providers and clients at risk of arrest, should they be found in possession of needles and syringes. The criminalization of drug use has been reported to directly impact access to and utilization of services by drug users who face repeated arrests, conviction, and incarceration .
Although policy dialogue between national AIDS authorities, drug control authorities, and public security is critical for effective HIV responses aimed at IDUs, such dialogue and collaboration between these authorities remain limited to a few countries where national taskforces bring these sectors together, such as Cambodia and Lao PDR, which have National Task Forces on Drugs and HIV. Ongoing high-level advocacy and sensitization of policy makers are required to promote a coordinated, effective national response to HIV epidemics.
Use of compulsory drug ‘treatment’ centres as a predominant approach to addressing illicit drug use
A significant population of drug users in east Asia are currently confined in compulsory drug ‘treatment’ centres or ‘rehabilitation through labour camps’. The annual reported number of drug users in these centres is over 2000 in Cambodia , 269 000 in China , 900 in Lao PDR , 7135 in Malaysia , 1500 in Myanmar , 16 445 in Thailand , and between 60 000 and 70 000 in Vietnam . Data on HIV prevalence among the centre residents are scarce, except in Vietnam, where HIV prevalence has been reported to range from 30 to 65% .
Concerns, in addition to those related to human rights violations, have been raised about absence or limited access to HIV services, including the availability and quality of evidence-based drug dependence treatment in such settings. In the absence of effective drug dependence treatment, reported relapse rates are high. For example, the government of Vietnam reports an 80–90% relapse rate following release from the centres . In Cambodia, China, and Malaysia, the relapse rates have been reported to be close to 100%, between 60 and 95%, and between 70 and 90%, respectively . High relapse rates combined with a reduced tolerance to the drug after treatment increase the risk of fatal and nonfatal overdose upon release. For example, a cross-sectional survey of self-reported nonfatal overdose found that doing time in such centres was one of the main predictors of nonfatal overdose in a cohort of IDUs in Vietnam . Furthermore, given the limited availability of prevention commodities, such as condoms , risk of HIV transmission among IDUs in these centres is of major concern.
It remains surprising that governments continue to sustain such centres despite evidence against its cost-effectiveness. For example, in Vietnam, the unit cost per inmate ranged from US$ 225 in rural areas to US$ 641 in urban areas. In urban areas, this cost was higher than the cost of providing community-based OST . Given the high relapse rates and the high unit cost, drug rehabilitation centres appear to be neither cost-effective nor able to achieve their intended objective of rehabilitating drug users. Therefore, such centres should be closed and evidence-based, voluntary, and community-based drug dependence treatment should be expanded as a matter of priority.
Resource requirements and availability
As countries in the region strive to meet their universal access and Millennium Development Goals targets, strategic allocation of financial resources for the HIV response is critical, especially within the context of increasingly scarce resource availability. Resource tracking and estimation of resource needs and gaps has become essential. In 2009, findings of an in-depth resource needs and gaps estimation of HIV prevention and treatment interventions, commissioned by the United Nations Regional Task Force on Injecting Drug Use and HIV/AIDS for Asia and the Pacific (UN RTF) indicated that US$ 0.5 billion in 2009 was required to finance implementation of the core elements of the WHO/UNODC/UNAIDS package of IDU interventions [70,71]. Of this, 69% (US$ 340 million) is required for NSP and OST, 17% (US$ 84 million) for ART, and 14% (US$ 69 million) for condoms, VCT, and STI diagnosis and treatment. One country, China, accounted for 60% of the resource need and the resource gap for NSP and OST was about 90% of the resource need in 2009. The total estimated resources needed for a scaled-up NSP, OST, VCT, ART, condom programme, and STI diagnosis and treatment package by 2015 was US$ 2.3 billion. Information is unavailable on the proportion of the prevention budget allocated to HIV prevention among IDUs in the majority of the countries in the region. Further research is needed on the availability of financial resources for prevention, treatment, and care for this population to estimate the total resource gap.
A large proportion of IDUs in the world reside in Asia and the burden of disease is disproportionately high in this population. The epidemic among IDUs has been documented to ‘kick-start’ epidemics among sex workers within a matter of years . Despite the high burden of HIV among IDUs, national responses aimed at prevention, treatment, and care for this population have been inadequate in scale. With few exceptions, the vast majority of IDUs in Asia, either do not have access to, or are not reached by programmes. The coverage of critical interventions such as NSP and OST remains limited in scale. In some countries, interventions are not available at all, despite the presence of IDUs and the increasing HIV prevalence in this population. The priority must be on introduction of the recommended services where they do not exist and expansion of access to these interventions in all countries reporting HIV epidemics associated with IDU.
Creation of a supportive legal and policy environment is required to enable introduction and expansion of access to the recommended interventions. Yet, as reported by the majority of governments, current laws and policies hamper access to and uptake of existing services. Review and amendment of laws that render NSPs, methadone, and buprenorphine illegal are urgently required. Similarly, harmonization of national drug control laws and national HIV policies and strategies is a priority in the region. Policy dialogue between national AIDS authorities and drug control authorities is critical to the resolution of the current legal and policy conflict between drug control and HIV/AIDS frameworks. As legal and policy frameworks alone will not suffice to ensure drug users’ access to vital services, sensitization and training of law enforcement staff at all levels is also needed.
Several countries, including Afghanistan, Bangladesh, and Pakistan, are yet to report an HIV epidemic among sex workers. These countries have a unique opportunity to reach IDUs with prevention measures at a sufficient scale and, therefore, avert an epidemic among sex workers with relatively modest financial resource allocation. In several other countries, including Indonesia, Thailand, and Vietnam, prevalence among IDUs has already risen to high levels and epidemics among sex workers have already taken off. These countries need to urgently increase access to and utilization of comprehensive services. Even by doing so, countries may not be able to reverse the already high prevalence levels for another 7–10 years from now . The resource requirement to scale-up interventions in the second group of countries, including Indonesia and Vietnam, is significant due to large populations of IDUs and the high burden of HIV in this population. Yet, investment in prevention, treatment, and care programmes with high coverage will be critical to controlling and reversing the trend of the national epidemics in Asia . Current drug rehabilitation programmes are ineffective and increase the risk of overdose among the estimated 90% who revert to drug use. More effective, evidence-informed, voluntary, and community-based drug dependence treatment approaches should be adopted as a matter of priority in the region.
Increased attention also needs to be given to strengthening the existing sentinel surveillance systems to improve availability of data on prevalence and trends in countries where IDUs are not yet included as a sentinel group. Better size estimation data are required to inform programme planning and for monitoring intervention coverage.
This review also unearthed significant gaps in data relevant to IDU and HIV, suggesting a need to build on the surveillance, size estimation, monitoring, and evaluation capacity related to IDU and HIV in the region. Finally, better use of existing and new data from surveillance, routine monitoring, and evaluation studies and operational studies is required to inform national policy, advocacy, and programming for IDUs.
Anne M. Bergenstrom conceptualized the review, conducted literature review and collection of data, developed the draft, and finalized the article. Abu S. Abdul-Quader made a significant contribution to the conceptualization of the review and reviewed the paper with specific focus on the epidemiology.
The authors are grateful to Professor Roger Detels and for review of the paper and to Sheena Sullivan for her review, inputs, and formatting of the final version of the paper.
The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the organizations and/or agencies the authors are affiliated with.
Conflicts of Interest: None.