Adam, Philippe C G PhD*†; de Wit, John B F PhD*‡; Toskin, Igor MD, PhD§; Mathers, Bradley M MBChB‖; Nashkhoev, Magomed MD, PhD§; Zablotska, Iryna MD, PhD*; Lyerla, Rob PhD§; Rugg, Deborah PhD§
Men who have sex with men (MSM) are recognized as being at high risk for HIV infection in the concentrated epidemics in high-income countries (HIC).1,2 Supported by community engagement and activism, research and prevention responses for MSM were quickly put in place in many of those countries, as were treatment and care programs.3 Concerted behavioral and structural prevention responses have contributed to substantial epidemiological changes,4,5 even though these seem imperfect or insufficient to entirely halt the HIV epidemics among MSM in HIC1,2,6 and their sustainability is not guaranteed. In contrast, and with only some national or regional exceptions,7 the HIV epidemics and prevention needs among MSM in low-income and middle-income countries (LMIC) have been relatively neglected.8-10 Compounded by limited resources, the quality and coverage of services and programs for MSM in (most) LMIC remain low,11 especially in contexts where the social stigma attached to male-to-male sex and the criminalization of this behavior is widespread.12-14
The urgency of adequately addressing HIV prevention needs among MSM in LMIC, however, has become most obvious in recent years. Same-sex behaviors in adult men have been documented in a variety of LMIC,8,9 including in regions such as sub-Saharan Africa where these behaviors have sometimes been considered to be nonexistent or highly infrequent.15,16 Available data across LMIC suggest that proportions of men who engaged in male-to-male sex in the past year may reach 7%-8% in parts of some regions.8 In addition, HIV prevalence data suggest that MSM in many LMIC are at markedly increased risk for HIV infection compared with their heterosexual counterparts.10 Furthermore, HIV epidemics among MSM may be increasing in some LMIC, including in China,17 where the sheer number of MSM holds the possibility of a sizable HIV epidemic.
Several studies suggest that HIV knowledge and the self-perceived risk for HIV infection may be low among MSM in LMIC,18,19 but such data are too scarce to allow for firm conclusions regarding the extent of HIV knowledge in MSM across LMIC. The availability and comparability of data concerning the frequency of unprotected and protected anal intercourse is also limited. As documented in a recent review of the literature,9 condom use during the last occasion of male-to-male anal sex was the indicator of protected sex reported most often, but this information was available for only 13 LMIC. Moreover, country estimates of this indicator of condom use were found to diverge widely, ranging from 0%-82%,9 and no overall average was computed from country estimates. Similarly, very little information is available on the number of male sex partners, the frequency and nature of sexual practices with male partners (penetrative sex or not), and the types of relationships among MSM in LMIC, including the prevalence of sex in exchange for money or other forms of compensation. Data concerning the uptake of HIV testing,20 and the coverage of HIV prevention policies and programs in MSM in LMIC are equally scarce and no published reviews are currently available. In addition to the scarcity and limitations of available data, the implications of findings for prevention policies and programs in the wider MSM communities beyond the study samples are unclear. Country data are generally not weighted by MSM population size, reflecting the limited information that is available regarding the prevalence of male-to-male sex in LMIC.
To adequately inform prevention policies and monitor prevention responses among MSM, it is vital that standardized and weighted information regarding the prevalence of HIV-related knowledge, HIV testing, and sexual behaviors (particularly condom use) among MSM in LMIC is regularly obtained. It is equally important to continually monitor the availability and coverage of HIV prevention services and programs for MSM in LMIC.21 The UNGASS country progress reports uniquely provide data against a set of standardized indicators to map and monitor the global and regional HIV prevention needs and responses among MSM in LIMC.22
This article presents data derived from the 2008 UNGASS reporting round. The 4 main indicators reflecting the HIV prevention needs and responses among MSM in LMIC covered by this article are the proportion of MSM who have tested for HIV and know the result, were reached by HIV prevention programs, hold correct knowledge regarding HIV transmission and prevention, and used condoms the last time they had anal intercourse with another man. For each of these indicators, the objective was to establish country estimates with 95% confidence intervals (CIs) and to calculate global and regional estimates weighted for the size of MSM populations in reporting LMIC.
All Member States of the United Nations are encouraged to report biennially to UNGASS concerning their country's progress in the fight against HIV in various populations. To enable comparison of information reported to UNGASS, a set of standardized indicators has been developed for country progress reports.23 However, heterogeneous data resulting from multiple sources may be available and used in any country to monitor the epidemic. In such cases, countries are advised to report the data that most closely reflect the specified indicators and that comply best with UNAIDS technical guidelines regarding internationally recommended methods of data collection and adequate sample size.23
Table 1 describes the 4 indicators related to the HIV prevention needs and responses among MSM that are addressed in this article. Data concerning these 4 indicators were taken from reports submitted by UN Member States during the 2008 UNGASS reporting round. Countries were considered to be low-income or middle-income when classified as such by the World Bank.24
Data Sources and Inclusion Criteria
For each of the indicator values reported, we assessed available information concerning measurement (the questions that were asked), methods of data collection (Behavioral Surveillance Survey or equivalent, AIDS Indicator Survey, population-based survey, mapping study, desk review, program data, or other studies) and sample size to determine whether or not it was consistent with UNAIDS guidelines.23 Data were considered to be consistent with technical guidelines when 2 criteria were met. First, if the questions asked were generally in accordance with specifications, although allowing for small variations to account for differences in cultures and contexts. Second, if the data were collected through eligible types of surveys and did not reflect programmatic data. Country reports were eligible for inclusion in the analysis of a specific indicator when data collection was consistent with UNAIDS guidelines and when sample size was provided and greater than 99.
For each of the 4 indicators, we present crude country estimates and 95% CI, derived from the calculated standard error of the indicator value provided in country reports. Eligible country data were weighted for the estimated size of the MSM population. The weighted estimates were then averaged to estimate the global proportion of MSM across reporting LMIC who have tested for HIV, were reached by HIV prevention programs, held correct HIV knowledge and used condoms. Particular attention was paid to identifying reported indicator values that seemed “too close to perfection” (ie, >95%) and the potential impact of these values on global estimates was assessed. In addition to global weighted estimates, regional estimates weighted by population size were calculated for regions for which sufficient numbers of country reports were available. For each of the indicators, many regions were represented by only 1 country and estimates for those regions were not calculated. For the other regions, and across the 4 indicators, the number of reporting countries ranged from 3 to 13. For each of the indicators, the number of reporting countries was sufficient to calculate regional estimates for 3 regions: Eastern Europe and Central Asia; Latin America; and South and Southeast Asia. Regarding condom use, the regional weighted estimates could also be calculated for the Caribbean and sub-Saharan Africa.
Estimating MSM Population Sizes
To estimate the size of the population of men who had sex with another man in the last year in each of the countries included, we used data from the published literature8,9 and made several assumptions, as explained below. Note, however, that estimates of MSM population sizes in LMIC, as reported in the literature, are often derived from small samples of relatively visible MSM, mostly from the major city or cities in a country. Although providing important and unique information, the representativeness of the samples used to derive estimates of the prevalence male-to-male sex may then be limited. Consequently, most currently available estimates of the prevalence of male-to-male sex should be considered as tentative and requiring further improvement.
Based on both the original and revised data reported by Caceres et al,8,9 we derived the potential lower and upper bounds of the percentage of men who had sex with another man in the last year in East Asia, South Asia, Southeast Asia, the Caribbean, Eastern Europe and Central Asia, and Latin America (Table 2). Because Caceres et al8,9 provided no estimates for sub-Saharan Africa, the Middle East and North Africa, the Pacific, and Western and Central Europe, we had to make several assumptions. For sub-Saharan Africa, the Middle East and North Africa, and the Pacific, we conservatively assumed that prevalence of male same-sex behaviors in the previous year was 1%-3% (reflecting the lowest values for the lower and upper bounds estimated for other regions). For Macedonia (the only LMIC included in Western and Central Europe), we assumed a prevalence of male same-sex behaviors of 1%-8%, which mirrored the estimated prevalence in Eastern Europe.
Due to the scarcity of country-specific data, the lower and upper bounds of the prevalence of same sex behaviors in the last year among males in a given country was considered equal to the estimate for the region in which the country is located. Furthermore, the midpoint of the regional estimate was taken to reflect the proportion of sexually active men in a country who engage in same sex behavior. Estimates of the numbers of men engaging in same sex behaviors were derived from estimates of the numbers of sexually active men in the general population. Men between the ages of 15 and 64 years were considered sexually active, and data regarding the size of this population in countries in 2007 were derived from the United Nations 2006 revision of World Population Prospects.25 SPSS (version 16.0) was used for all analyses.
UNGASS Reporting and Inclusion of Data
Of the 147 LMIC that are Member States of the United Nations, only 66 (45%) reported to UNGASS in 2008 on at least 1 of the 4 indicators capturing the HIV prevention needs and responses among MSM (extent of HIV testing, coverage by HIV prevention programs, HIV knowledge and condom use). Reporting rates ranged from approximately 75% in Eastern Europe and Central Asia, Latin America and South and Southeast Asia to 25% and lower in sub-Saharan Africa, Middle East and North Africa, and Oceania. Calculations using data derived from Caceres et al8,9 indicate that the 66 reporting countries represent about half (52%) of the estimated total MSM population in all of the 147 LMIC.
Fifty-one countries reported on HIV testing among MSM (Table 3). Of those countries, 9 (18%) reported inconsistent or programmatic data and 42 (82%) reported data consistent with the UNGASS guidelines. Of the countries reporting data that were consistent with the guidelines, 8 reports were ineligible because they reported data from surveys in small samples (n < 100) or with unknown sample sizes. One of those 8 excluded reports provided a population estimate without specifying the original data. Reports from 34 countries were eligible for analysis of the extent of HIV testing among MSM (the data were consistent with the guidelines and sample sizes were greater than 99). The estimated number of MSM in these 34 countries represented 42% of the MSM population in all of the 147 LMIC.
Among eligible countries, the range of HIV testing estimates was wide. Except for the extreme value reported by Argentina (98%), HIV testing estimates ranged between 5% in Lao and 76% in Panama. Across the 34 LMIC, the global estimate weighted by MSM population size indicates that less than a third of MSM on average were tested for HIV (mean = 31%, median = 33%, SD = 15, range: 5%-98%). Excluding the value reported by Argentina only slightly reduced this weighted global estimate (mean = 30%, median = 33, SD = 13, range: 5%-76%).
In 3 regions, enough countries reported data to allow for the calculation of regional estimates weighted by MSM population size (Table 4). HIV testing among MSM was found to be lowest in South and Southeast Asia (9 countries, mean = 20%, median = 16%, SD = 12, range: 5%-58%), intermediate in Eastern Europe and Central Asia (9 countries, mean = 33%, median = 31%, SD = 9, range: 5%-70%), and higher in Latin America (10 countries, mean = 57%, median = 54%, SD = 21, range: 21%-98%).
HIV Prevention Programs
Thirty-three countries reported on the coverage of HIV prevention programs among MSM (Table 3). Of those countries, 5 (15%) reported data that were inconsistent with the guidelines (including programmatic data from Côte d'Ivoire and Lebanon). Of the 28 countries (76%) that provided data consistent with the guidelines, 23 had drawn data from surveys with a sample size greater than 99, rendering their reports eligible for analysis of prevention coverage. The estimated number of MSM in these 23 countries represented 37% of the MSM population in all of the 147 LMIC.
Argentina also reported the highest, near perfect, estimate of HIV prevention coverage among MSM (98%). Among the other countries, estimated HIV prevention coverage ranged from 10% in Armenia to 90% in Belarus. Across the 23 LMIC, the global estimate weighted by MSM population size indicates that on average only a third of MSM in LMIC were reached by HIV prevention programs (mean = 33%, median = 38%, SD = 16, range: 10%-98%). Excluding the high value reported by Argentina only slightly reduced this global estimate (mean = 32%, median = 38, SD = 13, range: 10%-90%).
Comparing the 3 regions represented by a sufficient number of reporting countries to calculate regional weighted estimates (Table 4), HIV prevention coverage among MSM was found to be lower in South and Southeast Asia (5 countries, mean = 26%, median = 18%, SD = 13. range: 13%-47%), intermediate in Eastern Europe and Central Asia (6 countries, mean = 30%, median = 17, SD = 20, range: 10%-90%), and higher in Latin America (9 countries, mean = 44, median = 26, SD = 31, range: 17%-98%).
Of the 37 countries that reported on HIV knowledge among MSM (Table 5), 4 (11%) presented inconsistent data (no programmatic data were reported) and the remaining 33 countries (89%) reported consistent data. Data from 27 of the latter countries were eligible as they were derived from samples greater than 99. The estimated numbers of MSM in these 27 countries represented 37% of the MSM population in all of the 147 LMIC. Among eligible countries, the estimated level of correct HIV knowledge ranged from 10% in the Philippines to 89% in Kyrgyzstan.
The global estimate across LMIC weighted by MSM population size shows that less than half of MSM held correct HIV knowledge (mean = 44%, median = 47, SD = 15, range: 10%-89%). Assessed regionally (Table 4), the weighted estimates of HIV knowledge among MSM were lower in South and Southeast Asia (6 countries, mean = 30%, median = 27%, SD = 10, range: 10%-44%) and in Eastern Europe and Central Asia (8 countries, mean = 37%, median = 26%, SD = 15, range: 26%-89%), compared with Latin America (8 countries, mean = 57%, median = 65%, SD = 16, range: 21%-85%).
Of the 58 countries reporting on condom use among MSM (Table 5), 8 (14%) presented inconsistent data (including programmatic data for Côte d'Ivoire and Lebanon). Of the 50 countries (86%) presenting consistent data, 44 had extracted those data from surveys with a sample size larger than 99, making the data eligible for analysis of condom use in MSM. The estimated number of MSM in these 44 countries represented 47% of the MSM population in all of the 147 LMIC.
Among the eligible countries, the estimated use of condoms at last anal sex with another man ranged from 24% in Pakistan to 90% in Argentina. The global estimate weighted by population size indicates that, on average, slightly more than half of MSM across LMIC had used condoms the last time they had sex with another man (mean = 54%, median = 61%, SD = 19, range: 24%-90%). Weighted estimates of condom use among MSM were found to be lower in South and Southeast Asia (9 countries, mean = 38%, median = 24%, SD = 19, range: 24%-88%), intermediate in both sub-Saharan Africa (5 countries, mean = 57%, median = 53%, SD = 8, range: 53%-75%) and Eastern Europe and Central Asia (10 countries, mean = 58%, median = 60%, SD = 10, range: 39%-83%), and higher in the Caribbean (3 countries, mean = 74%, median = 73%, SD = 8, range: 47%-79%), and highest in Latin America (13 countries, mean = 73%, median = 79%, SD = 17, range: 29%-90%).
The aim of this study was to assess the HIV prevention needs and responses among MSM in LMIC. In particular, we derived country-level, global and regional estimates of HIV testing, coverage by HIV prevention programs, HIV knowledge and condom use in these communities from data provided during the 2008 UNGASS reporting round.22 To the best of our knowledge, the data presented in this article represent the largest set of country-level information to date on each of the 4 indicators of HIV prevention needs and responses among MSM in LMIC. The analyses of these country data can substantially enhance understanding of the status and monitoring of HIV prevention needs and responses among MSM in LMIC at national and global levels and provide valuable information for at least 3 specific regions.
The findings show that the extent to which LMIC report on MSM remains limited. Of 147 LMIC, only 51 reported on the prevalence of HIV testing in MSM, 33 reported on coverage by HIV prevention programs, 37 reported on HIV knowledge and 58 countries reported on condom use. Strong variations were observed in the proportion of reporting countries per region: the highest reporting rates were observed in Eastern Europe and Central Asia, in Latin America and in South and Southeast Asia and the lowest reporting rates were found in sub-Saharan Africa, North Africa and Middle East, and Oceania. Moreover, for each of the indicators, a substantial minority of country reports provided data that were not eligible because the measurement of indicators differed substantially from the UNAIDS guidelines or because sample sizes were small or unspecified. These observations corroborate commentaries in the literature noting that the HIV epidemics among MSM are neglected in many countries11,16,26 and that the quality, utility and comparability of available data are limited.8,9
For each of the indicators of HIV testing, coverage by HIV prevention programs, HIV knowledge and condom use, between two-thirds and three-quarters of available country reports provided eligible data for analysis (HIV testing 67%; coverage of HIV prevention programs 73%; HIV knowledge 73%; condom use 76%). In addition, the estimated number of MSM across countries reporting eligible data represented a substantial proportion of the total estimated MSM population in all of the 147 LMIC (HIV testing 42%; coverage of HIV prevention programs 37%; HIV knowledge 37%; condom use 47%). Using a calculation of standard errors based on sample sizes, this study estimated the CIs around indicator values provided in eligible country reports. For a substantial proportion of countries, these CIs were wide. This reflects the limited reliability of indicator values derived from smaller samples and suggests that understanding of the HIV prevention needs and responses among MSM in some of the limited number of countries that do collect MSM-pertinent data is still surrounded by substantial uncertainty. Larger studies in more countries will be needed in the future to extend the availability and strengthen the reliability of country indicator values.
Nevertheless, until now there has been no compilation of data describing levels of HIV testing, coverage of HIV prevention programs, HIV knowledge and condom use among MSM in LMIC. In addition, although UNGASS data previously have been used mainly without weighting for the size of MSM populations,27 the weighted indicator values presented here enable the calculation of more accurate global and regional indicator estimates of HIV services coverage.
For each of the 4 indicators, regional weighted estimates were found to be much higher in Latin America than in the other regions, including South and Southeast Asia, Eastern Europe and Central Asia. Global weighted estimates across all reporting LMIC suggest that the HIV prevention needs among MSM in LMIC are high. On average, less than half of MSM in LMIC have correct HIV knowledge, slightly more than half of MSM used a condom the last time they had anal intercourse with another man, and less than a third of MSM in LMIC have tested for HIV. Limited HIV knowledge, condom use and HIV testing increase the risk and vulnerability of MSM to HIV infection,28 and the 2008 UNGASS data provide clear indications that scaling-up of prevention responses among MSM in LMIC is urgently needed. Notably, 2008 UNGASS data also show that only one third of MSM may have been reached by HIV prevention programs. To be optimally effective, a comprehensive response is needed to ensure the availability, accessibility and acceptability of effective HIV prevention services and programs and to address social exclusion and other structural factors that increase the vulnerability of MSM in LMIC to HIV infection.5,11
The data used in our analyses have several limitations, in addition to those noted above. Importantly, the MSM population estimates that were used in this study remain imperfect due to the ongoing scarcity of data on the prevalence of male-male sexual behaviors in most LMIC. Other limitations are inherent in the methodologies of country surveys, in particular data collection and recruitment strategies. MSM data reported by countries generally cannot be considered representative of the MSM populations in the country because those data are generally collected in 1 or a limited number of localities, usually the capital city and/or other major cities. Strengthening the representativeness of samples is an important challenge for future surveillance and research among MSM in LMIC. Furthermore, some country indicator value reports to UNGASS, in particular those that are “close to perfection,” may have been inflated, reflecting a social desirability bias. A major additional concern in this respect was the difficulty of external validation and triangulation of reported data using independent sources. For many countries included in this study, no published data were available on the indicators. For other countries, published data were not consistent with UNGASS requirements and reflected a heterogeneous set of ad hoc indicators that prevented triangulation. In addition, as many country reports do not specify the source of the submitted data or the details of the survey from which they were extracted, it remains unclear whether other sources provide independent data.
Several further limitations are inherent in UNGASS indicator specifications and guidelines. The information that is requested regarding the prevalence of HIV testing among MSM, for example, does not enable differentiation between client-initiated, voluntary counseling and testing, and the impact of provider-initiated or compulsory HIV testing policies and programs. Similarly, the requested information regarding the coverage of HIV prevention programs should be specified to better reflect the diversity, intensity and quality of HIV prevention programs. Methodological improvements in measurement of HIV prevention coverage are also required to better monitor further progress at the country level and across LMIC. Also, questions to assess HIV knowledge are basic and they may have limited sensitivity for differentiating between individuals and communities with more or less relevant HIV knowledge. Questions that tap individuals' beliefs regarding the transmission and prevention of HIV may be more informative than factual knowledge questions. Furthermore, if adequate understanding of risky and protective behaviors is to be achieved, condom use should be assessed beyond the last occasion of anal sex. Condom use data need to be put in perspective by also assessing the proportion of men who engage in anal sex, the number of partners with whom anal sex is practiced and the frequency of anal sex in these male-male sexual interactions, which may vary between countries, regions and cultures. Ideally, information would also need to be collected on types of partnerships and the HIV status of partners to gauge the extent to which unprotected sex is part of a range of risk-reduction strategies that have been observed among MSM in HIC, such as negotiated safety, serosorting and strategic positioning.29,30
These limitations mean that the data from the 2008 UNGASS reporting round cannot be regarded as the final word on the HIV prevention needs and responses of MSM in LMIC. However, the data do provide important insights for HIV prevention and can serve as a base line against which future reporting rounds may be compared. The estimates presented here mark the first time that global level data addressing essential elements of HIV prevention programs for MSM have been compiled and analyzed. They constitute a potentially valuable resource for monitoring progress in HIV responses and can serve as a powerful tool for advocacy.
Our findings clearly support previous studies, which found that MSM are at high risk for HIV infection in many LMIC. Findings also unmistakably show that monitoring and prevention responses remain insufficient and do not reach satisfactory numbers of MSM. Increasing surveillance efforts and scaling up high quality evidence-based HIV prevention for MSM in LMIC is urgently needed to influence behaviors and alter epidemic trends. Extending and improving monitoring, research and prevention among MSM in LMIC is not only a matter of promoting sound public health practice, it also contributes to ensuring the human rights of MSM.31
1. Hall HI, Song R, Rhodes P, et al. Estimation of HIV Incidence in the United States. JAMA. 2008;300:520-529.
2. Hamers FF, Phillips AN. Diagnosed and undiagnosed HIV-infected populations in Europe. HIV Med. 2008;9(Suppl 2):S6-S12.
3. Merson MH, O'Malley J, Serwadda D, et al. The history and challenge of HIV prevention. Lancet. 2008;372:475-488.
4. Coates TJ, Richter L, Caceres C. Behavioural strategies to reduce HIV transmission: how to make them work better. Lancet. 2008;372:669-684.
5. Gupta GR, Parkhurst JO, Ogden JA, et al. Structural approaches to HIV prevention. Lancet. 2008;372:764-775.
6. Jaffe HW, Valdiserri RO, De Cock KM. The reemerging HIV/AIDS epidemic in men who have sex with men. JAMA. 2007;298:2412-2414.
7. Bastos FI, Cáceres C, Galvão J, et al. AIDS in Latin America: assessing the current status of the epidemic and the ongoing response. Int J Epidemiol. 2008;37:729-737.
8. Cáceres C, Konda K, Pecheny M, et al. Estimating the number of men who have sex with men in low and middle income countries. Sex Transm Infect. 2006;82(Suppl 3):S3-S9.
9. Cáceres CF, Konda K, Segura ER, et al. Epidemiology of male same-sex behaviour and associated sexual health indicators in low- and middle-income countries: 2003-2007 estimates. Sex Transm Infect. 2008;84(Suppl 1):S49-S56.
10. Baral S, Sifakis F, Cleghorn F, et al. Elevated risk for HIV infection among men who have sex with men in low- and middle-income countries 2000-2006: a systematic review. PLoS Med. 2007;4:e339.
11. Cáceres CF, Aggleton P, Galea JT. Sexual diversity, social inclusion and HIV/AIDS. AIDS. 2008;22(Suppl 2):S45-S55.
12. Neilands TB, Steward WT, Choi KH. Assessment of stigma towards homosexuality in China: a study of men who have sex with men. Arch Sex Behav. 2008;37:838-844.
13. Vu BN, Girault P, Do BV, et al. Male sexuality in Vietnam: the case of male-to-male sex. Sex Health. 2008;5:83-88.
14. Abell N, Rutledge SE, McCann TJ, et al. Examining HIV/AIDS provider stigma: assessing regional concerns in the islands of the Eastern Caribbean. AIDS Care. 2007;19:242-247.
15. Geibel EM, Van der Elst N, King'ola S, et al. ‘Are you on the market?’: a capture-recapture enumeration of men who sell sex to men in and around Mombasa, Kenya. AIDS. 2007;21:1349-1354.
16. Wade AS, Kane CT, Diallo PA, et al. HIV infection and sexually transmitted infections among men who have sex with men in Senegal. AIDS. 2005;19:2133-2140.
17. Ma X, Zhang Q, He X, et al. Trends in prevalence of HIV, syphilis, hepatitis C, hepatitis B, and sexual risk behavior among men who have sex with men: results of 3 consecutive respondent-driven sampling surveys in Beijing, 2004 through 2006. J Acquir Immune Defic Syndr. 2007;45:581-587.
18. Colby DJ. HIV knowledge and risk factors among men who have sex with men in Ho Chi Minh City, Vietnam. J Acquir Immune Defic Syndr. 2003;32:80-85.
19. Kelly JA, Amirkhanian YA, McAuliffe TL, et al. HIV risk characteristics and prevention needs in a community sample of bisexual men in St. Petersburg, Russia. AIDS Care. 2002;14:63-76.
20. Choi KH, Lui H, Guo YQ, et al. Lack of HIV testing and awareness of HIV infection among men who have sex with men, Beijing, China. AIDS Educ Prev. 2006;18:33-43.
21. Horton R, Das P. Putting prevention at the forefront of HIV/AIDS. Lancet. 2008;372:421-422.
22. Secretary General of the United Nations. Implementation of the Declaration of Commitment on HIV/AIDS and Political Declaration on HIV/AIDS: midway to the Millennium Development Goals. New York, NY: UN; 2008.
23. UNAIDS. Monitoring the Declaration of Commitment on HIV/AIDS: Guidelines for Construction of Core Indicators, 2008 Reporting. Geneva, Switzerland: Joint United Nations Program on HIV/AIDS; 2008.
25. United Nations Population Division. World Population Prospects: The 2006 Revision
. Available at: http://esa.un.org/unpp/
. Accessed October 26, 2008.
26. Sanders EJ, Graham SM, Okuku HS, et al. HIV-1 infection in high risk men who have sex with men in Mombasa, Kenya. AIDS. 2007;21:2513-2520.
27. AmfAR. Special Report: MSM, HIV, and the Road to Universal Access-How Far Have We Come? New York, NY: American Foundation for AIDS Research; 2008.
28. Nguyen TA, Nguyen HT, Le GT, et al. Prevalence and Risk Factors Associated with HIV Infection Among Men Having Sex with Men in Ho Chi Minh City, Vietnam. AIDS Behav. 2008;12:476-482.
29. Race KD. Revaluation of risk among gay men. AIDS Educ Prev. 2003;15:369-381.
30. Halkitis PN, Moeller RW, Pollock JA. Sexual practices of gay, bisexual, and other nonidentified MSM attending New York City gyms: Patterns of serosorting, strategic positioning, and context selection. J Sex Res. 2008;45:253-261.
© 2009 Lippincott Williams & Wilkins, Inc.