Background: The prevalence of HIV and sexually transmitted infections among men who have sex with men (MSM) has increased substantially in Vietnam. This study aimed to estimate the prevalence of HIV, syphilis, urethral gonorrhoea, and urethral chlamydia, and determined correlates of HIV infection among MSM in An Giang, Vietnam.
Methods: A group of 381 MSM were recruited in a community-based cross-sectional survey between August and December 2009. Face-to-face interviews were conducted for collecting data on sociodemographics, behaviors, and access to HIV prevention programs. Serological tests for HIV and syphilis and polymerase chain reaction for gonorrhoea/chlamydia were performed. Multivariate regression analyses were used to investigate the correlates of HIV infection.
Results: The prevalence of HIV, syphilis, gonorrhoea, chlamydia, and gonorrhoea/chlamydia were 6.3%, 1.3%, 1.8%, 3.2%, and 4.7%, respectively. HIV prevalence among 63 injecting MSM was significantly higher than that of 318 noninjectors (20.6% vs. 3.5%, P < 0.001). Approximately 40.4% identified as heterosexual and 42.8% had ever had sex with females. The rate of unprotected anal intercourse with another male in the last month was substantially high (75.3%). Injecting drugs (adjusted prevalence ratio [aPR] = 2.88, 95% confidence interval [CI]: 1.12–7.42), being transgender (aPR = 4.27, 95% CI, 1.17–15.57), and unprotected sex with a female sex worker (aPR = 4.88, 95% CI: 1.91–12.50) were significantly associated with HIV infection. The infection risk increased with age to a peak of 25 years and then decreased.
Conclusions: Although prevalence levels are lower in An Giang, Vietnam than in some other comparable locations, HIV/sexually transmitted infections prevention, and sexual health promotion targeting MSM are highly important in this location.
A community-based survey in An Giang, Vietnam, indicated moderately high prevalence levels of HIV/STIs among men who have sex with men (MSM) but lower than elsewhere in Vietnam.
From the *Pasteur Institute, Ho Chi Minh City, Vietnam; †Kirby Institute, The University of New South Wales, Sydney, NSW, Australia; ‡Oxford University Clinical Research Unit, Hospital for Tropical Diseases, Ho Chi Minh City, Viet Nam; §Vietnam Authority of HIV/AIDS Control, Hanoi, Vietnam; and ¶Provincial Centre for HIV/AIDS and Tuberculosis Control, An Giang, Vietnam
We thank the peer educators, data collectors, staffs, and leaders at the Preventive Medicine Centre in Long Xuyen and Chau Doc for their work and support; staff at the An Giang Provincial Centre for HIV/AIDS and Tuberculosis Control; Anh Tho Tran and Hung Viet Do and members in the provincial project management unit of the Vietnam HIV/AIDS Preventive Project in An Giang; A/Pro Lien Xuan Thi Truong, San Hoang Le, Ton Tran, Bao Quang Nguyen, Khanh Phu Dao, Cuc Thu Thi Cao, Phuong Kim Thi Tran, and other staff at the Pasteur Institute in HCMC. We would also like to thank Ms. Louisa Wright for her support in proofreading.
Conflicts of Interest and Source of Funding: This study was funded by the Vietnam HIV/AIDS Preventive Project in An Giang, Vietnam under the 2009 grant 04/HDTN/BQL-AG. The views expressed in this paper are those of the authors and do not necessarily represent the position of the Australian and Vietnamese Governments. We declare that no conflict of interest exists.
Correspondence: Quang Duy Pham, MD, Kirby Institute, The University of New South Wales, Sydney NSW 2052, Australia. E-mail: email@example.com; and Lei Zhang, PhD, Kirby Institute, The University of New South Wales, Sydney, NSW 2052, Australia. E-mail: firstname.lastname@example.org.
Received for publication February 19, 2012, and accepted May 13, 2012.
Approximately 254,000 people were living with HIV in Vietnam in 2010, leading to a national estimated prevalence among adults of ∼0.44%.1 A national HIV sentinel surveillance system has been implemented to monitor the epidemic for years.2 However, HIV prevalence data for men who have sex with men (MSM) have not been systematically collected, leading to a poor understanding about the pattern and extent of HIV transmission among Vietnamese MSM and the needs of public heath prevention programs targeting MSM.
There have been warning signs of a rapidly spreading HIV epidemic among MSM in major cities in Vietnam. Condom use by MSM is low in both homosexual and heterosexual encounters, and the proportion of MSM engaging in illicit drug use has substantially increased, from <10% in 2001–2005 to 20% to 30% in 2006–2009.3–7 Notably, <20% of Vietnamese MSM were tested for HIV, and 22% to 40% reported having sex with females in the past 12 months.3,4,7 Furthermore, the prevalence of sexually transmitted infections (STIs) among MSM has been estimated to be high in some Vietnamese cities. For instance, the pooled prevalence of syphilis, Neisseria gonorrhoea (NG), and Chlamydia trachomatis (CT) reached approximately 21% in Ho Chi Minh City (HCMC) in 2009.6 Using comparable sample sizes and similar methods of respondent-driven sampling, HIV prevalence among MSM between 2006 and 2009, respectively, increased markedly from 5.3% to 16.7% in HCMC as well as from 9.4% to 17.4% in Ha Noi.4,8 High levels of HIV have also been documented in other urban regions, such as Hai Phong (16.3% in 2009) and Can Tho (5.9% in 2009).8
An Giang is a southern Vietnam-Cambodia province with a population of 2 million people (General Statistics Office of Vietnam, 2010) (Fig. 1). In this province, high HIV prevalence was observed among injecting drug users (IDUs, 39%) and female sex workers (FSWs, 28%) in 2004.9 Despite a moderate-sized MSM population, numbered at 3800 to 11,500 individuals,1 there have not been any estimates of the prevalence of HIV and other STIs and risk behaviors in this population. Our descriptive cross-sectional study aimed to estimate the prevalence of HIV and STIs, including syphilis, urethral NG, and urethral CT among MSM in An Giang. It also determined correlates of HIV infection among this population.
MATERIALS AND METHODS
Participants and Data Collection
From August to December 2009, 389 potential male participants from 48 active MSM venues were recruited in a community-based survey with a target sample size of 380. Males were considered eligible for this survey if they were at least 15 years old, had been living in An Giang for at least 1 month, and self-reported having had oral and/or anal sex with another male in the past 12 months. After excluding people who did not meet the inclusion criteria or whose blood samples were not collected, the total sample size for this analysis was 381.
This survey included 3 steps. First, 2 local mapping teams were established and guided by “experienced” MSM to locate all known active MSM venues in the province. Second, peer educators accessed the selected venues, in which they invited potential males to participate the study by convenience sampling. Third, after completing informed consent forms, face-to-face interviews were held with the participants to elicit data on socio-demographics, sexual identity, sexual behaviors (for instance, the number of homosexual partners, types of sexual partnerships, condom and lubricant usage), history of STIs, alcohol and illicit drug use, knowledge related to HIV, and access to sexual health programs. Blood and urine samples were collected after finishing the interview.
In An Giang, sera were extracted daily and screened for syphilis using SD Bioline Syphilis 3.0 (Standard Diagnostics, Kyonggi-Do, Korea). All serum and urine samples were stored at –20°C and transported monthly to the Pasteur Institute, HCMC. In this institute, syphilis-positive sera with the screening test were confirmed using Treponema pallidum haemagglutination assay. The serum samples were tested for HIV antibodies using Genscreen HIV ½ V.2, and the HIV-positive sera were then confirmed with 2 other HIV antibody tests, including Murex HIV ½ Ag/Ab combination (Abbott, Kent, UK) and SFD HIV ½ PA. The Treponema pallidum haemagglutination assay, Genscreen, and SFD tests were made by Bio-Rad (Marnes La Coquette, France). Regarding diagnosis of urethral NG/CT, urine samples were analyzed by polymerase chain reaction with Amplicor NG/CT (Roche, NJ).
The test results were returned to the participants through local voluntary HIV counseling and testing clinics. Persons infected with syphilis, NG, and/or CT received free treatment according to the STI treatment syndrome guidelines.10 HIV-positive individuals were referred to local outpatient clinics. Ethical approval for this study was granted by the institutional review board at the Pasteur Institute, HCMC, Vietnam.
Statistical Data Analysis
Selected characteristics were stratified by noninjecting and injecting MSM, due to the prior studies which indicated significant differences between these populations.11,12 For 3 types of MSM sexual partnerships, including nonpaying partners, clients, and sex workers, we separately reported the average rate of reported condom use with 95% confidence interval (CI), which was estimated by using exact binomial method. Mann-Whitney U, χ2, and Fisher exact tests were used to determine the differences in stratified groups. The pooled prevalence of urethral NG and/or CT included those who were infected with either NG or CT or coinfected with both bacteria.
We used the prevalence ratio (PR) as a conservative, consistent, and interpretable measure of the magnitude of predictors for HIV in a multivariate regression model.13 We applied 2 different bivariate regression analyses, including Poisson regression analyses for categorical variables and fractional polynomials for continuous variables. In this process, if categorical variables yielded a P below 0.25 or were previously known to be an important risk factor (e.g., ever selling sex7 or history of inconsistent condom use in anal intercourse with male partners14), or continuous variables corresponded to a P < 0.05,15–17 these were included in multivariate analysis. In multivariate analysis, we used Poisson multivariate regression with backward elimination to identify the best-fitting model that described contributing variables. Data were entered using Epi-Data version 3.1 (EpiData Association, Odense, Denmark), and all statistical analyses were carried out in Stata version 12.0 (StataCrop, TX).
Demographics and Sexual Identity
The majority of the study participants were young MSM (median age: 20.4 years, interquartile range [IQR]: 18.0–25.1) and single, living alone, or living with their relatives (82.7%).
Up to 36.6% of respondents completed only primary education or were illiterate. Approximately one-half (51.4%) were unemployed or had a temporary occupation with low monthly income (median of 1.4 million [IQR: 0.9–2.0] Vietnam Dong ∼US $80). Sexual identity was reported as 39.6% homosexual, 40.4% heterosexual, and 20.0% transgender (Table 1).
Several significant differences in demographic characteristics and sexual orientation were observed between noninjecting and injecting MSM. Compared with the noninjectors, the injecting individuals had significantly higher median age (24.3 vs. 20.0), were more likely to be unemployed (81.0% vs. 45.6%), and more likely to identify themselves as heterosexual (63.5% vs. 35.9%); they were also less likely to obtain a secondary or higher level of education (49.2% vs. 66.3%), and less likely to be single, living alone, or living with relatives (74.6% vs. 84.3%) (Table 1).
Prevalence of HIV/STIs
The overall prevalence of HIV was 6.3% (95% CI: 4.1%–9.2%), and injecting MSM were more likely to be infected with HIV than noninjectors (20.6% vs. 3.5%, respectively, P < 0.001) (Table 1). Men who were unemployed, reported only insertive anal intercourse with male partners in the past 12 months, had ever had sex with a female, and had unprotected sex with an FSW in the past 12 months, had significantly higher HIV prevalence than other men. The seroprevalence rates among both heterosexual and transgender participants were higher than among homosexual males (9.1%, 7.9%, and 2.6%, respectively), approaching statistical significance (P = 0.056) (Table 2).
The prevalence of syphilis, urethral NG, CT, NG, and/or CT, and any of these STIs was 1.3% (95% CI: 0.4%–3.0%), 1.8% (95% CI: 0.7%–3.8%), 3.2% (95% CI: 1.6%–5.4%), 4.7% (95% CI: 2.8%–7.4%), and 6.0% (95% CI: 3.9%– 8.9%), respectively. No significant differences in the prevalence levels were observed between noninjecting and injecting subjects (Table 1). Similarly, these prevalence levels were not significantly different among 3 subgroups of MSM (reporting only insertive, both insertive and receptive, or only receptive anal intercourse in the past 12 months).
Substance Use, Sexual Risk Behaviors, and Condom Use
Substance use was remarkably high among the participants. The proportion of participants who reported ever using and injecting illicit drugs was 29.9% and 16.5%, respectively. Moreover, 73.2% of MSM surveyed used alcohol at least once per week, and this alcohol use was significantly higher among noninjecting MSM than among injecting MSM (75.5% vs. 61.9%, respectively, P = 0.026) (Table 1).
Selected sexual risk behaviors with both male and female partners are summarized in Table 1. The median age of sexual debut was 17 years (IQR: 15–18). None of the transgender participants reported a heterosexual encounter in their first sexual experience, whereas those who self-identified as homosexual or heterosexual males had a higher proportion, but still a minority, of first sexual experiences with a female (15.9% and 44.8%, respectively). The median number of male sexual partners in the past 3 months was only 2 (IQR: 1– 6). In the past 12 months, selling sex to male clients was reported by more than half of the participants (51.7%), and 326 (85.6%) study subjects had anal intercourse. In male-to-male sex partnerships, both insertive and receptive acts were commonly practised (44.4%), and the usage of water-based lubricant was low (12.6%). Moreover, 42.8% of MSM recruited had ever had heterosexual contact, and 13.4% reported having had sex with an FSW in the past 12 months. Across these behavioral indicators, the injecting respondents were more likely to have had sex with a female in their lifetime (73.0% vs. 36.8%, P < 0.001) and FSW in the past 12 months (27.0% vs. 10.7%, P = 0.001).
Condom use was considerably low in both heterosexual and homosexual sex. Only 45.9% of all respondents used condoms in their most recent homosexual encounter. Of 235 (61.7%) subjects who had anal sex with male partners in the past month, the overall rate of consistent condom use was only 24.7% (Table 1). The rate of reported condom use in the last anal intercourse and consistent condom use in the past month remained low across 3 types of homosexual contacts, including nonpaying male partners (45.5% and 23.0%), male clients (49.0% and 33.3%), and male sex workers (36.8% and 31.6%), respectively. Among those engaging in heterosexual practices in the past 12 months, condom use at the last vaginal and/or anal sex with female partners ranged between 50.4% and 80%, and the proportion reporting universal condom use was between 32.5% and 60% (Fig. 2).
History of HIV/STI Testing
The rate of HIV/STI testing among surveyed MSM remained relatively low. Approximately one-quarter of the participants had a history of ever being tested for HIV, and only 19.2% of these participants were tested in the past 12 months. Compared with the noninjecting participants, the injecting MSM were more likely to have ever been tested for HIV (41.3% vs. 21.7%, P = 0.001) or in the past 12 months (34.9% vs. 16.0%, P = 0.001) (Table 1). Furthermore, a low level (3.7%) of all participants reported ever having been diagnosed with an STI.
Associated Factors for HIV Infection
Bivariate regression analyses showed that 7 categorical variables (Table 2) plus (continuous) age were significant associated factors for HIV infection. However, after they were adjusted in a multivariate regression model, only being transgender (adjusted PR [aPR] = 4.27, 95% CI: 1.17–15.57), having had unprotected sex with an FSW in the past 12 months (aPR = 4.88, 95% CI: 1.91–12.50), and ever injecting illicit drugs (aPR = 2.88, 95% CI: 1.12–7.42) were significantly associated with HIV infection. Furthermore, the risk of HIV infection significantly increased with age peaked at age 25 years, then gradually decreased with age (Fig. 3).
The estimated prevalence of HIV among MSM in An Giang is considerably lower than the average level reported among MSM in Ha Noi, Hai Phong, HCMC, and Can Tho.5 Compared with these major Vietnamese cities, An Giang has a smaller IDU population (1000–2200 individuals) and also a lower HIV prevalence among IDUs.1,4,8 These factors probably limited HIV transmission among MSM in An Giang, as the HIV disease burden among IDUs is a key driving factor of the HIV epidemic in MSM.18 Further, the prevalence of STIs was substantially lower than the HIV prevalence among MSM in An Giang. This result is consistent with findings in a recent review of HIV and syphilis among MSM in southwest China, a border region with Vietnam, who also showed a much higher prevalence of HIV than of syphilis.19 Similar to Vietnam’s epidemic, the HIV epidemic in this region of China was also mainly driven by injecting drug use.20,21 These data suggest that the HIV epidemic among MSM in this region could be predominantly driven by sharing of injecting equipment more than unprotected sex. Consistently, we reported a high percentage of Vietnamese MSM having ever used illicit drugs and partaking in injecting behavior; injecting behavior was also found to be an associated factor for HIV infection in this study. Drug use has a strong link with male-to-male commercial sex.22,23
As shown in our study, MSM who self-identified as heterosexual males may play an important role in acting as a bridge for HIV/STIs from high-risk populations to the general population because of their prevalent bisexual behaviors and low condom usage in both male and female partnerships. Similar to findings among Cambodian MSM,24 our results strongly indicated that having unprotected sex with an FSW significantly increased the likelihood of HIV infection among Vietnamese MSM. These MSM participants, who had unprotected sex with an FSW, demonstrate lower risk behaviors with male partners than otherwise, including less frequent commercial sexual activities, lower overall number of male sex partner, and lower practice of anal-receptive intercourse. However, they reported a much higher rate of injecting drug use (35.7% vs. 15.0%), suggesting that sharing of injecting equipment is more likely to be the dominant transmission route in this subpopulation.
Being transgender was significantly associated with HIV infection in MSM. Although the transgender participants were less likely to inject drugs (5.3%) and have heterosexual partners in their lifetime (7.9%), they had significantly more frequent commercial sex activities, high number of male sexual partners, and frequent anal-receptive intercourse, which substantially increases their risk of acquiring HIV.7,25–27 In addition, our cross-sectional survey was not able to determine the MSM participant’s age at which they were infected with HIV. Nonetheless, we reported a nonlinear relationship between risk of HIV infection and the age, suggesting that MSM in their mid-20s have the highest risk of acquiring HIV. The risk decreased markedly with increasing age, indicating a reduction of risk behaviors in these age-groups.
This survey contained several limitations. First, our survey was possibly subjected to selection bias because of our recruitment method, as potential MSM were accessed and invited through convenience sampling in only mapped venues. Therefore, selection bias could occur if large heterogeneities of behaviors and HIV/STI prevalence exist between participants and nonparticipants. Second, some participants may have underreported their risk behaviors, such as anal sex and illicit drug use, due to the sensitive nature of the questions. Third, fellatio and receptive anal intercourse were commonly reported among the participating MSM, thus the lack of collection and testing of NG and CT in pharyngeal and anal swabs could result in an underestimate of the prevalence of these infections.28,29
This study has important implications for HIV prevention strategies for Vietnamese MSM. This survey indicated a substantial proportion of injection illicit drug usage among Vietnamese MSM, which is a primary risk factor for HIV infection. The expansion of existing needle/syringe and methadone maintenance therapy programs for injecting MSM should be one of the priorities to reduce their risk of HIV infection. In such programs, injecting MSM peer educators should be employed to distribute educational materials and harm reduction commodities, such as safe behavior booklets, free sterile needles, and information about methadone treatment. Because of the common practice of unprotected intercourse in both homosexual and heterosexual sex, there is an urgent need to promote the 100% condom use and lubricant program to all MSM to prevent further transmission of HIV/ STI among the population of MSM and also to their regular female partners. Given the higher prevalence of HIV found among MSM in other urban areas of Vietnam, it could be expected that HIV prevalence in An Giang will also rise toward those levels. It is important that prevention measures are established now. Special attention should be paid to young MSM and transgender people due to their higher risk of HIV infection. The social marketing campaign targeting MSM should also aim to increase the rate of uptake of HIV/STI testing and treatment services.30 Finally, we strongly recommend that MSM need to be included in the national sentinel HIV/STI surveillance system for better understanding and monitoring of the epidemiologic trends of the infections in this at-risk population.
2. Hien NT, Long NT, Huan TQ. HIV/AIDS Epidemics in Vietnam: Evolution and Responses. AIDS Educ Prev 2004; 16 (suppl_a): 137–154.
3. Colby DJ. HIV knowledge and risk factors among men who have sex with men in Ho Chi Minh City, Vietnam. J AIDS 2003; 32: 80–85.
4. National Institute of Hygiene and Epidemiology, Family Health International, Viet Nam Administration of HIV/AIDS Control, United States Agency for International Development. Results from the HIV/STI integrated biologic and behavioral surveillance (IBBS) in Vietnam, 2005–2006. Ha Noi, Vietnam: Ministry of Health, 2006. Available at: http://www.unaids.org.vn/resource/topic/evaluation/conference/documents/d3_e.pdf
6. Tuan NA, Hoang TV. HIV/STI Integrated Behavioral and Biological Surveillance in Vietnam (IBBS) 2009 (Round 2). Presented at: 4th National Scientific Conference on HIV/AIDS; December 1–3, 2010; Ha Noi, Vietnam.
7. 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 and Behavior 2008; 12: 476–482.
8. Hien NT. HIV Epidemic in Vietnam. Presented at: 4th National Scientific Conference on HIV/AIDS; December 1–3, 2010; Ha Noi, Vietnam.
9. Anh MH. Workshop on HIV/AIDS prevention and control in An Giang in 2008. An Giang, Vietnam: An Giang Provincial Center for AIDS and Tuberculosis Control and Prevention, 2009.
11. McKetin R, Ross J, Kelly E, et al.. Characteristics and harms associated with injecting versus smoking methamphetamine among methamphetamine treatment entrants. Drug Alcohol Rev 2008; 27: 277–285.
12. Semple SJ, Patterson TL, Grant I. A comparison of injection and non-injection methamphetamine-using HIV positive men who have sex with men. Drug Alcohol Depend 2004; 76: 203–212.
13. Thompson ML, Myers JE, Kriebel D. Prevalence odds ratio or prevalence ratio in the analysis of cross sectional data: What is to be done? Occup Environ Med 1998; 55: 272–277.
14. Xiao Y, Sun J, Li C, et al.. Prevalence and correlates of HIV and syphilis infections among men who have sex with men in seven provinces in China with historically low HIV prevalence. J AIDS 2010; 53: S66–S73.
15. Royston P, Altman DG. Regression using Fractional Polynomials of Continuous Covariates: Parsimonious Parametric Modelling. J R Stat Soc Ser C Appl Stat 1994; 43: 429.
16. Royston P, Ambler G, Sauerbrei W. The use of fractional polynomials to model continuous risk variables in epidemiology. Int J Epidemiol 1999; 28: 964–974.
17. Sauerbrei W, Royston P. Building multivariable prognostic and diagnostic models: Transformation of the predictors by using fractional polynomials. J R Stat Soc Ser A Stat Soc 1999; 162: 71–94.
18. Volz E, Frost SDW, Rothenberg R, et al.. Epidemiological bridging by injection drug use drives an early HIV epidemic. Epidemics 2010; 2: 155–164.
19. Chow EPF, Wilson DP, Zhang L. HIV and syphilis co-infection increasing among men who have sex with men in China: A systematic review and Meta-Analysis. PLoS ONE 2011; 6.
21. Jia M, Luo H, Ma Y, et al.. The HIV epidemic in Yunnan province, China, 1989–2007. J Acquir Immune Defic Syndr 2010; 53 (suppl 1): S34–S40.
22. Clatts MC, Giang LM, Goldsamt LA, et al.. Male sex work and HIV risk among young heroin users in Hanoi, Vietnam. Sex Health 2007; 4: 261–267.
23. Vu BN, Mulvey KP, Baldwin S, et al.. HIV risk among drug-using men who have sex with men, men selling sex, and transgender individuals in Vietnam. Cult Health Sex 2012; 14: 167–180.
24. Girault P, Saidel T, Song N, et al.. HIV, STIs, and sexual behaviors among men who have sex with men in Phnom Penh, Cambodia. AIDS Educ Prev 2004; 16: 31–44.
25. Chariyalertsak S, Kosachunhanan N, Saokhieo P, et al.. HIV incidence, risk factors, and motivation for biomedical intervention among gay, bisexual men, and transgender persons in Northern Thailand. PLoS ONE 2011; 6.
26. Feng Y, Wu Z, Detels R, et al.. HIV/STD prevalence among men who have sex with men in Chengdu, China and associated risk factors for HIV infection. J AIDS 2010; 53: S74–S80.
27. Zhong F, Lin P, Xu H, et al.. Possible increase in HIV and syphilis prevalence among men who have sex with men in Guangzhou, China: Results from a respondent-driven sampling survey. AIDS Behav 2011; 15: 1058–1066.
28. Ota KV, Tamari IE, Smieja M, et al.. Detection of Neisseria gonorrhoeae
and Chlamydia trachomatis
in pharyngeal and rectal specimens using the BD Probetec ET system, the Gen-Probe Aptima Combo 2 assay and culture. Sex Transm Infect 2009; 85: 182–186.
29. Peters R, Verweij S, Nijsten N, et al.. Evaluation of sexual history-based screening of anatomic sites for Chlamydia trachomatis
and Neisseria gonorrhoeae
infection in men having sex with men in routine practice. BMC Infect Dis 2011; 11: 203.
30. Wei C, Herrick A, Raymond HF, et al.. Social marketing interventions to increase HIV/STI testing uptake among men who have sex with men and male-to-female transgender women. Cochrane Database Syst Rev 2011; 9.