Characterizing the Individual,Social,and Structural Determinants of Condom Use Among Men Who Have Sex with Men in Swaziland

Abstract

Within the broadly generalized HIV epidemic in Swaziland, men who have sex with men (MSM) have specific HIV acquisition and transmission risks. In the current era of expanding antiretroviral therapy-based prevention and treatment approaches, condom use remains a core component of mitigating these risks. A cross-sectional study characterizing the vulnerabilities for HIV among MSM in Swaziland was analyzed to describe factors associated with condom use at last sex with a male partner. Bivariate and multivariate logistic regression analyses were conducted to assess correlates of condom use at last sex with both casual and regular male partners. Disclosure of sexual practices to a healthcare provider and being able to count on other MSM to support condom use were significantly associated with condom use at last sex with a causal and a regular partner, respectively. Reporting difficulty insisting on condom use was inversely associated with condom use at last sex with both regular and casual partners. In addition, having faced legal discrimination was similarly inversely associated with condom use at last sex with a regular partner. Condom use among MSM in Swaziland may increase with improved partner communication, provider sensitization to encourage sexual disclosure, and the promotion of safer sex norms within MSM communities. These approaches, in combination with existing and emerging evidence of informed and human rights affirming prevention and HIV treatment approaches may reduce the incidence of HIV among MSM in Swaziland and all those in their sexual networks.

Introduction

There are clear and consistent data highlighting that men who have sex with men (MSM) are at increased risk of HIV infection compared with other reproductive-age men across HIV epidemic patterns.¹ This disproportionate burden is due, in part, to the increased HIV transmission probability associated with condomless anal intercourse with serodiscordant and viremic partners compared with condomless vaginal intercourse.² Other important factors include increased HIV prevalence in sexual networks among MSM, limited coverage of effective HIV prevention, treatment, and care options for MSM, and widespread enacted and perceived stigma affecting these men.^1,3

A meta-analysis of 83 published studies reporting HIV prevalence in 38 low- and middle-income countries (LMICs) demonstrated that MSM were 19.3 times as likely to be living with HIV compared with all reproductive-age adults.⁴ Additional subgroup analyses showed that higher HIV prevalence in the MSM population is not limited to any one region or income level or to countries with any given population-level HIV prevalence. Even in settings where HIV prevalence among the general population is high, increased burden has been noted in key populations, including sex workers, people who inject drugs, and MSM.^5,6 In fact, a South African study among rural men in the Eastern Cape Province showed that HIV prevalence was 3.6 times higher among men reporting a history of sex with another man [95% confidence interval (CI) 1.0–13.0, p = .05].⁷ Despite indications of a disproportionate HIV burden in the MSM population throughout the world, epidemiologic data for HIV risk among MSM in southern Africa remain limited.⁸

Condom use is highly effective in reducing HIV transmission and acquisition among MSM. A U.S. study demonstrated that the per-act risk of HIV infection was reduced by 78% with condom use compared with nonuse for receptive anal intercourse with serodiscordant partners.⁹ However, a global report, including data from 147 LMICs, by the United Nations General Assembly Special Session (UNGASS) in 2009 showed that only 54% of MSM used condoms at last anal sex with a male partner.¹⁰ It is important to note that reporting in many of these settings is incomplete as only 44 of the 147 countries consistently presented data on condom use among MSM. One study examining factors related to condom use among MSM in Port Elizabeth and Cape Town, South Africa, demonstrated that in these urban centers, condom use was a function of individual behaviors, motivational beliefs, partnership characteristics, and interpersonal factors.¹¹ Factors promoting condom use included a high degree of self-worth and self-efficacy, ready-made condom negotiation scripts, a supportive family context, and partners preferring condoms.¹¹ Factors related to condom use among MSM in other non-South African settings include perception of HIV risk, type of sexual partnership (regular versus casual), condom negotiation practices, history of coercion, criminalization, and access to care and services.^12

–15 While there is controversy about valid indicators for condom use, condom use at last sex has been commonly used as a means of limiting the impact of recall bias.^16,17

Structural barriers to care and health access are especially important in contexts where sexual behaviors are criminalized and MSM reportedly face compounded stigma and discrimination due to both their HIV status and same-sex sexual behavior.¹⁸ Swaziland, located between South Africa and Mozambique, has the highest national estimated HIV prevalence in the world with 26.5% of all reproductive-age adults estimated to be living with HIV and 2.4% HIV incidence among adults.^19
–21 Same-sex practices are criminalized in Swaziland, and few resources are deployed to understand the risks and experiences of MSM living with HIV.^18,22 Although sex between heterosexual serodiscordant partnerships has been shown to be associated with the majority of HIV transmission events nationally,²³ evidence shows that MSM in Swaziland face unique HIV acquisition and transmission risks.²⁴ Moreover, the 2009 Swaziland Modes of Transmission study suggests that male–male sexual practices are reported infrequently and are assumed to represent a minority of the fraction of new HIV infections nationally. However, the prevalence of these risk factors has not been measured or reported in the HIV surveillance systems used to inform these surveys.^23
–25 In addition, a systematic review estimating the prevalence of same-sex practices in LMICs described hidden HIV epidemics among MSM in sub-Saharan Africa.²⁶ These findings support the idea that same-sex sexual behavior may be more common than previously assumed and that HIV risk among this population is disproportionately high compared with other men.²⁷

Given the high burden of HIV and limited study of specific key populations, data describing the sexual practices among MSM appear relevant in completely characterizing the epidemic in Swaziland and appropriate responses. Consequently, this study aims to describe the individual, social, and structural factors that are related to condom use at last sex with a male partner among MSM in Swaziland to better understand what factors influence condom use among these men and develop evidence-based and human rights, affirming HIV prevention, treatment, and care options. The research questions guiding these analyses include separately describing the correlates of condom use at last sex with a main and casual male partner among MSM in Swaziland.

Materials and Methods

Sampling and survey administration

A cross-sectional survey was administered among MSM in Swaziland from July to December 2011 recruited through respondent-driven sampling (RDS), and these methods have been previously described in detail.^24,28 Briefly, men who were at least 18 years old, reported having anal sex with a man in the past 12 months, had a valid RDS recruitment coupon, and gave informed consent in English or Siswati were eligible for inclusion in the study. The survey instrument was administered by trained staff and included questions on sociodemographics (i.e., age, education, and income), behavioral HIV-related risk factors (i.e., sexual behaviors, condom use, and number of sexual partners), and structural factors (i.e., stigma, discrimination, and social network strength).²⁴ All participants received HIV and syphilis testing in accordance with official Swazi guidelines, and counseling and treatment or referrals were provided as necessary. This article is a secondary analysis of these data.

Analytical methods

Primary outcomes were dichotomous and defined as condom use at last sex with a casual male partner and condom use at last sex with a regular male partner. Covariates for analysis were chosen to represent the modules in the structured study instrument and were informed by previous literature on factors related to condom use among MSM as described above.^12

–15 The covariates included were age; income; education; sexual orientation; gender identity; HIV status from biological test results; worry about HIV/AIDS in the past 12 months; history of being raped; difficulty insisting on condom use with a partner with whom you have not always used them, a partner who does not want to use one, and a partner who provides regular economic support; having told any health worker about sexual history with men; can count on other MSM in a group of friends to support the use of condoms (as part of a validated social capital module²⁹); type of condom access when needed; fear of seeking health services as a result of sexual orientation or practice; and experience of legal discrimination as a result of sexual orientation or practice. Ultimately, 326 men were included in the analyses.

Data were analyzed using Stata statistical software, version 12.1 (College Station, TX). The focus of these analyses is not representative and generalizable estimates. Rather, it is about characterizing the relationship between potential determinants of condom use and reported condom use at last sex with male partners. Thus, these data are unweighted to avoid the introduction of significant variance with RDS adjustments given that variable-specific weights may not be applicable in multivariate analyses.³⁰

Two-sample proportion tests (proportion of men who used a condom at last sex for both casual and regular partners) were used to determine completeness of outcome variables. Skip patterns resulted in 6% of values omitted for the outcome condom use at last sex with a regular partner (n = 21) for participants lacking reported regular partners, and 35% of participants omitted a response to condom use at last sex with a casual partner (n = 113) due to lack of reported casual partners. Unadjusted and adjusted odds ratios for condom use at last sex were analyzed for casual partners, and stepwise forward, backward, and akaike information criterion (AIC)-based model selection methods were employed.³¹ Pearson's goodness of fit testing on condom use at last sex with a casual partner indicated the need for a generalized linear model (GLM) regression. The model was fitted with the covariates of difficulty insisting on condom use with a partner with whom you have not always used them and having told any healthcare worker about sexual history with men. In the analyses for both casual and regular male partners, the covariates, age, education, income, sexual orientation, and gender identity, were included in the final models as these factors are likely to be related to condom use, despite being found nonsignificant in this context.

In the model selected by AIC-based methods for condom use at last sex with a regular partner, the five covariates chosen were difficulty insisting on condom use with a partner with whom you have not always used them, difficulty insisting on condom use if a partner does not want to use one, can count on other MSM in a group of friends to support the use of condoms, experienced legal discrimination as a result of sexual orientation or practice, and history of being raped, although history of being raped was not significant in this model (aOR = 0.40 [0.14–1.14], p = .09). A logistic regression model was fitted for condom use at last sex with a regular partner using stepwise forward, backward, and AIC-based model selection methods. The AIC-based method generated a model, including difficulty insisting on condom use with a partner with whom you have not always used them, difficulty insisting on condom use if partner does not want to use one, can count on other MSM in a group of friends to support the use of condoms, experienced legal discrimination as a result of sexual orientation or practice, and history of being raped, as covariates. Again, covariates, age, education, income, gender identity, and sexual orientation, were also included in the final model.

Collinearity for the logistic regression analyses was checked by performing a multiple regression analysis to calculate the variance inflation factors of all covariates in the model, including age, income, education, gender identity, and sexual identification. Regression diagnostics were conducted by calculating probability of outcome, standardized Pearson's residuals, and leverage scores.³² Scatterplots of residuals versus fitted values, leverage, and outliers were generated and supported the validity of the logistic regression model. Predictive properties of the model were analyzed with measures of sensitivity, specificity, and a receiver operating characteristic curve.³² An interaction term for difficulty insisting on condom use with a partner with whom you have not always used condoms and difficulty insisting on condom use if a partner does not want to use condoms was created and tested in the model to assess effect modification. These analyses were completed to confirm that although the two covariates may be related on an individual level with a participant potentially referring to the same partner for both statements, both covariates can be included alone in the model without an interaction term.

Ethics

The study received approval for research on human subjects from both the Scientific Ethics Committee of Swaziland and the Institutional Review Board of the Johns Hopkins Bloomberg School of Public Health.

Results

Demographics

Key demographic statistics and covariate descriptions are displayed in Table 1. Participants were on average 23 years old (SD 4.96); 42.0% (137/326) of the participants had completed high school or secondary school, 23.6% (77/326) had additional training beyond secondary school, and 34.4% (112/326) had not completed secondary school. Average monthly income among participants was 2929 EZ (SD = 4413) or about 280 USD. Most participants identified their gender as male and sexual orientation as homosexual. This sample had an HIV prevalence of 17.0% (55/324), which is consistent with the general population of reproductive-age men up until 24–26 years, at which point the HIV prevalence among age-matched MSM appears to be significantly higher than the men sampled for the Swaziland Demographic and Health Survey.²⁴ Experiences of sexual violence were high, with 6.0% (19/317) having been raped.

Table 1.

Characteristics of Men Who Have Sex with Men in Swaziland

Variable	Categories	n	Percentage
Individual factors
Mean age (years)		23.2
Mean monthly income (EZ)		2938.3
Educational attainment	Not completed high school	111	34.2
	Completed high school	137	42.2
	Some posthigh school	77	23.7
Self-identified sexual orientation	Gay/homosexual	205	63.1
	Other (including heterosexual, bisexual, and other)	120	36.9
Self-identified gender	Man	235	72.5
	Woman	80	24.7
	Both	6	1.9
	Do not know	3	0.9
Used condom at last sex with a casual male partner	No	79	33.5
	Yes	157	66.5
Used condom at last sex with a regular male partner	No	91	29.3
	Yes	220	70.7
HIV status	Negative	269	83.0
	Positive	55	17.0
Has been worried about HIV/AIDS in the last 12 months	No	207	63.9
	Yes	117	36.1
Ever been raped	No	296	94.0
	Yes	19	6.0
Social factors
Difficulty insisting on condom use with a partner with whom you have not always used condoms in the past	Not difficult	147	46.2
	Difficult	171	53.8
Difficulty insisting on condom use with a partner who does not want to use one	Not difficult	182	57.8
	Difficult	133	42.2
Difficulty insisting on condom use with a partner who provides regular economic support	Not difficult	141	44.3
	Difficult	177	55.7
Told any health worker about history of sex with men	No	224	68.9
	Yes	101	31.1
Can count on other MSM in social group to support the use of condoms	Disagree	48	15.0
	Agree	272	85.0
Structural factors
Access to condoms when needed	No access	3	0.9
	Difficult or little access	58	18.0
	Some access	37	11.5
	Very easy access	225	69.7
Ever afraid to seek health services as a result of sexual orientation or practice	No	143	44.4
	Yes	179	55.6
Felt legal discrimination as a result of sexual orientation or practice	No	220	68.3
	Yes	102	31.7

MSM, men who have sex with men.

In this sample, 66.5% (157/236) of MSM used a condom at last sex with a casual partner and 70.7% (220/311) at last sex with a regular partner; 42.4% (134/316) of participants cited difficulty insisting on a condom with a partner who did not want to use one, 53.9% (172/319) cited difficulty insisting on a condom with a partner with whom they had not always used one, and 31.3% (102/326) of participants had disclosed their sexual history with men to a healthcare worker.

Factors with casual partners

Men who had difficulty insisting on a condom with a partner with whom they had not always used them were 70% less likely to have used a condom at last sex with a casual partner (p < .01) (Table 2). For participants who told a healthcare worker about their same-sex sexual behavior, odds of condom use at last sex with a casual partner were 2.91 times higher than those who had not disclosed (p < .01).

Table 2.

Odds of Having Used a Condom at Last Sex with Casual Male Partner Among Men Who Have Sex with Men in Swaziland

Covariate	OR [95% CI]	p	aOR [95% CI]	p
Individual factors
Age	1.00 [0.94–1.06]	.999	1.01 [0.94–1.08]	.833
Monthly income	1.00 [1.00–1.00]	.780	1.00 [1.00–1.00]	.545
Educational attainment	0.85 [0.56–1.28]	.434	0.80 [0.50–1.27]	.337
REF = not completed high school
Self-identified sexual orientation	0.85 [0.45–1.62]	.628	0.90 [0.45–1.82]	.770
REF = other (including heterosexual, bisexual, and other)
Self-identified gender	1.03 [0.92–1.17]	.587	1.05 [0.91–1.21]	.529
REF = man
HIV positivity	1.85 [0.72–4.73]	.199	—	—
Worried about HIV/AIDS in the last 12 months	0.83 [0.44–1.55]	.556	—	—
Ever been raped	0.51 [0.16–1.63]	.254	—	—
Social factors
Difficulty insisting on condom use with a
partner with whom you have not always used condoms in the past	0.35 [0.17–0.70]	.003	0.30 [0.15–0.62]	.001
partner who does not want to use one	0.51 [0.27–0.96]	.038	—	—
partner who provides regular economic support	0.69 [0.35–1.34]	.267	—	—
Told any health worker about history of sex with men	2.54 [1.24–5.30]	.013	2.91 [1.35–6.29]	.006
Can count on other MSM in social group to support the use of condoms	1.83 [0.85–3.94]	.120	—	—
Structural factors
Access to condoms when needed	1.11 [0.87–1.42]	.403	—	—
Ever afraid to seek health services as a result of sexual orientation or practice	0.53 [0.28–1.02]	.057	—	—
Has felt legal discrimination as a result of sexual orientation or practice	0.57 [0.31–1.08]	.085	—	—

CI, confidence interval.

Factors with regular partners

Of the 11 variables tested in association with condom use at last sex with a regular male partner, four were shown to have a significant relationship with the odds of condom use at last sex with a regular partner (Table 3). Condom use at last sex with a regular partner appeared to be more than two times more likely among men not reporting a history of being raped, although this was ultimately found to be not statistically significant (p = .087). No effect modification was found when testing for an interaction between condom use difficulty with a partner who does not want to use one and with a partner with whom you have not always used them. Participants who found it difficult to insist on condom use with partners with whom they had not always used them were half as likely to have used a condom at last sex with a regular partner, although this reached only borderline statistical significance in our analysis (p = .058). Participants who found it difficult to insist on condom use with a partner who does not want to use one were also half as likely to have used a condom at last sex with a regular partner (p < .05). The odds of condom use at last sex for men who could count on other MSM in their social group to support the use of condoms were 2.64 times higher than those who could not (p < .05). Men who experienced legal discrimination as a result of their sexual orientation or practices were about half as likely to have used a condom at last sex with a regular partner as those who did not experience legal discrimination (p < .05).

Table 3.

Odds of Having Used a Condom at Last Sex with Regular Male Partner Among Men Who Have Sex with Men in Swaziland

Covariate	OR [95% CI]	p	aOR [95% CI]	p
Individual factors
Age	1.01 [0.96–1.06]	.737	1.04 [0.97–1.11]	.250
Monthly income	1.00 [1.00–1.00]	.592	1.00 [1.00–1.00]	.880
Educational attainment	0.90 [0.65–1.25]	.533	0.95 [0.62–1.45]	.808
REF = not completed high school
Self-identified sexual orientation	0.89 [0.52–1.51]	.661	0.90 [0.50–1.69]	.741
REF = other (including heterosexual, bisexual, and other)
Self-identified gender	1.06 [0.79–1.43]	.686	1.04 [0.89–1.21]	.595
REF = man
HIV positivity	1.02 [0.53–1.96]	.960	—	—
Worried about HIV/AIDS in the last 12 months	0.90 [0.53–1.50]	.675	—	—
Ever been raped	0.39 [0.15–0.99]	.048	0.40 [0.14–1.14]	.087
Social factors
Difficulty insisting on condom use with a
partner with whom you have not always used condoms in the past	0.28 [0.16–0.48]	<.001	0.50 [0.24–1.03]	.058
partner who does not want to use one	0.33 [0.20–0.57]	<.001	0.46 [0.24–0.90]	.023
partner who provides regular economic support	0.37 [0.22–0.65]	<.001	—	—
Told any health worker about history of sex with men	1.01 [0.59–1.73]	.973	—	—
Can count on other MSM in social group to support the use of condoms	2.98 [1.55–5.71]	.001	2.64 [1.20–5.81]	.016
Structural factors
Access to condoms when needed	1.09 [0.90–1.33]	.376	—	—
Ever afraid to seek health services as a result of sexual orientation or practice	0.45 [0.27–0.77]	.004	—	—
Has felt legal discrimination as a result of sexual orientation or practice	0.43 [0.25–0.73]	.002	0.48 [0.26–0.88]	.018

Discussion

The data presented here demonstrate that a complex variety of individual, social, and structural factors influence condom use among MSM in Swaziland. Modifiable social and structural factors are emphasized in this discussion. We have firmly entered the era of antiretroviral therapy (ART)-based prevention and treatment approaches. However, increasing condom use remains a relevant component of addressing the HIV pandemic in key populations around the world, including the broadly generalized HIV epidemic in Swaziland. In this study, participants were less likely to use condoms if they had difficulty insisting on condom use with partners who opposed their use or with whom patterns of nonuse had already been established. Moreover, disclosing of sexual orientation to a healthcare worker was positively associated with condom use, although data suggested that stigma appears to be associated with lower condom use among these men.²²

For both regular and casual partners, MSM in Swaziland reported difficulty insisting on condom use with partners. This highlights the importance of positive partner communication and condom negotiation, but counteracting norms of silence and changing established sexual practices with both regular and casual partners may be difficult.^13,33 Encouraging ready-made negotiation scripts can help in these scenarios; such scripts have been shown to be associated with condom use in other contexts, including in South Africa.¹¹ However, condom change within households and regular partnerships has proven difficult, and reviews of condom promotion have shown limited success in sub-Saharan Africa and other parts of the world.³⁴ In light of these barriers to condom use within regular partnerships, other means of HIV prevention could become increasingly important, including pre-exposure prophylaxis (PrEP), postexposure prophylaxis (PEP), and universal coverage of treatment for those living with HIV, as well as the potential for effective ART-based rectal microbicides, as they are developed.^3,15,35 Promotion of these alternative methods, in conjuncture with improved condom and lube access, could support prevention options in a context where condom use is difficult and condom promotion messages alone cannot curb HIV acquisition or transmission risk.

Drawing upon social support within the MSM community may be a way to promote safer sex practices and prevention methods.^36,37 Among the men included in this study, condom use was positively associated with being able to count on the MSM community to support their use. Social norms around condom use should be promoted and vocalized within the MSM community in an attempt to normalize safer sex behaviors.^37,38 Furthermore, as universal coverage of ART for those living with HIV and PrEP services expand, initiatives promoting these prevention strategies may similarly be served by utilizing the strength of MSM communities.³⁹ By tailoring HIV services and messages directly to local MSM communities to increase acceptance and support of these services, prevention impacts may be optimized.

The criminalization of adult consensual same-sex practices appears to hamper the ability of MSM in Swaziland to use condoms. In this sample, those who faced legal discrimination were half as likely to have used a condom at last sex. Yet, in the face of criminalized sexual practices, disclosure of these practices to healthcare workers was positively associated with condom use. This demonstrates the benefit of dialogue between MSM and providers around sexual practices and safer sex strategies, and finding ways to maximize the potential of this opportunity is important. Other studies in sub-Saharan African settings similarly show a positive link between same-sex sexual behavior disclosure and reduced HIV risk⁴⁰; however, sexual orientation disclosure has proven more difficult in a context of criminalized practices. A recent review on key populations showed that criminalization of same-sex practices limits access to care and HIV service uptake and reduces disclosure of risks.⁴¹

In light of these findings, which suggest the importance of sexual behavior disclosure and the difficulty of doing so in settings that criminalize same-sex practices, it is critical for health providers to be aware of legal repercussions facing MSM to provide confidential and comprehensive services. Additionally, sensitization of healthcare workers and facilities to create safe spaces for MSM could provide the support needed among this population to disclose sexual practices and encourage open and safe communication with health providers.²² There have been demonstrated improvements in health service provision for MSM through sensitization of healthcare workers in sub-Saharan Africa.^42
–44

There are several limitations with the approaches used to collect these data. One limitation with research involving hidden populations such as MSM is unbiased recruitment of participants; while recruitment is still biased with RDS, analytic approaches can be used to overcome some of this. Unweighted RDS data were reported, so generalizability of these results is limited to the sample of men included in the study. However, the primary purpose of these analyses was to characterize predictors of condom use among the men recruited, so unweighted data were deemed sufficient for these analyses. Responses to this instrument were self-reported and some biases may be introduced given the sensitive topics and participant unwillingness to disclose various activities or statuses. This would potentially result in underreporting of risks, which would diminish the effects of various factors in our model and could have led some factors to appear nonsignificant.

In this analysis, HIV status after testing was used as a covariate, while it is possible that perceived HIV status may be more relevant in shaping sexual risk behaviors. In this study, 5% (19/326) of participants reported ever having been told they were HIV positive, which is insufficient power to determine the impact of known HIV positivity on condom use, so further work should prioritize examining associations based on perceived status. However, chi-square analyses comparing known HIV-negative, newly diagnosed HIV-positive (through study participation), and known HIV-positive men with regard to condom use at last sex with a regular and with a casual male partner showed no significant differences by group, further justifying the use of a single HIV status variable. An additional limitation of these analyses was the diminished sample sizes for the outcome variables due to some participants not reporting either a regular or casual partner, somewhat decreasing the power of our findings; future investigations could be of a larger scale if more MSM can be identified. Additionally, access to condom-compatible lubricants is an important factor when discussing condom use for MSM; however, these lubricants were not widely available at the time of this study and were therefore not included in the analysis; future investigations should describe access to and use of safe lubricants for anal sex.

Given the increased risk of HIV transmission among MSM, it appears unlikely that condom promotion and behavior change programs and messages can independently change the trajectory of HIV transmission among these men. Consequently, new and innovative ways to incorporate condom use into a package of HIV prevention services for MSM should be considered with input from communities of MSM. In addition to condom promotion and increased condom accessibility, such a prevention package could include condom-compatible lubricants, options such as PrEP and PEP, as well as antiretroviral-based rectal microbicides, in the future. Determinants that limit targeted provision or uptake of condoms, such as stigma and discrimination within communities and healthcare settings, need to be addressed with culturally specific sensitization trainings. As long as same-sex behaviors are criminalized in Swaziland, these sensitization trainings should emphasize confidentiality to minimize social harm against MSM. Although no one intervention will likely prevent new HIV infections in Swaziland or globally, by combining proven and effective methods such as condom use into a package of services for MSM, new cases of HIV among a population disproportionately affected by HIV can be prevented.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Beyrer

, Baral

, van Griensven

, et al.: Global epidemiology of HIV infection in men who have sex with men. Lancet, 2012; 380:367–377.

Baggaley

, White

, Boily

M-C

: HIV transmission risk through anal intercourse: Systematic review, meta-analysis and implications for HIV prevention. Int J Epidemiol, 2010; 39:1048–1063.

Sullivan

, Carballo-Diéguez

, Coates

, et al.: Successes and challenges of HIV prevention in men who have sex with men. Lancet, 2012; 380:388–399.

Baral

, Sifakis

, Cleghorn

, Beyrer

: 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.

HIV/AIDS JUNPo, (ed.): Sub-Saharan Africa: AIDS Epidemic Update Regional Summary. Geneva, Switzerland, United Nations Programme on HIV/AIDS (UNAIDS) and the World Health Organization (WHO). 2008.

Beyrer

, Baral

, Walker

, et al.: The expanding epidemics of HIV type 1 among men who have sex with men in low-and middle-income countries: Diversity and consistency. Epidemiol Rev, 2010; 32:137–151.

Jewkes

, Dunkle

, Nduna

, et al.: Factors associated with HIV sero-positivity in young, rural South African men. Int J Epidemiol, 2006; 35:1455–1460.

Baral

, Trapence

, Motimedi

, et al.: HIV prevalence, risks for HIV infection, and human rights among men who have sex with men (MSM) in Malawi, Namibia, and Botswana. PLoS One, 2009; 4:e4997.

Vitinghoff

, Douglas

, Judon

, et al.: Per-contact risk of human immunodificiency virus tramnsmision between male sexual partners. Am J Epidemiol, 1999; 150:306–311.

10.

Adam

, de Wit

, Toskin

, et al.: Estimating levels of HIV testing, HIV prevention coverage, HIV knowledge, and condom use among men who have sex with men (MSM) in low-income and middle-income countries. J Acquir Immune Defic Syndr, 2009; 52:S143–S51.

11.

Siegler

, de Voux

, Phaswana-Mafuya

, et al.: Elements of condom-use decision making among South African men who have sex with men. J Int Assoc Provid AIDS Care. 2014; 13:414–423.

12.

Mansergh

, Naorat

, Jommaroeng

, et al.: Inconsistent condom use with steady and casual partners and associated factors among sexually-active men who have sex with men in Bangkok, Thailand. AIDS Behav, 2006; 10:743–751.

13.

Elwood

, Greene

, Carter

: Gentlemen don't speak: Communication norms and condom use in bathhouses. J Appl Commun Res, 2003; 31:277–297.

14.

Chemnasiri

, Netwong

, Visarutratana

, et al.: Inconsistent condom use among young men who have sex with men, male sex workers, and transgenders in Thailand. AIDS Educ Prev, 2010; 22:100–109.

15.

Beyrer

, Sullivan

, Sanchez

, et al.: A call to action for comprehensive HIV services for men who have sex with men. Lancet, 2012; 380:424–438.

16.

Younge

, Salazar

, Crosby

, et al.: Condom use at last sex as a proxy for other measures of condom use: Is it good enough?. Adolescence, 2008; 43:927.

17.

Grunbaum

, Kann

, Kinchen

, et al.: Youth risk behavior surveillance—United States, 2003. MMWR Surveill Summ, 2004; 53:1–96.

18.

Kennedy

, Baral

, Fielding-Miller

, et al.: “They are human beings, they are Swazi”: Intersecting stigmas and the positive health, dignity and prevention needs of HIV-positive men who have sex with men in Swaziland. J Int AIDS Soc, 2013; 16:18749.

19.

UNAIDS: HIV AIDS estimates (2012): Swaziland; 2012.

20.

Reed

, Justman

, Bicego

, et al.: Estimating national HIV incidence from directly observed seroconversions in the Swaziland HIV Incidence Measurement Survey (SHIMS) longitudinal cohort. J Int AIDS Soc; 2012: Int AIDS Soc Avenue De France 23, Geneva, 1202, Switzerland.

21.

Bank

: Gross national income per capita 2008. Atlas method and PPP. Geneva: World Bank; 2009.

22.

Risher

, Adams

, Sithole

, et al.: Sexual stigma and discrimination as barriers to seeking appropriate healthcare among men who have sex with men in Swaziland. J Int AIDS Soc, 2013; 16:18715.

23.

HIV/AIDS SNERCo: Executive Summary: In: Swaziland HIV Prevention Response and Modes of Transmission Analysis. (UNAIDS, ed.) Mbabane, Swaziland, World Bank; 2009.

24.

Baral

, Ketende

, Mnisi

, et al.: A cross-sectional assessment of the burden of HIV and associated individual-and structural-level characteristics among men who have sex with men in Swaziland. J Int AIDS Soc, 2013; 16:18768.

25.

Baral

, Phaswana-Mafuya

: Rewriting the narrative of the epidemiology of HIV in sub-Saharan Africa. SAHARA J, 2012; 9:127–130.

26.

Cáceres

, Konda

, Segura

, Lyerla

: 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:i49–i56.

27.

Baral

, Grosso

, Holland

, Papworth

: The epidemiology of HIV among men who have sex with men in countries with generalized HIV epidemics. Curr Opin HIV AIDS, 2014; 9:156–167.

28.

Heckathorn

: Respondent-driven sampling: A new approach to the study of hidden populations. Soc Probl, 1997; 44:174–199.

29.

Fonner

, Kerrigan

, Mnisi

, et al.: Social cohesion, social participation, and HIV related risk among female sex workers in Swaziland. PLoS One, 2014; 9:e87527.

30.

Johnston

, O'Bra

, Chopra

, et al.: The associations of voluntary counseling and testing acceptance and the perceived likelihood of being HIV-infected among men with multiple sex partners in a South African township. AIDS Behav, 2010; 14:922–931.

31.

Sakamoto

, Ishiguro

, Kitagawa

: Akaike information criterion statistics. Dordrecht, The Netherlands: D Reidel; 1986.

32.

Dept

: Regression Diagnostics: Biostatistics 624 Lecture Notes. In: Biostatistics (JHUDo, ed.) Baltimore, MD; 2014.

33.

Crepaz

, Marks

: Serostatus disclosure, sexual communication and safer sex in HIV-positive men. AIDS Care, 2003; 15:379–387.

34.

Foss

, Hossain

, Vickerman

, Watts

: A systematic review of published evidence on intervention impact on condom use in sub-Saharan Africa and Asia. Sex Transm Infect, 2007; 83:510–516.

35.

McGowan

: Rectal microbicides: Can we make them and will people use them?. AIDS Behav, 2011; 15:66–71.

36.

Wohl

, Galvan

, Myers

, et al.: Do social support, stress, disclosure and stigma influence retention in HIV care for Latino and African American men who have sex with men and women?. AIDS Behav, 2011; 15:1098–1110.

37.

Dandona

, Dandona

, Gutierrez

, et al.: Sex behaviour of men who have sex with men and risk of HIV in Andhra Pradesh, India. AIDS, 2005; 19:611–619.

38.

Harawa

, Williams

, Ramamurthi

, Bingham

: Perceptions towards condom use, sexual activity, and HIV disclosure among HIV-positive African American men who have sex with men: Implications for heterosexual transmission. J Urban Health, 2006; 83:682–694.

39.

Herbst

, Beeker

, Mathew

, et al.: The effectiveness of individual-, group-, and community-level HIV behavioral risk-reduction interventions for adult men who have sex with men: A systematic review. Am J Prev Med, 2007; 32:38–67.

40.

Henry

, Awondo

, Fugon

, Yomb

, Spire

: Coming out of the Nkuta: Disclosure of sexual orientation associated with reduced risk behavior among MSM in Cameroon. Arch Sex Behav, 2012; 41:525–527.

41.

Beyrer

, Baral

, Kerrigan

, et al.: Expanding the space: Inclusion of most-at-risk populations in HIV prevention, treatment, and care services. J Acquir Immune Defic Syndr, 2011; 57:S96–S99.

42.

van der Elst

, Evans Gichuru

, Kanungi

, et al.: Experiences of Kenyan healthcare workers providing services to men who have sex with men: Qualitative findings from a sensitivity training programme. J Int AIDS Soc, 2013; 16:18741.

43.

Wolf

, Cheng

, Kapesa

: The epidemiology of HIV and prevention needs among men who have sex with men in Africa. J Int AIDS Soc, 2013; 16:18972.

44.

van der Elst

, Smith

, Gichuru

, et al.: Men who have sex with men sensitivity training reduces homoprejudice and increases knowledge among Kenyan healthcare providers in coastal Kenya. J Int AIDS Soc, 2013; 16:18748.