HIV prevalence,sexual behaviours and engagement in HIV medical care among an online sample of sexually active MSM in Venezuela

Abstract

In Venezuela, members of a social and sexual partner networking site for men who have sex with men (MSM) completed an online survey regarding sexual behaviours and HIV medical care. Among the 2851 respondents, self-reported HIV prevalence was 6.6%. Of participants living with HIV, 73.2% reported taking antiretroviral medication and 56.6% reported complete adherence within the past month. Participants living with HIV were more likely to be older (aOR = 1.04 per one-year increase in age, 95% CI: 1.02, 1.06) and diagnosed with a sexually transmitted infection in the previous year (aOR 3.26, 95% CI: 2.11, 5.04). These data provide further understanding of the HIV epidemic among MSM in Venezuela, and potential targets for HIV prevention interventions.

Keywords

Sexual behaviour homosexual men Venezuela men who have sex with men HIV AIDS sexually transmitted infection STI Latin America

Introduction

In Latin America, men who have sex with men (MSM) are estimated to be 30 times more likely to contract HIV compared to the general population.^1–5 Moreover, the disproportionate vulnerability for HIV infection among MSM is reported to be higher in Latin America and the Caribbean than in any other region of the world.⁴ Country-specific data are necessary to inform HIV-related research, policy and programmatic work; yet an in-depth understanding of the epidemiology of HIV among key populations is limited for some Latin American countries.

With an estimated generalised HIV prevalence of 0.6% among adults aged 15–49, the República Bolivariana de Venezuela (Venezuela) has one of the highest HIV prevalences in the Andean Region and third highest in Latin America.^1,3,6 As such, characterising epidemiologic profiles of HIV and understanding factors related to engagement in HIV medical care among those most impacted by the epidemic is important for curbing continued HIV spread. Key populations in Venezuela are believed to be MSM and female sex workers (FSWs).^3,6–8 Compared to other countries in Latin America where evidence suggests heightened vulnerability to HIV among transgender women and sex workers in addition to MSM, there is a dearth of research to support these claims in Venezuela.⁹ The scarcity of evidence regarding HIV risk factors for key populations may be due in part to the social, political and economic climate in Venezuelan over the past few decades.^10,11 Specifically, the change in political ideology and growth in economic and social instability may have posed unique barriers to describing behavioural and other health-related characteristics associated with HIV vulnerability in Venezuela.¹²

To address this gap, members of one of the largest online social and sexual partner seeking websites for MSM were invited to participate in an online survey regarding sexual behaviours and engagement in HIV medical care in Venezuela. Due to an increase in popularity of online and mobile applications for social and sexual partner seeking among MSM, there are over six million active MSM users globally,¹³ these sites provide an opportunity to recruit members of a group potentially at high risk for HIV with relative ease where conventional methods may not be feasible. Seeking to contribute to the limited available data regarding MSM in Venezuela, here we describe the HIV testing behaviours, prevalence and engagement in medical care among sexually active MSM in Venezuela.

Methods

Participants and procedures

Between October and November 2012, an anonymous online survey was administered to members of a social and sexual networking site for MSM in Spanish- and Portuguese-speaking countries and territories in Latin America and the Caribbean, and in Spain and Portugal. The present analysis was restricted to individuals who reported currently living in Venezuela, who reported both a male sex assigned at birth and current male gender identity (i.e. cisgender men). To protect anonymity of participants the website name cannot be disclosed and IP addresses were not recorded in the database. Nonetheless, to minimise the likelihood that the same individual would complete the survey more than once, the survey software (Qualtrics) was programmed to only allow one response per IP address. Although it is possible that a participant could have completed the survey more than once from a different IP address, the length of the survey and lack of incentive for participation make this unlikely. An email recruitment message was sent to all users at the time of the study that had indicated their residence was in Latin America. The recruitment email provided a description of the study and included a link to the study website. Individuals who visited the study website were able to read a more detailed description of the study procedures and, if interested, proceed to the study consent form. Those who decided to participate were directly sent from the consent form to the study questionnaire, which took approximately 30 min to complete. All participants were ≥18 years of age. Detailed methods have been described elsewhere.¹⁴ In Venezuela specifically, a total of 20,342 emails were opened, 4935 individuals clicked the provided link to the survey, and 3175 completed the survey. Ethical approval was obtained from the Institutional Review Board at the Fenway Institute at Fenway Health in Boston, MA.

Measures

Demographics

This included age, urbanicity (coded as currently living in an urban versus rural area), and education (coded as at least some university-education or above versus less than university education). Given that the original survey included a wide range of countries and economies, a relative measure of income/class was chosen to assess socioeconomic status (coded as no income, low income/low class, middle income/middle class, high income/upper class). Sexual orientation was obtained by asking participants to select terms that best described their sexual orientation and responses were coded as homosexual/gay identity versus any other sexual orientation. Sexual role was measured by asking participants to describe their sexual role during intercourse and participants were able select multiple roles, if applicable. For this analysis, if participants endorsed top (i.e. insertive sexual role) and not bottom (i.e. receptive sexual role) and not versatile (i.e. endorsing both insertive and reception sexual roles), they were coded as primarily ‘top’, similar coding was conducted for bottom and versatile sexual roles.

History of HIV/STI testing and diagnosis

This was measured by asking participants to self-report if they had ever been tested for HIV and if they had ever been told by a healthcare professional that they had HIV, if they had ever been tested for sexually transmitted infections (STIs) and if they had been told by an healthcare provider that they had syphilis, gonorrhoea, chlamydia, HPV/genital warts, or genital herpes, or if they had ever been told by a healthcare professional that they had hepatitis A, hepatitis B, or hepatitis C. Individuals who reported that they did not know or preferred not to respond to questions related to HIV and STI were excluded from the analysis.

Sexual behaviour

This was measured by asking participants if they had been sexually active in the previous three months, how many male partners they had in the past three months, and how often they used condoms with these partners in the same time period. Participants who reported any condomless anal intercourse in the previous three months were coded as having engaged in condomless anal intercourse.

HIV medical care

Data were collected among participants who reported being diagnosed with HIV. Participants were asked if they were currently engaged in medical care for their HIV, if they were currently prescribed antiretroviral therapy (ART), and adherence for those prescribed ART. Using a scale from 0% to 100%, participants were asked to rate their own level of adherence in the previous month.

Outcomes

The primary outcomes of interest were (1) having ever been HIV tested, (2) self-reporting having been diagnosed with an HIV infection and (3) current engagement in HIV-related medical care among those with self-reported HIV infection. Participants were asked if they had ever had an HIV test. Participants who responded ‘yes’ were then asked if a healthcare professional had ever diagnosed them with HIV infection; those who responded ‘yes’ were coded as living with HIV. Those who responded ‘no’ were coded as not living with HIV. Individuals who reported no history of HIV testing, or who refused to answer questions regarding HIV serostatus were not included in analyses of HIV serostatus. Finally, participants who self-reported living with HIV were asked if they were currently being seen by a doctor for HIV treatment. Those who responded ‘yes’ were coded as currently engaged in HIV-related medical care.

Data analysis

Proportions for categorical variables and means and standard deviations for continuous variables were calculated overall and by HIV serostatus of respondent. Separate bivariate logistic regression models were used to assess the associations between each outcome, i.e. (1) having ever been tested for HIV; (2) having been diagnosed with HIV; and (3) engagement in HIV medical care (among those reporting HIV infection) with demographic and sexual behaviour variables, including age, urbanicity, sexual orientation (coded as gay/homosexual or other sexual orientation), education, income, STI diagnosis in the past year, sexual role (top/insertive, bottom/receptive, or versatile), and condomless anal intercourse in the previous three months.

Then, three separate multivariable logistic regression models were constructed using backwards selection with a p-value of 0.2 as the cutoff to assess factors associated with: (1) having ever been tested for HIV; (2) having been diagnosed with HIV; and (3) engagement in HIV medical care among those reporting HIV infection. The backward selection procedure begins with all candidate variables in the model, and removes them one by one until all final variables have a p-value of 0.2 or less. Models assessing factors related to HIV serostatus and engagement in HIV medical care were restricted to individuals with a history of HIV testing. Because the focus is on strength of associations, analyses were not adjusted for multiple comparisons.¹⁵ A complete case analysis was conducted for all analyses. All analyses were conducted in Stata 12.0 (StataCorp, College Station, TX).

Results

Table 1 lists the sample characteristics of the 2851 respondents including demographics, sexual behaviours and HIV and STI testing history. Participants who self-reported living with HIV were significantly older, mean of 33.4 years (SD = 9.0) compared to 30.3 (SD = 9.3) for HIV uninfected (p = 0.003) men. Those who self-reported living with HIV were more likely to report having ever been diagnosed with any STI (27%) compared to 9.8% among participants not living with HIV (p < 0.001). Among participants who self-reported living with HIV a significantly higher prevalence of the following STIs was also reported: syphilis (7.4% versus 3.1%; p = 0.005), HPV (15.8% versus 4.2%; p < 0.001), HSV (6.3% versus 2.2%; p = 0.002) and hepatitis B (5.9% versus 1.2%; p < 0.001).

Table 1.

Descriptive characteristics of full sample of MSM who use the Internet to find partners in Venezuela and by self-reported serostatus.

Characteristic	HIV⁺ (N = 188)	HIV⁻ (N = 2663)	Total (N = 2851)
Age (years, mean [SD])	33.4 (9.0)	30.3 (9.3)	30.5 (9.3)
Urban-dwelling	181 (96.3%)	2519 (94.6%)	2700 (94.7%)
Sexual orientation^b
Homosexual/gay	175 (93.1%)	1853 (70.0%)	2028 (71.5%)
Bisexual	13 (6.9%)	675 (25.5%)	688 (24.3%)
Heterosexual	0	21 (0.8%)	21 (0.7%)
Unsure	0	77 (2.9%)	77 (2.7%)
Other	0	21 (0.8%)	21 (0.7%)
University education or above	156 (83.4%)	2128 (80.4%)	2284 (80.6%)
Income
No income	7 (3.9%)	162 (6.3%)	169 (6.2%)
Low income	15 (8.2%)	261 (10.2%)	276 (10.1%)
Middle income	151 (83.0%)	2005 (78.5%)	2156 (78.8%)
High income	9 (5.0%)	126 (4.9%)	135 (4.9%)
Any sexually transmitted infection diagnosis, past 12 months^b	51 (27.6%)	256 (9.8%)	307 (10.9%)
Sexual role
Top	15 (8.7%)	487 (18.8%)	502 (18.1%)
Bottom	46 (26.6%)	458 (17.6%)	504 (18.2%)
Versatile	112 (64.7%)	1651 (63.6%)	1763 (63.7%)
Any condomless anal sex in the past three months	59 (43.4%)	1041 (52.9%)	1100 (52.3%)

Participants were instructed to ‘check all that apply’, percentages will not add up to 100%.

Including syphilis, gonorrhoea, chlamydia, human papillomavirus, and/or herpes simplex virus.

Of those who self-reported living with HIV (N = 188), 96.2% answered questions related to their HIV medical care. Of these, 135 (75.4%) reported being engaged in HIV medical care, 131 (73.2%) were currently taking ART, and 73 (56.6%) of those who were on ART reported 100% adherence to their antiretroviral regimen in the past month. Of 44 (42.3%) participants engaged in HIV medical care reported condomless anal intercourse in the previous three months compared to 15 (48.4%) participants who did not report engagement in HIV medical care (p = 0.68). Participants who self-reported currently taking ART reported less frequent condomless anal intercourse as compared to participants who reported not being currently on ART (42.0% versus 48.6%, respectively [p = 0.56]). Similarly, condomless anal intercourse is less frequently reported among participants who self-reported 100% ART adherence in the past month (N = 19, 35.9%) compared to those who self-reported less than 100% ART adherence in the past month (N = 22, 47.8%; p = 0.23).

Table 2 lists results of bivariate and multivariable models assessing factors associated with self-reported HIV testing history, living with HIV and engagement in HIV medical care. In a final multivariable model, factors associated with increased odds of reporting having had an HIV test (Model 1) included older age (aOR = 1.08, 95% CI: 1.06, 1.10), homosexual or gay sexual orientation (aOR = 2.10, 95% CI: 1.59, 2.79), middle income compared to no income (aOR = 1.80, 95% CI: 1.11, 2.90), and an STI diagnosis in the past year (aOR = 2.38, 95% CI: 1.37, 4.14).

Table 2.

Factors associated with self-reported HIV testing, living with HIV and engagement in HIV medical care among Venezuelan MSM.

	HIV testing		Participants self-reporting living with HIV		Engaged in HIV medical care
	Bivariate OR (95% CI)	Multivariable^a aOR (95% CI)	Bivariate OR (95% CI)	Multivariable^a aOR (95% CI)	Bivariate OR (95% CI)	Multivariable^a aOR (95% CI)
Age (per year)	1.10 (1.08 to 1.11)	1.08 (1.06 to 1.10)	1.02 (1.01 to 1.04)	1.04 (1.02 to 1.06)	1.17 (1.10 to 1.24)	1.18 (1.09 to 1.27)
Urban (vs. rural)	1.90 (1.30 to 2.79)	^b	1.27 (0.58 to 2.78)	^b	1.26 (0.24 to 6.72)	^b
Homosexual/gay identity (vs. any other sexual orientation)	1.80 (1.45 to 2.23)	2.10 (1.59 to 2.79)	5.21 (2.94 to 9.23)	4.88 (2.35 to 10.2)	3.45 (1.05 to 11.3)	^b
Education (university or above vs. secondary or below)	2.49 (1.98 to 3.13)	1.32 (0.94 to 1.84)	1.00 (0.67 to 1.49)	^b	2.05 (0.89 to 4.74)	^b
Income
No income	Ref	Ref	Ref	^b	Ref	^b
Low income	2.31 (1.50 to 3.55)	1.72 (0.96 to 3.11)	1.06 (0.42 to 2.67)		1.50 (0.20 to 11.54)
Middle income	4.01 (2.86 to 5.62)	1.80 (1.11 to 2.90)	1.24 (0.57 to 2.72)		1.62 (0.28 to 9.22)
High income	4.30 (2.34 to 7.91)	1.99 (0.88 to 4.54)	1.17 (0.42 to 3.25)		4.0 (0.27 to 58.56)
Diagnosed with an STI in the last year^c	2.58 (1.62 to 4.12)	2.38 (1.37 to 4.14)	3.25 (2.29 to 4.62)	3.26 (2.11 to 5.04)	2.32 (0.96 to 5.64)	4.91 (1.37 to 17.6)
Sexual role
Top	Ref	^b	Ref	Ref	Ref	^b
Bottom	0.72 (0.51 to 1.01)		3.80 (2.05 to 7.02)	2.85 (1.38 to 5.88)	0.80 (0.19 to 3.34)
Versatile	1.01 (0.76 to 1.34)		2.38 (1.35 to 4.20)	2.18 (1.13 to 4.20)	0.84 (0.22 to 3.23)
Any condomless anal intercourse in past three months	0.67 (0.52 to 0.87)	0.76 (0.57 to 1.00)	0.72 (0.51 to 1.03)	0.78 (0.53 to 1.13)	0.78 (0.35 to 1.72)	^b

OR: odds ratio; aOR: adjusted odds ratio; CI: confidence interval; STI: sexually transmitted infection.

Final multivariable logistic regression model selected via backwards elimination with a p-value cutoff of 0.2.

Not included in final multivariable model.

Including syphilis, gonorrhoea, chlamydia, human papillomavirus, and/or herpes simplex virus.

Factors associated with increased odds of self-reporting living with HIV (Model 2) included older age (aOR = 1.04 per one year increase in age, 95% CI: 1.02, 1.06), homosexual or gay sexual orientation (aOR = 4.88, 95% CI 2.35, 10.2), STI diagnosis in the previous year (aOR = 3.26, 95% CI 2.11, 5.04), bottom/receptive sexual role (versus top/insertive) (aOR = 2.85, 95% CI: 1.38, 5.88), and versatile sexual role (versus top/insertive) (aOR = 2.18, 95% CI: 1.13, 4.20).

Finally, we assessed factors associated with increased odds of self-reported engagement in HIV medical care (Model 3). These factors included older age (aOR = 1.18 per one year increase in age, 95% CI: 1.09, 1.27) and an STI diagnosis in the previous year (aOR = 4.91, 95% CI: 1.37, 17.6).

Discussion

To our knowledge, this is the first study to assess self-reported HIV prevalence, sexual behaviours and engagement in HIV medical care among MSM in Venezuela.⁹ These preliminary data provide an important contribution to the limited available literature in Venezuela and support the continued need for HIV and STI epidemiologic research among MSM in this context. Moreover, with close to 3000 respondents, our study also provides support for the use of online surveying as a low-cost and novel method to access often stigmatised members of a population at high risk for HIV acquisition. This sampling technique may be particularly salient in a country with political and/or economic constraints, such as Venezuela, which have previously presented barriers for studies seeking to assess sexual health among MSM.

Our results are generally consistent with a well-established epidemiologic profile of HIV and associated factors among MSM in Latin America.^1,3,4,6,9 For example, sexual role was associated with increased odds of self-reported HIV infection but not with increased odds of HIV testing. These findings support previous work in Latin America highlighting the relevance of sexual role in HIV-transmission networks.^1,2,5,6 Notably, education and income were only significantly associated with increased odds of HIV testing practices and not HIV infection itself. This may be because our study included men with Internet access, potentially representing a higher socioeconomic status group. However, these findings also suggest that factors associated with HIV testing and HIV prevalence may be qualitatively different. Although there may be similarities between HIV testing and vulnerability to HIV, among our sample, they are characterised as distinct phenomena with different associated factors. The heterogeneity within this sample of MSM also supports previous literature in Latin America bringing attention to the fact that the population included in the ‘MSM’ category is a very diverse group.^16,17 Jointly, these results highlight that there is a need to address a range of factors that contribute to testing and engagement in HIV medical care to ameliorate vulnerability to HIV.

In contrast to studies reporting HIV prevalence around 15% among MSM in Central and South America,¹ our study reports a 6.6% HIV prevalence among Venezuelan MSM. The lower self-reported HIV prevalence in our sample may point to the success of the Venezuelan National HIV/AIDS Program, operational for the past five years, in providing ART free of charge.^7,18,19 Though scarce information exists on engagement in HIV medical care among MSM, the available literature indicates that among the general population currently living with HIV more than 50% are prescribed treatment.^18,20 Similarly, over three-fourths of MSM in our sample who reported living with HIV were engaged in HIV-related medical care, further suggesting the strength of existing comprehensive HIV services. However, since this study presents only preliminary data of a specific sub-sample of MSM, reserved enthusiasm should be employed when pointing to these results as evidence of the successes of current efforts of HIV treatment and care in Venezuela.

MSM who self-reported living with HIV disclosed three times as many recent STI diagnoses as compared to HIV-uninfected MSM. However, it is likely that individuals who are engaged in care were screened more frequently for STIs, and therefore participants engaged in care were more likely to report an STI diagnosis. Nonetheless, a higher proportion (∼50%) of the participants who reported living with HIV also described inconsistent condom use in past three months compared to those who are not living with HIV. These results parallel similar associations between STIs and HIV co-infection documented in South America,^21–24 suggesting that HIV transmission among MSM in this region is associated with STI co-transmission.²⁵ Accordingly, in this context, particular attention should be paid to secondary prevention efforts to also incorporate STI screening and treatment to mitigate increased HIV transmission and acquisition risk associated with co-occurring STIs.

For participants who were prescribed ART but non-adherent in the past month, slightly less than half were engaging in condomless intercourse. Condomless anal intercourse among individuals who are living with HIV but are not fully adherent to their ART regimen may be of public health concern due to decreased likelihood of viral suppression.²⁶ As a result, secondary prevention interventions must incorporate adherence counselling and support to reduce HIV transmission.

The results of this study must be interpreted in the context of several limitations. Our sample comprised of MSM recruited from an online social and sexual networking website for MSM and as such, results may not be generalisable to other Venezuelan MSM. As such, there may have been characteristics of non-respondents that would make them different than respondents and would limit generalisability. All collected data were self-reported by participants and may be subject to bias, including social desirability bias. HIV serostatus and history of STIs were determined based on self-report, and participants were not tested for HIV or STIs. Participants may be affected by social desirability bias and thus potentially underreport HIV and STI diagnoses, and some of the results may be reflective of access to HIV and STI screening services. As expected, given that this was an online, self-administered survey, there was attrition that resulted in missing data for some variables. However, this attrition rate did not differ greatly from previous online surveys.²⁷ Additionally because these analyses were restricted to individuals who reported natal sex as male and self-identified with a male gender identity (i.e. cisgender men), this paper provides a limited epidemiologic profile of key populations in Venezuela. This approach was chosen because of the small number of transgender and/or gender non-conforming individuals that participated in our study. Though this analysis was restricted to MSM, it is nonetheless important to highlight that transgender women may also be members of a population at heightened vulnerability to HIV in South America.^9,28,29 Constituting a different population compared to cisgender MSM, future work is needed to assess factors most salient for improving HIV prevention and access to medical care among transgender women in Venezuela.

Despite these limitations, given the current lack of data on sexual behaviours and HIV prevalence among Venezuelan MSM, our analysis provides a much needed preliminary assessment on aspects of the HIV continuum of care among MSM in Venezuela and highlights that care is not a continuous, uninterrupted flow from testing to treatment. Rather, barriers and motivations at each step of the process are likely to be distinct, calling for interventions at each step to address the HIV epidemic in Venezuela. Highlighting the heterogeneity of MSM in our sample, future epidemiological research in this area would benefit from varied recruitment strategies to better inform our understanding of the HIV epidemic in Venezuela and implications for prevention interventions to address varied needs among Venezuelan MSM. The results of this study further highlight the need to understand the mechanisms by which Venezuelan MSM engage in behaviours that place them at risk for HIV, and how socioeconomic and demographic factors affect their vulnerability.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: APB is supported by a Eunice Kennedy Shriver National Institute of Child Health & Human Development T32 NRSA grant (T32 HD049339; PI: Nathanson). CEO is supported by a National Institute of Allergy & Infectious Disease T32 NRSA grant (T32AI007535, PI: Seage).

References

Beyrer

Baral

van Griensven

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

Goodreau

Goicochea

Sánchez

. Sexual role and transmission of HIV Type 1 among men who have sex with men, in Peru. J Infect Dis 2005; 191(Suppl 1): S147–S158.

Pan American Health Organization (PAHO). Blueprint for the provision of comprehensive care to gay men and other men who have sex with men (MSM) in Latin America and the Caribbean, Washington, D.C.: OPS, 2010.

Beyrer

Baral

Walker

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

Oldenburg

Perez-Brumer

Reisner

. global burden of HIV among men who engage in transactional sex: a systematic review and meta-analysis. PLoS One 2014; 9: e103549–e103549.

UNAIDS. The Gap Report. Joint United Nations programme on HIV and AIDS, http://www.unaids.org/en/resources/campaigns/2014/2014gapreport/gapreport (2014, accessed 3 July 2015) pp.1–422.

Munoz

Pacheco

Aponte

. Assessment of human immunodeficiency virus type 1 and risk practices among female commercial sex workers in Isla Margarita, Venezuela. Mil Med 2006; 171: 1–6.

Faas

Rodriguez-Acosta

Echeverría de Pérez

. HIV/STD transmission in gold-mining areas of Bolívar State, Venezuela: interventions for diagnosis, treatment, and prevention. Rev Panam Salud Publica 1999; 5: 58–65.

De Boni

Veloso

Grinsztejn

. Epidemiology of HIV in Latin America and the Caribbean. Curr Opin HIV AIDS 2014; 9: 192–198.

10.

Torres

. Human right to health, national courts, and access to HIV/AIDS treatment: a case study from Venezuela. Chic J Int Law 2002; 3: 105–114.

11.

Lind

. Governing intimacy, struggling for sexual rights: challenging heteronormativity in the global development industry. Development 2009; 52: 34–42.

12.

Mahmood

Muntaner

. Politics, class actors, and health sector reform in Brazil and Venezuela. Glob Health Promot 2013; 20: 59–67.

13.

Grov

Breslow

Newcomb

. gay and bisexual men’s use of the internet: research from the 1990s through 2013. J Sex Res 2013; 51: 390–409.

14.

Biello

Rosenberger

Novak

. Epidemiology of sexual health in the virtual environment: a multinational online survey of Spanish- and Portuguese-speaking men who use an internet sexual networking site. AIDS Behav 2014; 18: 1675–1685.

15.

Nakagawa

. A farewell to Bonferroni: the problems of low statistical power and publication bias. Behav Ecol 2004; 15: 1044–1045.

16.

Bastos

Caceres

Galvao

. AIDS in Latin America: assessing the current status of the epidemic and the ongoing response. Int J Epidemiol 2008; 37: 729–737.

17.

Kim

Creswell

Guardado

. Correlates of bisexual behaviors among men who have sex with men in El Salvador. AIDS Behav 2013; 17: 1279–1287.

18.

Castillo

Comegna

Quijada

. Surveillance of HIV type 1 drug resistance among naive patients from Venezuela. AIDS Res Hum Retroviruses 2009; 25: 1329–1333.

19.

Arán-Matero

Amico

Arán-Fernandez

. Levels of spending and resource allocation to HIV programs and services in Latin America and the Caribbean. PLoS One 2011; 6: e22373–e22373.

20.

Abreu

Noguer

Cowgill

. HIV/AIDS in Latin American countries: the challenges ahead, Washington, DC: The World Bank, 2004.

21.

Perez-Brumer

Konda

Salvatierra

. Prevalence of HIV, STIs, and risk behaviors in a cross-sectional community- and clinic-based sample of men who have sex with men (MSM) in Lima, Peru. PLoS ONE 2013; 8: e59072–e59072.

22.

Clark

Konda

Segura

. Risk factors for the spread of HIV and other sexually transmitted infections among men who have sex with men infected with HIV in Lima, Peru. Sex Transm Infect 2008; 84: 449–454.

23.

Lama

Lucchetti

Suarez

. Association of herpes simplex virus type 2 infection and syphilis with human immunodeficiency virus infection among men who have sex with men in Peru. J Infect Dis 2006; 194: 1459–1466.

24.

Konda

Klausner

Lescano

. The epidemiology of herpes simplex virus type 2 infection in low-income urban populations in Coastal Peru. Sex Transm Dis 2005; 32: 534–541.

25.

Solomon

Mayer

Glidden

. Syphilis predicts HIV incidence among men and transgender women who have sex with men in a preexposure prophylaxis trial. Clin Infect Dis 2014; 59: 1020–1026.

26.

Das

Chu

Santos

G-M

. Decreases in community viral load are accompanied by reductions in new HIV infections in San Francisco. PLoS One 2010; 5: e11068–e11068.

27.

Weatherburn

Schmidt

Hickson

. The European men-who-have-sex-with-men internet survey (EMIS): design and methods. Sex Res Soc Policy 2013; 10: 243–257.

28.

Silva-Santisteban

Raymond

Salazar

. Understanding the HIV/AIDS epidemic in transgender women of Lima, Peru: results from a sero-epidemiologic study using respondent driven sampling. AIDS Behav 2012; 16: 872–881.

29.

Jacobson

Sánchez-Gómez

Montoya

. A continuing HIV epidemic and differential patterns of HIV-STI risk among MSM in Quito, Ecuador: an urgent need to scale up HIV testing and prevention. AIDS Behav 2014; 18: 88–98.