Brief Report: Self-Reported Knowledge of HIV Status Among Cisgender Male Sex Partners of Transgender Women in Lima, Peru

Abstract

Background: 

Knowledge of HIV status is a critical first step in the HIV care cascade. Cisgender male sex partners of transgender women (MSTW) are at a disproportionately high risk of HIV, but little is known about their access to HIV testing or knowledge of HIV status.

Methods: 

We used cross-sectional data from a respondent-driven sampling study to analyze self-reported HIV status and predictors of knowledge of HIV status among MSTW in Lima, Peru. Mixed-effects models were used to generate crude and adjusted prevalence ratios (aPRs) and 95% confidence intervals (CIs) for the association between knowledge of HIV status and predictors of interest, including sociodemographics and recent sexual behavior.

Results: 

From February to July 2018, 196 eligible MSTW enrolled, of which 90 (46%) reported not knowing their HIV status. Recent casual or 1-time partners were reported by 84% of MSTW and 54% reported purchasing sex. In adjusted analyses, MSTW participants were less likely to know their HIV status if they reported buying sex (aPR 0.43, 95% CI: 0.32 to 0.59) or reported ≥16 recent sex partners compared with ≤5 partners [aPR 0.32, (0.20 to 0.50)]. Those who reported male sex partners were 80% more likely to know their status [aPR 1.80, (1.33 to 2.44)].

Conclusions: 

Reported knowledge of HIV status was low among MSTW in Lima, and unknown HIV status was associated with behaviors linked to HIV acquisition. MSTW who reported male partners were more likely to know their status, potentially indicating that HIV testing is more accessible to men who have sex with men.

INTRODUCTION

The United Nations Joint Program on HIV/AIDS (UNAIDS) 95-95-95 goals call for 95% knowledge of HIV status among people living with HIV, 95% antiretroviral therapy (ART) use among those diagnosed with HIV, and 95% viral load suppression among those on ART.¹ The critical first step in this cascade is effective outreach and testing to increase knowledge of HIV status, particularly among populations most at risk for HIV acquisition.

Transgender women are a most affected population in the HIV pandemic, with an HIV prevalence of 19% globally.² In Peru, research combining transgender women and men who have sex with men (MSM) has estimated that only 24% of HIV-seropositive transgender women and MSM know their HIV status, and only 14% are on ART.³ In populations with low ART and/or PrEP coverage, sex partners of people living with HIV are at high risk of acquisition; this applies to the cisgender men who have sex with transgender women (MSTW).^4,5 Participation in transactional sex, which has been reported among MSTW in previous studies in South America, could provide an additional contribution to HIV acquisition risk.^6,7 Little is known about HIV among MSTW, including their HIV prevalence, HIV risk behaviors, and likelihood of accessing HIV testing,⁴ because few studies have intentionally enrolled MSTW. Data from South America suggest that MSTW may seek HIV testing less often than transgender women, potentially because of lower risk perception.⁶ Understanding HIV testing behaviors among MSTW and predictors for low HIV testing uptake could inform outreach efforts aimed at achieving the first of the 95-95-95 goals.

In a previously reported study,⁸ we found that few MSTW in Lima reported knowing their HIV status. The objective of this analysis was to further investigate this finding to assess whether any behaviors or characteristics were associated with self-reported knowledge of HIV status among MSTW.

METHODS

Design, Setting, and Population

The design of this study has been previously described.⁸ In brief, we used a modified respondent-driven sampling (RDS) recruitment method to collect cross-sectional data from transgender women and MSTW in Lima, Peru. RDS is a chain-based recruitment design that uses peer-to-peer recruitment to better reach populations that are underrepresented in research.⁹ In this study, we modified this approach to recruit through sex partners because we hypothesized that this would be the best method to find MSTW. Recruitment began with transgender women who were purposively selected as seeds (initial study participants) based on behavioral and demographic characteristics. Each seed was asked to complete a survey and invite up to 3 recent sex partners, defined as anyone the participant had sex with in the previous 3 months. These partners were asked to complete the study survey and recruit 3 of their sex partners to do the same, and enrollment continued through chain recruitment of sex partners until the sample size was met. Recruitment chains that did not include transgender women or MSTW for 2 consecutive waves were not permitted to recruit further, truncating these chains, to prevent sample drift to nontarget communities (eg, MSM).

Participants were eligible for inclusion if they were 18 years or older, were a sex partner of their recruiter, and had not previously completed the questionnaire.

Data Collection

Data were collected using a REDCap online survey (ITHS, Seattle, WA)¹⁰ that could be remotely self-completed by participants using any internet-enabled device. Once a survey was completed, study staff contacted the participant by phone to verify eligibility and provide further information about recruitment and compensation. Eligible participants were sent 3 recruitment “coupons” in the form of a WhatsApp message (WhatsApp Inc, Mountain View, CA) that included survey links to send to each new recruit. If a survey respondent reported that their recruiter was not a sex partner or if they had previously completed the survey, the results were excluded from analysis, and the respondent was not invited to recruit. Surveys collected information on sociodemographics, sex identity, recent sexual behavior, and alcohol use. Knowledge of HIV status was self-reported and assessed through the question “Do you know your HIV status?” with options of “I do not have HIV,” “I am HIV positive,” “I do not know my HIV status,” and “I prefer not to answer.”

Consent was obtained through a checkbox at the beginning of the survey. Participants received 40 soles (approximately $12 USD) for survey completion and 20 soles ($6 USD) for each successful recruitment. Participant names were not collected in the survey. All study activities were approved by the IMPACTA ethics committee (Lima) and the Fred Hutchinson Cancer Research Center IRB (Seattle).

Statistical Analysis

Analyses used cross-sectional self-reported survey data. The primary outcome was self-reported knowledge of HIV status. Mixed-effects models were used to generate crude and adjusted prevalence ratios (aPRs) and 95% confidence intervals (CIs) for the association between knowledge of HIV status and predictors of interest, including sociodemographic characteristics, sexual behavior, and alcohol use. Multilevel modeling adjusted for cluster correlation (clustering by seed group) and wave correlation (categorized into wave groups 1–2, 3–6, and 7–10). For adjusted models, age, education, and number of sexual partners were selected a priori as potential confounding variables and were assessed for their association with knowledge of HIV status. Those associated at P < 0.10 in univariate models were entered into multivariable models alongside random effects (seed, wave group). Bonferroni corrected CIs and P values were calculated in adjusted analyses to account for multiple comparisons. The modified RDS recruitment design violated core assumptions used to derive RDS sampling weights, so results are reported without RDS sampling weights applied.⁸ All analyses were performed in Stata version 15.0 (College Station, TX).

RESULTS

Recruitment

In total, 543 surveys were completed between February and July 2018 beginning with 13 seed participants. Fifty-one ineligible surveys were excluded (n = 29 not a sex partner of recruiter, n = 21 duplicate questionnaire, and n = 1 younger than 18 years), as were seed participants, leaving 479 eligible participants. Of these, 196 participants were MSTW and are included in the present analysis.

Participant Characteristics

Average age among participants was 32 years, and 34% had any postsecondary education (Table 1). Over two-thirds of MSTW identified as bisexual, 15% identified as heterosexual, and 6% as homosexual. MSTW primarily reported sexual attraction to transgender women and cisgender women.

TABLE 1. - Characteristics of MSTW (N = 196)

Characteristic	No Response Provided,* n (%)	Responses, n (%)
Age, mean (SD)	0 (0%)	32.2 (9.2)
Education (any postsecondary school)	9 (4.6%)	66 (33.7%)
Work status	17 (8.7%)
Full or part time		110 (56.1%)
Informal		50 (25.5%)
No work		19 (9.7%)
Housing status	8 (4.1%)
Rent or own		59 (30.1%)
With sex partner		7 (3.6%)
With parent		104 (53.1%)
With friend		18 (9.1%)
Street or car		0 (0.0%)
Sexual orientation	14 (7.1%)
Heterosexual		29 (14.8%)
Bisexual		135 (68.9%)
Homosexual		11 (5.6%)
Pansexual		4 (2.0%)
Sexual attraction	0 (0%)
Cisgender women		133 (67.9%)
Cisgender men		13 (6.6%)
Transgender women		165 (84.2%)
Transgender men		17 (8.7%)
Sexual role	7 (3.6%)
Insertive		168 (85.7%)
Receptive		5 (2.6%)
Versatile		16 (8.2%)
Gender of partners (past 3 mos)	0 (0%)
Cisgender woman		109 (55.6%)
Cisgender man		12 (6.1%)
Transgender woman		196 (100%)
Transgender man		1 (0.5%)
Reported partnership type (past 3 mos)	7 (3.6%)
Stable/spouse		35 (17.9%)
Casual/1 time		164 (83.7%)
Client (sold sex)†		13 (6.6%)
Purchased (bought sex)†		106 (54.1%)
Number of partners (past 3 mos), median (interquartile range)	0 (0%)	8 (4–16)
Sought sex at venues (past 3 mos)	0 (0%)	144 (73.5%)
Ever met partners online	11 (5.6%)	168 (85.7%)
Insertive CAI (past 3 mos)	0 (0%)	119 (60.7%)
Receptive CAI (past 3 mos)	0 (0%)	15 (7.7%)
HIV status	29 (14.8%)
Negative		72 (37.8%)
Positive		5 (2.6%)
Do not know status		90 (45.9%)
Ever use of drugs‡	2 (1.0%)	59 (30.1%)
Ever injected drugs or medicine	0 (0%)	5 (2.6%)
Alcohol use frequency§	11 (5.6%)
≤1 per month		66 (33.7%)
2-3 per month		65 (33.2%)
1+ per week		38 (19.4%)
Binge drinking (past 3 mos)‖	5 (2.6%)	91 (46.4%)

^*Participants were given the option “prefer not to answer” for topics that were deemed to be sensitive. Missing responses include both questions that were not answered and questions in which participants selected this option.

^†Buying and selling sex defined as trading sex for money, goods, or services.

^‡Drug use included ever recreational use of marijuana, cocaine, ketamine, amphetamines, inhalants, sedatives, hallucinogens, MDMA, and opioids.

^§Alcohol use frequency was defined as “consuming an alcoholic beverage.”

^‖Binge drinking was defined as having 6 or more standard drinks in 1 day. A standard drink was defined as one 12 oz. bottle or can of beer, one 5 oz. glass of wine, one 1.5 oz. shot of hard alcohol, or 1 drink mixed with hard alcohol such as rum or vodka.

Sexual Behavior and Transactional Sex

When asked about sexual partnerships in the prior 3 months, 18% of MSTW reported a stable partner or spouse, 84% reported casual or 1-time partners, and 54% reported having purchased sex (Table 1). Selling sex was less common, with only 7% reporting selling sex in the prior 3 months. The median number of sex partners in the previous 3 months was 8 (interquartile range 4–16). Nearly three-quarters of MSTW reported seeking sex at social venues (eg, bars and plazas), whereas 86% reported ever meeting a partner online. Condomless anal intercourse (CAI) was common, with 61% of MSTW respondents reporting insertive CAI in the previous 3 months. By study definition, all MSTW had recent transgender women partners; 56% also reported cisgender women partners, and 6% reported cisgender male partners.

Knowledge of HIV Status

Ninety MSTW (46%) reported that they did not know their HIV status (Table 1). HIV-positive serostatus was reported by only 5 MSTW (3%), whereas 72 (38%) reported HIV-negative status (n = 29 chose not to answer).

MSTW participants who reported knowing their HIV status were less likely to engage in some behaviors associated with HIV acquisition, including purchasing sex (aPR 0.43, 95% CI: 0.32 to 0.59) and reporting ≥16 sex partners in the previous 3 months compared with ≤5 partners (aPR 0.32, 95% CI: 0.20 to 0.50) (Table 2). By contrast, male partnerships were associated with knowledge of HIV status; MSTW who reported having a male partner were 80% more likely to know their status (aPR 1.80, 95% CI: 1.33 to 2.44). Reported knowledge of HIV status was also associated with both reporting sexual role as receptive compared with insertive (aPR 1.72, 95% CI: 1.13 to 2.61) and with reporting receptive CAI (aPR 1.28, 95% CI: 1.13 to 1.44). No association was seen between binge drinking and knowledge of HIV status.

TABLE 2. - Characteristics Associated With Self-Reported Knowledge of HIV Status Among MSTW

Reported Demographic Characteristic or Behavior	HIV Status Known (n = 77)	HIV Status Unknown (n = 90)	Association of Reported Knowing HIV Status and Correlate
n (%)	n (%)	PR* (95% CI)	aPR† (95% CI)
Age‡
<25 yrs	24 (31.2%)	22 (24.4%)	1
25–35 yrs	31 (40.3%)	30 (33.3%)	0.97 (0.71 to 1.33)	0.99 (0.51 to 1.95)
>35 yrs	22 (28.6%)	38 (42.2%)	0.73 (0.38 to 1.40)	0.84 (0.30 to 2.35)
Education‡
>Secondary school	42 (54.5%)	18 (20.0%)	2.25 (1.39 to 3.64)	1.72 (0.94 to 3.15)
Total partners (past 3 mos)‡
1–5	47 (61.0%)	17 (18.9%)	1
6–15	21 (27.3%)	35 (38.9%)	0.51 (0.36 to 0.73)	0.55 (0.23 to 1.30)
16+	9 (11.7%)	38 (42.2%)	0.26 (0.19 to 0.36)	0.32 (0.20 to 0.50)***
Sexual role
Insertive	61 (79.2%)	84 (93.3%)	1
Receptive	4 (5.2%)	1 (1.1%)	1.94 (0.90 to 4.20)	1.65 (0.91 to 2.98)
Versatile	9 (11.7%)	4 (4.4%)	1.75 (0.92 to 3.20)	1.72 (1.13 to 2.61)***
Sold§ sex (past 3 mos)	7 (9.1%)	5 (5.6%)	1.30 (0.67 to 2.53)	1.82 (0.69 to 4.78)
Bought§ sex (past 3 mos)	22 (28.6%)	70 (77.8%)	0.33 (0.26 to 0.41)	0.43 (0.32 to 0.59)***
Reported a stable partner (past 3 mos)	21 (27.3%)	6 (6.7%)	1.83 (1.10 to 3.03)	1.36 (0.94 to 1.95)
Reported cisgender male partner	9 (11.7%)	3 (3.3%)	1.66 (1.07 to 2.58)	1.80 (1.33 to 2.44)***
Insertive CAI (past 3 mos)	46 (59.7%)	60 (66.7%)	0.94 (0.61 to 1.43)	1.03 (0.56 to 1.93)
Receptive CAI (past 3 mos)	10 (13.0%)	3 (3.3%)	1.38 (0.80 to 2.39)	1.28 (1.13 to 1.44)***
Sought sex at social venue	46 (59.7%)	79 (87.8%)	0.54 (0.28 to 1.04)	0.80 (0.59 to 1.09)
Met partner online	64 (83.1%)	82 (91.1%)	0.74 (0.57 to 0.95)	0.97 (0.53 to 1.77)
Binge drinking (past 3 mos)	34 (44.2%)	47 (52.2%)	0.87 (0.63 to 1.20)	1.05 (0.62 to 1.78)

^*All univariate analyses adjusted for random effects (seed group, wave category).

^†All adjusted models include a priori confounders of age (continuous), number of partners (log transformed), secondary education (completed any university or other secondary education), and random effects (seed group, wave category). The listed 95% confidence intervals in adjusted analyses are Bonferroni corrected CIs. Stars denote statistical significance in adjusted models, based on Bonferroni adjusted P value: *<0.05, ** <0.01, and *** <0.001.

^‡Variables selected a priori as confounders are presented in the table in their descriptive form for reference. Each is adjusted for the other a priori confounders in adjusted analysis.

^§Buying and selling sex defined as trading sex for money, goods, or services.

DISCUSSION

Nearly half of the MSTW in this study reported not knowing their HIV status. Transactional sex and multiple sex partners were common in this population, and both were associated with reduced self-reported knowledge of HIV serostatus. These associations and the low overall reported knowledge of HIV status could reflect low perceived risk of HIV infection and/or low access to HIV testing among MSTW in Lima. By contrast, participants who reported male sex partners and those engaging in receptive CAI or report their sex role as versatile were more likely to report knowing their HIV status. These findings suggest that MSTW who have more overlap with the MSM community may have greater knowledge of or access to HIV testing or, alternatively, may have higher self-perceived HIV risk.

Very little work has been conducted to understand HIV testing access and serostatus knowledge among MSTW in Peru. Only 1 previous analysis assessed partner testing among transgender women and found that 78% reported that their male partner's HIV status was unknown.¹¹ In comparison, studies conducted with MSM in Peru have demonstrated >70% of MSM report knowing their status.^12,13 Thirty-six percent of transgender women and 18% of MSM enrolled in our study reported not knowing their HIV status, both considerably lower than the 46% of MSTW examined in this analysis.⁸ Our findings among MSTW in Peru contrast sharply with data on knowledge of HIV status among MSTW in the United States (US). Data collected between 2005 and 2016 in the United States suggest that only ∼9% of MSTW did not know their HIV status.^5,14–16 Although this difference likely suggests a disparity in access or uptake of HIV testing, more research is needed to identify the reason for this testing disparity. Research among minoritized populations in low-income and middle-income countries suggests low testing uptake has been associated with low-risk perception, little knowledge of HIV or HIV prevention tools, reduced access to HIV testing resources, and HIV stigma.^17–19 The extent to which these factors may affect testing among MSTW in Peru is largely unknown. Some existing research among MSTW in Lima, including in-depth interviews conducted as part of this study, suggested that HIV risk perception in this group is moderate to low,^6,20 but no studies to date have evaluated testing access or the impact of stigma on testing in this population.

The association between knowledge of HIV status and reporting male partners has important public health implications. These findings suggest that MSTW who identify as MSM may have more access to HIV testing and by extension other HIV prevention tools. HIV prevention interventions and peer outreach in Peru are heavily targeted to gay men and to a lesser extent transgender women, so it is possible that MSTW without male partners are less likely to access these services and thus are less aware of HIV risk factors because these interventions do not reach them.

Outreach and testing are a critical next step in this population. One possibility is to intervene with MSTW seeking sexual partners at bars, clubs, and plazas. Most of the MSTW in our study reported seeking sex partners at social venues, so interventions using peer outreach at these sites, through transgender women partners or other MSTW, could be effective in reaching MSTW. Although our previous interview data suggested that venue-based testing may have low acceptability among MSTW in Lima,²⁰ a larger study conducted among MSTW in Latin America reported that 88% considered HIV home testing to be acceptable.²¹ Thus, distribution of self-tests for use at home could provide an alternative approach, particularly if stigma is a barrier to HIV testing in this population.

MSTW in Latin America are an understudied population, with only a few previous analyses collecting data directly from this group despite what seems to be high HIV risk. A strength of our study is the direct data collection from partners of transgender women in Peru, which provides some of the first evidence of low knowledge of HIV status and predictors of knowing HIV status.

Our study had several limitations. We were not able to conduct HIV testing, so we only have self-reported data on HIV status. As a result, HIV positivity in our sample is likely underestimated, and unknown status may be inflated if participants selected this instead of disclosing HIV-positive status. Underreporting of HIV-positive status is well-documented,²² and future research should include HIV testing to get a more accurate assessment. The gender classifications and sexual orientation categories used in this research may not represent a full spectrum of identities, which could result in misclassification. Owing to the sensitivity of the topics, participants were given the option not to answer questions, which resulted in some missing data. Owing to our use of recruitment through sexual networks, only MSTW with recent sex partners who are transgender women were recruited into this study, and those with more partners were more likely to be recruited. Finally, we truncated recruitment chains that did not include partners of transgender women in 2 consecutive waves. It is possible that this biased our sample if allowing these chains to continue could have resulted in enrolling MSTW with different behavior than those included in this study.

Our results add to a growing literature on MSTW in Latin America. Our findings suggest that MSTW are likely not being reached by HIV testing interventions, despite sexual behavior and sexual partnerships that could put them at high risk of HIV acquisition. More research to better characterize barriers to effective testing and prevention interventions among MSTW is needed to engage them more fully and achieve the 95-95-95 goals by 2030.