Objectives Disclosure of sexually transmitted infections (STI)/HIV diagnoses to sexual partners is not mandated by public health guidelines in Mexico. To assess the feasibility of couples-based STI/HIV testing with facilitated disclosure as a risk-reduction strategy within female sex workers’ (FSW) primary partnerships, we examined STI/HIV test result disclosure patterns between FSWs and their primary, non-commercial male partners in two Mexico–US border cities.
Methods From 2010 to 2013, 335 participants (181 FSWs and 154 primary male partners) were followed for 24 months. At semiannual visits, participants were tested for STIs/HIV and reported on their disclosure of test results from the previous visit. Multilevel logistic regression was used to identify individual-level and partnership-level predictors of cumulative (1) non-disclosure of ≥1 STI test result and (2) non-disclosure of ≥1 HIV test result within couples during follow-up.
Results Eighty-seven percent of participants reported disclosing all STI/HIV test results to their primary partners. Non-disclosure of ≥1 STI test result was more common among participants who reported an STI diagnosis as part of the study (adjusted OR=3.05, 95% CI 1.13 to 8.25), while non-disclosure of ≥1 HIV test result was more common among participants in longer-duration partnerships (AOR=1.15 per year, 95% CI 1.03 to 1.28). Drug use before/during sex within partnerships was associated with non-disclosure of both STI (AOR=5.06, 95% CI 1.64 to 15.62) and HIV (AOR=4.51, 95% CI: 1.32 to 15.39) test results.
Conclusions STI/HIV test result disclosure was highly prevalent within FSWs’ primary partnerships, suggesting couples-based STI/HIV testing with facilitated disclosure may be feasible for these and potentially other high-risk, socially marginalised couples.
- CHLAMYDIA INFECTION
Statistics from Altmetric.com
Globally, female sex workers (FSW) are disproportionately affected by sexually transmitted infections (STI), including HIV.1 Although FSWs’ elevated risk of STI/HIV acquisition has been attributed to various individual, interpersonal and structural factors, research documenting lower rates of condom use during sex with primary, non-commercial male partners than with clients2 ,3 has underscored FSWs’ need for couples-based STI/HIV prevention interventions.4 Couples HIV counselling and testing (CHCT) is an intervention promoted by the US Centers for Disease Control and Prevention that involves risk assessment, pre-test and post-test counselling, the development of personalised risk reduction plans and facilitated test result disclosure.5 The efficacy of CHCT in reducing STI/HIV-related risk behaviours has been demonstrated among heterosexual couples in Africa and the Caribbean,6 as well as partnerships between substance-using women and their primary male partners in the US.7 However, couples-based STI/HIV counselling and testing interventions with facilitated test result disclosure have not been evaluated within FSWs’ primary partnerships in resource-constrained settings.
Mexico's nation-wide HIV prevalence is 0.3%; however, a dynamic subepidemic has emerged among FSWs along the Mexico–US border.8 In Tijuana and Ciudad Juárez, the largest Mexico–US border cities, sex work is concentrated in zona rojas (red light districts) where it is legally tolerated.8 The Tijuana zona roja is situated in a neighbourhood adjacent to the US border, while sex work is more dispersed throughout the downtown area in Ciudad Juárez. In both cities, FSWs work in bars/nightclubs, cantinas, motels and on the street.8 In a 2006 study conducted among FSWs in the Mexico–US border region, prevalence of HIV, gonorrhoea, chlamydia and active syphilis (titres ≥1:8) were 6%, 6%, 13% and 14%, respectively,9 while HIV prevalence among FSWs who inject drugs was 12%.10 In Tijuana, 30%–50% of FSWs have primary, non-commercial male partners with whom condom use is rare.11 ,12 To investigate the potential contribution of FSWs’ primary, non-commercial male partners to this growing subepidemic among FSWs, Proyecto Parejas (Couples Project) was conducted between 2010 and 2013 to examine the context and epidemiology of STIs/HIV among FSWs and their primary, non-commercial male partners in Tijuana and Ciudad Juárez.13
To date, Proyecto Parejas findings suggest that STI/HIV prevalence may be lower among FSWs with primary, non-commercial male partners (HIV=3.8%, gonorrhoea=0.9%, chlamydia=7.5% and active syphilis=1.4%) than previously reported for FSWs overall, and even lower among their primary male partners (HIV=1.4%, gonorrhoea=1.4%, chlamydia=4.3% and active syphilis=1.4%).14 However, elevated rates of STI/HIV-related risk behaviours were observed within this cohort.14 For example, at baseline, 64% of FSWs reported always having unprotected sex with their primary male partners (past month).15 Yet 16% of FSWs and their primary male partners reported concurrent partners with whom inconsistent condom use was common,15 and >50% of primary male partners reported injection drug use, which was associated with ever having sex with other men and exchanging money or goods for sex.16 Moreover, these couples expressed difficulty disclosing extra-dyadic risk behaviours, which may exacerbate their risk.17
Disclosure of STI/HIV diagnoses to sexual partners is not mandated by public health guidelines in Mexico (Dr. Carlos Magis-Rodriguez, personal communication, 2014). Thus, little is known about current patterns of STI/HIV test result disclosure between FSWs and their primary, non-commercial male partners and whether couples-based STI/HIV counselling and testing with facilitated disclosure would be a feasible risk-reduction strategy for this population. To inform the development of such interventions, we investigated the prevalence and correlates of STI/HIV test result disclosure between FSWs and their primary, non-commercial male partners participating in Proyecto Parejas. We hypothesised that non-disclosure would be associated with both individual-level (e.g., being STI/HIV positive and having concurrent partners) and partnership-level (e.g., partnership duration and drug use before/during sex) characteristics.
As previously described,13 FSWs were recruited via targeted sampling in areas where sex work visibly occurs (e.g., bars, motels and street corners) and snowball sampling (i.e., enrolled FSWs referred other women they believed to be involved in sex work). Eligible women were asked to return to the study site with their primary, non-commercial male partners for a rigorous screening process to verify their status as a couple prior to enrolment. Eligible FSWs had to be ≥18 years of age; report ever using heroin, cocaine, crack or methamphetamine (due to the elevated rate of HIV previously observed among substance-using FSWs); report being in a partnership with a primary (i.e., someone with whom she lives or spends a considerable amount of time), non-commercial (i.e., someone with whom she does not exchange sex for money, drugs or other goods) male partner for ≥6 months; and report sex with their primary, non-commercial male partner as well as ≥1 client in the past month. FSWs were ineligible if they planned to imminently end their partnership, anticipated moving to another city, refused treatment for STIs or feared that participation would result in life-threatening intimate partner violence (IPV). Eligible men had to be ≥18 years of age and report sex with their primary, FSW partner in the past month. While not required for enrolment, all men knew their primary partners engaged in sex work. Participants provided written informed consent and institutional review boards at the University of California, San Diego, Tijuana's Hospital General, El Colegio de la Frontera Norte, and the Universidad Autónoma de Ciudad Juárez, approved all study protocols.
Participants completed study visits every 6 months for 24 months (5 visits total) and were compensated US$20 for each visit. All study procedures were conducted individually. At each visit, participants received pre-test and post-test counselling and underwent HIV, syphilis, chlamydia and gonorrhoea testing. Positive Advance Quality HIV rapid tests (InTech Products) were confirmed via western blot. Syphilis infection was identified via rapid plasma reagin (RPR) testing (Macro Vue, Becton Dickenson; Cockeysville, Maryland, USA) and the Treponema pallidum particle agglutination (TPPA) assay (Fujirebo; Wilmington, Delaware, USA). Active syphilis cases had positive RPR and TPPA test results and antibody titres ≥1:8. Urine samples were tested for Chlamydia trachomatis and Neisseria gonorrhoeae using transcription-mediated amplification assays (Genprobe; San Diego, California, USA). Participants received rapid test results at each visit. The San Diego County Health Department conducted all chlamydia and gonorrhoea testing and confirmatory testing of rapid positive HIV and syphilis blood samples. Within 1 month, participants received confirmatory HIV and syphilis and all chlamydia and gonorrhoea test results. At that time, participants were encouraged to disclose their results to their primary partners. STI-positive participants were offered free treatment according to Mexican and US guidelines and HIV-infected participants were referred to municipal clinics for free care and treatment.
Participants also completed interviewer-administered computer questionnaires at each visit, which collected information on sociodemographics, substance use, sexual behaviours, sexual and physical abuse, and primary partnership characteristics, including: partnership duration, partnership trust, partnership satisfaction, receipt of financial support from partners, conflict within partnerships, needle/syringe sharing within partnerships (past 6 months), substance use before/during sex within partnerships (past 6 months) and sexual behaviours within partnerships (past month). Partnership trust was measured using a 10-point scale (1=‘I do not trust my partner at all’ to 10=‘I trust my partner with my life’).18 The sum of Likert scale responses (1=strongly disagree to 4=strongly agree) to 5 items of the Satisfaction with Married Life scale,19 such as ‘In most ways, my relationship with my partner is close to ideal’ (Cronbach's α: FSWs=0.92; men=0.87), defined partnership satisfaction. Conflict (psychological aggression, physical assault, injury and sexual coercion) victimisation and perpetration within partnerships was measured using 8 items of the Revised Conflict Tactic scale—Short Form (CTS2S).20 For each item, participants were asked if they had ever been the victim/perpetrator of a given behaviour within their partnership. For example: ‘My partner pushed, shoved, or slapped me’ (i.e., victimisation; Cronbachs’ α: FSWs=0.89; men=0.76) and ‘I pushed, shoved, or slapped my partner’ (i.e., perpetration; Cronbach's α: FSWs=0.86; men=0.76). Binary variables for conflict victimisation and perpetration were created based on responses to the CTS2S.
Beginning at visit 3, questionnaires collected information on participants’ STI/HIV test results from the previous visit: what their results were and whether they disclosed their results to their primary partners. Based on their disclosure history during the study period, regardless of whether results were positive or negative, two variables were created: (1) cumulative non-disclosure of ≥1 STI test result and (2) cumulative non-disclosure of ≥1 HIV test result. These binary variables (0=disclosed all STI/HIV test results vs. did not disclose ≥1 STI/HIV test result) were used as the outcomes in our analyses.
Sample selection and follow-up
Of the 428 Proyecto Parejas participants, 370 completed ≥1 visit during which data on disclosure of test results from the prior visit were ascertained (visit 3, 4 or 5). Among those participants, 335 (181 FSWs and 154 male partners) provided data that could be used to determine their cumulative STI/HIV test result disclosure status, and were included in our sample. Our sample contains an unequal number of FSWs and male partners because one member of the couple was lost to follow-up or did not provide disclosure data (FSWs=8 and men=21) or the partnership dissolved during follow-up (men=14) (follow-up of men was discontinued in the event of partnership dissolution). Compared to those with complete follow-up, those who were lost to follow-up, did not provide disclosure data, or whose partnerships dissolved did not differ significantly with respect to STI/HIV diagnoses during the study period. The 335 participants represented 189 primary partnerships overall: 146 in which both partners reported on disclosure, 35 in which only FSWs reported on disclosure and 8 in which only male partners reported on disclosure. In the analysis, participants were followed to their last study visit, partnership dissolution or the end of the study period, whichever came first.
We calculated the Pearson-type pairwise interclass correlation coefficient (PICC) to determine the degree of within-dyad interdependence in cumulative STI/HIV test result disclosure.21 Given the potential for Type II errors, we used an α level of 0.20 to interpret the PICC and justify the use of a multilevel modelling approach for dyadic data.21
We used multilevel logistic regression to examine the effect of individual-level and partnership-level characteristics on (1) cumulative non-disclosure of ≥1 STI test result and (2) cumulative non-disclosure of ≥1 HIV test result. Characteristics significant at an α level of 0.10 in bivariate models were assessed for collinearity and included in multivariate models. If non-disclosure of ≥1 STI/HIV test result was reported by <5 participants for a single level of a covariate (e.g., city of residence), that covariate was excluded from multivariate models to minimise the potential for unstable estimates due to small cell sizes. Due to our limited sample size, we were unable to examine interactions. To eliminate temporal ambiguity, we examined characteristics measured at visit 2 in all regression models. Because self-reported STI/HIV test results reflect what participants actually disclosed to their primary partners, we examined self-reported test results in our analysis. Many partnership characteristics represent shared behaviours or experiences (e.g., partnership duration) and do not vary within dyads. However, because data were ascertained from both partners, values for these characteristics varied slightly within dyads. For continuous measures, we averaged values within dyads and for binary measures where only one partner reported the presence of a particular characteristic, that characteristic was considered present for both members of the dyad.21
Our sample (n=335) contributed data from 813 follow-up visits during which disclosure data were collected (visits 3, 4 and 5) to the analysis (median number of follow-up visits=3; IQR=2–3). Participants had a mean age of 35.8 years (SD=9.2), 37% lived in Tijuana and 36% reported an average monthly income <2500 pesos (US$∼200) (table 1). Among the 189 couples in our sample, mean partnership duration was 5.0 years (SD=4.2), 43% used drugs before/during sex (past 6 months) and 89% never used condoms during vaginal sex (past month).
Thirteen per cent (42/335) of participants did not disclose ≥1 STI/HIV test result obtained as part of the study; 71% (30/42) and 29% (12/42) did not disclose negative and positive results, respectively. Among participants who reported an HIV positive test result (n=15), 20% (3/15) did not disclose that result. Among participants who reported an STI positive test result (n=58), 17% (10/58; syphilis=6; gonorrhoea=2; chlamydia=2) did not disclose that result. Reasons for not disclosing positive results included: being scared that their partner would become violent/angry (3/11), their result was positive (3/11), not trusting their partner (1/11), not thinking it was important for their partner to know (1/11) and breaking up before being able to tell their partner (3/11).
There was significant interdependence between outcomes within dyads (PICC=0.18, 80% CI 0.03 to 0.34), thus, we used a multilevel modelling approach. In our multivariate model for cumulative non-disclosure of ≥1 STI test result (table 2), non-disclosure was more common among participants who reported ≥1 positive STI test result obtained as part of the study (adjusted OR (AOR)=3.05, 95% CI 1.13 to 8.25). In our multivariate model for cumulative non-disclosure of ≥1 HIV test result (table 3), non-disclosure was more common among those in longer-duration partnerships (AOR=1.15 per year, 95% CI 1.03 to 1.28). Drug use before/during sex within partnerships (past 6 months) was associated with non-disclosure of both STI (AOR=5.06, 95% CI 1.64 to 15.62) and HIV (AOR=4.51, 95% CI 1.32 to 15.39) test results.
To our knowledge, this is the first study to examine STI/HIV test result disclosure within FSWs’ primary partnerships. Nearly 90% of FSWs and their primary, non-commercial male partners reported disclosing all STI/HIV test results to one another. While disclosure may have been facilitated by participants’ knowledge that their primary partners were also tested as part of the study, our findings highlight an existing form of resiliency that couples-based STI/HIV testing interventions could build on to reduce transmission risk within FSWs’ primary partnerships.
However, several individual-level and partnership-level characteristics were associated with non-disclosure and require consideration in the design of STI/HIV prevention interventions for this population. First, participants reporting an STI diagnosis as part of the study were less likely to disclose all test results. Moreover, of those who did not disclose positive test results, some cited a lack of trust or fear of violence as their reason for not disclosing. These findings are consistent with those from previous research suggesting that perceived stigma and the fear of rejection or abuse from sexual partners limit STI/HIV disclosure.22–24 Thus, couples-based interventions that provide a safe and non-judgmental environment may reduce potential conflict and promote risk communication and STI/HIV disclosure within FSWs’ primary partnerships.
Non-disclosure of HIV test results was more common among participants in longer-term partnerships, which may be explained by contextual factors specific to resource-constrained settings that shape the lives of vulnerable populations. In Mexico–US border cities, employment opportunities are often inaccessible to socially marginalised women who use drugs, which drives many to sex work to support themselves and their families.25 Their marginalised position in society is then exacerbated by the stigma associated with sex work.17 ,26 While FSWs’ relationships with their primary, non-commercial male partners vary in terms of their emotional intensity, FSWs’ long-term and more committed relationships are often characterised by love, trust, mutual respect and a deep emotional connection.27 As a result, many FSWs depend on their intimate relationships for their emotional well-being and the sense of inclusion they provide.26 Qualitative data from this study suggest substance using, primary male partners of FSWs in this setting experience similar difficulty securing employment and often depend on their partners’ sex work earnings for their survival.17 As a result, both FSWs and their primary, non-commercial male partners avoid the discussion of sex work to minimise the stress associated with sex work on their relationships.17 Thus, STI/HIV test result disclosure, which could lead to the discussion of sex work and risk behaviours with clients, may be more difficult in longer-term partnerships due to a need to protect their relationships and maintain the emotional intimacy and financial stability they provide.
Non-disclosure of STI/HIV test results was also associated with drug use before/during sex within partnerships, which might be explained by the reported frequency of methamphetamine use in this context. Given the well-documented association between methamphetamine use and the practice of high-risk sexual behaviours,28 drug use before/during sex is likely a marker for higher-risk behaviours within our sample. Individuals within these high-risk couples may be less likely to disclose their STI/HIV test results if they also engage in high-risk behaviours with clients or casual partners, but are not open with their primary partners about their extra-dyadic risk behaviours. Thus, interventions that address sexual and drug-related risk behaviours may facilitate STI/HIV test result disclosure and the development of effective risk reduction plans within FSWs’ primary partnerships.
Our study has several limitations. First, due to the recruitment strategies and eligibility criteria implemented in Proyecto Parejas, our findings may not be generalisable to FSWs’ casual, non-commercial partnerships or the primary partnerships of FSWs without a history of substance use, who experience extreme IPV or whose primary partners are unaware of their sex work. Second, data were collected via face-to-face interviews conducted by study staff who encouraged participants to disclose their STI/HIV test results to their primary partners. However, study staff were encouraged to develop a rapport with participants and earn their trust to minimise the potential for social desirability bias. Finally, because our study did not include a control group unexposed to a couples-based study that encourages STI/HIV test result disclosure, we cannot assess the effect of such encouragement on disclosure rates. Future research with this population should investigate the impact of couples-based STI/HIV counselling and testing on test result disclosure and sexual risk behaviours in the context of a controlled trial.
Despite these limitations, as one of the first studies of STI/HIV test result disclosure within FSWs’ primary partnerships, our findings have important implications for future research and prevention efforts within this population, and potentially other socially marginalised couples. STI/HIV test result disclosure was highly prevalent within our sample, suggesting couples-based STI/HIV counselling and testing with facilitated test result disclosure may be a feasible risk-reduction strategy within FSWs’ primary partnerships. Although additional research on barriers to STI/HIV test result disclosure is needed to inform the development of effective interventions, our findings suggest interventions that minimise potential conflict, promote effective communication and encourage risk-reduction planning within partnerships may facilitate STI/HIV test result disclosure.
Couples-based STI/HIV testing with facilitated disclosure, which reduces STI/HIV-related risk behaviours in some settings, remains underexplored within female sex workers’ (FSWs) primary partnerships.
STI/HIV test result disclosure was common between FSWs and their primary, non-commercial male partners in Tijuana and Ciudad Juárez, two Mexico–US border cities.
Couples-based interventions that minimise conflict, promote effective communication and encourage risk-reduction planning may facilitate STI/HIV test result disclosure within FSWs’ primary partnerships.
The authors thank the study participants and staff without whom this study would not have been possible.
Correction notice The title has been changed since published Online First.
Handling editor Jackie A Cassell
Contributors HAP, SAS and TLP contributed to the design of the present study. HAP conducted all analyses and wrote the initial draft of the manuscript. All authors contributed to the interpretation of the results, manuscript revisions and approved the final manuscript.
Funding This work was supported by grants from the National Institute on Drug Abuse: R01-DA0277772 and T32-DA023356.
Competing interests None.
Ethics approval Human Research Protections Programme of the University of California, San Diego (#090570).
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.