Studies show that trans women in low- and middle- income countries (LMICs) have the highest population-specific rates of HIV with the least access to HIV prevention and care. Brazil is a resource constrained country with a large and rising HIV epidemic among key populations.1 Recent research with adult trans women found an HIV prevalence of 30-31.2%, making this population one of the most highly impacted in Brazil.2,3 Youth are also highly affected by HIV. A recent study found that 24.2% of young trans women ages 18-24 years old in Brazil were living with HIV, higher than any other population of young trans women published.4,5 In the same study, young trans women had three times
as many unrecognized HIV infections as adults, lower HIV risk perception
and twice as many condomless sex encounters, speaking to an urgent need for
Intervention.3,6 Previous research has identified anti-trans stigma in the Brazilian public health care system (SUS) as the main barrier to HIV prevention and care among adult trans women in Brazil.7 For youth, additional barriers related to low risk perception, low health literacy and lack of systems navigation skills may exacerbate existing discrimination-based HIV prevention and care service barriers.8,9

Interventions to address HIV among young trans women are limited. Only three HIV prevention interventions were found in the literature that have been specifically designed and tested with young trans women.10–12 Two were multi-session group-based interventions requiring frequent in-person visits while another was to promote home HIV testing. In person interventions present challenges for young trans women who face competing needs and access barriers due to stigma and discrimination.13 These disparities are acute in Brazil where discrimination against trans people in health care where HIV prevention and care is provided is pervasive.7,14 Researchers have pointed out that disenfranchised youth are at especially high risk of HIV and interventions are needed in LMICs like Brazil.15

Our project, “Brilhar e Transcender” is one of eight sponsored research projects chosen to be a part of this NICHD-funded PATC3H consortium. BeT combines knowledge and experience gained through the first population-based study of HIV risk and PrEP awareness among trans women in Brazil (Transcender), and capitalizes on the youth development related needs in prior observational research with young trans women in the U.S. (SHINE).16 We adapted an evidence-based peer and mHealth delivered systems navigation intervention for implementation digitally using social media and text messaging to interact with participants..17 Our approach was designed in response to the enormous challenges young trans women face in accessing HIV prevention and care in Brazil and leverages prior successes with mHealth to engage young trans women in research and interventions.18,19 This analysis describes the results of our pilot BeT intervention to improve engagement of Brazilian young trans women in HIV prevention and HIV care.


Participants and procedures

Participants were referred by word of mouth, peers or by our community education team and all youth resided in Rio de Janeiro city or the larger metropolitan area. The pilot study took place at the Evandro Chagas National Institute of Infectious Diseases (INI)-FIOCRUZ, in Rio de Janeiro, Brazil, between 19 September 2019 and 26 November 2020. Inclusion criteria for the pilot intervention were self-identification as trans women, aged 18-24 years, and living in Rio de Janeiro, Brazil. Criteria for HIV prevention was that participants were not living with HIV and not on PrEP, had PrEP indication, and had no contraindication to use PrEP. Participants living with HIV had an HIV diagnosis and at least one of the following: 1) not on ART, 2) detectable viral load (self-report), 3) not linked to health care, or 4) a missing visit in the past six months. The study provided mobile phones for digital interactions for participants that did not have one.

This study was a collaboration between researchers at the San Francisco Department of Public Health in the U.S. and FIOCRUZ in Brazil. The pilot protocol was approved by each site’s institutional review board (IRB), and a certificate of confidentiality was obtained from the National Institutes of Health. INI-FIOCRUZ IRB (CAAE 05018818.0.0000.5262) and the University of California, San Francisco Committee on Human Research (#18-26770) reviewed and approved the study. Peer navigators explained the study to potential participants and obtained written informed consent from those interested in participating. Retention strategies included reminder calls at different times of the day and outreach on social media platforms. This intervention was still in the follow-up phase when social distancing measures due to coronavirus disease (COVID-19) were put into place in Brazil. Ride share transportation was arranged for participants who needed visits onsite and all other visits were moved to being conducted remotely.


Using data from a formative phase with 10 young trans women and with advisory group input, we adapted a US Center for Disease Control and Prevention (CDC) evidence-based, individual-level, multi-session, time-limited, system navigation intervention for use in BeT.17 The intervention, Anti-Retroviral Treatment and Access to Services (ARTAS), was adapted to the cultural and healthcare system context and to include both people living with HIV and those vulnerable to HIV. We also adapted it for delivery by peers and using mHealth methods, to build upon ways in which young people engage with the world using mobile phones. Our 3-month intervention included 3 behavioral sessions (2 in person and 1 mHealth) with peer navigators, weekly digital interactions with peer-navigators, and asynchronous text messages in between sessions.

Data collection

Surveys were conducted at baseline and follow-up visits at 3, 6, 9, and 12 months. Participants received tiered incentives over the three months to motivate continued participation in the intervention activities and retention in the study, starting at US$10 in mobile phone credits, increasing US$5 every two weeks (US$10 for week 1-2, US$20 for week 2-4, US$25 for week 5-6, etc.) for a total of US$160 in mobile phone credits over the course of three months. Incentives were also given for completion of questionnaires at baseline, three, six, nine and 12- months in the form of paying for transportation, providing a meal and entering them into a raffle for a computer or other technology gift worth US$100.


Surveys contained questions on sociodemographic data, sexual behavior, sex work engagement, violence, alcohol and drug use, and mental health. Trans women were defined as using feminizing hormone therapy (FHT) if they reported any estrogen-containing regimen, which could have included hormones they acquired in non-medical settings and without a prescription. Participants were asked specifically if they had ever used non-prescribed FHT, whether they were currently on FHT of any type, whether they had ever used fillers such as “industrial silicone” for feminization and whether or not they had even undergone a gender-affirming surgery and if it occurred outside a medical setting by an unlicensed provider.

We assessed gender-based discrimination using the General discrimination scale.20 Gender-based discrimination in a healthcare setting was assessed as proposed by Costa et. al.14 We used a trans-adapted version of the Group-Based Medical Mistrust Scale (GBMMS), a 12-item scale, to measure suspicion of health care services, health professionals, and the treatment provided.21 Intimate partner violence (IPV) was evaluated asking concerns about her physical safety due to anger or threats from a partner, sexual assault by partner or non-consensual sexual act, or attitudes not-compatible with her gender identity (e.g., hiding makeup) forced by partner. We screened psychological distress with the Patient Health Questionnaire (PHQ)-4, composed by two subscales, one for depression and the other for anxiety.22 Participants were asked if they had ever had suicidal thoughts and were also asked if they had ever attempted suicide. We used the Alcohol Use Disorders Identification Test-Concise (AUDIT-C),23 as well as the Alcohol to assess alcohol intake, as well as the Smoking and Substance Involvement Screening Test (ASSIST) to screen for all levels of problem or risky substance use in adults.24

We assessed HIV knowledge using the HIV Knowledge Scale (HKQ-10).25 Evaluation of negative attitudes at HIV testing used the components as proposed by Kalichman & Simbayi.26 HIV testing self-efficacy, which refers to the level of confidence of a person to have HIV testing, was measured as proposed by Zhou et al,26 adapted to trans women. Scores ranged from 1-4; higher scores indicated higher HIV testing self-efficacy. Risk perception was estimated using a 5-point response scale as previously described.27 Scores ranged from 4-20, and higher scores indicate higher self-perceived risk. We used the HIV Risk Behavior score,28 adapted to trans women. The score assessed sexual behavior (number of cisgender men partners and condomless anal sex) for each type of sexual partnership (stable, casual, and commercial). Overall score was the sum of the score for each type of sexual partner and ranged from 0-48 points. Individuals were classified as low (0-2), medium (3-8) and high risk (9+). We also assessed sexually transmitted infections (STI) using laboratory results for syphilis, rectal CT and NG at the baseline and final visits.


For participants not living with HIV, our main outcomes were HIV testing and PrEP use in the last 3 months. For participants living with HIV, we focused on HIV care engagement, a combined outcome composed of HIV linkage to care, HIV care attendance, ART use, and undetectable viral load (VL). All outcomes were self-reported, except VL, which was collected after enrollment. HIV care engagement is defined as ‘yes’ if youth changed their status from baseline to final visit regarding any of the following: reported that they were linked to HIV care, had not missed a HIV care appointment in the last 6 months, were currently on ART or were virologically suppressed. We considered study retention as attending to the 9-month visit.

Statistical analysis

We described baseline characteristics and compared study outcomes (HIV prevention and care) and other behaviors at pre- and post-intervention (i.e., baseline vs. 12-month visit). Descriptive analysis used medians and interquartile ranges (IQR) for numeric variables and absolute number and proportions for categorical variables. We used Fisher exact test, Ranksum test, or t-test used, as appropriate, to compare study outcomes and other behaviors at pre- and post-intervention. Pre/post data analysis was conducted with a per protocol approach. All analyses were performed using R Software version 4.0.3.


We screened 27 individuals; 4 were ineligible (2 without PrEP indication, one was already using PrEP, and one had an ineligible gender identity), 3 were discontinued (2 withdrawn consent, and 1 moved to another region), and 20 were enrolled as participants (13 not living with HIV, 7 living with HIV). Among these, 19 completed the intervention (13 not living with HIV, 6 living with HIV) and 18 were retained until the last study visit (13 not living with HIV, 5 living with HIV). Retention in the study took place during the Covid-19 pandemic and was 90% to 9 months.

At baseline, overall median age was 21 years (IQR 19-23); most participants were Black or Parda (n=16, 80.0%), 45.0% (n=9) of participants had 8 years of schooling or less, 80.0% (n=16) were experiencing food insecurity, and all had an extremely low monthly income (Table 1). Almost one-third (40.0%, n=8) were currently engaged in sex work. All participants reported experiencing discrimination due to their gender identity in general (n=20, 100%) and by healthcare professionals (n=16, 80.0%). 60% had suffered IPV (n=12), and half had suicidal thoughts (n=10, 50.0%). A quarter had suffered sexual violence (n=5, 25.0%), and 57.9% reported that the impact of the political context in Brazil had a negative impact on them (n=11). Use of non-prescribed FHT ever was high (90.0%, 16 out of 18 ever on hormones), and 55.0% (n=11) were currently on FHT at baseline. Only two participants (10.0%) had undergone a gender-affirming procedure, both of them outside of medical care. Only one participant (5.0%) reported having injected illicit silicone. About one-third of young trans women were living with HIV (n=7, 35.0%) at baseline. At baseline, 66.7% of young trans women had ever been diagnosed with syphilis (n=11/18) and 7 (38.9%) had an active case of syphilis, 17.6% (n=3/17) had rectal chlamydia trachomatis, and 11.8% (n=2/17) had Neisseria gonorrhea.

Table 1.Characteristics of young trans women ages 18-24 years old enrolled in the BeT pilot study. Rio de Janeiro, Brazil, 2019-2020.
Characteristics N%
Total 20 (100)
Social Determinants of Health
Age 21.0 (19.0,23.0)
Black 6 (30.0)
Parda 10 (50.0)
White 4 (10.0)
Monthly per capita income¹ 84.13 (64.27,115.25)
Schooling, years
0-8 9 (45.0)
9+ 11 (55.0)
Cannot get by on current income 14 (70.0)
Live with Family
4 (20.0)
9 (45.0)
7 (35.0)
Food insecurity 16 (80.0)
Currently doing sex work 8 (40.0)
Prior arrested 4 (20.0)
Gender-based discrimination 20 (100.0)
Gender-based discrimination in healthcare 16 (80.0)
Group-based medical mistrust 29.0 (20.0,33.0)
Negative Impact of Brazilian political context 11 (57.9)
Ever suffered sexual violence 5 (25.0)
Ever suffered intimate partner violence 12 (60.0)
Mental Health
PH-4 score 6.0 (2.5, 8.5)
Positive anxiety screening³ 11 (57.9)
Positive depression screening³ 9 (45.0)
Ever had suicidal thoughts 10 (50.0)
Ever attempted suicide 5 (25.0)
Medical care access
Ever used hormones 18 (90.0)
Ever used non-prescribed hormones² 16 (94.1)
Currently on hormones² 11 (68.8)
Filler use 1 (5.0)
Feminizing surgery by unlicensed providers 2 (10.0)
Living with HIV (HIV positive) 7 (35.0)
  1. Conversion rate: USD1.00=5.64, 2. Denominator: participants that reported ever using hormones.

At baseline, all participants not living with HIV were PrEP naïve and eligible for PrEP (n=13), of which 61.5% (n=8) were knowledgeable about PrEP and all (100.0%) reported high willingness to use PrEP after a brief explanation (Table 2). Most (84.6%) young trans women not living with HIV had been tested for HIV previously (n=11), but only 46.2% had been tested for HIV in the last three months (n=6). The majority (76.9%) of participants had negative attitudes about HIV testing (n=10), although the median HIV testing self-efficacy score for the sample was 3.5 (3.1,4.0), and median self-perceived risk score was 14.0 (11.0,16.0). Most participants (11, 84.6%) had a medium/high risk behavior score at baseline. Condomless receptive anal sex was frequent at baseline (92.3%, n=12). Among those living with HIV, 71.4% (n=5) had been linked to care, 14.3% (n=1) had recently missed a HIV care visit, 57.1% (n=4) were currently using ART, and only 1 (14.3%) of those living with HIV had an undetectable VL.

Table 2.HIV prevention, care, knowledge and related behaviors
Characteristics N (%)
Ever tested for HIV¹ 11 (84.6)
Tested for HIV in the last 3 months¹ 6 (46.2)
Currently using PrEP¹* 0 (0.0)
Engaged in HIV care2 5 (71.4)
Missed any HIV care visit (prior 6 mos)2 1 (14.3)
Current ART use2 4 (57.1)
Undetectable HIV viral load2 1 (14.3)
Negative attitudes at HIV testing¹ 10 (76.9)
HIV testing self-efficacy score¹ 3.5 (3.1,4.0)
HIV knowledge score¹ 5.0 (5.0,7.0)
PrEP knowledge¹ 8 (61.5)
PEP knowledge¹ 5 (38.5)
Risk perception score¹ 14.0 (11.0,16.0)
Sexual risk score¹
Low 2 (15.4)
Medium 7 (53.8)
High 4 (30.8)
Condomless receptive anal sex¹ 12 (92.3)

PrEP: Pre-Exposure Prophylaxis, ART: Antiretroviral therapy, 1 Denominator: HIV-negative participants, 2 Denominator: Participants living with HIV. *Participants were not eligible for the intervention if they were already on PrEP.

Post intervention, all participants not living with HIV (n=13) were regularly tested for HIV resulting in a change from 46.2% having been tested for HIV prior to 100% being recently tested (Table 3). We observed 100% uptake of PrEP among participants who were not living with HIV, and no participants seroconverted over the course of the study. All participants living with HIV (n=5) were engaged in HIV care by post intervention, with the most notable improvements being in ART use improving from 40% to 100% and having an undetectable VL improving from 20% to 80% of participants. The VL of the one participant with a detectable result was 201 copies/mL. At the final study visit, no young trans women had active syphilis, 11.8% (n=2/17) tested positive for rectal chlamydia trachomatis, and 11.8% (n=2/17) had Neisseria gonorrhea.

Table 3.Primary outcomes of BeT pilot intervention
Characteristics Pre-intervention
N (%)
N (%)
HIV test in the last 3 months² 6 (46.2) 13 (100) <0.01
Current PrEP use² 0 (0.0) 13 (100.0) <0.001
HIV care engagement³ 0 (0.0) 5 (100.0) <0.010
HIV care³ 3 (60.0) 5 (100.0) 0.440
Missed any HIV care visit³ 1 (20.0) 1 (20.0) 1
Current ART use³ 2 (40.0) 5 (100.0) 0.170
Undetectable HIV viral load³ 1 (20.0) 4 (80.0) 0.210

PrEP: Pre-Exposure Prophylaxis, ART: Antiretroviral therapy, 1 Fisher exact test, Ranksum test, or t-test used as appropriate, 2 Denominator: HIV-negative participants, 3 Denominator: Participants living with HIV.

Results from the pre/post analysis show that significantly fewer young trans women experienced discrimination overall post intervention (100.0% pre-intervention and 55.6% post-intervention, p<0.01) and had significantly less medical discrimination (77.8% pre-intervention and 11.1% post-intervention, p<0.001) post intervention (Table 4). HIV knowledge score significantly increased among young trans women from pre to post intervention (5.5 pre-intervention to 7 post-intervention, p<0.01), and HIV risk perception was significantly lower post intervention (13.2 per intervention to 8.5 post intervention, p<0.001). Lastly, there were significantly fewer young trans women who screened positive for anxiety post intervention than pre-intervention (55.6% pre-intervention and 22.2% post-intervention, p=0.04).

Table 4.Pre- and post-intervention assessments among trans women in the BeT pilot study. Rio de Janeiro, Brazil, 2019-2020 (N=18).
Characteristics Pre-intervention N (%) Post-intervention N (%) p-value¹
Experienced/reported discrimination
Gender-based discrimination 17 (100) 10 (55.6) <0.010
Discrimination score 9 (3,13) 1 (0,2.8) <0.001
Gender-based discrimination in healthcare 14 (77.8) 2 (11.1) <0.001
Group-based medical mistrust 30.0 (21.2,33.0) 18.5 (14.5,33.2) 0.160
HIV-related knowledge and behavior
Negative attitudes at HIV testing2 8 (72.7) 6 (54.5) 0.660
HIV testing self-efficacy score2 3.5 (3.2,4) 3.7 (3.3,4.0) 0.600
HIV knowledge score 5.5 (5,7) 7 (7,8) <0.010
PrEP knowledge
PEP knowledge
Risk perception score2 13.2 (3.8) 8.5 (2.3) <0.001
Sexual risk score² 0.260
Low 2 (15.4) 5 (38.5)
Medium 7 (53.8) 7 (53.8)
High 4 (30.8) 1 (7.7)
Condomless receptive anal sex2 12 (100) 9 (69.2) 0.100
Mental Health
PH-4 score 5.5 (1.8,8.8) 2 (0,5.8) 0.140
Positive anxiety screening³ 10 (55.6) 4 (22.2) 0.040
Positive depression screening³ 9 (50) 6 (33.3) 0.310
Medical care access
Current hormone use4 11 (61.1) 16 (88.9) 0.120

1 Fisher exact test, Rank sum test, or t-test used as appropriate, 2 Denominator: HIV-negative participants, 3 Derived from PH-4 score, 4 Denominator: participants that reported ever using hormones.


The “Brilhar e Transcender” (BeT) intervention resulted in improvements in all three areas of intervention, including HIV testing, PrEP use and HIV care engagement with lasting effects nine months post intervention. All young trans women participants not living with HIV engaged in PrEP, and all but one participant who was living with HIV had an undetectable VL at follow-up. Moreover, evidence of similar rates of STIs at baseline to follow up provide evidence that young trans women continued having condomless sex while in the intervention, but risk for HIV was likely mitigated by PrEP use.

This is the first intervention, to our knowledge, that has demonstrated success in improving HIV prevention and care outcomes using peer navigation with young trans women in a LMIC. Peer navigators have been central to reducing health care barriers related to stigma and discrimination, like medical mistrust and unfamiliarity with systems, in many marginalized populations.29,30 Peer-based delivery has also been related to improved HIV care and prevention outcomes among trans women adults.31–33 Our application presents preliminary evidence of the positive impact of peer delivery in systems navigation to improve HIV prevention and care behaviors among young trans women.

Our intervention may have also impacted factors known to influence HIV risk.34 We found less anxiety in the pilot sample post intervention, lower HIV risk perception scores and higher HIV knowledge scores among participants. We also found significantly less experienced discrimination overall and in healthcare settings among young trans women. Co-location of HIV care services onsite with digital peer navigation has been found to facilitate engagement in biomedical prevention and care in another intervention implementing ARTAS.35 The BeT intervention offered co-location of health care services that also were also trans-affirming. We believe that offering services convenient to youth and in a de-stigmatized and trans-affirming manner was a key component of our pilot intervention. The use of mobile phones for sexual and reproductive health interventions is an increasingly effective way to reach adolescents and young adults in LMICs.36 The digital delivery of a health intervention is a new approach in HIV prevention with young trans women that may overcome stigma-related obstacles to participation like fears of taking public transportation and anticipation of discrimination.

Our study was also able to achieve 90% retention of participants. Follow up of trans women participants in HIV research is usually 3 months or less.37 Strong study retention could be attributed to the availability of digital peer navigators asynchronously to help maintain engagement and support for young trans women participants. Our ability to maintain high study retention in the midst of a global pandemic and maintain a lasting effect on HIV prevention and care behaviors among young trans women is promising evidence of impact of the BeT intervention.


The primary limitation to this study was the small sample size overall limiting the ability to calculate tests for significance in changes pre- and post-intervention, although the number of participants we enrolled is consistent with pilot interventions. Our design was also not a randomized clinical trial, which was not appropriate at this stage of the intervention development. Although this is the only HIV prevention and care intervention with young trans women in a LMIC context, findings may not be generalizable to other countries.


This pilot points to the BeT intervention as a promising approach for improving HIV prevention and care behaviors of young trans women in Brazil and fills a major gap in interventions for young trans women in LMICs. An important next step in this research is to test the intervention in a larger efficacy trial in Brazil and testing it in other countries with similar epidemics and stigma-based barriers to engagement in HIV prevention and care.


The content of the current manuscript was partially presented in the International AIDS Conference 2020.

Data Access, Responsibility and Analysis

Dr. Emilia Jalil and Dr. Beatriz Grinsztejn had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.


This work was supported by the National Institute of Child Health and Human Development/National Institutes of Health as the PATC3H sponsor with grant # UG3HD096914. The sponsor had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Authorship contributions

ECW, BG, EMJ, and VGV conceived the study and interpreted the findings. EW and EMJ drafted the manuscript. EMJ did the statistical analyses with aid from ECW and BG. CMJ, CRVC, ALNF, NFM, IM, CO, DB, ECN, and LM helped with data acquisition, interpretation of the findings, and drafting the manuscript. VGV was involved in revising the manuscript for important intellectual content. All authors read and approved the final manuscript.

Competing interests

The authors completed the Unified Competing Interest form at (available upon request from the corresponding author) and declare no conflicts of interest.

Correspondence to:

Erin C. Wilson
25 Van Ness Ave, suíte 500
San Francisco, CA 94131
[email protected]