TransPrEP: Results from the Pilot Study of a Social Network-Based Intervention to Support PrEP Adherence Among Transgender Women in Lima, Peru

Abstract

We conducted a pilot randomized controlled trial of a social network-based intervention to promote PrEP adherence among transgender women (TW) in Lima, Peru. We enrolled 89 TW from six social networks and cluster-randomized them 1:1 to standard of care (n = 44) or the TransPrEP intervention (n = 45). Core workshops discussed strategies to support PrEP adherence and defined group adherence objectives. Maintenance workshops discussed participants’ experiences taking PrEP and collective adherence goals. At 3-month follow-up, we evaluated 40 participants and obtained 29 hair samples for tenofovir level measurements. Though no significant differences were observed, 36.4% (4/11) of participants of TransPrEP participants and 10.0% (1/10) of control participants had tenofovir levels > 0.023 ng/mg, consistent with ≥ 4 doses per week. 81.8% (9/11) of intervention and 40.0% (4/10) of control participants had any detectable tenofovir in their hair. Pilot assessment of our network-based intervention suggested a trend towards improved PrEP adherence, measured objectively, for TW in Peru.

Introduction

Despite substantial advances in biomedical HIV prevention, the incidence of HIV infection among transgender women (TW) worldwide remains high. Prior meta-analyses have estimated the global HIV prevalence among TW as 19.1%, with the odds of HIV infection among TW compared to other adults almost 50 [1–4]. Although calculations of HIV prevalence among TW in Latin America are complicated by inconsistent sampling methodologies and the previously common conflation of TW with men who have sex with men (MSM), the lower boundary of most estimates begins at 20% [5–10]. In Peru, the HIV prevalence among TW is approximately 30%, with a higher frequency observed among TW engaging in transactional sex practices [11–15]. Factors contributing to vulnerability for HIV among Peruvian TW include social and economic marginalization that restrict work opportunities other than compensated sex, gender-based sexual roles that typically define TW as the receptive/feminine partner during anal intercourse with cis-male partners, power dynamics that limit willingness to enact HIV prevention modalities by TW during sex with cisgender men, and high prevalences of undiagnosed HIV and STIs in the sexual networks of TW [16–21].

The iPrEx trial established pre-exposure prophylaxis (PrEP) as an HIV prevention strategy for MSM and TW, though suboptimal medication adherence resulted in no observed effect of PrEP among TW in stratified analyses [22, 23]. Other PrEP trials enrolling TW have reported similarly disappointing outcomes, though with some recent improvements [24–27]. Drug-drug interactions between antiretroviral (ART) medications and exogenous feminizing hormones may partially explain the low levels of tenofovir (i.e., the metabolite of tenofovir disoproxil fumarate, TDF) or emtricitabine (FTC) observed in analyses of blood samples among TW [28–30]. Yet, this fact does not explain the absence of any detectable medication in samples from many TW participants. Other factors that may negatively influence PrEP adherence among TW include mistrust of medical professionals and “experimental therapies,” concerns that ART can negatively affect blood levels of feminizing hormones, and difficulties in situating PrEP within traditional community approaches to HIV prevention [31–36]. To address these concerns, efforts to promote PrEP use and reduce HIV transmission among TW must include community-based strategies that acknowledge and address the complexity of social, cultural, and structural factors impacting delivery of HIV prevention and treatment services available to TW in resource limited settings.

Previous research has highlighted the importance of social networks and community support in promoting the health of TW in Peru. Excluded from traditional social and economic structures, TW often construct alternative systems to organize social interactions, family relationships, and work environments [37–41]. In response to frequent experiences of harassment, discrimination, and abuse in healthcare settings, Peruvian TW have also created alternative methods to access the care they need in the ways they want it provided [42–44]. Leveraging these organic, community-based alternatives to traditional healthcare systems, we designed the TransPrEP intervention to use existing social network structures to support the introduction, uptake of, and adherence to PrEP among TW in Lima, Peru.

Methods

Between September, 2017 and July, 2018, we conducted a pilot randomized controlled trial (RCT) of the “TransPrEP” a social network-based intervention to promote PrEP adherence among TW in Lima, Peru. The primary goal of the TransPrEP intervention was to promote PrEP uptake and adherence within existing TW social networks. Through a combination of individual counseling and group workshops, investigators collaborated with facilitators hired from local TW communities to provide information on PrEP as HIV prevention, to collaboratively define norms of PrEP use and adherence within the group, and to facilitate an interactive environment where participants could offer mutual advice and support to define and maintain collective adherence goals. The theoretical framework for the intervention borrowed elements from Fisher’s Information-Motivation-Behavioral Skills model, Rogers’ Diffusion of Innovations Theory, and network principles on the dissemination of influence [45–50]. Objectives of the intervention were to: (1) Promote acceptability and uptake of PrEP among TW in Peru; (2) Provide information on the importance of PrEP adherence and strategies to maintain adherence; (3) Encourage the formation and consolidation of behavioral norms of PrEP use in TW social networks; and (4) Promote an understanding of PrEP as part of a comprehensive approach to health by addressing other issues of health and well-being for TW in Peru.

A total of 89 TW were divided into six clusters based on geographic area and pre-existing social network ties. Clusters were stratified according to three geographic regions (Northern Lima, Central Lima, and Southern Lima) and randomly assigned on a 1:1 basis within each stratum to intervention and control arms. Participants in the intervention arm attended group workshops addressing issues of PrEP use and adherence while control arm participants received the standard of care. All participants were asked to attend 1- and 3-month Medication Refill and Follow-up visits.

Screening Procedures

Potential participants were identified through network-driven snowball sampling by staff from Asociacion Civil Impacta Salud y Educacion, a not-for-profit HIV research site, and Epicentro, a grassroots LGBT community wellness center, both located in Lima, Peru. “Seed” participants were identified by peer recruiters and asked to refer other HIV-uninfected TW from their social network. Screening was limited to individuals 18 years and older who: (1) Identified as “transgender, transsexual, travesti, or a similar term;” (2) Reported condomless anal intercourse with ≥ 1 male or transwoman partner in the previous 6 months; and (3) Denied previously testing positive for HIV. After providing informed consent, screened participants completed a computer-assisted self-administered (CASI) survey of demographics, gender identity, sexual identity, transgender adaptation and integration, history of HIV testing, history of HIV/STIs, characteristics of recent sexual partners, and knowledge and acceptability of PrEP.

Screened participants underwent a medical evaluation including history and physical examination. Blood was collected to test for syphilis by RPR assay (RPRnosticon™; Biomérieux, Marcy l’Etoile, France) with positive results confirmed by MHA-TP (MHA-TP; Organon Teknika, Durham, NC) and serial dilution of RPR titers; for Hepatitis B infection using a standard Hepatitis B panel; for renal insufficiency through Cockcroft-Gault equation estimation of creatinine clearance; and for HIV infection using the Determine rapid assay (Alere; Waltham, MA), with IFI confirmation of positive results performed by the Peruvian Ministry of Health Laboratory.

Participants with syphilis were provided with appropriate antibiotic therapy according to physician review of their prior RPR titer(s) and treatment history. Participants with HIV were referred to local treatment centers designated by the Ministry of Health. Participants diagnosed with renal insufficiency, active or chronic hepatitis B infection, or other medical issues during screening were referred to local healthcare providers for ongoing management. Individuals with HIV, active Hepatitis B, or renal insufficiency (creatinine clearance < 50 mg/dL) were excluded from enrollment.

Individual Counseling Sessions

Prior to randomization and enrollment, potential participants were asked to attend two individual counseling sessions. The goal of individual counseling was to introduce PrEP as an HIV prevention tool and to address basic adherence strategies. Individual counseling was modeled after the LifeSteps intervention, which uses cognitive behavioral therapy to identify and manage common obstacles to ART adherence [51].

Enrollment and Randomization

After completing individual counseling, potential participants were asked whether they were interested in starting PrEP. Those who confirmed interest then underwent additional rapid HIV testing. If negative, they then received an initial 30-day supply of TDF 300 mg/FTC 200 mg, and were asked to schedule a 30-day medication refill visit.

After fully enrolling both clusters from a given geographic area, intervention allocations were randomly assigned by cluster on a 1:1 basis using a coin flip. Participants in the intervention arm were then informed of the time and place of their first workshop.

TransPrEP Workshops

Participants in intervention clusters were invited to attend eight group workshops emphasizing a collective approach to HIV prevention and PrEP adherence. Building on concepts introduced during individual counseling, workshops were designed to establish a common understanding of PrEP as a part of comprehensive HIV prevention, to address the importance of adherence in maintaining PrEP efficacy, to develop practical strategies for supporting adherence, to identify barriers to PrEP adherence and problem-solve solutions to these obstacles, and to generate and maintain norms of HIV prevention in the group. Drawing on formative research highlighting the importance of community leaders for introducing and defining innovations in TW networks, workshops were facilitated by community leaders who had previously participated in an “open pilot” of the TransPrEP intervention [37].

Core Workshops

Four weekly core workshops sought to provide a common knowledge and understanding of PrEP and HIV prevention within the group. Core sessions introduced four fundamental concepts: (1) basic principles of adherence; (2) sexual risk behavior, sexual risk-taking, and the role of PrEP; (3) balancing alcohol and substance use with medication use; and (4) collective approaches to HIV prevention in the group.

Maintenance Workshops

Four biweekly maintenance meetings addressed ongoing issues related to PrEP use and adherence. The goal of the maintenance workshops was to reinforce and maintain principles introduced during the core workshops while collectively encouraging sustained PrEP adherence among group members. Meetings used a semi-structured format, blending social activities with free-form group discussions about participants’ experiences taking PrEP, managing side effects, and maintaining adherence to a daily medication schedule. To encourage attendance and address HIV prevention as part of a comprehensive approach to wellness, discussions also addressed other topics of interest to the group, like legal procedures for name changes, management of hormones for gender affirmation, sexual and romantic partnerships with cisgender men, and transgender community empowerment.

Follow-up Evaluations

30-Day Medication Refill Visit

30-day medication refill visits included a clinician evaluation for signs or symptoms of acute HIV seroconversion and laboratory testing for HIV infection and serum creatinine levels using the procedures described above. After completing these procedures, participants were provided with a 60-day supply of TDF/FTC and scheduled for a 3-month Follow-up visit.

3-Month Follow-up Evaluation

3-month follow-up visits included physical examination, laboratory testing for HIV, syphilis, and kidney function, and PrEP adherence assessments. Participants were evaluated by a clinician and provided blood samples for HIV, syphilis, and renal function testing, as described above. Hair samples were collected to measure tenofovir levels from TDF over the previous 30-days. Participants also completed a CASI survey assessing recent sexual behavior, substance use, knowledge and acceptability of PrEP, self-reported PrEP adherence, and satisfaction with the TransPrEP workshops.

Analysis of TFV Levels in Hair

For participants who consented to hair collection, a small amount of hair (~ 150–200 strands) was cut close to the scalp in the occipital region, and the distal portion of the hair strands labeled. Tenofovir levels were measured using previously described, validated methods [52, 53]. A cut-off of > 0.038 ng/mg TFV in hair was used to define daily PrEP use during the previous 30-day period, while levels between 0023 and 0.038 ng/mg hair suggested regular use (4–5 days per week), and detectable levels < 0.023 ng/mg hair suggested occasional use (1–3 days per week).

Sample Size and Power Calculations

For power calculations, we assumed an adherence rate of 90% in the intervention condition and 60% in the control group. We estimated that three clusters per arm, each with ten individuals, would be needed for 80% power, giving a total sample of six clusters and 60 individuals. We planned to enroll 90 TW from three geographic areas, with two network clusters from each area, each network with 15 participants, and 67% (n = 60) of enrolled participants completing follow-up procedures.

Statistical Analysis

Descriptive characteristics were calculated with medians and interquartile ranges (IQR) for continuous variables and proportions for categorical variables. The primary outcome was adherence over the prior 30-day period, defined by levels of TFV in hair samples. Due to the small number of participants who provided hair samples for evaluation, only descriptive comparisons were made between intervention and control arms without regard to social network clusters. The proportion of participants in each arm with levels of TFV in hair >0.023 ng/mg (consistent with >4 days per week medication adherence) was calculated and compared using contingency tables and Chi-square tests. All analyses were conducted in Stata 14.1 (StataCorp, College Station, TX).

Human Subjects Protections

All study procedures were approved by the UCLA Office for Human Research Participant Protection (IRB #16-001418) and Asociacion Civil Impacta Salud y Education’s Comite Institutional de Bioetica (Certificate #0089-2014-CE) and registered with www.clinicaltrials.gov (NCT02710032) prior to any participant enrollment.

Results

Demographics

Between September 2017 and July 2018, 205 individuals were screened and 89 TW enrolled from pre-existing social network clusters in three geographic areas of Lima, Peru (Fig. 1). Among potential participants, 33 were excluded due to HIV infection, 34 for other medical reasons (lab abnormality or other medical condition precluding use of TDF/FTC), 33 were lost to follow-up during screening, and 16 chose not to start PrEP. The median age of enrolled participants was 24–25 years, with most participants reporting less than a complete secondary school education and 300–1500 Nuevos Soles (approximately $100-$500 USD) monthly income (Table 1). Use of feminizing hormones was described by 81.8% of participants in the intervention and 51.1% of participants in control clusters.

Fig. 1.

Participant screening, enrollment, and follow-up

Table 1.

Baseline demographic and behavioral characteristics of participants in TransPrEP study; Lima, Peru 2017–2018

	Intervention arm N = 45	Control arm N = 44
Agea	24 (22–35)	25 (23–28)
Education level
Primary school only	10 (22.2%)	9 (21.4%)
Incomplete secondary school	19 (42.2%)	21 (50.0%)
≥ Secondary school graduate	16 (35.6%)	12 (28.6%)
Monthly household income
< 300 soles	4 (19.0%)	7 (26.9%)
301–500 soles	8 (38.1%)	6 (23.1%)
501–1500 soles	7 (33.3%)	8 (30.8%)
> 1500 soles	2 (9.5%)	5 (19.2%)
Feminizing hormone use	36 (80.0%)	23 (51.1%)
Number of sexual partners (30 days)a	9 (2–25)	10 (4–25)
Baseline PrEP knowledge scorea	23 (18–25)	23 (19–26)
Baseline PrEP attitudes scorea	21 (18–24)	20 (15–24)

^a(Median ± IQR)

Workshop Attendance

For participants assigned to the intervention arm, meeting attendance ranged from 4–13 participants for each Core Workshop and 5–9 participants for each Maintenance Workshop (Table 2). Individuals reported attending an average of 2.8 (out of 4) Core Workshops and 2.4 Maintenance Workshops. Workshop attendance correlated with intervention satisfaction, with the majority of respondents stating they “Strongly Agreed” that the sessions were helpful, enjoyable, and a comfortable space to discuss issues of HIV prevention and PrEP adherence with other TW.

Table 2.

TransPrEP workshop attendance and satisfaction outcomes; Lima, Peru 2017–2018

Number of participants (facilitator reported)	Core workshop 1	Core workshop 2	Core workshop 3	Core workshop 4	Maintenance workshops
	Mean: 9.3 (range: 4–13)	Mean: 8.1 (range: 5–12)	Mean: 8.7 (range: 4–11)	Mean: 9.0 (range: 6–10)	Mean: 7 (range: 5–9)
Number of workshops attended (participant self-report)	Core workshops (4 sessions)		Maintenance workshops (4 sessions)
	Mean: 2.8 (range: 0–4)		Mean: 2.4 (range: 0–4)
Satisfaction with workshops	Strongly agree	Partially agree	Partially disagree	Strongly disagree	Prefer not to answer
“It was helpful to learn and talk about PrEP, HIV, and sexual health in a group”	13 (62%)	2 (9%)	0	2 (9%)	4 (19%)
“It was enjoyable to learn and talk about PrEP, HIV, and sexual health in a group”	12 (57%)	5 (24%)	0	1 (5%)	3 (14%)
“I felt comfortable discussing these topics with other transgender women”	12 (57%)	3 (14%)	0	1 (5%)	5 (24%)
Frequency of outside meetings with other workshop participants	Never	Less than monthly	Monthly	Weekly	Daily
To discuss general topics	5 (23.8%)	0	4 (19.0%)	6 (28.6%)	6 (28.6%)
To discuss PrEP adherence and/or HIV prevention topics	3 (14.3%)	2 (9.5%)	4 (19.0%)	7 (33.3%)	5 (23.8%)

PrEP Knowledge and Acceptability

All participants who completed both of the individual counseling sessions demonstrated a strong baseline knowledge of PrEP use as well as a high level of acceptability for using PrEP to prevent HIV (Table 3). There was no significant difference or change in PrEP knowledge or acceptability between Control and Intervention arms at either Baseline or Follow-up evaluations.

Table 3.

PrEP knowledge and attitude scores of TransPrEP participants; Lima, Peru 2017–2018

	Intervention arm N = 44	Control arm N = 42
PrEP knowledge score (median ± IQR) (5-point likert score; each item 1 = completely disagree to 5 = completely agree)
1. “I know that taking PrEP (Truvada) can reduce my risk of becoming infected with HIV.”
2. “I know how often I should take PrEP.”
3. “I know how PrEP should be taken (e.g., with or without food, other medicines, hormones, etc.).”
4. “I know how PrEP interacts with feminizing hormones.”
5. “I know the possible side effects of taking PrEP.”
6. “I know what to do if I forget to take a dose of PrEP.”
7. “I know how often I should see my doctor while I am taking PrEP.”
8. “I know how PrEP interacts with alcohol and drugs.”
Baseline	23 (18–25)	23 (19–26)
3-months	23 (20–24)	23 (19–25)
PrEP attitudes score (median ± IQR) (5-point likert score; each item 1 = completely disagree to 5 = completely agree)
1. “Taking PrEP, reduces my risk of acquiring HIV.”
2. “Taking PrEP is a positive change in my life.”
3. “Taking PrEP gives me more control over my risk for acquiring HIV.”
4. “I am sure that I can take PrEP every day without forgetting a dose.”
5. “I believe that the support of my friends and others will be important in helping me remember to take PrEP every day.”
Baseline	21 (18–24)	20 (15–24)
3-months	22 (17–25)	20 (16–25)

PrEP Adherence

All participants reported a high frequency of PrEP adherence, though self-report correlated poorly with measured levels of tenofovir in hair samples (Table 4). Regular adherence was reported by 90.4% of participants in the Intervention arm and 73.6% of TW in the Control arm. However, tenofovir levels consistent with > 4 days per week PrEP adherence during the previous 30-day period (TFV > 0.023 ng/mg) were only identified in 29.4% (5/17) of participants in the Intervention and 8.3% (1/12) of TW in the Control arm (p = 0.17). In addition, 26.7% (12/45) of TW in Intervention Arm network clusters and 31.8% (14/44) of participants assigned to the Control Arm were lost to follow-up immediately after enrollment, never refilling their initial 30-day supply of medication.

Table 4.

Biological and behavioral markers of PrEP adherence during prior 30 days among transgender women; Lima, Peru 2017–2018

	Intervention arm		Control arm
	N = 17		N = 12
Tenofovir levels in hair from prior 30-day period
> 0.038 ng/mg (7 days/week)	1	5.9%	0	0%
0.023–0.038 ng/mg (4–5 days/week)	4	23.5%	1	8.3%
0–0.023 ng/mg (1–3 days/week)	8	47.0%	3	25.0%
Not detected (0 days/week)	4	23.5%	8	66.7%
	N = 21		N = 19
Self-report of doses taken in prior 30-day period
All days	16	76.2%	10	52.6%
Most days	3	14.2%	4	21.0%
Some days	0	0%	0	0%
No days	2	9.5%	5	26.3%

Sexual Behavior and HIV/STI Incidence

At Baseline, participants in both arms reported a median of ten sexual partners during the previous 30 days. Participants in the control arm reported receptive anal intercourse (RAI) with a median of ten partners (IQR 4, 10) and condomless RAI (CRAI) with two partners (IQR 0, 4). In the intervention arm, participants reported RAI and CRAI with six (IQR 4, 10) and two (IQR 0, 3) partners, respectively. At 3-Month Follow-up, both groups reported dramatic reductions in their number of sexual partners, reporting a median of ffour partners total (IQR 1, 10) and four RAI partners (IQR 1, 10) in the prior 30-days. CRAI was reported slightly more often by TW in the control [median 1 (IQR 0, 4)] than the intervention group [median: 1 (IQR 0, 2)].

Untreated syphilis (RPR ≥ 1:8) was diagnosed at Baseline more often among TW in the control than the intervention arm [25.0% (11/44) vs. 11.1% (5/45); p = 0.09]. Two other participants in the Control arm (and none in the Intervention arm) were diagnosed with incident syphilis at 3-Months. One Control arm participant, who had no detectable amount of tenofovir in her hair, was diagnosed with new HIV infection at the 3-month Follow-up visit.

Discussion

The TransPrEP social network-based adherence intervention was found to be feasible and acceptable by Peruvian TW, suggesting its potential impact for promoting PrEP uptake and use among TW. Participants who completed the program reported a high degree of satisfaction with the workshop structure, describing the sessions as an enjoyable, comfortable space, and a necessary resource to discuss health promotion and HIV prevention with other TW. Despite difficulties in maintaining and measuring PrEP adherence, similar to the experiences of other studies with TW, TransPrEP participants had a higher frequency of tenofovir in hair samples (consistent with both any recent PrEP use and with regular PrEP adherence) compared with participants receiving standard of care. Although additional work is needed to further refine the intervention and improve participant retention and workshop attendance, these findings suggest the potential benefits of the TransPrEP model for addressing the critical problem of PrEP adherence among TW in LMIC settings.

This pilot RCT was designed to assess the logistics, feasibility, and acceptability of the TransPrEP intervention, and to develop a preliminary estimate of effect on PrEP adherence. The research team screened 205 potential candidates to enroll 89 HIV-uninfected TW interested and eligible to initiate PrEP. Among 89 participants, only 60 (66.7%) attended their initial 30-day refill visit and 40 (44.4%) attended the 3-Month Follow-up. In-person workshop attendance by TW in the intervention arm was similarly sporadic, with participants attending an average of 2.4 (of four) core workshops and 2.8 (of four) maintenance sessions. Reasons cited for missing workshops and study visits included distance from home and/or work, timing of workshops, and conflicts with established work and/or school schedules. Other HIV prevention studies enrolling TW have encountered similar difficulties with recruitment and retention, highlighting a central challenge for future research [23, 24]. One potential advantage to our program was that participants who attended the workshops described this community-based health promotion strategy as an important source of affirmation for healthy trans identities, providing a space where they could address a range of issues that included, but were not limited to, HIV prevention and sexual health [32, 54].

Intermediate objectives of the intervention were to promote PrEP acceptability and community-disseminated knowledge translation among TW. After completing individual counseling sessions providing basic information on PrEP use and adherence, all participants indicated confidence in their knowledge of how PrEP works, the importance of adherence, side effects, and potential interactions with feminizing hormones. More importantly, participants expressed a high level of PrEP acceptability—an important predisposing factor to starting treatment and engaging with TransPrEP program’s objectives. Although there were no significant changes in PrEP knowledge or acceptability over the course of the intervention, these findings provide important information to help guide future programs to introduce TW in Peru to the idea of PrEP as an HIV prevention tool.

The ultimate objective of the TransPrEP intervention was to leverage existing social network formations to support PrEP adherence among Peruvian TW. Based on the available hair sample data, in which 29.4% of participants in the intervention had levels of tenofovir consistent with dosing at least 4 days per week over the past 30 days (compared with 8.3% of TW in the control arm), our findings suggest that the TransPrEP framework may dramatically improve PrEP adherence among TW in Peru. While 66.7% of hair samples from TW in the control arm had no detectable medication, tenofovir was present in 76.4% of samples from intervention participants. Self-reported adherence was similarly higher in the TransPrEP arm, though the poor correlation of self-reported behavior with laboratory-confirmed adherence further undermines the utility of self-report in assessing medication adherence [55]. Participants in the control arm also showed improvement in PrEP adherence, though not at the level seen with the intervention participants, suggesting a Hawthorne effect, i.e., that supportive contact with the study staff constituted an intervention in and of itself [56]. It should also be noted that the frequency of regular PrEP adherence remained disappointingly low throughout both arms of the trial. In addition, due to the limited duration of our follow-up period, it is not clear whether the initially observed improvements in adherence would be sustained over an extended period of time. While these issues have been noted in other studies of PrEP use among TW in diverse global settings, they suggest that there is substantial room for improvement in efforts to promote PrEP acceptability, uptake, and adherence by TW. Though necessarily limited in scope, results from our pilot RCT suggest that social networks of TW involved in the workshop-based adherence intervention took PrEP medication more often and were more likely to maintain optimal levels of PrEP adherence compared to TW offered standard of care. Future studies need to be sufficiently powered to adequately demonstrate differences in the impact of the intervention compared to the standard of care.

There are several limitations to our findings that should be acknowledged. Among 89 individuals enrolled, only 40 (45%) returned for 3-month Follow-up evaluation and 29 (32%) provided hair samples for drug level measurements. Given these issues, we are not able to make any definitive conclusions on the impact of the TransPrEP intervention on PrEP adherence. Results of our hair sample analysis may have been affected by participants’ concomitant use of exogenous feminizing hormones, which has been shown to reduce circulating levels of tenofovir [30]. However, since the baseline frequency of hormone use was higher in the intervention arm (80.0% vs. 51.1%; p <0.01), any impact on observed levels of tenofovir would have been more likely to diminish the estimated effect of the intervention. To address the reluctance to allow hair sample collection observed in this cohort, future studies may benefit from measuring Tenofovir levels in blood and/or urine instead [57, 58]. Due to the limited amount of data available, we were not able to complete the originally planned cluster analysis of adherence by social network groups and can only provide descriptive data on the intervention outcomes. Similarly, the large number of participants who were lost to follow-up, and presumably discontinued PrEP use, highlight the difficulties in developing and introducing effective PrEP adherence interventions for TW. While these limitations are necessary to acknowledge and address in future research, they reflect the challenges encountered by other investigators working with socially and economically marginalized populations of TW, and highlight the critical importance of interventions like TransPrEP to promote the health and well-being of TW communities worldwide.

In our pilot evaluation, the TransPrEP intervention was logistically feasible and suggested effectiveness in promoting PrEP uptake and adherence through the social networks of TW in Lima, Peru. Reflecting principles of collective resilience, community-building, and alternative institutional frameworks to promote the health and wellness of TW, TransPrEP provided a safe space for TW to learn about, gain experience with, and ideally to adopt and maintain PrEP as an HIV prevention strategy together with their social network peers. Results from this pilot RCT should be used to inform future modifications to the intervention, including reducing demands on participant schedules and increasing service delivery within participants’ home neighborhoods. Building on these lessons, the TransPrEP program can provide a model for the community-based delivery of HIV prevention services through the existing social networks of TW in LMIC settings.