Pre-exposure prophylaxis (PrEP) for men who have sex with men and transgender women in a clinic and a community-based setting: results from the first demonstration project in India

Abstract

Background

Understanding the ‘real-world’ challenges of delivery of pre-exposure prophylaxis (PrEP) among men who have sex with men (MSM) and transgender women (TGW) provides critical insights to guide and optimise future roll-out of the PrEP in India.

Methods

Between 18 December 2019 and 31 July 2023, self-identified MSM and TGW, ≥18 years, at substantial risk for HIV were followed for 12 months for PrEP acceptability, adherence, adverse events and condom use at community-based and clinic-based settings in Jalandhar (Punjab) and Pune (Maharashtra), respectively. Data were analysed for correlates of retention and adherence. Exit interviews were conducted.

Results

The median (IQR) age of the 650 participants was 26 (22, 30) years; the majority were literate and unmarried. The overall 1-year retention was 63.2%, with 84% and 41.7% in the community-based and clinic-based settings respectively. Overall, 64.8% participants had ≥80% adherence. Being from a community-based setting, having received secondary/higher secondary/graduate level education and living with family were significantly associated with ≥80% adherence. Overall, 9.7% showed grade-2 decline in creatinine clearance. Among 411 retained, condomless anal sex (CAS) with any partner, anytime in the last 6 months was 84.4% and 78.1% at baseline and 12 months respectively. Overall, seven seroconversions occurred.

Conclusion

Daily oral PrEP was acceptable and safe among MSM and TGW. Inclusion of PrEP as an additional prevention option would be critical for controlling HIV. Baseline risk assessment is important before initiating PrEP. Recurrent condom and risk reduction counselling during oral PrEP uptake is crucial for preventing CAS. Self-risk assessment and planned usage/stoppage of PrEP observed in the study indicates a need for counselling and assessment for potential interruptions of daily oral PrEP. Additionally, adherence would be affected, with implications for evaluating the programme’s success.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Pre-exposure prophylaxis (PrEP) is an efficacious biomedical intervention for HIV prevention among men who have sex with men (MSM) and transgender women (TGW). Based on the results of efficacy and effectiveness trials, the WHO recommended oral PrEP containing TDF as an additional prevention option for people at substantial risk of HIV infection. A fixed-dose combination of 300 mg of Tenofovir Disoproxil Fumarate and 200 mg of Emtricitabine for PrEP received marketing approval by CDSCO in February 2016 in India.

WHAT THIS STUDY ADDS

• Despite growing evidence for PrEP as an HIV prevention tool, significant gaps remained in the Indian context regarding its acceptability, feasible service delivery models, adherence and change in condom use behaviour among MSM and TGW. We have conducted the first demonstration project among MSM and TGW in India to study the feasibility of administering daily oral TDF-based PrEP. Oral PrEP was acceptable and safe among MSM and TGW and feasible to be delivered through clinic and community-based settings.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• India is committed to achieving the global targets of ending AIDS by 2030. These findings are relevant for informing policies on inclusion of PrEP in the basket of HIV prevention among MSM and TGW in India. The emerging behaviours of perceived/planned low-risk periods for taking oral PrEP suggest further investigation into event-driven and/or situation-driven and long-acting PrEP.

Introduction

India witnesses the second largest HIV epidemic globally, with the highest infection rates concentrated among key populations of men who have sex with men (MSM) and transgender women (TGW). Between 2018–19 and 2021–22, people living with HIV (PLHIV) who knew their HIV status increased from 72% to 77%.¹ Among the TGW population, HIV prevalence has remained relatively stable, fluctuating between 3.14% in 2017 and 3.78% in 2021. The HIV prevalence among MSM showed a sharp decline from 7.41% in 2007 to 4.43% in 2011, followed by a more gradual decrease to 2.69% by 2019, and eventually stabilised at 3.26% during 2021–2023.² Between 2010 and 2019, in the Asia Pacific region, 44% of new infections among MSM fell short of UNAIDS’ fast track targets of 2020.³ The situation calls for comprehensive HIV prevention programmes to curb the epidemic. However, the major financial cuts for HIV programmes in early 2025 bode a crisis for HIV treatment and prevention globally, with potential risk for reversal of the epidemic, with projections showing an additional 4.43–10.75 million new HIV infections.⁴ To meet the 2030 ‘End AIDS Epidemic’ targets, preventing new infections among MSM and TGW is crucial. Introducing highly effective prevention options, such as pre-exposure prophylaxis (PrEP), could be critical to address the sustained epidemic driven by the key population in India.^{5 6}

The oral PrEP has the potential to close the HIV prevention gap for MSM and TGW. It can also serve as an entry point for linking high-risk populations to HIV care in India. Following the promising results of clinical trials7,9 in 2012 and approvals by the U.S. Food and Drug Administration (FDA) and Centre for Disease Control (CDC) for using co-formulated Tenofovir Disoproxil Fumarate (TDF) and Emtricitabine (FTC) for HIV PrEP among adults at high risk of HIV infection, the Central Drugs Standard Control Organisation (CDSCO), India, granted marketing approval for a fixed dose combination of 300 mg of TDF and 200 mg of FTC to reduce the risk of sexually acquired HIV in high-risk individuals.¹⁰ Despite the proven efficacy in clinical trials, the effectiveness of PrEP depends on community willingness to use it. A pooled analysis across cities in India and other studies reports high willingness for PrEP use among MSM, including research from India.11,13 A demonstration project among female sex workers in India found PrEP integration feasible, effective and safe, with no reported seroconversions.¹⁴ However, it was highlighted that this willingness is contingent on effectively addressing individual, social and structural barriers.¹⁵ It is important to understand that traditionally, the TGW community is organised into gharanas (household), headed by a guru (head), and there is a chela (disciple TGW), who is introduced to the gharana through a ritualistic ceremony.¹⁶ Guru-chela relationships are non-egalitarian; rather, there is a hierarchical power relationship.¹⁷

Additionally, current funding disruptions have affected community-led grassroots work, thus negatively impacting the efforts to reduce stigma, discrimination and gender-based violence.¹⁸

Evidence for PrEP feasibility has become an urgency, especially with evidence of stabilising epidemic among MSM/TGW in India.^{2 19}

The demonstration project aimed to study the access, uptake, retention and adherence of daily oral PrEP among MSM and TGW in real-world settings in the states of Maharashtra and Punjab, where HIV prevalence rates were 4.18% and 11.6%, respectively, among MSM, exceeding the national average in India.²⁰ The study also identifies the implementation challenges and informs the strategies for scale-up of oral PrEP in real-world settings.

Methods

A prospective cohort demonstration project among MSM and TGW was conducted in community-based and clinic-based settings in Jalandhar (Punjab) and Pune (Maharashtra), respectively, in India between 18 December 2019 and 31 July 2023. Pune and Jalandhar, two major cities in the western and northern states of Maharashtra and Punjab, respectively, were the sites selected based on higher than the national HIV prevalence among MSM²¹ and the available infrastructural setup. Pune is a large metropolitan education, IT and industrial hub with diverse populations and a well-developed healthcare infrastructure. At Pune, a ‘clinic-based setting’ for PrEP delivery was established at an existing research clinic. The second site, Jalandhar, is a commercial city. Despite healthcare facilities available, healthcare access by the surrounding rural communities is inequitable²² because of high catastrophic costs.²³ The city has a more close-knit community structure, having a vibrant Punjabi culture, with fun-loving, joyful people. Here, community-based organisations play a central role in HIV prevention outreach. At Jalandhar, a ‘community-based setting’ for PrEP delivery was established, and the study was implemented by a well-established community-based organisation (CBO), working for the LGBTQI and PLHIV people and key populations in the districts of Punjab.

Both the study sites had a counselling room, a clinical examination room, a blood collection room/space and a dedicated pharmacy managed by a team of one clinician, two counsellors, field workers and a pharmacist. The community-based setting was managed by the team that primarily belonged to the MSM/TGW community, while in the clinic-based setting, only one team member belonged to the MSM/TGW community, who was the guide for ‘community literacy and sensitive community engagement’ for the Pune team. The services provided through both settings were the same. Owing to funding limitations, we could not establish both types of study settings together. The overall sample size for the demonstration project and selection of participants at substantial risk were as per the WHO guidelines.²⁴

A total of 650 consenting, 18 years and above, self-identified TGW/MSM at substantial HIV risk in the last 6 months were enrolled. Participants were screened for HIV, HBV, HCV, liver and renal functions in 7 working days before initiation of oral PrEP. All eligible and enrolled participants were dispensed PrEP consisting of Tenofovir Disoproxil Fumarate 300 mg and Emtricitabine 200 mg, along with adherence counselling. Participants were followed up twice a month for the first 3 months and quarterly thereafter for 12 months for side effects, adverse events, adherence assessment and HIV. Risk reduction counselling was provided at each follow-up visit and at any contact that was established during the COVID-19 lockdown period. Exit interviews were conducted among 20% of enrolled, willing and available participants (MSM=131 & TGW=17). The interview focused on understanding willingness to continue/discontinue PrEP, barriers and facilitators of PrEP use. Trained social scientists holding Masters level education at the clinic-based setting, while at the community-based setting, community workers with Bachelors/Masters degree conducted the exit interviews. All the interviews were noted verbatim in local language (Hindi/Marathi/Gurumukhi) and translated verbatim in English. Adherence was evaluated using questionnaires, pill counts conducted at each visit, and measurements of TDF levels in a randomly selected 20% of participants and in individuals who seroconverted. HIV seroconversions and creatinine clearance levels were monitored. Safety and adverse events were monitored by the clinicians and discussed with SB. Detailed study eligibility criteria and study procedures are described in the protocol.²¹

Data analysis

Selected variables and regression analyses were done in Stata 16.1. The data were assessed independently for the two of the service delivery models but they were not compared because the study was not powered. Demographic variables included age, religion, education, marital status, living with and occupation. The demographic variables were segregated by behavioural risk group, viz. MSM and self-reported gender identity (TGW), as well as by study-site settings. Retention was defined as study follow-up visits. Those who completed the final follow-up visit were categorised as ‘retained in the study’. The retention rate, acceptability and adherence, along with associated 95% CIs, were calculated overall and by the study settings. Factors influencing retention were analysed using a logistic regression model. The variables for the multivariable logistic regression models were selected based on significance in univariate analysis at p≤0.2. We used a higher threshold of p≤0.2 to avoid excluding important predictors and confounders of clinical relevance and literature.^{25 26}

We analysed ‘adherence to PrEP medicines’ data of participants whose adherence levels were available for the complete study duration. Participants who completed the study follow-up were comparable in age and group classification (MSM or TGW) with those ‘who were lost to follow-up’. They differed significantly in study settings, marital status, education and occupation. We calculated the percent adherence of participants to PrEP medicines and recoded them with the cut-off value of 80%, comparable to published PrEP efficacy trials earlier.^{24 27} The binary logistic regression model was used to identify correlates of adherence. Data were analysed to identify correlates of retention and adherence by adjusting for study settings and socio-demographic, sexual, psychosocial and other cofactors. The incidence rate of HIV among PrEP users was calculated against person-time follow-up.

The study was approved by the Institutional Ethics Committees of ICMR National AIDS Research Institute (EC Protocol Number: NARI EC/2017-07), HIV/AIDS India Alliance and presented to the CDSCO, India. All study participants provided voluntary written informed consent before participating in the study.

Patient and Public Involvement

Continued discussions with CBOs and the Community Advisory Board (CAB) were held for community literacy, mobilisation and advocacy for PrEP.

Results

Screening and enrolment

Of the 910 screening and 58 re-screening, a total of 813 MSM/TGW were found eligible (online supplemental figure 1). Reasons for ineligibility included positive results for HIV (42), HCV (26), HBV (12), HBV-HCV co-infection (02) and HIV-HCV coinfection (01). Of those who were re-screened, they were ineligible temporarily because of low creatinine clearance or the presence of STI; some remained ineligible as they continued to present abnormal laboratory values or had medical/surgical reasons. A total of 163 eligible potential participants did not get enrolled for multiple psychosocial reasons. Finally, 650 were enrolled, of which 529 were MSM and 121 were TGW. A total of 331 and 319 MSM/TGW were enrolled at the community-based and clinic-based sites, respectively.

Demographic characteristics

The median age of participants was 26 years (IQR 22–30), with 363 (55.8%) aged ≥26 years, 536 (82.5%) were unmarried and 615 (94.6%) were literate (online supplemental table 1). At the clinic-based site, nearly two-thirds (62.7%) of the participants reported having completed graduation and one-fifth (20.4%) were students. A total of 224/529 (42.3%) MSM and 29/121 (24.0%) TGW were graduates. Overall, 367 (56.5%) participants were living with their family, while 335 (63.3%) of MSM and 52 (43.0%) of TGW reported living with their family and guru/chela, respectively. More than half (54.5%) of the TGW participants reported sex work as their primary occupation. Among the 650 enrolled participants, the median age at first sex was 17 (IQR 15–19) years, with the youngest reported age as 6 years (lower, upper limit: 6, 37 years). (Data not shown in table 1).

Retention

A total of 411 participants (63.2%) were retained at the end of the study. Retention rates after 1 year were 84.0% and 41.7% at community-based and clinic-based sites, respectively. In the community-based site, retention was significantly associated with having moderate [AOR 7.9 (1.4, 45.0) p=0.019] and high [AOR 10.2 (1.2, 85.4) p=0.032] HIV risk perception. Conversely, the odds of retention were 70% lower among participants who had a regular partner [AOR 0.3 (0.1, 0.9) p=0.037] online supplemental table 2.

Online supplemental table 2 also shows that retention was significantly associated with moderate (AOR 2.1 (1.1, 3.8) p=0.019) and high (AOR 2.7 (1.2, 6.1) p=0.020) HIV risk perception at the clinic-based site. Conversely, the odds of retention were 50% and 60% lower among participants who were living with family (AOR 0.5 (0.3, 0.96) p=0.036) and guru/chela (AOR 0.4 (0.2, 0.9) p=0.022), respectively. Retention was also negatively associated with report of ‘having some special support (for anything) from parents/siblings’ (AOR 0.4 (0.2, 0.9) p=0.034), girlfriend/boyfriend (AOR 0.3 (0.2, 0.6) p=0.001), and Guru/Chela (AOR 0.4 (0.2, 0.99) p=0.048) in the clinic-based site.

Adherence

Adherence data were available for 355 (86.4%) of 411 participants. Of these, 230 (64.8%) participants showed >80% adherence, with 186 (69.1%) and 44 (51.2%) participants reporting >80% adherence in community-based and clinic-based site respectively. Online supplemental table 3 shows the correlates of adherence at the end of the study. Being from community-based setting (AOR 2.3 (1.1, 4.9) p=0.032), up to secondary (AOR 6.5 (2.2, 19.8) p=0.001), higher secondary (AOR 6.5 (1.7, 15.0) p=0.001) and graduate level (AOR 4.1 (1.2, 12.2) p=0.020) education and living with family (AOR 3.4 (1.6, 8.3) p=0.002) were found to be significantly associated with >80% adherence, respectively. 20% of the blood samples and all samples of seroconverted participants (n=7) were tested for TDF level using LC-MS/MS test. A detectable concentration of >40 ng/mL was found in 119 (72.6%) participants (data not shown in the tables).

Safety

There were two reported serious adverse events, namely, hospitalisation for (1) gender-affirming surgery and (2) COVID-19 care; these were not related to PrEP. A drop in creatinine clearance of grade 2 severity from baseline or their last visit was recorded for 63 (9.7%) participants (Data not shown in tables). Of these, the creatinine clearance in five participants was below 60 mL/min, but none reported any symptoms. PrEP was withheld in these five participants and reinitiated in four participants after improved creatinine clearance. All participants recovered by the end of the study. Gastrointestinal system disorders were the most reported ‘Adverse Event’ (AE) in the first 15 days among 26 MSM and 7 TGW. All AEs were resolved with symptomatic treatment and dietary counselling.

HIV seroconversion

A total of seven seroconversions were detected among the MSM PrEP users only, with an estimated HIV incidence of 1.57 per 100 PY (95% CI 0.6 to 3.2). There was no seroconversion among TGW in this study. Of the seven seroconversions among MSM, three were from the community-based setting, who were 30–35 years, reporting multiple sexual partners, and they had seroconverted at 3-month and 6-month follow-up visits. There were four seroconversions at the clinic-based setting, with all except one below 30 years of age (20–37 years). Two of them reported a history of substance abuse, and they had seroconverted between enrolment to 3-month follow-up visit. A history of missed pill doses and high-risk behaviour was reported by all the seroconverted participants. The TDF level at their last visit ranged between 25.6 ng/mL and 230.33 ng/mL (Data not shown in the tables).

MSM versus TGW overview

The study was not powered for comparison, and some of the differences between MSM and TGW are described. HIV low/no risk perception was reported by 33% MSM and 26% TGW. The difference in the levels of HIV risk perception was significant (p<0.001). Retention was found non-significant between MSM (65%) and TGW (56%) (p=0.075). Adherence of >80% was significantly different between MSM (69%) and TGW (41%) (p<0.001). There was a significant difference in condomless anal sex (CAS) between MSM (82%) and TGW (61%) (p<0.001) at baseline. At the end line, among the retained participants, this difference was non-significant, with 77% MSM and 82% TGW reporting CAS. In both sites, participants showed a willingness to continue with PrEP at subsidised rates (n=488, 75%), of which 405/529 (77%) and 83/121 (69%) were MSM and TGW, respectively (Data not shown in tables).

Acceptability for laboratory investigations

Acceptability for laboratory investigations for HIV and creatinine clearance was felt important, and anxiety was reportedly reduced.

It is good that every 3 months testing is done, use rapid test more… an immediate report should be there… of HIV. Anxiety will be less. [Clinic-based site-MSM-279]

PrEP use and other behaviours

Most of the participants, 471 (72.5%), reported willingness to take PrEP. At the end of 1 year, 341 of 411 (83.0%) retained participants continued to be willing for PrEP (p<0.001) (Data not shown in the tables). Comprehending the risk from a partner/s was often cited as fear and therefore a reason to continue with PrEP in the future.

When your partner isn't telling you everything, and you don't know how risky he is, you can be sure that you're safe because you're taking PrEP… I started PrEP because friends were testing HIV positive. [Clinic-based site-TGW-146]

I have sexual relations with people who further might have sexual relations [/with others/], so I need PrEP to be safe from the disease. [Community-based site-MSM-238]

TGW feared being stigmatised by their community if seen at a clinic-based setting, reflecting the non-acceptability of the clinic-based setting among the ‘visible’ community.

If our community people see us, they will tell me why you are going there [Clinic-based setting]? What will they [community] think? That she has HIV infection… [Clinic-based site-MSM-277]

Both MSM and TGW shared belief that PrEP could save them.

This medicine will protect me from the dangerous disease spreading in the community. [Community-based site-MSM-227]

Because it will save me from HIV and set me free. [Community-based site-TGW-28]

Protection from HIV, people who are in sex work will feel assured, I will not get HIV in the future because I am taking PrEP. [Clinic-based site-TGW-286]

Among both MSM and TGW participants, side effects were tolerable enough to continue PrEP.

No side effects of medicine, initially only cold and cough, later no effect. I was not afraid of HIV because of PrEP. [Clinic-based site-MSM-279]

My experience with PrEP tablet is very good. Initially, I had some problems [/side effects/], but after that everything was fine. [Community-based site-TGW-05]

In the exit interviews, participants’ willingness to purchase PrEP emerged, while some TGW expected PrEP to be free of cost.

For HIV prevention, I would like to purchase. I want to prevent myself and my wife also. [Clinic-based site-MSM-self-identified bisexual-257]

The cost of this [/PrEP/] should be as low as it can be, cost should be around [/INR/] 500- 600 [/6-7$/]. [Community-based site-MSM-227]

I want to get this PrEP free of cost. [Community-based site-TGW-41]

Episodic PrEP use was reported by the participants:

Because I am not at risk right now, I am currently having only one partner, so I will not take this pill. [Clinic-based site-TGW-286]

If I am more active in sexual activity or if I am in the HIV risk zone, then there is a need [/for PrEP/], otherwise not. [Community-based site-MSM-242]

Of all the 411 retained participants, CAS with any partner, anytime in the past 6 months, was reported among 84.4% (n=347) at baseline, while at 12 months follow-up visit, it was reported among 78.1% (n=321) (p<0.001) (Data not shown in the tables). The exit interview showed the impact of risk-reduction counselling among the PrEP users.

Such a thought [/not use condom/] came to my mind, but it was told here earlier that these pills only prevent HIV and do not prevent other sexually transmitted diseases, so I have always used condoms. [Clinic-based site-MSM-295]

In my mind, it was there that if I had started taking PrEP medicine, then I would be safe from HIV. I did not know that it does not prevent STIs. However, PrEP staff told me about it, and then I started using a condom. [Community-based site-MSM-223]

Our community outreach workers highlighted the low retention among TGW because the decision to take/continue PrEP was in the hands of the Guru.

Transgender people listen to their Gurus. Only they [/those who were permitted/] used to come for the follow-ups. Because of this, there were problems with retention. They used to follow the orders given by the Guru. [Clinic-based site, community outreach worker]

Traditional income generation, such as ‘singing/ dancing and blessings’ [/at weddings and birth in exchange for gifts and money/] and ‘mangati’ [/begging/], got hampered among TGW, hence Gurus’ dictum posed barriers:

In ‘Dera’ [/residence of TGW and Guru/], the community [/TGW/] used to go around people’s houses asking Badhai [/gift money from people on family celebrations/], and they used to go daily in the morning. And if they come here [/study site/] then their [/TGW’s/] timings used to get disturbed, and they were unable to take the badhai to the Guru that day, and the Guru used to get angry. And the Gurus did not allow all these things [/Participation/]. [Community-based site, community outreach worker]

Impact of COVID-19 on enrolment/follow-ups

The national lockdown between March and May 2020 halted the new screenings and enrolments, but 54 MSM participants were already initiated on PrEP in the clinic-based site. The enrolments in the community-based site had not begun. The prevalent fear of COVID-19 and limited transportation resulted in the loss to follow-ups at the clinic-based site. Eighteen participants did not return from their hometowns during the lockdown. With prior regulatory approvals and following appropriate COVID-19 prevention guidelines, PrEP tablets were dispensed to willing enrolled participants at confidential, covert, and acceptable venues.

Discussion

Daily oral co-formulated Tenofovir Disoproxil Fumarate and Emtricitabine (TDF/FTC), HIV PrEP was found to be feasible to deliver, acceptable, and safe in a clinic-based setting in Pune and a community-based setting in Jalandhar in the first PrEP demonstration project among MSM/TGW in India. While a substantial proportion of participants demonstrated high adherence (>80% adherence) to PrEP, adherence to PrEP remained dependent on perceived self-risk of HIV infection, as the participants stopped taking PrEP when they perceived themselves as low-risk, indicating the need for future research on ‘event-based and/or long-acting PrEP’ in the Indian context.

A decrease in CAS was observed in our study, which concurs with findings of the IPERGAY trial²⁸ showing a decrease in behavioural disinhibition among PrEP users. Several other studies predicted the converse.^{29 30} Our findings are similar to the demonstration project conducted among women involved in criminal justice systems (WICJ)³¹ and studies conducted among MSM globally.^{32 33} Reduction in CAS in our study might be attributed to enhanced risk-reduction counselling; we recommend risk assessment, especially focusing on CAS, before the initiation of PrEP among MSM and TGW. The programme would need to institute repeated risk-reduction counselling for this population when initiated on PrEP. Counselling for condom use and STI prevention/treatment counselling, along with PrEP was evidently successful and therefore strongly recommended.

Low retention in our study could be due to episodic usage of PrEP, namely, self-deferment of PrEP due to low self-risk perception, hiatus, or end of a sexual relationship, or events like examination, vacation, or travel. These resulted in PrEP interruption showing a recurring pattern or cyclical order of PrEP avoidance for some period and further planning sexual encounters and mitigating the risk by using PrEP again. A similar argument might hold for situations of the COVID-19 pandemic. In a pharmacokinetic modelling study, utilising data from the iPrEx trial,³⁴ participants taking four doses of PrEP weekly could provide up to 96% protection against HIV infection. Several other studies have also reported the PrEP decision-making process to be linked to HIV risk perception.35,37 Understanding the stochastic nature of PrEP intake, which depended on the participants’ self-risk perception/plan, episodic PrEP intake needs to be studied further among MSM and TGW in India. To address the stochastic nature of PrEP intake, we recommend continuous ‘Informed Choice Counselling (ICC)’ among PrEP users to monitor the continuous change process for PrEP medication adherence. Additionally, allied behavioural practices may be developed, studied, and implemented to address seasons/cycles of risk observed in this study.

Gurus, in the ‘Hijra’ (Transgender person) community in India, provide a guardian’s role, a sense of community³⁸ with Gurus also having centralised power and sometimes sociologically repressive authority over their chela (TGW).³⁹ Lower retention among MSM living with family and TGW living with Gurus reflects the exercise of traditional sovereign powers on the chelas (disciples)/children/wards. Retention among TGW and younger MSM is being peripherally determined by these traditional sovereign powers. Tailored approaches are recommended to maximise effectiveness and impact;³ among others, India would need the strong engagement of the Gurus and families.

The changeable nature of oral PrEP usage indicated in this study informs the programme to evaluate adherence in alignment with the user’s self-risk perception and ensure effective usage of oral PrEP. This would prevent underestimation of monthly adherence, which has implications for rating programme outcomes. Risk assessment and messages around PrEP need to be aligned with the end users’ needs and usage patterns based on their plans/seasons. The findings provide encouraging evidence for new long-acting antiretrovirals, which should be further investigated in India.⁴⁰ Lessons from the suboptimal uptake of PrEP in Asia and the Pacific region among key populations and their partners,⁴¹ is applicable for India, which has its own concentrated epidemic. India, thus, needs to urgently include PrEP in its national programme. PrEP rollout under the national programme has not occurred. It is to be noted that India has conducted demonstration projects^{14 21} and has developed technical guidelines for PrEP for the key population at substantial risk.⁶ The major concerns expressed in the guidelines were acceptability, adherence, and behavioural disinhibition after using PrEP. This study is the first to document that PrEP was acceptable and safe to the majority of MSM and TGW participants in both service delivery settings in two states of India.

Despite the recent financial cuts in HIV treatment and prevention services, India has been able to sustain the treatment for all infected individuals through self-funding.⁴² In view of recent financial cuts, another possible constraint envisaged for inclusion of PrEP in the guidelines by the programme could be financial burden on the programme in addition to providing anti-retroviral treatment to infected individuals. Our study provides evidence on the willingness of many participants to purchase PrEP medicines at subsidised rates, which needs to be planned. Similar preferences for free or subsidised PrEP have been reported by demonstration projects in female sex workers in India⁴³ and MSM in the United States⁴⁴ and in India.^{12 45} In view of limited funding and to ensure equitable distribution of PrEP to those most at risk, the programme needs to prioritise MSM and TGW at ‘substantial risk’ of HIV as per WHO criteria for PrEP roll-out in the countries. In India, too, most of the funding for HIV prevention was external. Although the National AIDS Control Programme, India, had indicated self-sufficiency with domestic funds,⁴² the government, communities, civil society, and donors need to unite and seize this opportunity to move the HIV prevention programme domestically. The community-based setting showed better retention and adherence, indicating acceptance of community-managed PrEP dispensing, thus opening the idea of empowering and engaging community volunteers to help in task shifting to accommodate drying resources.

The two delivery models successfully helped serve the invisible and visible key populations, respectively. The acceptance of PrEP in each site shows that stigma, discrimination, and criminalisation that prevent people from accessing health services could be addressed for diverse MSM/TGW. The diverse community accessing a typical delivery setting emphasises the need for both delivery models in the country to facilitate a smooth PrEP roll-out.

In our clinic-based delivery model, community liaisons were involved to engage with the community to improve outreach, as reported in other countries like Australia and New Zealand.46,48 As reported in a review,⁴⁹ the community setting model was most adopted in the Asia‐Pacific region. Community outreach using peer educators would be a suitable model for both clinic and community-based delivery settings in India also.

In the clinic-based model in our study, students, college-based, elite-educated MSM and TGW were reached; digital tools and respondent-driven referring are recommended along with community engagement and focused adherence and risk-reduction counselling. To institute a client-centred dynamic choice⁵⁰ for optimal HIV prevention intervention, both the sites, namely, Pune and Jalandhar might need both delivery models, which can be tested in the future. However, in both cities, there were specific barriers, namely, Guru chela dynamics; invisibility needs of different socio-economic strata and different categories of key populations, affordability, or social class, stigma, parental control, etc., and these indicate a need for adaptable health services.

With the emerging evidence of the community-based setting serving mostly to visible MSM/TGW, and the clinic-based delivery model reaching educated elite hidden MSM who preferred the clinic because of its exclusivity and anonymised and confidential ambience, models of differentiated testing and treatment services in Kenya and Uganda can be adapted for optimal PrEP coverage.50,52 The differentiated PrEP delivery model⁵³ is ideally suited for a vast country like India, which serves the diverse needs of its diverse key populations. Adopting ‘Differentiated Service Delivery’ (DSD) with prioritisation of MSM and TGW with ‘substantial risk’ for oral PrEP might be cost-effective in this resource-limited setting.

Nearly a quarter of the participants in our clinic-based setting were students with initiation of sex at a young age, much before the targeted intervention under the programme could have reached them.^{54 55} This emphasises the pressing need for early biomedical, behavioural, and structural approaches, including PrEP interventions among adolescent boys and young adults. A rigorous evaluation of combination prevention intervention packages for adolescents and young adults should be performed. Alluding to the findings that different categories of MSM and TGW have different needs, we recommend a DSD model for PrEP through both community and clinic-based settings. Under the DSD model, the ‘who’ component to be prioritised are: (1) MSM and TGW who are young, (2) students, (3) those with low or medium self-risk perception, and (4) those reporting sex work as their primary occupation.⁵⁶ Confirmation of eligibility using substantial risk parameters for PrEP, periodically/quarterly by trained experts, for more rational, efficient, and cost-effective usage of PrEP is recommended.⁵³ It will be critical to include continued assessment of risk for the possibility of self-stoppage of PrEP. Quarterly monitoring was acceptable and became essential to address sudden drops in creatinine clearance, and for HIV and HBV/HCV observed in the study. Continued self-risk perception assessments need to be included in the PrEP counselling.

Strengths and limitations

The study followed a cohort of a cumulative of 650 MSM/TGW at community-based and clinic-based settings at Jalandhar and Pune, respectively, which forms the strength of the study. Although data from two study settings may not be considered generalisable, the differentiated service delivery method used in our study provides evidence on actual implementation issues like acceptability, adherence assessment in the programme setting and identification of changing risk among MSM and TGW. Employing multiple recruitment strategies across two delivery settings was a key strength.⁵⁷ However, the differences between the MSM and TGW populations at two settings could have influenced participant retention. The adherence in our study was dependent on cycles of perceived self-risk and COVID-19 pandemic, which could have affected the follow-up visits.

Conclusions

Daily oral PrEP is found acceptable with high adherence among the retained MSM and TGW and it should be included in the basket of HIV prevention options. Lessons from other demonstration projects globally and regionally substantiate the need for differential delivery modes of community and clinic-based setting that would be needed to target different categories of MSM and TGW in India. The emerging behaviours of perceived/planned low-risk periods of taking oral PrEP suggest need for further investigations for event and/or situation-driven uptake. Risk reduction counselling and periodic monitoring throughout the usage of PrEP is strongly recommended.

Data availability statement

Data are available upon reasonable request.