Abstract

Introduction

Travestis and transgender women (TrTGW) are among the key populations with the highest HIV prevalence, accounting for 70% of HIV infections worldwide in 2021 ¹1. Joint United Nations Programme on HIV/AIDS. In danger: UNAIDS Global AIDS Update 2022. Genebra: Joint United Nations Programme on HIV/AIDS; 2022.. TrTGW not only have a disproportionate probability of infection compared with the general population, with a 14-fold increased risk compared with cisgender women ¹1. Joint United Nations Programme on HIV/AIDS. In danger: UNAIDS Global AIDS Update 2022. Genebra: Joint United Nations Programme on HIV/AIDS; 2022.^,22. Baral SD, Poteat T, Strömdahl S, Wirtz AL, Guadamuz TE, Beyrer C. Worldwide burden of HIV in transgender women: a systematic review and meta-analysis. Lancet Infect Dis 2013; 13:214-22., but also a lower adherence to prevention and treatment strategies than other vulnerable groups ³3. Sevelius JM, Patouhas E, Keatley JG, Johnson MO. Barriers and facilitators to engagement and retention in care among transgender women living with human immunodeficiency virus. Ann Behav Med 2014; 47:5-16.^,44. Baguso GN, Turner CM, Santos G-M, Raymond HF, Dawson-Rose C, Lin J, et al. Successes and final challenges along the HIV care continuum with transwomen in San Francisco. J Int AIDS Soc 2019; 22:e25270.. Their experiences, permeated by stigma and discrimination, are often crossed by social markers, such as class and race, that make belonging to a minority group even more challenging and promote generally less favorable health outcomes ⁵5. Hughto JMW, Resiner SL, Pachankis JE. Transgender stigma and health: a critical review of stigma. Soc Sci Med 2015; 147:222-31..

Besides the stigma faced daily, TrTGW have specificities that have not been sufficiently met, such as care related to gender issues, the need to be called by social name, and care at alternative periods in health services ⁶6. Rocon PC, Rodrigues A, Zamboni J, Pedrini MD. Difficulties experienced by trans people in accessing the Unified Health System. Ciênc Saúde Colet 2016; 21:2517-25.^,77. Nadal KL, Davidoff KCBS, Fujii-Doe WMA. Transgender women and the sex work industry: roots in systemic, institutional, and interpersonal discrimination. J Trauma Dissociation 2014; 15:169-83.. Another problem concerns the other health issues experienced, leading this group not to consider HIV diagnosis as a priority ⁸8. Pelúcio L. Abjeção e desejo: uma etnografia travesti sobre o modelo preventivo de aids. São Paulo: Annablume Editora/Fundação de Amparo à Pesquisa do Estado de São Paulo; 2009.. Therefore, HIV infection in TrTGW should be seen as a syndemic ⁹9. Poteat T, Scheim BA, Xavier J, Reisner S, Baral S. Global epidemiology of HIV infection and related syndemics affecting transgender people. J Acquir Immune Defic Syndr 2016; 72 Suppl 3:S210-9..

In Brazil, the AIDS detection rate fell from 21.4 cases per 100,000 population in 2012 to 18.0 in 2019 ¹⁰10. Secretaria de Vigilância em Saúde, Ministério da Saúde. Boletim Epidemiológico HIV/AIDS 2021; número especial. https://www.gov.br/saude/pt-br/centrais-de-conteudo/publicacoes/boletins/epidemiologicos/especiais/2021/boletim-epidemiologico-especial-hiv-aids-2021.pdf/view (acessado em 27/Mar/2023).
https://www.gov.br/saude/pt-br/centrais-... . Since 1996, the Brazilian Unified National Health System (SUS) guarantees universal and free access to HIV treatment, including appointments with infectious disease specialists and other specialized professionals ¹¹11. Departamento de Condições Crônicas e Infecções Sexualmente Transmissíveis, Ministério da Saúde. Sou +, estou indetectável. http://antigo.aids.gov.br/indetectavel/ (acessado em 27/Mar/2023).
http://antigo.aids.gov.br/indetectavel/... ^,1212. Montenegro L, Velasque L, LeGrand S, Whetten K, Rafael RMR, Malta M. Public health, HIV care and prevention, human rights and democracy at a crossroad in Brazil. AIDS Behav 2020; 24:1-4.. This initiative made Brazil, at the time, a pioneer in the fight against the AIDS epidemic ¹³13. Berkman A, Garcia J, Muñoz-Laboy M, Paiva V, Parker R. A critical analysis of the Brazilian response to HIV/AIDS: lessons learned for controlling and mitigating the epidemic in developing countries. Am J Public Health 2005; 95:1162-72.. However, despite the reduction in the AIDS detection rate in the general population, some vulnerable groups, such as TrTGW, still have difficulty initiating treatment, and, when they do, they often fail to maintain adherence ¹⁴14. Baguso GN, Gay CL, Lee KA. Medication adherence among transgender women living with HIV. AIDS Care 2016; 28:976-81., partly due to a combination of factors, such as the stigma related to gender identity, extreme social inequality, and the insufficiency of inclusive public policies in education and health. This unfavorable context promotes high school dropout rates and affects access, linkage, and retention to health services ³3. Sevelius JM, Patouhas E, Keatley JG, Johnson MO. Barriers and facilitators to engagement and retention in care among transgender women living with human immunodeficiency virus. Ann Behav Med 2014; 47:5-16.^,1515. Amorim ACPS, Rocha TS, Diniz DMFA. Evasão escolar de mulheres trans e travestis: uma análise acerca da (in)existência de políticas públicas educacionais. In: VI Congresso Nacional de Educação. https://editorarealize.com.br/artigo/visualizar/60450 (acessado em 04/Set/2022).
https://editorarealize.com.br/artigo/vis... .

Some strategies to overcome these barriers have been proposed, with the potential to meet, even if only partially, the specific needs of TrTGW, especially regarding HIV prevention and treatment. As this group has difficulty accessing and low adherence and retention to health services, combined social-structural interventions that considered what makes this population less likely to adhere to treatment have been developed. Among these interventions, we highlight sensitization and training of health professionals for the development of specific skills, social support, and interventions focused on gender affirmation, such as the expansion of access to hormone therapy integrated with HIV treatment, aiming to promote autonomy and reduce stigma and discrimination in health services ³3. Sevelius JM, Patouhas E, Keatley JG, Johnson MO. Barriers and facilitators to engagement and retention in care among transgender women living with human immunodeficiency virus. Ann Behav Med 2014; 47:5-16.^,1616. Bockting W, MacCrate C, Israel H, Mantell JE, Remien RH. Engagement and retention in HIV care for transgender women: perspectives of medical and social service providers in New York City. AIDS Patient Care STDS 2020; 34:16-26.^,1717. Sevelius J, Murray LR, Fernandes NM, Veras MA, Grinsztejn B, Lippman S. Optimising HIV programming for transgender women in Brazil. Cult Health Sex 2019; 21:543-58.^,1818. Sevelius JM, Carrico A, Johnson MO. Antiretroviral therapy adherence among transgender women living with HIV. J Assoc Nurses AIDS Care 2010; 21:256-64..

Systematic reviews conducted in several continents with TrTGW living with HIV ²2. Baral SD, Poteat T, Strömdahl S, Wirtz AL, Guadamuz TE, Beyrer C. Worldwide burden of HIV in transgender women: a systematic review and meta-analysis. Lancet Infect Dis 2013; 13:214-22. and, especially, a study with rural populations living with HIV in South Africa ¹⁹19. Lippman SA, Shade SB, Sumitani J, DeKadt J, Gilvydis JM, Ratlhagana MJ, et al. Evaluation of short message service and peer navigation to improve engagement in HIV care in South Africa: study protocol for a three-arm cluster randomized controlled trial. Trials 2016; 17:68. served as the basis for the development of the TransAmigas project. This project aimed to evaluate the effectiveness of peer navigation in increasing adherence to antiretroviral therapy (ART) among TrTGW. Peer navigation interventions are strategies in which individuals who are part of the same key population, with similar profiles, follow up the participants for a certain period to help them with their health-related issues. This strategy has proven important to improve HIV-related outcomes by providing an opportunity to focus on patients’ personal realities beyond health services, which contributes to overcome barriers and facilitates treatment adherence and retention ²⁰20. Karver TS, Barrington C, Donastorg Y, Perez M, Gomez H, Page KR, et al. Exploring peer navigation and support in the quality of HIV care experiences of female sex workers in the Dominican Republic. BMC Health Serv Res 2022; 22:56..

This project was conducted in São Paulo, the most populous city in Brazil, with more than 12 million inhabitants. This city has a high Human Development Index (HDI) ²¹21. Instituto Brasileiro de Geografia e Estatística. São Paulo. https://cidades.ibge.gov.br/brasil/sp/sao-paulo/panorama (acessado em 10/Nov/2022).
https://cidades.ibge.gov.br/brasil/sp/sa... , but about 85% of its inhabitants live in areas of social exclusion ²²22. Sposati A, Monteiro M, Koga D, Ramos FR, Anazawa TM, Coelho G. Desigualdades nos territórios da cidade: métricas sociais intraurbanas em São Paulo. São Paulo: Educ; 2017..

São Paulo is home to an important community of homosexual, bisexual, and transgender people. Since 1997, in partnership with public authorities and private institutions, the social movement annually holds one of the largest LGBT+ pride parades in the world, which surpasses one million participants since its seventh edition in 2003 ²³23. Simões J, Facchini R. Na trilha do arco-íris: do movimento homossexual ao LGBT. São Paulo: Editora Fundação Perseu Abramo; 2009.. Same-sex marriage was legalized in the country in 2013 ²⁴24. Conselho Nacional de Justiça. Resolução nº 175, de 14 de maio de 2013. Dispõe sobre a habilitação, celebração de casamento civil, ou de conversão de união estável em casamento, entre pessoas de mesmo sexo. https://atos.cnj.jus.br/atos/detalhar/1754 (acessado em 10/Nov/2022).
https://atos.cnj.jus.br/atos/detalhar/17... and, since 2018, transgender people have the right to legally change their names ²⁵25. Conselho Nacional de Justiça. Provimento nº 73, de 28 de junho de 2018. Dispõe sobre a averbação da alteração do prenome e do gênero nos assentos de nascimento e casamento de pessoa transgênero no Registro Civil das Pessoas Naturais (RCPN). https://atos.cnj.jus.br/atos/detalhar/2623 (acessado em 10/Nov/2022).
https://atos.cnj.jus.br/atos/detalhar/26... . However, as in the rest of Brazil, TrTGW living in São Paulo face great social marginalization and have difficulty adhering to health care ²⁶26. Pinto TP, Teixeira FB, Barros CRS, Martins RB, Saggese GSR, Barros DD, et al. Silicone líquido industrial para transformar o corpo: prevalência e fatores associados ao seu uso entre travestis e mulheres transexuais em São Paulo, Brasil. Cad Saúde Pública 2017; 33:e00113316..

The living and health conditions of this group have promoted the development of studies in Brazil ²⁷27. Ferreira ACG, Coelho LC, Jalil EM, Luz PM, Friedman RK, Guimarães MRC, et al. Transcendendo: a cohort study of HIV-infected and uninfected transgender women in Rio de Janeiro, Brazil. Transgend Health 2019; 4:107-17.^,2828. Silva RAD, Silva LAVD, Soares F, Dourado I. Uso de hormônios não prescritos na modificação corporal de travestis e mulheres transexuais de Salvador/Bahia, Brasil. Ciênc Saúde Colet 2022; 27:503-14., including in São Paulo ²⁶26. Pinto TP, Teixeira FB, Barros CRS, Martins RB, Saggese GSR, Barros DD, et al. Silicone líquido industrial para transformar o corpo: prevalência e fatores associados ao seu uso entre travestis e mulheres transexuais em São Paulo, Brasil. Cad Saúde Pública 2017; 33:e00113316.^,2929. Maschião LF, Bastos FI, Wilson E, McFarland W, Turner C, Pestana T, et al. Nonprescribed sex hormone use among trans women: the complex interplay of public policies, social context, and discrimination. Transgend Health 2020; 5:205-15.. Aiming to expand the knowledge about the characteristics and needs of TrTGW, TransAmigas project innovated by providing a personalized and culturally targeted intervention for this group, using as a reference the gender affirmative model, developed by Sevelius et al. ³⁰30. Sevelius J, Chakravarty D, Neilands TB, Keatley J, Shade SB, Johnson MO, et al. Evidence for the Model of Gender Affirmation: the role of gender affirmation and healthcare empowerment in viral suppression among transgender women of color living with HIV. AIDS Behav 2021; 25 Suppl 1:64-71. (p. S65), which defines gender affirmation as “an interpersonal, interactive process whereby a person receives social recognition and support for their gender identity and expression”. The gender affirmative model was developed to understand why TrTGW are disproportionately affected in their living conditions and health and how the gender affirmation process relates to risk behaviors, self-care, and search for health services. Thus, this project considers that multidimensional phenomena, such as the deprivation of rights, racism, sexism, and transphobia, converge to widen social inequalities and bring potentially preventable damage to health. On the other hand, negative health outcomes decrease as gender affirmation needs are met by adequate support ³⁰30. Sevelius J, Chakravarty D, Neilands TB, Keatley J, Shade SB, Johnson MO, et al. Evidence for the Model of Gender Affirmation: the role of gender affirmation and healthcare empowerment in viral suppression among transgender women of color living with HIV. AIDS Behav 2021; 25 Suppl 1:64-71.^,3131. Reisner SL, Poteat T, Keatley J, Cabral M, Mothopeng T, Dunham E, et al. Global health burden and needs of transgender populations: a review. Lancet 2016; 388:412-36..

Considering these challenges, this study presents the factors associated with the retention of TrTGW living with HIV to the TransAmigas project.

Methods

The project Improving Health Outcomes for HIV-positive Trans Women in Brazil (TransAmigas project) was conducted as a pilot to test the feasibility and acceptability of a peer navigation intervention aimed at improving adherence to ART among TrTGW ³²32. Lippman S, Sevelius J, Saggese GSR, Gilmore H, Bassichetto KC, Barros DD, et al. Peer navigation to support transgender women’s engagement in HIV care: findings from the Trans Amigas Pilot Trial in São Paulo, Brazil. AIDS Behav 2022; 26:2588-99.. This cohort study was performed in the city of São Paulo in a SUS health service called STD/AIDS Reference and Training Center (CRT-DST/Aids SP). The expected sample size was 150 participants.

Participants were recruited from April 2018 to September 2019, using a convenience sample. In total, 194 potential participants were identified by different initiatives, of which the most effective was inviting CRT-DST/Aids SP users and participants of previous research coordinated by the Center for Research on Human Rights and Health of the LGBT+ Population (NUDHES). External actions were also performed, such as the distribution of printed invitations in spaces of sociability of TrTGW by the peer navigators (PNs) themselves (TrTGW who also lived with HIV). Finally, health professionals who worked with homeless people helped disseminate the research. Among the potential participants, 32.4% (45/194) were excluded for several reasons, such as: not agreeing to schedule an interview; not attending the eligibility confirmation; not being located after telephone contact; or not meeting all eligibility criteria at the time of the interview (being under 18 years old; not living in municipalities in Greater São Paulo; and/or not wanting to enroll in the CRT-DST/Aids SP for HIV treatment).

Of the remaining 149, 14 missed the face-to-face screening, 14 were not eligible, and eight refused to participate in the study. Finally, 113 TrTGW ³²32. Lippman S, Sevelius J, Saggese GSR, Gilmore H, Bassichetto KC, Barros DD, et al. Peer navigation to support transgender women’s engagement in HIV care: findings from the Trans Amigas Pilot Trial in São Paulo, Brazil. AIDS Behav 2022; 26:2588-99. were randomly assigned to either the peer navigation intervention (75) or the control group (38), in a 2:1 ratio.

Procedures

The protocol established that the participants in the intervention group would attend at least one monthly face-to-face meeting with their respective PN, holding individual thematic sessions on self-esteem, self-care, communication strategies, and ART adherence, over nine months. All participants had to be undergoing follow-up for HIV at their partner health service.

At enrollment, a baseline questionnaire was applied face-to-face by trained interviewers. After three months, participants received a phone call to confirm contact, check the quality of the intervention, and be reminded to return at nine months to complete the follow-up questionnaire, either face-to-face or remotely, ending their participation in the project. The initial protocol estimated 70% retention in this stage ³²32. Lippman S, Sevelius J, Saggese GSR, Gilmore H, Bassichetto KC, Barros DD, et al. Peer navigation to support transgender women’s engagement in HIV care: findings from the Trans Amigas Pilot Trial in São Paulo, Brazil. AIDS Behav 2022; 26:2588-99..

To facilitate retention to the project, all participants received a BRL 30.00 compensation at the time of the interviews, regardless of the group to which they were randomized, to cover transportation and food expenses. Moreover, at enrollment, they received a small snack and a kit with beauty products, condoms, an enema, a lubricant gel, and a guide to services for TrTGW, besides BRL 20.00 per month to add minutes on their cell phones to maintain contact with the research team. As an additional incentive for the participants of the intervention group to attend the face-to-face meetings, PNs distributed small gifts, such as pill boxes, appointment books, stylized panties, among others.

Analysis

A quantitative analysis was performed and its outcome was “retention at nine months”, regardless of telephone contact at three months. Retention to the project was defined by the “full completion of the final questionnaire” applied at nine months, which could be answered face-to-face or remotely. Therefore, it is an object cut-out of the original study. This definition is restricted to the contact between the participants and the research team, without considering retention to health care or peer navigation procedures. However, to meet the main objective of the project, clinical data were collected to identify HIV treatment adherence, defined by at least two medical visits or ART abandonment, with a maximum six-month interval during the research period ³²32. Lippman S, Sevelius J, Saggese GSR, Gilmore H, Bassichetto KC, Barros DD, et al. Peer navigation to support transgender women’s engagement in HIV care: findings from the Trans Amigas Pilot Trial in São Paulo, Brazil. AIDS Behav 2022; 26:2588-99.. These data are not part of the scope of this study.

To determine the variables associated with retention to the project, several variables of the baseline questionnaire included in the gender affirmative model were considered, such as sociodemographic and behavioral information and data on gender affirmation, mental health, substance use, and experiences of transphobia.

“Name change on personal documents” was the gender affirmation variable (“Yes, I did”; “I tried, but I couldn’t”; “I’m trying; I’ve already filed the lawsuit”; “I haven’t tried, but I intend to try”; “I haven’t tried”). A normalized four-item score was also created to synthesize a measure of internalized gender affirmation: (1) “Generally speaking, I feel good about myself and my gender identity”; (2) “I fully accept myself in my gender identity”; (3) “I’m proud of being a travesti/trans women”; and (4) “My feelings about my gender identity are mostly positive”. These items describe feelings regarding participants’ own gender identity, which may affect their decision or ability to remain engaged in both the project and health care. The answer options were based on the Likert scale, ranging from 1 (“strongly disagree”) to 5 (“strongly agree”).

The following variables of mental health and use of alcohol and other substances were selected: “psychic distress”, measured using the K10 scale ³³33. Kessler RC, Barker PR, Colpe LJ, Epstein JF, Gfroerer JC, Hiripi E, et al. Screening for serious mental illness in the general population. Arch Gen Psychiatry 2003; 60:184-9. (likely to be well; likely to have a mild disorder; likely to have a moderate disorder; likely to have a severe disorder); “frequency of alcohol use” (never; once a month; 2-4 times a week; 2-3 times a week; ≥ 4 times a week); “illicit substance use in the last six months” (no; yes); “prior suicide attempt” (no; yes); and “HIV serostatus disclosure to someone other than health professionals” (no; yes).

Regarding experiences of transphobia, the following variables were selected: “previous experience of verbal violence for being trans” (no; yes) and “previous experience of sexual violence for being trans” (no; yes). The participants’ “allocation group” (control group; peer navigation intervention group) and their response to the “check telephone call at three months” (no; yes) were also outcome predictor variables.

In order to evaluate the effect of the stigmatization process on the retention of TrTGW, peer contact forms were qualitatively analyzed, considering the questions previously used in the quantitative component as search criteria. The forms were completed by the PNs to record both the attempts to contact the participants and the topics discussed in the meetings, such as their location and the participant’s mental state and appearance. This information was typed and exported to a spreadsheet, allowing the identification of the following thematic categories: schooling level; financial situation; marital status/relationships; occupation/professional situation; general health; mental health; alcohol and other drug use; HIV-related issues; and family and housing situation. The category “additional information” was also included, with other themes that caught the attention of the PNs.

These categories were later divided into analytical subcategories, such as abusive relationships, financial difficulties, and lack of family support regarding gender identity. In turn, these subcategories were evaluated for the recurrence in which they appeared in the forms. Since the information reflected a follow-up history of up to nine months, the transformations experienced by the participants throughout the process could be observed.

Statistical analysis

For quantitative variables, the absolute and relative frequency of the selected categories were estimated and the mean and standard deviation were used as measures of central tendency and dispersion for continuous variables, respectively.

To analyze the association between the selected variables and the outcome, bivariate and multivariate logistic regression models were used. Analyses were performed in Stata 13 (https://www.stata.com). The construction of the final model included variables based on the pre-existing literature, regardless of statistical adjustment criteria, variables with significant association in the bivariate analysis (p < 0.05), and the variable “name change on personal documents,” evidenced by qualitative analysis ³⁴34. Weng H-Y, Hsueh Y-H, Messam LLM, Hertz-Picciotto I. Methods of covariate selection: directed acyclic graphs and the change-in-estimate procedure. Am J Epidemiol 2009; 169:1182-90.^,3535. Miles MB, Huberman MA, Saldaña J. Qualitative data analysis: a methods sourcebook. Los Angeles: SAGE; 2020..

For bivariate regression models, odds ratios (OR) were estimated for each category of predictor variables, while multivariate models considered adjusted odds ratios (aOR), 95% confidence intervals (95%CI), and p-value. The following categories of predictor variables were aggregated: schooling level (< 12 and ≥ 12 years of complete education); age (< 35 and ≥ 35 years old); and race/skin color (“other” referring to Asian and indigenous women). For statistical decision-making, in the final model, variables with p < 0.05 were considered significant.

Ethical aspects

The TransAmigas project was approved by the Research Ethics Committee of the University of California, San Francisco (USSF), United States (16-19922), on October 4, 2016, and the Brazilian National Research Ethics Commission (CAAE: 61338116.0.0000.5375), on July 13, 2017. Potential participants were previously interviewed to confirm their eligibility and informed about the planned procedures before signing the informed consent form, thus, they had the right to withdraw from the study at any time, without any losses.

Results

The mean age of participants was 32.9 years old (standard deviation - SD = 10.2); 44.3% were aged 25-34 years. Most participants (59.3%) were black or mixed-race. In total, 84.1% had disclosed their serostatus to someone other than health professionals; 63.9% reported moderate or severe mental health problems; 66.4% had used illicit substances in the last six months; 39.8% had already attempted suicide; 45.1% had experienced sexual violence; and 92% had experienced verbal violence for being TrTGW. Regarding gender affirmation, 88.5% agreed or strongly agreed that their appearance corresponded to their gender identity (Table 1).

Of the 113 participants, 83.2% (94/113) were located for the scheduling of the interview after nine months. Of these, 90.4% (85/94) agreed, but did not attend, even after multiple rescheduling attempts. Finally, 92.9% (79/85) actually participated in the interview, corresponding to 69.9% retention to the study. Of the 79 respondents, 72% (54/75) were from the intervention group and 66% (25/38) from the control group. About half of the interviews were face-to-face and the rest were via the Internet or by telephone.

In the bivariate logistic regression model, retention at nine months was associated with: “response to the check telephone call at three months” (OR = 7.78; 95%CI: 2.96-20.44); “higher schooling level” (≥ 12 years of complete education) (OR = 3.53; 95%CI: 1.31-9.47); and “illicit drug use in the last six months” (OR = 0.4; 95%CI: 0.16-1.03) (Table 2).

In the final multivariate model, “contact at three months” (aOR = 6.15; 95%CI: 2.16-17.51) and “higher schooling level” (≥ 12 years) (aOR = 3.26; 95%CI: 1.02-10.42) remained associated with the outcome, adjusted by race/skin color, age under 35 years, and HIV serostatus disclosure. The pseudo R² value was 0.1880 (Table 3).

Discussion

Considering the challenges of recruiting and retaining TrTGW living with HIV, which is a highly vulnerable population, the retention indicator in the TransAmigas project reached the expected value, according to what was hypothesized during its preparation. Several intervention studies have been using these indicators to assess factors that may interfere in the acceptability and feasibility analyses, especially in stigmatized populations ³⁶36. Siskind RL, Andrasik M, Karuna ST, Broder GB, Collins C, Liu A, et al. Engaging transgender people in NIH-Funded HIV/AIDS clinical trials research. J Acquir Immune Defic Syndr 2016; 72 Suppl 3:S243-7.. Studies conducted in sub-Saharan Africa ³⁷37. Wahome EW, Graham SM, Thiong’o AN, Mohamed K, Oduor T, Gichuru E, et al. Risk factors for loss to follow-up among at-risk HIV negative men who have sex with men participating in a research cohort with access to pre-exposure prophylaxis in coastal Kenya. J Int AIDS Soc 2020; 23 Suppl 6:e25593.^,3838. Sandfort TGM, Hamilton EL, Marais A, Guo X, Sugarman J, Chen YQ, et al. The feasibility of recruiting and retaining men who have sex with men and transgender women in a multinational prospective HIV prevention research cohort study in sub-Saharan Africa (HPTN 075). J Int AIDS Soc 2020; 23 Suppl 6:e25600.^,3939. Kayode BO, Mitchell A, Ndembi N, Kokogho A, Ramadhani HO, Adebajo S, et al. Retention of a cohort of men who have sex with men and transgender women at risk for and living with HIV in Abuja and Lagos, Nigeria: a longitudinal analysis. J Int AIDS Soc 2020; 23 Suppl 6:e25592.^,4040. Crowell TA, Fast PE, Bekker LG, Sanders EJ. Involvement of African men and transgender women who have sex with men in HIV research: progress, but much more must be done. J Int AIDS Soc 2020; 23 Suppl 6:e25596., the United States ⁴¹41. Kuhns LM, Vazquez R, Ramirez-Valles J. Researching special populations: retention of Latino gay and bisexual men and transgender persons in longitudinal health research. Health Educ Res 2008; 23:814-25.^, and Brazil ⁴²42. Grinsztejn B, Hoagland B, Moreira RI, Kallas EG, Madruga JV, Goulart S, et al. Retention, engagement, and adherence to pre-exposure prophylaxis for men who have sex with men and transgender women in PrEP Brasil: 48 week results of a demonstration study. Lancet HIV 2018; 5:e136-45. with men who have sex with men (MSM) and TrTGW presented great variability (50-93%) in these indicators, depending on the design, type, and characteristic of the population, location, and context. As a result, estimation methods vary from methods that consider intermediate visits to approaches with non-binary categorical outcomes.

In this study, we found a statistically significant association between retention at nine months and schooling level greater than or equal to 12 years and response to the check telephone call at three months. In studies with diverse populations, schooling is consistently associated with retention ⁴¹41. Kuhns LM, Vazquez R, Ramirez-Valles J. Researching special populations: retention of Latino gay and bisexual men and transgender persons in longitudinal health research. Health Educ Res 2008; 23:814-25., suggesting the existence of structural factors that do not exclusively affect TrTGW. In Brazil, basic education, offered free of charge by the government, comprises 12 years of complete education (primary and secondary). However, this factor should be interpreted in the light of other structural determinants. Especially in the case of TrTGW, low access and permanence are influenced by the effects of biographical background, family support, experiences of transphobia, and the need for early work ⁴³43. Reisner S, Keatley J, Baral S. Transgender community voices: a participatory population perspective. Lancet 2016; 388:327-30.^,4444. Gonzalez Júnior JLG. Experiências escolares de transexuais e travestis participantes do Projeto Muriel em São Paulo: um estudo qualitativo [Tese de Doutorado]. São Paulo: Faculdade de Ciências Médicas da Santa Casa de São Paulo; 2019.. Therefore, schooling is an independent predictor of retention for a better understanding of procedures, greater organization for return, and a social position that allows the expenditure of time and displacement that retention requires. At the same time, it is among the factors that explain greater retention in studies with TrTGW and MSM and characterizes the perception of individuals of greater safety and openness provided by the research environment to talk about their own sexuality ³⁷37. Wahome EW, Graham SM, Thiong’o AN, Mohamed K, Oduor T, Gichuru E, et al. Risk factors for loss to follow-up among at-risk HIV negative men who have sex with men participating in a research cohort with access to pre-exposure prophylaxis in coastal Kenya. J Int AIDS Soc 2020; 23 Suppl 6:e25593.^,3838. Sandfort TGM, Hamilton EL, Marais A, Guo X, Sugarman J, Chen YQ, et al. The feasibility of recruiting and retaining men who have sex with men and transgender women in a multinational prospective HIV prevention research cohort study in sub-Saharan Africa (HPTN 075). J Int AIDS Soc 2020; 23 Suppl 6:e25600.^,4242. Grinsztejn B, Hoagland B, Moreira RI, Kallas EG, Madruga JV, Goulart S, et al. Retention, engagement, and adherence to pre-exposure prophylaxis for men who have sex with men and transgender women in PrEP Brasil: 48 week results of a demonstration study. Lancet HIV 2018; 5:e136-45..

The association between the outcome and response to the check telephone call at three months shows that this strategy increases retention in longitudinal studies, considering the greater contact with the research team, regardless of the intervention. On the other hand, it may suggest a confluence of other factors that influence the response to the telephone call, such as housing stability, fixed income, participation in the labor market, access to a personal cell phone, and property security, even if these conditions have not been associated with the outcome. Some studies point having a cell phone and using social media, as well as the existence of safe spaces and researchers properly trained to establish bonds, important factors for greater retention of hard-to-reach populations ³⁶36. Siskind RL, Andrasik M, Karuna ST, Broder GB, Collins C, Liu A, et al. Engaging transgender people in NIH-Funded HIV/AIDS clinical trials research. J Acquir Immune Defic Syndr 2016; 72 Suppl 3:S243-7.^,3939. Kayode BO, Mitchell A, Ndembi N, Kokogho A, Ramadhani HO, Adebajo S, et al. Retention of a cohort of men who have sex with men and transgender women at risk for and living with HIV in Abuja and Lagos, Nigeria: a longitudinal analysis. J Int AIDS Soc 2020; 23 Suppl 6:e25592.^,4545. Kunzweiler C, Bailey RC, Okall DO, Graham SM, Mehta SD, Otieno-Nyunya B, et al. Enrolment characteristics associated with retention among HIV negative Kenyan gay, bisexual and other men who have sex with men enrolled in the Anza Mapema cohort study. J Int AIDS Soc 2020; 23 Suppl 6:e25598.. A study on pre-exposure prophylaxis (PrEP) among MSM and TrTGW, with multiple visits, observed that early contact was a significant predictor of long-term returns ⁴²42. Grinsztejn B, Hoagland B, Moreira RI, Kallas EG, Madruga JV, Goulart S, et al. Retention, engagement, and adherence to pre-exposure prophylaxis for men who have sex with men and transgender women in PrEP Brasil: 48 week results of a demonstration study. Lancet HIV 2018; 5:e136-45.. This finding may be analogous to the effect of early telephone contact in the TransAmigas project.

We found no association between the outcome (“retention to the study at nine months”) and the variables study type, race/skin color, age, and name change on personal documents. The non-association between being part of the intervention group and greater retention suggests the need to treat retention to the project and retention to the intervention differently. Although the calculation of retention to the intervention and the evaluation of associated factors are outside of the scope of this study, the different contexts in which the intervention took place and the research procedures support this distinction. While the intervention was conducted by TrTGW living with HIV, the recruitment, enrollment, and interviews were performed in a health service by a team of cisgender interviewers with higher education. Moreover, only visits that occurred within the research period (baseline and at nine months) were reimbursed. On the other hand, peer navigation requires creating bonds and, therefore, more frequent contacts. They also allow the sharing of experiences potentially unique to this population.

The variables age and race/skin color, in turn, had no significant effect in the TransAmigas project, although they were classically associated with higher retention in longitudinal studies, possibly due to the sample size. They were also not significantly associated with the outcome “name changed on personal documents”, which is a right legally guaranteed in Brazil. Discrimination and disrespect to the changed/preferred name exists at all levels of health care ¹⁷17. Sevelius J, Murray LR, Fernandes NM, Veras MA, Grinsztejn B, Lippman S. Optimising HIV programming for transgender women in Brazil. Cult Health Sex 2019; 21:543-58., which makes them barriers for this group to access services and a predictor of worse mental health ⁴⁶46. Scheim AI, Perez-Brumer AG, Bauer GR. Gender-concordant identity documents and mental health among transgender adults in the USA: a cross-sectional study. Lancet Public Health 2020; 5:e196-203.. The welcoming environment and the research team, which was sensitive and capable of receiving the participants, probably avoided transphobic practices.

We performed a qualitative analysis in order to evaluate the effect of stigma on the retention of TrTGW. This process is expressed, mainly, when social expectations based on heteronormativity conflict with gender identity and expression, leading to situations of discrimination and violence ⁴⁷47. Magno L, Dourado I, Silva LAV. Estigma e resistência entre travestis e mulheres transexuais em Salvador, Bahia, Brasil. Cad Saúde Pública 2018; 34:e00135917.. Moreover, other factors, such as social class and race/skin color, act together and exacerbate each other, reinforcing social inequalities and health disparities, with consequences even for access to health services ⁴⁸48. Lacombe-Duncan A. An intersectional perspective on access to HIV-related healthcare for transgender women. Transgend Health 2016; 1:137-41..

Although the qualitative approach brings challenges for the operationalization of phenomena such as gender, race/skin color, and class, which are incorporated in the analysis, it allows a more comprehensive understanding and better reading and interpretation of the research object, expanding the capacity to act and transform reality ⁴⁹49. Mello L, Gonçalves E. Diferença e interseccionalidade: notas para pensar práticas em saúde. Revista Cronos 2010; 11:163-73..

The research team sought strategies to understand the reasons why some participants did not regularly engage in the intervention. In these cases, they sought ways to intervene, considering the particular characteristics and demands of the participants, even offering the possibility of changing their PN if necessary.

Limitations

The choice of convenience sampling instead of methods traditionally used in studies with hard-to-reach populations, such as respondent-driven sampling (RDS) or snowball, in which pre-selected participants recruit others in order to constitute the desired sample, was a limitation of the study ⁵⁰50. Heckatorn D. Respondent-driven sampling: a new approach to the study of hidden populations. Social Problems 1997; 44:174-99.^,5151. Volz E, Heckathorn D. Probability based estimation theory for respondent-driven sampling. Journal of Official Statistics 2008; 24:79-97.^,5252. Goodman L. Snowball sampling. Annals of Mathematical Statistics 1961; 32:148-70.. Since TransAmigas was a pilot project, aimed at assessing the feasibility and acceptability of peer navigation, and considering the limitation of resources and time, which resulted in a small sample, recruiting participants in the same health service would ensure the elimination of potential biases related to the organization and quality of care. Moreover, individuals with a recent HIV diagnosis who have not started or abandoned treatment are difficult to include in samples based on the participants’ own social relationship networks, such as RDS, or recruitment in spaces of sociability, such as time-location sampling (TLS). On the other hand, this option may have influenced the final sample, since a larger number of participants already linked to health services were recruited and they were possibly more likely to adhere to the treatment and procedures of the study.

Conclusions

Identifying the factors associated with retention is essential to achieve more promising results. In the TransAmigas projects, besides the satisfactory results obtained, maintaining contact with the research team and having higher schooling level were associated with retention. Moreover, evidence shows that drug use may also play a role.

The strategies implemented, such as collective activities in public spaces, promoted experiences that increased the bonds among the team, the participants, and the PNs. Thus, these strategies should be included in the guidelines that contribute to strengthen retention.

Future studies with TrTGW should include contact at regular intervals, with additional efforts aimed at participants with lower schooling level. To measure and reduce potential losses, using monitoring tools that provide data in a timely manner is also important, helping adapt strategies of retention to the study. In line with the current literature, in the light of the gender affirmative model ³⁰30. Sevelius J, Chakravarty D, Neilands TB, Keatley J, Shade SB, Johnson MO, et al. Evidence for the Model of Gender Affirmation: the role of gender affirmation and healthcare empowerment in viral suppression among transgender women of color living with HIV. AIDS Behav 2021; 25 Suppl 1:64-71.^,3131. Reisner SL, Poteat T, Keatley J, Cabral M, Mothopeng T, Dunham E, et al. Global health burden and needs of transgender populations: a review. Lancet 2016; 388:412-36., the multidimensionality of phenomena that can be harmful to health should be considered, seeking the support of existing intersectoral policies and encouraging the development of new policies, if necessary. Finally, showing the results of the studies to the populations involved in less formal environments, using accessible language, is important to engage and build community leaders and organizations that become part of the construction of solutions.

Acknowledgments

We thank Prof. Dr. Mark Guimarães (Federal University of Minas Gerais, School of Medicine, Department of Preventive and Social Medicine) for teaching the discipline “Special Seminars in Epidemiology: Data Analysis and Interpretation” in the Graduate Program in Collective Health, Faculty of Medical Sciences of the Santa Casa of São Paulo and guide the initial conception of the analysis of this article. To the National Institutes of Health (NIH; project n. R34, process MH112177/MH/NIMH) for funding.

References

Publication Dates

History