HIV prevention and HIV care among transgender and gender diverse youth: design and implementation of a multisite mixed-methods study protocol in the U.S.

Engagement of TGD youth across the HIV prevention and care continua was assessed using a multiphasic mixed-methods study design. Both quantitative survey data and in-depth qualitative interview data were collected from TGD youth as well as from healthcare providers who had experience working with TGD youth. The mixed-methods design was utilized to gain greater insight into the research questions and to allow for data triangulation, lessening the limitations and biases inherent in one single research methodology [26, 27]. Qualitative and quantitative data on TGD youth were collected concurrently, reflecting a convergent parallel design, which permitted the collection of different but complementary data on the same topic in order to gain greater insight into a research problem than would be gained by using only one of the two methods in isolation [28]. The multiphasic study design allowed for an initial phase where secondary data analyses from a previous qualitative study with young transgender women (ATN 039 [29];) and a previous quantitative study on the HIV continuum among adolescents (ATN 086/106 [30];) were conducted and informed the development of the quantitative and qualitative measures for data collection. The qualitative portion of the study used a phenomenological investigative approach in order to help understand the sociocultural behaviors, language, roles, and interactions within a culture-sharing group (i.e., TGD youth). Phenomenology is specifically focused on describing what a given group of people have in common as they experience a particular phenomenon, and it is an inductive analytic approach that allows the patterns, themes, and categories of analysis to emerge from the data [27, 31].

The study was conducted in line with the community-based participatory research principles of cooperative and participatory engagement from stakeholders, collaborative participation, representation from community members, and dissemination of findings to stake-holders [32‐34]. A Youth Advisory Board (YAB) was convened during initial protocol development with members at three geographically diverse sites: Detroit, Los Angeles, and Boston (3–4 youth per site). Advisory board members were ages 19–26 and were diverse in terms of racial, ethnic, and gender identities (e.g., genderqueer, trans, trans masculine, trans woman), and included youth living with HIV and youth not living with HIV. The advisory board was active throughout the duration of the study, providing insight and feedback on aspects of study design, measures and interview guide development, recruitment, and data interpretation. The advisory board participated in professional development and capacity-building trainings, and received payments for all study-related time and activities. The study also utilized the broader ATN youth advisory board to aid in the interpretation of the qualitative data and assure a culturally sensitive approach to the study of TGD youth. The employment of community feedback is especially beneficial when conducting research with understudied groups such as TGD youth as it ensures that the views of the culture-sharing group are integrated into the design, execution, and interpretation of the findings, as opposed to sole reliance on the researchers’ interpretations of the culture-sharing group’s experiences [26].

A sample of TGD youth and healthcare and social service providers serving TGD youth were recruited for the study from across the 14 U.S. sites: Tampa, FL; Los Angeles, CA; Washington, DC; Philadelphia, PA; Chicago, IL; New York, NY; New Orleans, LA; Miami, FL; Memphis, TN; Houston, TX; Detroit, MI; Baltimore, MD; Boston, MA; and Denver, CO.

Inclusion criteria for healthcare and social service providers interviewed included: 1) work experience as a medical provider, mental health professional, case manager, HIV test counselor, or health educator/outreach worker; 2) provides services at one of the 14 enrolling study sites/cities; and 3) works directly with or has formerly worked directly with TGD youth.

Affirming Voices for Action was conducted through the ATN, which consisted of a network of researchers and 14 adolescent clinical sites located across the U.S. The study was centrally coordinated by the core research team located at the University of Michigan with assistance from the ATN Coordinating Center, and staff at each of the 14 study sites. The University of Michigan research team provided extensive training on qualitative interviewing to staff at each site, as well as training on cultural humility practices in working with TGD youth, use of the study screening tools, and implementation of the computer-based survey. Each site was in charge of local recruitment efforts, with the goal of enrolling 5–10 TGD youth currently in HIV care, 5–10 TGD youth not in HIV care, and 3–8 healthcare providers who worked with TGD youth. The involvement of all 14 sites in the study allowed for an adequate sample of TGD youth, a hard-to-reach population, without overburdening any particular site and allowing for a more geographically-diverse sample including regions of the country not often included in research on TGD populations.

Staff on the core research team included a Transgender Community Specialist, who identifies as a woman of trans experience and is a local and national advocate for transgender communities as well as people living with HIV. The Transgender Community Specialist assisted the sites with youth recruitment by identifying clinics and organizations that had programs or services specific to TGD youth in each city. These clinics and organizations were identified through her existing network of advocates across the U.S., conversations with members of advisory boards, additional colleagues from the TGD community, and extensive online research followed by phone and/or email follow up with identified organizations. In instances where study sites were not already connected with the identified transgender organizations in their area, the Transgender Community Specialist served as a liaison to provide information about the study and gain the organizations’ buy-in, then connecting them to staff at the local study site. TGD youth are a vulnerable population who often have interactions that are stigmatizing and not culturally sensitive; understandably, clinics and organizations serving TGD youth are sometimes hesitant to permit study recruitment of their clients. Having a member of our staff who was transgender-identified and who had extensive connections with transgender communities across the U.S. helped these clinics and organizations feel comfortable promoting the study to their TGD youth clients. The involvement of an advisory board of TGD-identified young people who assisted with the development of and approved final versions of the study survey and interview guides also helped local TGD-serving organizations feel comfortable working to promote the study.

In addition to facilitating new connections with TGD-serving organizations in each of the 14 cities, each site recruited potential youth participants through their patient population, existing community partners, existing youth community advisory boards, local HIV prevention coalitions, and outreach at local community events. Members of the study Youth Advisory Board and existing advisory boards at each site were asked to promote the study to TGD youth in their communities, and sites received Institutional Review Board (IRB) approval to provide a $10 recruitment incentive for each young person they referred. When permitted by their local IRB, sites also promoted the study via social media. The coordinating site created developmentally and culturally appropriate promotional materials including print and digital flyers and social media advertisements. Materials featured images of young people that were diverse in terms of gender expression and race/ethnicity with a youth-friendly design (e.g., vibrant colors and updated font style) and a study logo designed by YAB members, tailored with local contact information for each study site. The promotional materials were designed with advisory board input and feedback. These materials were disseminated to each site and local TGD-serving organizations. Promotional materials were also shared with the networks of the research team, the TAG, and national transgender organizations to increase awareness of the study.

To recruit healthcare providers who work with TGD youth, site staff provided information about the study to healthcare providers in their clinic who worked with TGD youth as well as other healthcare providers in their city who were known to provide services to TGD youth. The core research team and TAG provided site staff with additional suggestions of healthcare providers to reach out to in their city. The contact information of healthcare providers who met eligibility criteria and expressed an interest in participating in the study was provided to the Project Director.

The study was approved by the IRB of each study site, as well as the University of Michigan. All but one site received permission from their IRB to waive parental consent from participants ages 16 or 17; the site that was not able to waive parental permission did not enroll participants under age 18. Youth participants completed all study activities at their local site or at an alternate community-based site with study staff. Youth who were approached by study staff or who contacted staff about the study were asked if they were interested in learning about the study; if interested, they were informed of the nature of the study, the information to be collected, and the assessments involved. Those who were interested in participating were asked to give verbal consent to undergo a brief screening to determine eligibility before enrolling into the study (written consent for the screening procedure was waived by all involved IRBs).

Written informed consent was obtained from all youth participants who screened eligible to participate. TGD youth completed the audio computer-assisted quantitative survey first, which took approximately 45 min, before participating in the in-depth qualitative interview, which lasted 1.5 h on average. All TGD youth participants completed both the quantitative and qualitative portions of the study. A short debriefing interview was conducted with each youth participant at the end of the visit to assess participant distress and, if necessary, connect them to any needed resources. TGD youth participants were provided transportation to and from the study site as well as incentives for their participation; the amount varied by study site according to their usual levels of compensation for research participation. Transportation costs for youth participants were also covered.

The Project Director or Study Coordinator at each site contacted interested healthcare provider participants by phone and screened them for eligibility. Written informed consent was waived for provider participants; verbal consent was obtained all providers who screened eligible to participate. Provider participants were interviewed by phone by one of two University of Michigan staff members; they then answered a brief demographic questionnaire. Qualitative interviews with healthcare providers lasted 45 min on average. Debriefing questions at the end of each interview assessed participants for potential distress related to the interview and, if desired, healthcare providers were provided resources related to providing services to TGD youth. Provider participants were compensated for their time with a $25 gift card.

The web-based quantitative survey for youth participants was hosted and stored on a secure University of Michigan server. The data were encrypted and included no identifiable information beyond a unique identifier to link the quantitative data to the qualitative data from the same participant. The data for completed surveys was downloaded by the Project Director each week and stored in a restricted folder in the University of Michigan School of Public Health file server (with firewall protection). Once downloaded, the data were expunged from the web server. Quantitative data were reviewed weekly to ensure that the survey software was working properly.

All interviews were recorded and transcribed. Audio recorded interviews were transmitted to the core research team via the secure server of Westat, the institution that handled study data management, where they were downloaded and stored in a restricted folder in the University of Michigan School of Public Health file server (with firewall protection). Throughout the collection of qualitative data, portions of each interview were listened to and checked for quality and consistency; additional supervision and feedback was provided to interviewers as needed. Once transcriptions were thoroughly reviewed for accuracy, the recordings were destroyed.

The quantitative survey for TGD youth elicited information regarding the facilitators and barriers to engagement with each stage of the HIV care continua (i.e., prevention, diagnosis, linkage to care, engagement in care, retention in care, initiation of ART, and viral suppression). The quantitative survey also included several new measures developed for this study to assess gender-specific constructs related to TGD youth’s needs and experiences in healthcare settings. The quantitative survey was reviewed by Protocol Team experts in survey design with transgender and adolescent populations, and piloted with our Youth Advisory Board, to ensure the feasibility and acceptability of the instrument. When possible, we used measures that had been developed for and validated with adolescent populations. The constructs assessed on the quantitative youth survey are shown in Table 1; the quantitative survey in its entirety is available as supplemental material (see Additional file 1: ‘ATN 130 Quantitative Survey’).

The semi-structured qualitative interview guide for TGD youth was developed in collaboration with the Youth Advisory Board and Protocol Team using an iterative process of community feedback and revisions, as well as a secondary analysis of qualitative data from ATN 039 (see Table 2 for TGD Youth interview guide sections and example questions). Grounded in phenomenological and constructivist frameworks, the interview guide first asked participants to discuss their general life experiences as a transgender or gender diverse individual (using their own definitions, identity labels, and conceptualizations). Participants were prompted to explore their perceptions of the positive aspects of their gender identity that may contribute to resilience, followed by a discussion of their successes and challenges when seeking out and obtaining social, legal, and medical forms of gender affirmation. Participants were then guided through an in-depth exploration of their experiences engaging in the full continua of HIV prevention and care using a socio-ecological framework, (i.e., exploration of experiences at the intrapersonal, interpersonal, institutional/community, and socio-cultural/policy levels). A visual aid (see Fig. 1) was developed by the research team with feedback from the Youth Advisory Board to use in the qualitative interviews with TGD youth to describe the socioecological levels of relevance.

The semi-structured qualitative interview guide for providers was also developed in collaboration with the Youth Advisory Board and Protocol Team using an iterative process of community feedback and revisions. See Table 3 for Healthcare Provider Interview Guide sections and example questions. The guide, grounded in phenomenological and constructivist frameworks, began by eliciting information regarding participants’ experiences working with TGD youth. The interviewer then probed participants for their understanding of TGD youth’s barriers and facilitators to accessing each stage of the HIV prevention and care continua across four socioecological levels (i.e., intrapersonal, interpersonal, institutional/community, and socio-cultural/policy). The interview concluded by asking participants to provide recommendations on programs that are needed for TGD youth as well as recommendations for other providers on how to better serve TGD youth.

Primary quantitative analyses will focus on descriptively characterizing the study sample of TGD youth in order to enhance qualitative data. This will include descriptively examining experiences of engagement in the HIV prevention and care continua in the sample by HIV serostatus. The Protocol Team will examine whether any subgroup differences exist by gender identity (TF vs. TM spectrum). Demographic, psychosocial, and other self-reported health-related indicators will be examined alongside the HIV prevention and care continua with an eye toward intervention development and delivery. Next, the Protocol Team will examine the specific barriers, facilitators, and correlates TGD youth face by HIV serostatus in relation to the HIV prevention and care continua. This will include elucidating factors unique to TGD youth in accordance with Gender Affirmation and Gender Minority Stress Frameworks—i.e., factors such as hormone use, gender-related developmental milestones (age first became aware of their trans identity, age of first hormone use), surgeries and body modification (pumping, risks with needle-sharing), transgender minority-related stigma and discrimination within healthcare and in multiple other domains. Differences will also be examined by age given developmental differences in healthcare and service engagement, and by race/ethnicity given racial/ethnic disparities in HIV and the HIV care continuum. For example, we will explore whether experiences of stigma in healthcare differ according to HIV status.

All quantitative analyses will be generated with a statistical significance at the alpha 0.05 level using SAS software, Version 9.4.1 of the SAS System for Microsoft Windows. Copyright© 2013 SAS Institute Inc. SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc., Cary, NC, USA. Descriptive statistics (mean, standard deviation, frequencies, and proportions) will be estimated for all variables of interest. Distributions of individual items will be assessed, including missingness. The sample will also be stratified according to HIV status (positive vs. negative or unknown). Descriptive statistics on engagement in the HIV prevention continua will be examined for participants with an HIV negative or unknown status, while descriptive statistics for engagement in the HIV care continuum will be examined for HIV positive participants. Cross-tabs by gender identity category will be used to descriptively examine the distribution of variables across the HIV prevention and care continua. For outcomes that are independent of HIV status (e.g., stigma in healthcare, sexual risk behavior), cross tabs by HIV status (HIV positive vs. HIV negative or unknown status) will be explored. Bivariate and multivariable regression analyses will appropriately adjust for clustering due to recruitment across multiple geographic locations and study sites. Analyses will account for both natal sex (i.e., assigned sex at birth) and current gender identity.

Since this study seeks to learn about the range of facilitators and barriers that impact TGD youth’s engagement across the HIV care continuum, the primary qualitative analysis will utilize a psychological phenomenological framework [26, 27]. The research team will also employ aspects of deductive analysis that will involve the use of a priori codes, which represent critical constructs in the Socioecological model, Gender Affirmation Framework, and Gender Minority Stress Framework.

Data coding and analysis will be an iterative and interactive processes directed by the Protocol Chair and Project Director with participation from the broader research team. The qualitative data analysis team will hold regular meetings to review the developing set of themes. The team will first read all interview transcripts in order to increase familiarity with the data. Next, the team will assign a priori codes and create emergent codes. Through regular meetings, the qualitative data analysis team will develop a coding structure (i.e., a hierarchical set of constructs that account for the phenomena seen in the data). Once a codebook has been assembled, team members will apply to codes to transcripts via Dedoose software [55]. Team members will meet throughout the coding process to ensure the consistent application of codes.

Once the transcripts have been coded, team members will meet to identify consistent patterns in meaning, concepts, and themes across all interviews [31]. Data matrices will then be created as visual representations of the findings. Comparative analyses will be conducted to clarify differences that may exist for any subgroups of youth (e.g., by gender identity) in order to determine if and how recommendations and resources will need to be tailored to address sub-group specific topics. In addition, the research team will examine potential differences based on type of informant (i.e., youth vs. provider) to see if there are unique understandings of the lives of TGD youth. These comparative analyses will permit the qualitative data analysis team to specify resulting recommendations and resources that incorporate differences between groups as well as similarities among groups. Youth Advisory Board and community members who are not members of the Protocol Team will also be asked to review and provide feedback regarding the credibility of findings.

The study’s mixed-methods, convergent parallel design will allow for the triangulation of both qualitative and quantitative data, bringing together the strengths of these types of data by providing different but complementary perspectives on the same topic. The Protocol Team will use several approaches to integrate the qualitative and quantitative data. One method that will be used is data triangulation, whereby findings from one method will be validated or corroborated by using other methods, and the goal is to achieve a more accurate representation of a social phenomenon through this convergence [56]. This convergence may not always occur (such as in the case of conflicting or contradictory data), especially when working with understudied and oppressed/marginalized populations for whom quantitative methods may not be sufficient to capture a particular social phenomenon due to the lack of valid and reliable instruments. If confronted with a lack of data convergence, the research team will assess the underlying assumptions about the population and the types of data collected—a reflective process that will be beneficial in studying TGD youth given the multiple layers of oppression often experienced by these young people and the lack of empirical data documenting their lived experiences. Another method that will be used to address data divergence is to integrate multiple forms of data using data merging [57]. In cases where the qualitative and quantitative findings are divergent, the Protocol Team will also explore the possibility of using a complementary approach to data integration, whereby the goal is to attempt to connect and integrate parts, segments, or layers of a social phenomenon to validate or corroborate each other [58]. This is based on an assumption that the different methods do not provide the same representations of social realities, so there is not a desire to determine which method is a more accurate representation of reality than the other. Another approach will be to construct a multi-dimensional explanation of a phenomenon, based on the idea that social phenomena are multi-dimensional and thus may not be best represented by a single dimension alone [58]. This would allow the Protocol Team to ask different but intersecting questions with the qualitative and quantitative data, and allow for interdisciplinary perspectives on the social phenomena to be considered.

For the purposes of this article, analyses focused on summarizing demographic and related characteristics of TGD youth and healthcare providers who participated in the AVA study protocol. Descriptive characteristics were tabulated in SAS 9.4.1.

A total of 186 TGD youth were enrolled from the 14 U.S. participating ATN sites. Five youth did not complete key survey items (e.g., assigned sex at birth or current gender identity) and were therefore not included in the final quantitative data analytic sample. Demographic and related characteristics of TGD youth in the analytic sample (n = 181) are shown in Table 4. TGD youth had a mean age of 20.69 years (SD = 2.23 years) and were diverse in terms of sex assigned at birth (76.8% male, 23.2% female), gender identity spectrum (59.7% transgender women, 15.5% transgender men, 7.7% non-binary assigned female sex at birth, 17.1% non-binary assigned male sex at birth), race (77.3% youth of color), ethnicity (28.7% Latinx), sexual orientation (72.9% sexual minority), socioeconomic status (53.6% family income under poverty level, 45.3% lifetime participation in sex work, 50% lifetime experience of homelessness), and educational attainment (48.6% current students, 38.1% had high school degree or GED; 7.2% had college degree or more). Fifty-nine providers of TGD youth were enrolled and participated in an interview during this same 5-month period. Characteristics of healthcare and social services providers are presented in Table 5. Providers had a mean age of 41.8 years (SD = 11.2 years). 64.4% were cisgender female, 16.9% cisgender male, and 18.7% TGD. 30.5% were people of color. Providers were from various healthcare and social service professions, including medical providers (42.4%), mental health professionals (30.5%), case managers/care coordinators (18.6%), health educators/outreach workers (15.3%), and HIV test counselors (6.8%).

Many factors contributed to the successful recruitment of TGD youth over a several month period. Simultaneous enrollment and implementation at 14 participating ATN sites was one factor that allowed the recruitment of this diverse sample of 186 TGD youth from July – December 2015. Sites each enrolled an average of 13.3 TGD youth participants (range = 6–21). The average total time each site spent enrolling participants was 2.9 months (range = 1.0–5.3 months), with sites enrolling an average of 5.4 TGD youth per month (range = 2.5–12.5).

Involvement of the transgender community from the very initial stages of the research was integral to the success of the study. Involvement included: TGD-identified study staff, including the Transgender Community Specialist who was a national advocate for transgender communities; a multi-site advisory board of TGD youth who provided feedback on all aspects of the study, including promotional materials and recruitment strategies; and guidance from the Youth Advisory Board and Protocol Team as well as study staff on survey and interview guide design. Recruitment was aided through the use of a multipronged recruitment approach that combined the study sites’ usual modes of recruitment with local outreach. Study sites’ usual modes of recruitment included recruiting from their patient population, existing community partners, and site-specific community advisory boards. Additionally, local outreach was conducted by the core study team to connect study sites with additional TGD-specific community organizations near each site; the study was promoted through social media channels; and study team, Youth Advisory Board, and Protocol Team networks were used to raise awareness of and enhance buy-in for the study. Sites varied in terms of their extant TGD patient population and established community connection(s) with TGD organizations; thus, these recruitment methods provided the needed assistance and resources for sites with less access to TGD youth populations.

Additionally, the present study took steps to ensure that participants would have an affirming experience participating in the study. Affirming participant experiences were achieved through the involvement of TGD youth advisers and community members, and TGD-identified staff in the development and refinement of the survey and interview guides. Additionally, study interviewers received training on cultural sensitivity practices with TGD youth and received ongoing feedback regarding language used and interactions with participants during practice interviews with mock participants and throughout the study (see description below). By creating an affirming research experience for TGD youth, participants felt more comfortable sharing information about the study with their networks after having participated, aiding in study recruitment efforts by spreading the word. The most impactful recruitment method was word of mouth from TGD youth who had participated in the research themselves and could attest to the gender-affirming research procedures and study staff.

This study demonstrated the feasibility of training and overseeing mixed-methods data collection on sensitive TGD health topics from a large cohort of interviewers across 14 sites. Increasing the number of study sites is beneficial when working with hard-to-reach populations like TGD youth because it decreases the recruitment burden on any one individual site. However, it also poses potential challenges to collecting high-quality data, particularly qualitative data, which relies heavily on the individual skills of each interviewer. Interviewers for each site were selected from existing staff, and some site staff selected had little or no experience conducting qualitative interviews or working with TGD youth. To address these challenges, the study team developed an extensive training on qualitative research methods, qualitative in-depth interviewing, and working with TGD youth populations.

The initial training was delivered via two live webinar sessions, 2 h each, which allowed for interactive activities and answering trainees’ questions. The sessions were recorded for later reference and used to train several replacement interviewers due to personnel turnover at study sites. After the initial webinar training, interviewers received additional individual training and feedback on mock interview sessions conducted. Interviewers continued practicing and receiving detailed feedback from the Project Director and/or Project Coordinator until recorded mock interviews showed that the interviewer was consistently proficient in both the principles of qualitative interviewing and the use of affirming language with TGD youth. This required a minimum of 2 and a maximum of 5 mock interviews per interviewer, with training feedback calls with after each mock interview that lasted 30–60 min. Finally, as the interviews were conducted, portions of the audio recordings of interviews were reviewed for quality assurance, and additional feedback and training was provided to interviewers as needed throughout the study via telephone or video conference. This method of providing virtual training and supervision in both group and individual settings allowed for the cost-effective collection of high-quality qualitative data across multiple study sites and can be used to collect sensitive health information with other hard-to-reach and highly stigmatized populations.

This novel study addresses critical gaps in current knowledge of factors affecting U.S. TGD youth’s experiences across the HIV prevention and care continua. This research is the first, to our knowledge, to study the experiences of TGD youth within the HIV prevention and care continua and to identify barriers and facilitators to engaging TGD youth in HIV services in the U.S. Given the high prevalence of HIV and other sexually transmitted infections among TGD youth [6‐8], and lack of access to inclusive, affirming quality healthcare reported by TGD [6, 59‐61] the study findings could prove critical for improving HIV prevention and care services for TGD youth in the U.S., including increasing retention in care and viral load suppression for at-risk youth. This community-engaged research will ultimately assist providers to become better informed on how to better provide supportive and inclusive HIV services to TGD youth across the full range of HIV services. Findings from this study can also be used to inform the development of policy-, clinic-, and individual-level interventions to support the health and well-being of TGD youth including but not limited to HIV prevention and treatment.

Much of the research on transgender populations has focused on large metropolitan areas on the East and West Coasts of the U.S. [62‐66], limiting our understanding of transgender communities from other geographic regions. Additionally, most studies with TGD populations that have focused on engagement in care have been comprised of adults over age 18, frequently with average age of the sample in the mid-30s or older, limiting our understanding of TGD youth populations’ experiences at this developmental stage [59, 67‐70]. This study demonstrates the feasibility of collecting data from a relatively large sample of transgender and gender diverse youth in a short period from diverse geographic areas using a multisite design informed by community members at each decision-point of the protocol.