This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information must be included.
HIV epidemic among young men who have sex with men (YMSM) is characterized by strong racial disparities and concerns about the availability and access to culturally appropriate HIV prevention and care service delivery. Get Connected, a Web-based intervention that employs individual- and system-level tailoring technology to reduce barriers to HIV prevention care (eg, HIV or sexually transmitted infection [STI] testing, pre-exposure prophylaxis [PrEP]), was developed for YMSM (age 15-24 years). This protocol details the design and procedures of a 2-phase project that includes mystery shopping and a randomized controlled trial (RCT) to test the efficacy of Get Connected among YMSM in Philadelphia, Atlanta, and Houston.
The objective of mystery shopping is to examine the quality of HIV test counseling and PrEP-related referrals for YMSM within local HIV or STI testing sites. The objective of the RCT is to test the efficacy of Get Connected for increasing HIV-negative or HIV-unknown YMSM’s successful uptake of HIV prevention services (eg, routine HIV or STI testing), PrEP awareness, and likelihood to start PrEP (PrEP willingness), compared with those in the control condition, over a 12-month period.
For Phase 1, we will create a master list of HIV and STI testing sites in each city. We will enroll and train 10-15 mystery shoppers per city; each testing site will be separately visited and assessed by 2 mystery shoppers. After each site visit, the mystery shoppers will complete a site evaluation to record their perceptions of various measures including lesbian, gay, bisexual, transgender, queer visibility and inclusivity, privacy and confidentiality, provider-patient interactions, and clinic environment. For Phase 2, we will enroll 480 YMSM for 12 months across the 3 iTech cities into a 2-arm prospective RCT. Participants randomized to the control condition are directed to the AIDSVu.org testing site locator. Participants randomized to the intervention condition will be granted access to a Web app with content tailored to their specific demographic characteristics (eg, age, race or ethnicity, location, and relationship status), HIV and STI risk behaviors (eg, HIV and STI testing history, substance use, communication with partners regarding status) and sociocultural context (eg, homelessness, incarceration). Study assessments will occur at enrollment and at 1, 3, 6, 9, and 12 months postenrollment.
Get Connected research activities began in September 2016 and are ongoing. To date, institutional review board (IRB) submission is complete and IRB authorization agreements are pending at several other universities.
The deployment of Get Connected through a mobile-optimized Web app seeks to optimize the intervention’s acceptability, accessibility, availability, and long-term affordability among YMSM.
ClinicalTrials.gov (NCT03132415); https://clinicaltrials.gov/ct2/show/NCT03132415 (Archived by WebCite at http://www.webcitation.org/70j4gSFbZ)
RR1-10.2196/10444
Young men who have sex with men (YMSM) now account for 72% of new HIV infections among people aged 13-24 years and 30% of all new infections among men who have sex with men (MSM) [
Successful engagement in HIV prevention for HIV-negative youth (routine HIV testing, consistent condom use, PrEP adoption) requires that YMSM overcome a series of multilevel barriers at the individual (eg, risk awareness, self-efficacy to get tested), system (eg, costs, medical mistrust, lack of culturally competent care), and structural (eg, homelessness, stigma) levels [
HIV prevention tools that are culturally and developmentally appropriate for YMSM are needed [
Researchers and practitioners have sought to encourage routine HIV or STI testing by creating Web-based tools that provide the physical location of testing centers in a given geographic area (ie, testing locators). These testing locator interventions have demonstrated a wide reach when evaluated (eg, AIDS.gov test locator had over 16,000 searches and was adopted by over 100 websites in its first year [
Building on the efficacy of the CDC’s project Connect Health Systems Intervention to link heterosexual adolescents to competent comprehensive sexual health care services [
Participants in the pilot trial were randomized to receive a full GC intervention or an attention-control condition. Data [
As a step toward filling the current gap in efficacious Web-based interventions for HIV prevention and care among YMSM, we propose to implement and test the efficacy of GC 2.0 across 3 iTech cities heavily impacted by HIV: Philadelphia, Atlanta, and Houston. This protocol describes the methods for the testing of the intervention.
The research and ethics presented in this study were approved by the IRB of the University of North Carolina at Chapel Hill (16-3183). A Certificate of Confidentiality has been obtained from the National Institute of Child Health and Human Development, and a waiver of parental consent or assent has been obtained for participants who are 15-17 years old. This study is also registered on ClinicalTrials.gov (NCT03132415).
We will enroll mystery shopping participants (10-15 mystery shoppers per city) to conduct the mystery shopper assessment in 3 iTech cities: Philadelphia, Atlanta, and Houston. This approach follows best practices suggesting that youth involvement is vital when designing relevant and appropriate HIV interventions for the target population. We will work with iTech subject recruitment venues (SRV) in each city to recruit and enroll HIV-negative YMSM (age 18-24 years) who are interested in serving as mystery shoppers. We will apply a multimodal recruitment strategy, including ads in Web-based LGBTQ listservs, flyers in HIV or AIDS community-based organizations, local coffee shops and bars, college listservs, and Web-based advertisements on social media sites such as Facebook.
Eligible mystery shoppers are participants assigned male sex at birth and who currently identify as male; must be 18-24 years old (inclusive) at the time of screening; self-report as HIV-negative, speak and read English, live in Philadelphia, Atlanta, or Houston; must be able to travel to and from HIV or STI testing sites; report same-sex attraction; and have access to the internet via a computer or mobile phone.
We will recruit and enroll 10-15 mystery shoppers per city. Each participant can visit up to 10 unique testing sites in their city, with 2 mystery shoppers visiting each testing site, separately. Testing sites will be identified in collaboration with each city’s health department and by crosschecking sites with AIDSVu.org. We will employ a stratified purposive sampling strategy to ensure age and racial or ethnic diversity across mystery shoppers. Of the 10-15 mystery shoppers per city, 5-8 will be aged 18-20 years (2-3 Black or African American, 2-3 White, and 1-2 Hispanic or Latino) and 5-8 will be aged 21-24 years (2-3 Black or African American, 2-3 White, and 1-2 Hispanic or Latino).
The mystery shoppers will receive a maximum of US $600: US $100 for attending the 1-day training session and US $50 for each testing site visit (10 maximum site visits).
Once the mystery shoppers are consented and enrolled, they will attend a 1-day training at the iTech SRV where they will learn about the fundamentals of HIV or STI transmission, the guidelines and protocols surrounding HIV or STI testing and PrEP eligibility, and how to use the Web-based site assessment survey to evaluate their site visits. State-specific guidelines and policies will also be discussed in each city. Additionally, they will receive training to strengthen their self-efficacy to feel empowered as clients. Specifically, we will conduct role-plays with scenarios and interactions that might occur during a visit. We will underscore the importance of being well-versed in their rights and procedures and provide skills on how to respond to worst-case scenarios (eg, how to turn down any unwanted procedures), were they ever to occur. The mystery shoppers will be instructed to be honest about their sexual behaviors during their visits. By avoiding creating “personas” or “scripts,” shoppers will increase the social validity of the assessment and avoid arousing suspicion due to exaggerated or unrealistic scenarios.
The study staff will create and use a secure database to manage the mystery shoppers, site assignments, and testing schedules. Upon completion of the 1-day training, study staff will assign the mystery shoppers a specific day and time for their initial testing site visit. The mystery shoppers will report to the iTech SRV before each scheduled site visit to check in with a staff member and receive their site assignment. They will be loaned a mobile phone equipped with a car share app to use for travel to and from testing sites. All car share trips will be tracked and paid for by the study, so no transportation costs will be incurred by the mystery shoppers.
Once at the testing site, the mystery shoppers will state that they have no income, health insurance, or any proof of identification. In doing so, we will be able to ascertain whether these would be potential barriers to testing at a given location and ascertain the lowest possible fees that would be charged to YMSM. As in the original study [
The site assessment survey will record shoppers’ perceptions of their testing experience, specifically LGBTQ visibility, medical form inclusivity, clinic environment, privacy and confidentiality, PrEP information and dialogue, patient-provider relationship context, patient-provider counseling, safer sex education, perceived provider competency, and participant-provider interactions (
In addition to the site assessment survey, the mystery shoppers will complete a secure video chat session with study staff to discuss their testing experience and have the opportunity to share any adverse interactions. These video chat sessions will not be recorded; their purpose will be to ensure mystery shoppers’ safety and prevent subsequent mystery shoppers from being exposed to a site reported to be risky or unsafe (physically or emotionally). Following the video chat session, the mystery shoppers will be given their incentive for the visit and their next visit with study staff will be scheduled before leaving the iTech SRV.
The mystery shoppers’ site assessment scores will be aggregated for use in Phase 2 of the research activities: RCT to test the efficacy of GC. Specifically, the scores for each site will be averaged and embedded in the intervention condition of the GC Web app: when participants search for testing sites they will only see sites that rank in the top 50% for that city, sorted from highest to lowest ranking.
The research activities involve a 2-arm 12-month prospective RCT enrolling 480 HIV-negative or status-unaware YMSM (age 15-24 years). After assent or consent and completion of a baseline survey, YMSM will be randomized on a 1:1 basis to either the control or intervention condition (intervention, n=240; control, n=240). Participants randomized to the control condition will be directed to the AIDSVU.org testing site locator. While the provision of a test locator is a low intensity intervention, we felt that withholding referrals to testing and care services would be unethical given YMSM’s vulnerability to HIV and STIs. Furthermore, given the availability of search engines to locate HIV or STI testing sites, the test locator condition may be considered usual care. Nevertheless, by providing the existing testing site locator only, we will still be able to test the effect of GC (ie, user-tailored content focused on HIV or STI testing and PrEP referral and the linkage to high-quality agencies). Web-based study assessments are conducted every 3 months across the intervention and control conditions, with a total follow-up period of 12 months. At the end of RCT, we will make the intervention accessible to YMSM in the control condition.
The GC intervention was developed by customizing content based on YMSM’s psychosocial and sexual profiles (eg, sociodemographics, HIV and STI testing history and testing motivations, recent sexual behavior, sources of support, self-reported values), as reported by participants’ answers to their baseline assessment. At the individual level, GC delivers tailored Web-based content specific to each user’s demographic characteristics (eg, age, race or ethnicity, location, relationship status), HIV and STI risk behaviors (eg, HIV and STI testing history, substance use, communication with partners regarding status), and sociocultural context (eg, homelessness, incarceration). GC also employs tailoring at the system level using mystery shopper scores. Participants across both conditions who have been tested will be asked to rate their visit at their quarterly follow-up assessment using the same mystery shopper criteria. Sites will receive biannual summaries, including the aggregated user reviews and brief technical assistance reports, to help sites understand their performance based on quality assurance evaluations from YMSM clients and to optimize service delivery, if needed.
For the participants in the intervention condition, the tailored Web app has 4 sections of content: “What,” “Why,” “How,” and “Where.” The “What” section is split into 3 pages: “Facts,” “STIs,” and “Tests.” On each of those pages, topics are displayed in boxes that are randomly organized and open to display additional information if the user clicks or presses. The Facts page (
Session speed (min)
LGBT visibility (Cronbach alpha=0.84)
The clinic has symbols aimed at lesbian, gay, bisexual, transgender (LGBT) people (eg, equal sign, rainbow flag)
The clinic has printed materials (eg, brochures) aimed at LGBT people
The clinic has LGBT welcoming symbols
Medical forms (Cronbach alpha=0.59)
The clinic uses LGBT-inclusive language on medical forms
The clinic uses transgender-inclusive language on medical forms
Clinic environment (Cronbach alpha=0.76)
The office staff were generally friendly
The office staff were judgmental (Reverse coded)
The office staff were not lesbian, gay, bisexual, transgender, questioning or queer (LGBTQ)–sensitive (Reverse coded)
I felt uncomfortable in the waiting room (Reverse coded)
The clinic used LGBT-affirming language when speaking to me
Privacy and confidentiality
The clinic staff kept patient information confidential
Interactions between clients and staff were kept private
The provider explained confidentiality (either verbally or via a document)
Pre-exposure prophylaxis (PrEP)–specific indicators
The clinic had information about PrEP
The clinic offers PrEP or PrEP referrals
Relationship context (Cronbach alpha=0.89)
The provider asked me about my sexual orientation
The provider asked me about my relationship status
The provider asked if I had experienced intimate partner violence
Counselling session (Cronbach alpha=0.76)
The provider explored my motivation for testing
The provider offered to help me set goals
The provider offered to help me set action steps to meet safer sex goals
The provider offered me risk reduction options
The provider’s recommendations were valuable
Safer sex education (Cronbach alpha=0.88)
The provider made sure I knew how to use a condom
The provider helped me identify a condom that works for me
The provider helped me identify a lube that works for me
The provider discussed PrEP as a prevention strategy with me
Perceived provider competency (Cronbach alpha=0.65)
The provider or test counsellor appeared knowledgeable about HIV and STIs
The provider appeared knowledgeable about LGBTQ health issues
Negative provider interactions (Cronbach alpha=0.89)
The provider made me feel comfortable (Reverse coded)
I felt pressured by the provider to adopt specific risk reduction options
The provider was judgmental about the kind of sex I have (eg, anal, receptive, or penetrative, etc)
The provider was judgmental about how many partners I have had
The provider was judgmental about how I met my partners
HIV and STIs (sexually transmitted infections) facts.
The second section focuses on the “Why” on 2 pages: “Values” and “Pros and Cons.” The basic design and functionality is the same as described above in the “What” section. The Values page (
The third section is about the “How” of testing and includes pages on potential “Barriers” to getting tested and “Supports” that may help a participant decide to get tested. Barriers (
The final section is the “Where” of testing and includes a page where a user can “Customize” their search for nearby testing sites (
Eligible participants will be those assigned male sex at birth who currently identify as male, aged 15-24 years (inclusive) at the time of screening, have had consensual anal sex with another man in the past 6 months, self-report as HIV-negative or unsure of their HIV status, have access to a computer or mobile phone, can read and speak English, and live within the city limits of Philadelphia, Atlanta, or Houston.
Our target enrollment across both conditions is 480 participants (intervention, n=240; control, n=240). This number allows for 20% loss to follow-up rate and a final analytic sample of 400 YMSM across the 3 cities. Participants may continue the study even if they miss assessments intermittently over the data collection period. We will compare those who completed different follow-up assessments with those who did not based on key predictors from the baseline assessment to check for possible bias due to missing data and informative censoring. When appropriate, we will use expectation-maximization algorithm-based imputation methods in our analyses [
Values page.
Barriers page.
Customization page.
In order to have 80% power to the intervention and control groups, we require at least 400 participants to find an absolute difference of 13% in cross-sectional analyses. Assuming a within-person correlation of 0.25, we can detect an 8.8% difference, indicating we have power to detect the smallest possible difference between arms. A less favorable within-person correlation of 0.75 allows us to detect an 11.3% difference. For mean differences across continuous outcomes (eg, PrEP willingness), our sample size calculations are based on a 2-sample
Participants can earn up to US $155 total: Baseline survey=US $20, month 1 survey=US $20, month 3 survey=US $25, month 6 survey=US $30, month 9 survey=US $30, and month 12 survey=US $30.
After assent or consent and completion of the baseline survey, YMSM will be randomized by city 1:1 to either the control or intervention condition (intervention, n=240; control, n=240) [
The primary outcomes relate to the successful uptake of HIV prevention services among our sample of self-reported HIV-negative or serostatus-unaware YMSM. We have considered 3 prevention outcomes: HIV testing, STI testing, and PrEP awareness and willingness.
The baseline survey will include questions on lifetime HIV testing history. Follow-up surveys will repeat the questions from the baseline and will also include questions on HIV testing in the prior 3-month period, including test results. The HIV testing outcome will be the proportion of YMSM tested for HIV 2 or more times at least 3 months apart in the 12-month follow-up period (“frequent tester”). As an additional analysis, we will also examine the proportions of participants who receive 1 HIV test.
The baseline survey will include separate questions on lifetime testing history of gonorrhea, chlamydia, and syphilis, respectively, as well as questions about ever having a genital exam, an anal pap smear, or a vaccination for Hepatitis A and B, human papilloma virus, and meningitis. Follow-up surveys will repeat the questions from the baseline but will ask about STI testing behavior in the prior 3-month period, including test results if a participant indicates they received a test. The STI testing outcome will be the proportion of YMSM tested for any STI 2 or more times, at least 3 months apart, in the 12-month follow-up period (“frequent tester”). As an additional analysis, we will also examine the proportions of participants who receive 1 STI test.
The survey will contain a brief description of PrEP to orient the participant. Most questions were adapted from recent studies of PrEP attitudes with YMSM [
As secondary outcomes, we will examine the uptake of PrEP, changes in sexual risk behavior, and the linkage and retention in care among newly diagnosed HIV-positive cases. While we expect a small number of newly diagnosed HIV infections, we will measure initiation of antiretroviral therapy and self-reported adherence as a secondary outcome. We are not powered to measure differences in engagement in HIV care across trial arms, so we include this as an exploratory analysis.
Consistent with our theoretical framework, we will assess YMSM’s psychosocial correlates predicting adoption of HIV services (ie, attitudes, norms, self-efficacy, and behavioral intentions to get HIV tested). Integrated Behavioral Model constructs will be assessed with subscales assessing YMSM’s attitudes, social norms, and behavioral intentions [
At each follow-up assessment, PrEP-eligible (per CDC guidelines), HIV-negative YMSM will be asked whether they have begun using PrEP [
Sexual risk behavior will be assessed using the Sexual Practices Assessment Schedule used in previous Web-based studies with YMSM [
Among newly diagnosed HIV-positive cases, we will measure participants’ linkage and engagement with appropriate medical care after initial diagnosis, using criteria employed in prior ATN protocols with youth [
We will also measure the following constructs as potential predictors or moderators in our analyses.
We will include questions on participants’ race or ethnicity, educational attainment, employment status, place of birth, housing status, and history of incarceration, sexual identity, and “outness” to their social network.
Across both trial arms, YMSM who report testing in the prior 3 months will complete site assessments of their testing experiences to measure comfort, quality, and concerns after visiting a site for HIV or STI testing. The site assessment form is the same form used by the mystery shoppers. We will use these assessments to send aggregate data of YMSM’s satisfaction with services to agencies biannually.
Previous studies have demonstrated higher vulnerability to HIV risk behaviors and engagement in prevention and care among YMSM who report alcohol, tobacco, and other drug (ATOD) use and psychological distress; therefore, we will measure both ATOD and psychological distress as potential effect moderators.
We will assess the frequency of ATOD use (as measured in the National Survey on Drug Use and Health) over the past 3 months in the baseline survey and follow-up surveys for alcohol, tobacco products, marijuana, nonprescription drugs, cocaine, amphetamines, inhalants, opioids (including heroin), hallucinogens, and depressants [
We will measure psychological distress using existing, well-validated scales: the Patient Health Questionnaire-8 (PHQ-8) [
At each follow-up, participants will report on the acceptability of their assigned study arm. We will use the Systems Usability Scale [
We will measure intervention exposure using paradata from the intervention, including counts of user sessions, length of sessions, pages visited, and functions utilized. This information will assist in examining whether intervention dosage influences the overall efficacy of the intervention and in informing the cost analysis and wider implementation and scalability [
We will include Pew Internet survey questions [
Descriptive statistics of the psychosocial and demographic characteristics of the participants will be used to describe all participants. These will be compared between treatment groups using
We will use the general framework of generalized linear mixed models (GLMM) to test for intervention effects over time. Note that some of our outcomes are binary, some are count, and some continuous traits and thus need to be treated differently. The general form of GLMM will be g(
Models will be compared according to the information criteria such as Akaike Information Criterion and Bayesian Information Criterion. For some binary outcomes, such as HIV testing, we will perform an aggregate analysis after collapsing across the repeated measures using simple logistic regression, comparing whether the probability of having tested at least once over the entire follow-up period is different across treatment groups, after adjusting for baseline values. To ensure robustness, we will also apply an exchangeable working correlation structure to its corresponding generalized estimating equation model. We will conduct exploratory regression analyses to examine regional differences. These regressions will be run with group assignment and region in the model, controlling for sociodemographic characteristics. Interactions between group assignment and region will be tested to explore potential site-specific moderators of treatment effect.
As a secondary analysis, we will build on our GLMM framework to examine whether the intervention effects in the theoretical mediators (eg, attitudes, norms, and self-efficacy) are associated with our outcomes. We will also test whether these relationships vary as a function of YMSM’s varying engagement with the intervention (intervention acceptability, use of intervention over time). Interactions between group assignment and these characteristics will test for potential moderators of treatment effect.
In order to inform the eventual scale-up of GC, we will also conduct a cost analysis of GC and control conditions to inform discussions of sustainability and roll out of the GC intervention. We will collect information on costs associated with the delivery of the intervention. No costs associated with research data collection will be included. These components of cost will be summed over the 12-month study period for each participant to generate an estimated per person cost. Effectiveness will be measured by examining HIV-related outcomes reported by YMSM over the 12-month period. Incremental cost effectiveness ratio (ICER) across treatment arms will be defined as delta C or delta E, where delta C denotes the estimated difference in mean cost of the intervention and delta E reflects the estimated difference in mean effectiveness between the intervention and control groups. Nonparametric bootstrap resampling will be used to estimate the 95% CI of incremental cost effectiveness ratio [
We will qualitatively assess testing sites’ satisfaction with the biannual performance assessments and their improvements in service delivery when working with YMSM across the 3 regions. Ten site directors will be randomly selected from testing sites in each city. Eligible participants will be able to read and speak English and serve as the site director of an HIV or STI testing site in Philadelphia, Atlanta, or Houston. We will conduct semistructured qualitative in-depth interviews (60-90 minutes) that focus on 4 domains: (1) existing prevention services used and promoted by the agency, (2) agency (internal) resources currently missing, that if identified and addressed, could improve the delivery of HIV, STI, and PrEP services to YMSM, (3) feedback on the biannual performance assessments and their use for service delivery improvements, and (4) the advantages and disadvantages of GC rollout within AIDS Service Organizations.
Interviews will occur via teleconference to maximize candidness and privacy while decreasing travel-related costs. We will use VSee, a simple and low-cost video chat platform that requires no server infrastructure to set up or maintain and allows providers to be HIPAA-compliant. Interviews will be audio-recorded to allow for verbatim transcription, and then checked for accuracy and completion. Initial reading and coding of the transcripts will be reviewed, compared, and refined in team meetings. This systematic process will lead to the creation of a coding structure that includes a hierarchical set of constructs seen in the data. We will analyze several transcripts jointly to establish intercoder reliability. The team will then code all transcripts using our coding structure and add inductive codes during the iterative analysis process. Throughout, we will discuss emerging themes, resolve difficulties or concerns that may arise, and adapt the codebook as necessary.
Since we seek to gain a multilevel understanding of the structural, organizational, and interpersonal barriers and facilitators of implementing GC, our analysis will utilize a phenomenological framework [
GC research activities began in September 2016 and are ongoing. Institutional review board (IRB) submission is complete, with IRB authorization agreements being finalized across the participating universities and SRVs.
There are several potential challenges and limitations to the proposed clinical trial. First, we will rely on self-reported outcomes. We will not include biological measures (eg, presence of HIV or STI), as we would have to dramatically increase our sample size to detect significant effects in biomarkers among newly diagnosed HIV or STI cases and it would be inefficient to collect biomarkers in a Web-based study. We will frame the presentation of results as self-reported outcomes. Second, we propose to recruit a diverse (in terms of race or ethnicity and age) sample of 15-24 year-old participants. It is possible that we may experience more success in recruiting older YMSM (those aged >18 years). To counteract this, we will include a broad range of social media outlets in our recruitment, allowing us the potential to recruit our full age range. Collectively, the team has a vast experience of recruiting youth into HIV research efforts and substantial experience in recruiting online samples of urban race or ethnic minority YMSM. Third, we are unable to untangle race from Latino ethnicity, as it would require a much larger sample size to examine race by ethnicity subgroup differences. Because we propose to quota sample across race or ethnicity in each of the regions, the breakdown of Latino race would create some small sample sizes. Fourth, we recognize that socioeconomically disadvantaged participants may require access to a computer or secure Wi-Fi connection to participate fully in the study. YMSM who are interested in participating but require access or who prefer to complete assessments at a study location will be able to complete intervention activities at their local iTech SRV. Finally, to minimize potential risks, all iTech SRVs have specific policies governing the treatment of human participants, including the referral to medical and psychological services in the event a participant should report a need for these services or experience any adverse reactions resulting from study procedures.
With increasingly promising evidence of the efficacy of biomedical prevention tools, such as PrEP, for reducing the risk of HIV infection among MSM [
Efforts to encourage and motivate YMSM to engage in repeat HIV or STI testing or to adopt other prevention efforts (eg, PrEP [
alcohol, tobacco, and other drug
Centers for Disease Control and Prevention
effect sizes
Get Connected
generalized linear mixed models
institutional review board
lesbian, gay, bisexual, transgender, questioning or queer
men who have sex with men
pre-exposure prophylaxis
randomized controlled trial
subject recruitment venues
sexually transmitted infection
young men who have sex with men
This work was supported by the National Institutes of Health Adolescent Medicine Trials Network for HIV/AIDS Interventions (ATN 139; MPI: Bauermeister & Stephenson) as part of and the UNC/Emory Center for Innovative Technology (iTech; PIs: Drs Hightow-Weidman/Sullivan, 1U19HD089881). The content is solely the responsibility of the authors and does not represent the official views of the funding agencies. The authors would like to thank Adi Ferrara, MS, ELS, for help in the preparation of this manuscript.
None declared.