Epic Allies, a Gamified Mobile Phone App to Improve Engagement in Care, Antiretroviral Uptake, and Adherence Among Young Men Who Have Sex With Men and Young Transgender Women Who Have Sex With Men: Protocol for a Randomized Controlled Trial

Background In the United States, young men who have sex with men (YMSM) and transgender women who have sex with men (YTWSM) bear a disproportionate burden of prevalent and incident HIV infections. Once diagnosed, many YMSM and YTWSM struggle to engage in HIV care, adhere to antiretroviral therapy (ART), and achieve viral suppression. Computer-based interventions, including those focused on behavior change, are recognized as effective tools for engaging youth. Objective The purpose of the study described in this protocol is to evaluate the efficacy of Epic Allies, a theory-based mobile phone app that utilizes game mechanics and social networking features to improve engagement in HIV care, ART uptake, ART adherence, and viral suppression among HIV-positive YMSM and YTWSM. The study also qualitatively assesses intervention acceptability, perceived impact, and sustainability. Methods This is a two-group, active-control randomized controlled trial of the Epic Allies app. YMSM and YTWSM aged 16 to 24 inclusive, with detectable HIV viral load are randomized 1:1 within strata of new to care (newly entered HIV medical care ≤12 months of baseline visit) or ART-nonadherent (first entered HIV medical care >12 months before baseline visit) to intervention or control conditions. The intervention condition addresses ART adherence barriers through medication reminders and adherence monitoring, tracking of select adherence-related behaviors (eg, alcohol and marijuana use), an interactive dashboard that displays the participant’s adherence-related behaviors and provides tailored feedback, encouragement messages from other users, daily HIV/ART educational articles, and gamification features (eg, mini-games, points, badges) to increase motivation for behavior change and app engagement. The control condition features weekly phone-based notifications to encourage participants to view educational information in the control app. Follow-up assessments are administered at 13, 26, and 39 weeks for each arm. The primary outcome measure is viral suppression. Secondary outcome measures include engagement in care, ART uptake, ART adherence, and psychosocial barriers to engagement in care and ART adherence, including psychological distress, stigma, and social support. Results Baseline enrollment began in September 2015 and was completed in September 2016 (n=146), and assessment of intervention outcomes continued through August 2017. Results for primary and secondary outcome measures are expected to be reported in ClinicalTrials.gov by April 30, 2018. Conclusions If successful, Epic Allies will represent a novel adherence intervention for a group disproportionately impacted by HIV in the United States. Adherent patients would require less frequent clinic visits and experience fewer HIV-related secondary infections, thereby reducing health care costs and HIV transmission. Epic Allies could easily be expanded and adopted for use among larger populations of YMSM and YTWSM, other HIV-positive populations, and for those diagnosed with other chronic diseases such as diabetes and hypertension. Trial Registration ClinicalTrials.gov NCT02782130; https://clinicaltrials.gov/ct2/show/NCT02782130 (Archived by Webcite at http://www.webcitation.org/6yGODyerk)


Background
Men who have sex with men (MSM) account for nearly two-thirds of all new HIV infections in the United States and young MSM (YMSM) are the only risk group experiencing an increase in HIV incidence [1][2][3].Regional studies suggest that HIV prevalence among transgender women is among the highest of all risk groups, especially among transgender women of color, and African American transgender women in particular [4][5][6].Although likely underestimated, HIV prevalence among young transgender women, including young transgender women who have sex with men (YTWSM), ranges from 4.5% to 15.9% [7].Youth diagnosed with HIV must adjust to living with a highly stigmatized health condition that requires lifelong medical management.Due to structural, developmental, and psychosocial barriers, many youth struggle to enter medical care, initiate antiretroviral therapy (ART), adhere to ART, or achieve viral suppression (VS) [8,9].For YMSM and YTWSM who may already be ostracized from families and friends because of their sexual identity, receiving an HIV diagnosis can lead to an increase in social isolation, as well as negative affective states such as depression and anxiety, which may create additional barriers to HIV treatment [10][11][12][13].Interventions for YMSM and YTWSM that increase engagement in care, ART uptake, ART adherence, and VS are needed to maximize the individual and public health benefits of treatment [14].
As of January 2017, 88% of US adults are online, 95% have a cell phone, and 77% have a smartphone [34,35].Youth (ages 18-29 years) have the highest levels of smartphone ownership at 92% [35].US lesbian, gay, bisexual, and transgender individuals under the age of 35 years have had consistently higher rates of smartphone ownership than their general population counterparts [36,37].In addition to increased smartphone ownership, the use of mobile phone apps is on the rise [38].
Serious games (games designed to accomplish a purpose, such as influencing learning, civic engagement, or health behavior change) are increasingly being used to address behavioral and psychological factors that inhibit adherence to medical treatment regimens [39][40][41].Such games are intended to be goal-oriented, immersive, challenging, and motivating [42].Games designed to improve health can influence health attitudes and improve behavior change self-efficacy [43][44][45][46].As a result, games are an ideal platform to engage youth in behavior change as they have the ability to attract and maintain attention, avoiding the development of boredom and attrition [42].The ability to add "fun" into design and game play serves to enhance overall motivation.
Social networking sites are also extremely popular among young adults.As of July 2015, 90% of black and 95% of white youth aged between 18 and 29 years use social networking sites [47].MSM and transgender women have high rates of social networking use [6], in part, because online venues often represent one of a limited number of venues for connecting with one another.Social networking has been used successfully to change behaviors, increase social support, and reduce social isolation in HIV prevention and care interventions [48,49].
Epic Allies was developed based on the information, motivation, and behavioral (IMB) skills model [50] to address the urgent need for interventions that improve engagement in care, ART uptake, and ART adherence among YMSM and YTWSM.The app was created using an iterative process with input from the target population at each stage of development to ensure acceptability, relevance, and appeal [16].We anticipate that the gaming features will enhance motivation for behavior changes related to engagement in care and ART adherence.Furthermore, social networking features will increase motivation by providing users with a sense of community and social support.Funded by the National Institutes of Health, we worked with programmers

XSL • FO
RenderX and designers at Caktus Consulting Group, LLC to develop and test the Epic Allies prototype and found it to be acceptable among a sample of HIV-positive YMSM [16].

Aims and Objectives
The aim of this paper is to describe the study protocol for the randomized controlled trial (RCT) of the Epic Allies intervention.The first objective of the study is to test the efficacy of the Epic Allies intervention among HIV-positive YMSM and YTWSM by conducting a two-arm RCT.The primary outcome measure is VS.Secondary outcomes include engagement in HIV care (ie, completion of HIV-related care clinic visit in last 3 months), ART uptake (ie, initiation of ART in the last 3 months), ART adherence (ie, >90% of doses taken in previous week), and psychosocial barriers to engagement in care and ART adherence such as psychological distress, stigma, and social support.
The second objective is to qualitatively assess intervention acceptability, perceived impact, and potential for long-term sustainability.In-depth interviews with a subset of intervention arm participants conducted after the intervention period will evaluate acceptability of Epic Allies and examine participants' perspectives on the relationship between app use and study outcomes and potential for long-term sustainability of app use.

Trial Design
This study is a two-arm parallel RCT that will test the 26-week Epic Allies intervention against a control condition that includes weekly phone-based notifications to encourage participants to view educational information in the control app (Figure 1).Approximately 200 YMSM and YTWSM will be enrolled from 5 participating sites that provide HIV medical care for youth.Participants will be randomized 1:1 to intervention or control arms that are balanced by new to care (newly entered HIV medical care within 12 months of baseline visit) or ART-nonadherent status (first entered HIV medical care more than 12 months before baseline visit).Outcomes of interest will be measured at baseline, week 13 (during intervention phase), week 26 (end of intervention phase), and week 39 (postintervention phase).In-depth qualitative app satisfaction interviews will be conducted with approximately 20 intervention arm participants at the end of intervention use at week 26 to assess intervention experiences, acceptability, perceptions of associations between app use and study outcomes, and potential for long-term sustainability of using the app to support ART adherence.

Ethics
The study protocol was approved by the institutional review boards (IRB) at the University of North Carolina at Chapel Hill and all participating study sites, including University of South Florida, Tampa, FL; Stroger Hospital of Cook County, Chicago, IL; Montefiore Medical Center, Bronx, NY; Tulane Medical Center, New Orleans, LA; and University of North Carolina Hospital, Chapel Hill, NC (also includes Regional AIDS Interfaith Network, Charlotte, NC).Individuals who express interest in the study will be required to provide signed informed consent before medical records are abstracted to confirm eligibility or study procedures are performed.The informed consent documents will describe all study procedures in detail.During the informed consent process, site study staff will go over the consent documents and answer any questions that may arise.A waiver of parental consent for individuals younger than 18 years has been obtained for all sites given that this is a minimal-risk study.

Participants
Individuals participating in this study must meet the following eligibility criteria: (1) HIV-1 infected; (2) aged from 16 to 24 years; (3) assigned male sex at birth, of any gender identity, and self-reports a desire to engage or is engaging in sex with men; (4) at least one VL collected within the 12 weeks before the baseline visit, and the VL collected closest to the baseline visit is greater than the lower limit of detection for the site-specific assay used to test the specimen; (5) reliable daily access to an Android-or iOS-based mobile phone with a data plan; and (6) able to speak and read English.Self-reported eligibility criteria will be verified through an in-person screening with site study staff.Medical eligibility criteria will be verified through medical chart abstraction by site study staff.Individuals who cannot provide consent due to active substance use or psychological condition will be considered ineligible.

Recruitment and Enrollment
Potential study participants will be identified through medical chart reviews and/or medical provider referrals at participating sites.Identified individuals will be informed of the nature of the study, the information to be collected, and the evaluations and assessments that are involved.For individuals interested in study participation, self-reported eligibility criteria will be verified.Before confirmation of medical eligibility criteria, a signed informed consent will be obtained.Individuals who provide informed consent and meet all study eligibility criteria will be enrolled in the study and complete a baseline computer-assisted self-interviewing (CASI) survey.

Randomization
Study participants will be classified as either new to care (newly entered HIV medical care within the 12 months before the baseline visit) or ART nonadherent (first entered HIV medical care more than 12 months before the baseline visit).Randomization will occur in a 1:1 ratio within each of the 2 classification strata separately, with an equal number of participants assigned to the intervention and control arms.Due to rolling enrollment, block randomization will be used to help ensure balance within strata.Status as new to care vs nonadherent (eg, eligibility group) will be chosen as the primary stratum and randomized separately in blocks of 4 participants.Study statisticians will not be blinded to study arm assignment because they will be involved in data quality control and quality assurance.

Incentives
The amount of participant compensation for study participation is determined separately by each site and approved by each site's IRB.Participants will receive US $40 to $60 for completion of each RCT assessment at baseline, week 13, week XSL • FO RenderX 26, and week 39.Intervention arm participants who log on to the Epic Allies app 12 out of the first 14 days after the baseline visit will receive US $20 as a one-time sign-on bonus and those who log on at least once in each 30-day period will receive US $10 to help defray costs associated with smartphone data usage during that period.Participants in the intervention arm who are selected for and complete the in-depth qualitative interview will receive US $20 to $50.

Intervention Theoretical Model and Features
The design of Epic Allies was informed by the IMB skills model, which conceptualizes health behavior change (eg, medication adherence) as a product of mediators, including information about the behavior, motivation to change, and the skills needed to achieve change [50].Studies testing the IMB model of ART adherence support relationships between information, motivation, and behavioral skills and medication adherence [51][52][53][54][55]. Group-and individual-level IMB-based interventions improve ART adherence [56][57][58][59], though none have been designed specifically for YMSM and YTWSM.Epic Allies' features (see Multimedia Appendix 1) address numerous elements of the IMB model (Figures 2-11).For example, the gaming features are designed to enhance sustained app use and motivate patterning new adherence behaviors [60].The social networking features allow users to give and receive support, a relevant need for many YMSM and YTWSM who experience social isolation due to HIV-related stigma and homophobia [61][62][63].

Intervention and Control Conditions
At baseline visit, participants assigned to the intervention arm will download and install Epic Allies, create a 4-digit app password, and receive a guided tour of the app by site study staff.Intervention arm participants will have full access to all features of Epic Allies during the 26-week intervention period.Participants assigned to the control group will download and install the Epic Allies control app (phone notification messages only), create a 4-digit app password, and be provided with instructions on using the app.During the 26-week trial, control participants will receive weekly phone notifications that inform users that new content is available and one brief informational article will be provided.Control group articles are a subset of Daily Dose articles focused on ART adherence and HIV disease self-management.

Data Collection Study Objective 1: Efficacy
Baseline, week 13, week 26, and week 39 assessments will be conducted in person.At each time point, participants will complete a CASI survey.Clinical data will include data collected via chart abstraction and/or laboratory values (VL only).If a participant does not have a VL value recorded in their chart in the 6-week window before the study visit, VL testing will be conducted on the day of the study visit as part of standard of care (ie, participant is scheduled for a medical care visit that includes VL testing) or by the study (ie, the study visit does not coincide with a medical care visit with VL testing).Table 1 lists primary and secondary outcomes and the source, collection points, and a description of each measure.
App usage data will be transmitted from the participant's smartphone to a secure server any time the participant is connected to the Internet via broadband or Wi-Fi.App data metrics include log-ins/log-outs, use of app features, and app progress.

Data Collection Study Objective 2: Acceptability
In-depth qualitative app satisfaction interviews will be conducted via Skype with approximately 20 intervention arm participants at the end of the intervention.We will attempt to enroll equal numbers of participants into one of four cells in Table 2 based on their ART status at study entry and app usage during the intervention.As each participant finishes the 26-week intervention period, they will be asked if they are willing to participate in the in-depth app qualitative satisfaction interview.Once a given care-usage cell has reached its quota, that cell will be "closed," and interviews will be offered to only those participants who fall within the remaining open cells.
The in-depth qualitative app satisfaction interview will last between 45 and 60 min and will be recorded with the participant's consent.Participants who prefer can opt to use video during the interview, but there will be no video recording.All interviews will be conducted by one of 3 trained qualitative interviewers from the study team using a semistructured interview guide.Following each interview, the recording will be transcribed by Verbal Ink (a division of Ubiqus, Los Angeles, California), checked for accuracy by study staff, and uploaded to the UNC-CH secure server.The transcripts of the first three interviews will be reviewed by the study team to assess for quality and content before completing the remaining interviews.Primary study outcomes <40 copies/mL or lower limit of detection for site-specific assay used to test the specimen X X Chart review (any value in 6week window before scheduled visit) OR laboratory value collected at study visit VS defined as VL below the lower limit of detection in the 6-week window before the scheduled study visits

Secondary study outcomes
<40 copies/mL or lower limit of detection for site-specific assay used to test the specimen X Chart review (any value in 6week window before scheduled visit) OR laboratory value collected at study visit VL suppression defined as VL below the lower limit of detection in the 6-week window before the scheduled study visits Completion of HIV-related care clinic visit in last 3 months X X X CASI survey Engagement in care "Are you currently taking medication to treat your HIV (Y/N)?"X X X CASI survey ART uptake (for participants not on ART) (1) "How many times during the day has your doctor told you to take a dose of medicine (pills or other medicines) to treat your HIV?" and (2) "Thinking about the last 7 days, how many times did you miss taking a dose of pills?" [64] X X X CASI survey ART adherence a (for participants on ART) a Outcome of >90% adherence is comprised 2 components (1) is the denominator, indicating the frequency of doses prescribed (multiplied x7 to represent total weekly doses); (2) is the numerator, indicating the number of times, total, a dose was missed.

Tracking Participant Follow-Up
All participants will be contacted before each follow-up study visit (ie, 13, 26, and 39 weeks after baseline).Multiple contact methods will be used for youth who are difficult to reach (eg, mail, alternate phone numbers, email, text message, Facebook).Participants will be asked whether or not messages can be left for each of the phone numbers that they provide.They will be informed that messages will not contain any information regarding the nature of the project.

Study Visit Management
The preferred time frame for all follow-up visits is within 4 weeks before or after the target study visit date.If the participant is unable to attend a visit within this time frame, the site staff will work with the participant to identify a day closest to the scheduled visit to perform the visit.
Participants in the intervention arm will be reminded by the Epic Allies app via a discreet phone notification (eg, "Your allies need you-log in to Epic Allies") to log on to the app every week.If a participant does not log on for 4 weeks, study staff will notify site staff and ask that they reach out to the participant.

Completing Web-Based Computer-Assisted Self-Interviewing Surveys
CASI surveys will ideally be completed at the clinic site during each study visit.Participants will be provided with a quiet, private area to complete the survey.The survey may be completed on the participant's smartphone, but a computer with Wi-Fi connection should be made available in case the participant prefers to complete the survey on a computer.
If a participant is unable to attend a follow-up study visit, the participant may complete the survey on his or her own.The survey should be completed within 4 weeks before or after the study visit target date.If a participant is unable to complete the survey within this window due to extenuating circumstances, the window may be lengthened to 7 weeks.

Epic Allies App Data
Caktus Consulting Group will store app usage data on a secure Web server for the duration of the study.At the end of the study, Caktus Consulting Group will send app usage data to the study team, destroy the data on the server, and then shut down the server.Protected health information is neither collected nor stored on the Web server.

In-Depth Qualitative App Satisfaction Interview
For the in-depth qualitative app satisfaction interview with intervention arm participants, the audio recording as well as the transcript will be marked with the participant's study ID only.Any identifying information mentioned in the interviews will be redacted in the transcripts, thus the transcripts will be deidentified.Both files will be uploaded and stored on a secure server.

Sample Size and Power Estimates
We estimated a sample size of 200 as feasible enrollment.Power calculations are estimated to detect between-group differences in the primary outcome (viral suppression) in a parallel two-group repeated measures design with equal allocation, based on a generalized estimating equation (GEE) framework assuming an exchangeable covariance structure, measurements at 3 follow-up points, and correlation among same participant repeat measures (rho) of .4.For all estimates, we used a two-sided test of significance and an alpha level of .05.Assuming a 20% loss to follow-up, we will have 80% power to detect absolute differences in viral suppression of 16.3% between the intervention and control groups in the proportion of participants with viral suppression when the proportion in the control group is 27%.

Quantitative Data Analysis
We will compare within-and between-group differences in primary and secondary outcomes for each follow-up time period.The 13-and 26-week follow-up will be considered our primary endpoint for the primary outcome, thus P values will only be computed for these time periods for the primary outcome.P values will be adjusted for multiple comparisons using the Benjamini-Hochberg procedure [65].Estimates for all outcomes will be presented with 95% CI.Intervention and control groups will be compared on baseline characteristics to assess balance.Patterns of missing data for our primary outcome of VS will be examined and baseline characteristics of participants with complete vs incomplete follow-up will be compared with assess nonresponse and attrition biases.Intervention effects will be evaluated using an intention-to-treat (ITT) approach.The primary study outcome (VS, defined as the lower limit of detection of site assay) will be compared at 13, 26, and 39 weeks in the intervention and control groups using generalized linear models (GLM), which can be used for dependent variables with normal, binary, poisson, and negative binomial distributions.Link functions will be selected as appropriate based on the distribution of the dependent variable.We will apply the GEE extension of GLM to account for within-participant correlation associated with repeated measures.

RenderX
GEEs allow for inclusion of categorical and count-dependent variables and appropriate specification of working covariance structures for observations that are correlated within groups and across time.Fixed main effects parameters for study site and eligibility group will be fitted to data to account for the nature of the randomized design.Intervention efficacy will be assessed in terms of the main effect for overall group differences.Use of a GEE framework means that inference will be made to the marginal effect of the Epic Allies treatment on the outcome, averaged across the study population.Secondary analyses will be performed similar to the methods described above to identify potential mediators and moderators of the intervention impact on primary outcome.

Missing, Unused, and Spurious Data
Several procedures will be used to conduct data analysis when data for either outcomes or baseline covariates are missing.The first step will be to assess the extent and pattern of missing data.If data are missing for only a few cases, then data analysis will be conducted only on study participants with complete data.If the pattern of missing outcome data is monotone, then inverse probability weighting will be performed to adjust the available data for loss to follow-up [66].If substantial nonmonotone missing outcome data are present, then a multiple imputation approach will be used.Unused or spurious data will be documented and discussed when disseminating results of this study.Baseline covariates will be compared between participants with complete follow-up vs those who have incomplete follow-up in order to assess the presence of informative missingness.

Qualitative Approach
The interview and analysis structure will follow a phenomenological approach to optimize our ability to capture and understand the study's experience-based topics of interest (eg, experience of HIV diagnosis and acceptance, experience of engaging with Epic Allies intervention and participants).Phenomenology is an ideal theoretical approach for this component as it is focused on describing both what a given group of participants experience and how they experience this particular phenomenon [67][68][69].Data are presented through textual descriptions of the phenomena based on summaries of the experiences described by respondents.The composite descriptions offer an explanation of the underlying structure which exists across the participants' experiences [69,70].This will focus on individual and shared experiences and meanings.

Qualitative Data Analysis
For the analysis, process interviews will be transcribed and then we will begin with our a priori list of themes (experience using Epic Allies, recent ART adherence challenges, etc).Study team members will read all transcripts and identify emergent themes from participants' experiences.These themes will be discussed as a group, and a final list of themes will be developed with brief descriptions, relationships between themes, and supporting quotes.For the qualitative research component, the Atlas.tiqualitative data analysis software (version 8, Scientific Software Development, Berlin, Germany) will be used to assist with theme identification and building, as well as coding textual data [69,71].Coding and analytic activities will be discussed during weekly team meetings.

Interim Analysis
No interim analysis will be performed for this study.The study team determined that this study does not involve greater than minimal risk (45 CFR Part 46.404 and 21 CFR Part 50.51).Participation in this study poses no more harm or discomfort to participants than they may experience in normal daily life or during routine physical or psychological examinations or tests.

Protection Against Harms
All sites have specific policies governing the treatment of human subjects.These policies specify that medical and psychological assistance will be available in the immediate environment in the event a participant should experience any adverse reactions resulting from study procedures.

Results
A total of 146 YMSM and YTWSM were enrolled in Epic Allies between September 2015 and 2016.Demographic characteristics of study participants are shown in Table 3.Although we estimated 200 as feasible for enrollment, study sites had fewer individuals eligible for participation in the study than expected.As a result, our ability to detect differences in our primary outcome (viral suppression) with 80% power assuming 20% loss to follow-up and the proportion of viral suppression in the control group is 27%, decreases by 3.1% (>16.3% to >19.4%).

Epic Allies Summary
Epic Allies addresses ART uptake and adherence, a critical need among a disproportionately affected patient population via familiar technologies using engaging, theory-based components.The app targets the most common ART adherence barriers among youth, addresses specific behavior outcomes, and is tailored for the target population and customizable for individual users.The social support, encouragement, and informational features listed above are designed to help youth overcome barriers to adherence across various stages of engagement in HIV care, ranging from lack of understanding and low health literacy, coping with side effects and drug toxicities, to the impact of drug and alcohol use on ART adherence.This novel intervention app, Epic Allies, targets HIV-positive YMSM and YTWSM, aged between 16 and 24 years (inclusive), with a detectable HIV VL.Epic Allies utilizes self-management tools, social support, and gamification to increase ART information, motivation, and behavioral skills and improve ART adherence, including (1) real-time data tracking of adherence with graphic visualizations; (2) tailored reminders and motivational messages; (3) connection to a network of other HIV-positive YMSM and YTWSM; and (4) a gaming approach engineered to reinforce daily adherence tracking, promote social networking support among users, encourage learning and skill building, and maintain user engagement.

Limitations
As with all longitudinal studies, a loss of participants to follow-up may induce a selection bias if missingness is informative and is related to both treatment arm and the study outcome.Furthermore, if compliance with the treatment assignment is less than 100% in either study arm, the ITT estimate, the study's primary estimate, will differ from the compliance-averaged causal estimate [72,73].In this case, the ITT estimate will still validly measure the efficacy of being randomized to the treatment arm but may not estimate the efficacy of the treatment itself.Data for secondary outcomes will be collected primarily from self-report survey, which is prone to both exposure and outcome misclassification.This misclassification could bias our study results either toward or away from the null hypothesis.Contamination may also be an issue, as participants at each of the study sites can be randomized to the intervention or control arm, and participants in the intervention arm may show Epic Allies to those in the control arm.All study participants are recruited from sites that provide HIV medical care and have procedures to monitor and address poor retention in care.Although many study participants were not regularly engaged in care at enrollment, retention in care outcomes among the study sample may be inflated when compared with a community-recruited sample.Thus, caution should be exercised regarding generalizability of retention in care outcomes.Finally, a modest sign-on bonus for regular use of the app in the first 2 weeks of the study and nominal monthly data use reimbursements are offered to those in the intervention group but not to those in the control group.The purpose of the sign-on bonus is to encourage regular use of the app early in the study to try to increase the likelihood that app use becomes a daily habit.The bonus is intentionally modest and time-limited to decrease the likelihood that money alone influences differences in study outcomes between the arms.Reimbursement for data use is intended to ensure that intervention arm participants do not intentionally avoid the app due to concerns about data usage.This is only warranted for intervention participants because the amount of data used for the control app is extremely low.While it is important to acknowledge the differences in incentives for the intervention and control groups, XSL • FO RenderX we believe they are unlikely to explain differences in outcomes between the arms.

Conclusions
If successful, Epic Allies will represent a novel adherence intervention for a group disproportionately impacted by HIV in the United States.Epic Allies would be clinically attractive, as adherent patients would require less frequent clinic visits and experience fewer HIV-related secondary infections [74][75][76].Reducing clinic visits and secondary infections could make the intervention financially attractive by reducing health care costs.Epic Allies could also greatly impact public health as ART adherence reduces HIV infectivity and subsequently reduces HIV transmission [74].Epic Allies could be used during times of adherence vulnerability (eg, when initiating ART or changing medication regimens) and could easily be expanded and adopted for use among larger populations of YMSM and YTWSM, other HIV-positive populations, and for those diagnosed with other chronic diseases such as diabetes and hypertension.©Sara LeGrand, Kathryn E Muessig, Alyssa Platt, Karina Soni, Joseph R Egger, Nkechinyere Nwoko, Tobias McNulty, Lisa B Hightow-Weidman.Originally published in JMIR Research Protocols (http://www.researchprotocols.org),05.04.2018.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.

Table 1 .
Primary and secondary outcomes."X" indicates that this outcome was assessed at the time point indicated in the column above.VS: viral suppression.VL: viral load.CASI: computer-assisted self-interviewing.ART: antiretroviral therapy.

Table 2 .
In-depth qualitative app satisfaction interview enrollment.