Developing an mHealth program to improve HIV care continuum outcomes among young Black gay and bisexual men

Young Black gay and bisexual men (YBGBM) face substantial disparities in HIV diagnosis and care outcomes. While gay, bisexual, and other men who have sex with men (MSM) constitute approximately 3.9% of the total U.S. population [1], they accounted for a disproportionate 67% of new HIV diagnoses in 2021 [2]. Among MSM, Black MSM experience the highest rate of HIV diagnoses by race/ethnicity, with 37% of new HIV diagnoses in 2021 attributed to this group [2]. Among 13–24-year-olds, Black youth experienced a rate of HIV diagnosis 14.3 times that of their White counterparts [2]. Furthermore, youth between the ages of 13–24 made up 24% of HIV diagnoses among MSM, but 53% of new diagnoses among Black MSM [2].

Enhancing engagement in the HIV care continuum, from diagnosis to viral suppression [3], is crucial to reducing HIV morbidity, mortality, and transmission [4‐7]. However, disparities persist, as Black MSM have poorer linkage to care, retention in care, and viral suppression outcomes compared to their White counterparts, even after controlling for access to care [8, 9]. Of particular concern is that, among Black MSM living with HIV, only 65% of those ages 13–24 and only 61% of those 25–34 have achieved viral suppression [10].

Multiple factors negatively impact HIV care continuum outcomes among YBGBM, including stigma related to HIV and sexual orientation [11‐16]; low self-efficacy related to health management and provider communication [17, 18]; mental health issues, such as depression, anxiety, and post-traumatic stress disorder [18, 19]; substance use [17, 18, 20‐22]; and insufficient social support [23, 24]. Moreover, many of these factors intersect, creating syndemics that magnify negative outcomes for YBGBM [24‐29]. In light of this spectrum of challenges, it is essential to support YBGBM using holistic approaches [30, 31] that engage them in the HIV care continuum through acceptable and accessible solutions [32, 33].

Mobile health (mHealth) interventions can positively impact HIV prevention, testing, and treatment outcomes [33], including those related to HIV self-management, and have been found to be feasible and acceptable for YBGBM [32‐37]. These interventions can also be very convenient and accessible, as 96% of Americans ages 18–29 own a smartphone [38], and they can be used in rural areas where in-person HIV interventions might otherwise be unavailable [39‐41]. Though evidence on the effects of mHealth strategies for stigma reduction among youth is nascent [42], emerging research suggests that these approaches can help build social support and resilience among youth who face discrimination due to their race/ethnicity or sexual orientation [43, 44].

In addition to the wide array of mHealth tools and text-based information that can be provided via smartphones, these devices enable the delivery of video-based media [32]. Prior research has found that targeted HIV prevention videos for Black MSM increased understanding and recall of messages [45]. Furthermore, online personal testimonial videos of people with chronic conditions, including people living with HIV, may improve education and social support, and reduce stigma [46]. Several studies have demonstrated the effectiveness of video-based interventions, including some specifically created for young MSM living with HIV [47‐51]. These interventions have demonstrated impacts on HIV testing [47], risky sexual behaviors [47, 49], HIV disclosure [47, 49], and HIV treatment initiation [51].

Studies have identified preferences for mHealth HIV prevention and care programs in general, as well as among youth and YBGBM specifically. Desired characteristics include confidentiality protections, motivational content, both video and text content, adherence and lab results tracking, an online community, resource lists, and an overall focus on holistic health rather than only on HIV [31, 52, 53]. However, few if any mHealth programs include most or all of these features [52, 53]. In this article, we detail the human-centered design (HCD) approach used to create an mHealth app, called Amp, intended to enhance HIV care continuum outcomes and to holistically support the health and well-being of YBGBM. Amp was developed in close collaboration with YBGBM living with HIV and with HIV service providers. It includes comprehensive, video-based information, a range of mHealth tools to support HIV care outcomes, resources, and a community of people living with HIV. In the subsequent sections, we discuss the impetus for this app, the methods used to generate initial insights, the transformation of these insights into testable prototypes, and the iterative refinement of the intervention concept.

In July 2019, we conducted 12 key informant interviews in person or by telephone with HIV service providers at five different organizations in Alabama (Huntsville), Louisiana (New Orleans), Texas (Dallas), and California (Los Angeles and Riverside; see Table 1). Providers gave either written or verbal consent and received a $75 incentive for a 60-min interview. All worked at HIV service organizations that provide case management and linkage to care to YBGBM, and 10 worked at organizations that also offer HIV treatment. We interviewed providers from different levels within the organization, including executive, senior, and frontline staff. We developed a semi-structured interview guide to solicit both general and specific input on program content, tools, and features that could be useful to YBGBM living with HIV and that would help staff to better support clients along the HIV care continuum. Provider interviews were audio recorded and professionally transcribed.

Rapid prototyping is a methodology used in HCD that involves the iterative testing and refining of low-fidelity prototypes to efficiently optimize program elements. YBGBM and providers were recruited for rapid prototyping through our team’s HIV service organization contacts. Rapid prototyping sessions were carried out over two days each in California (Los Angeles), Alabama (Huntsville), and Mississippi (Jackson), with a total of 31 YBGBM ages 20–30 and 12 HIV service providers (see Table 1). Sessions took place from August–September 2019. Each session was conducted in-person with individual participants by one of three study staff. Participants provided written informed consent and received a $100 incentive for a 60-min session.

During these sessions, low-fidelity hand-drawn and digital prototypes were used to gather participants’ input on: the user interface; content, tools, and resources to be included in the program; how YBGBM might want to interact in a community of peers; ways to maximize program engagement; and whether the program should include an app, mobile website, or both. At the end of each day of prototyping, responses were integrated into revised prototypes and tested again the next day.

Prior to the rapid prototyping, we produced 13 videos on eight topics based on needs expressed by YBGBM participants and providers in the ideation stage. We varied the settings (a person’s home or a film studio), speakers (YBGBM or HIV clinician), and video characteristics (e.g., subject looking directly into the camera or off camera, and heavy or light editing) in order to give participants the opportunity to provide feedback on both video content and style. Each participant was shown at least two videos, and videos were evenly divided between participants.

Using data from stages 1 and 2 and the online survey, we worked with a website developer to create a working digital prototype of the app. The objective of the online usability testing was to identify potential functionality issues and use the findings to maximize program acceptability and effectiveness. Participants were sent a link to the prototype and asked to view it on their smartphone. The prototype included the proposed topics, tools, and videos developed in stage 2, but was designed in grayscale without any branding or images. This approach allowed us to test usability of the interface without distracting users with design elements.

Usability testing participants (n = 21; see Table 1) were recruited through our team’s network of HIV service providers and from the online survey, with participants contacted in the order in which they completed the survey. Usability testing sessions were conducted in October 2021, with study team members observing participants interacting with the prototype over Zoom. During the session, participants were presented with several tasks related to finding information, tools, or resources on the prototype (e.g., “find and use the medication reminder” or “find a video about disclosing HIV status to sex partners”). Participants were also asked open-ended questions about topics, community features, tools, and resources; ease and predicted frequency with which they would use the app; and whether most of the content should be video- or text-based. Finally, each participant orally completed the System Usability Scale, or SUS [58], a reliable tool for measuring usability comprised of 10 questions with five response options, from “strongly agree” to “strongly disagree.” Each testing session lasted approximately one hour. Participants provided verbal consent and received a $100 incentive.

Stage 1 focus group and interview transcripts were coded using deductive and inductive approaches by two of the team members experienced with qualitative data analysis. They developed codes based on the interview and discussion guides and added codes as they emerged from the data. Transcripts were coded to identify themes relevant to developing the app and its components (e.g., content, tools, community features, and resources). In stage 2, written notes of participants’ feedback during rapid prototyping sessions were entered into a digital spreadsheet after each session and were categorized into prototype elements that received positive feedback and those needing improvement. In stage 3, survey data and the SUS results were assessed using descriptive statistics (e.g., frequencies) in SPSS v.26. During usability testing of the functional prototype, task completion times were captured by the interviewer using a timer and were used to calculate mean and range values. Qualitative responses were analyzed for common themes with regard to satisfaction with content, tools, usability, and likely uptake of the program.

This study details the use of HCD to create Amp (Fig. 1), a holistic mobile app created to improve the health and well-being of YBGBM living with HIV. Insights gathered across three stages of research—ideation, rapid prototyping, and user feedback—were iteratively incorporated to develop a program prototype that received high acceptability and usability ratings. HCD is a comprehensive approach that has been increasingly utilized in the development of mHealth programs [54, 59]. As demonstrated in this study, HCD is an approach that can yield high usability, feasibility, acceptability, and potential uptake from end users. The use of HCD in this study, which included engaging YBGBM living with HIV and HIV service providers as co-creators, resulted in a desirable program with a combination of content, tools, and resources that participants had not previously seen all in the same place. Further, the Amp concept was described by providers as promising for engaging YBGBM living with HIV across the care continuum, which is imperative to reducing HIV-related transmission, morbidity, and mortality [4, 5, 7].

Findings from our HCD research reinforced those identified in prior research regarding the needs of people living with HIV, including: social support [60], adherence support [53], HIV knowledge [61], programs to address stigma and discrimination [6, 11, 25, 62], and the general demand for more holistic approaches to HIV care [30, 31, 53]. Amp includes content, tools, and features to address all of these needs. Amp offers several alternatives to the ways that YBGBM living with HIV usually receive support—including community features that allow users to create a profile, communicate with peers via forums and direct messaging, professionally produced video narratives, trained mentors who can be messaged on-demand, and a range of tools to support medication adherence, retention in care, and viral suppression. This wide variety of tools and resources is designed to meet the needs of YBGBM where they are, and as their needs change over time. Further, HIV service providers were very enthusiastic about using Amp within their organization, to improve their services and help clients reach their HIV care continuum goals. While several existing mHealth HIV care continuum interventions take a comprehensive approach [63], Amp is unique with regard to the holistic range of tools and content, and in its extensive use of custom-created personal narrative video content.

Our study has strengths stemming from the mixed methods approach it utilized. Intensive involvement of program end users and stakeholders is fundamental to designing interventions using HCD. During every stage of our research, we collected data from geographically diverse groups of YBGBM and providers, validating and triangulating findings at each stage. This is likely the reason that Amp was so well received in the final user feedback stage.

While this study outlines a design process that was effective, there are several limitations to our findings that should be noted. Recruiting a sample of 200 HIV positive YBGBM for participation in a survey is logistically challenging, and it was not possible to ensure that this sample is representative of the broader population. Thus, our survey findings may not be generalizable to all YBGBM living with HIV in the U.S. Similarly, relying on our professional network of HIV service providers for much of the recruitment may have introduced some bias, especially as this meant that most YBGBM participants were linked to care. We attempted to mitigate this potential bias by working with multiple HIV service providers from diverse organizations who had direct connections to local populations, as well as by recruiting from a geographically diverse sample of YBGBM in stage 3, mostly via geospatial apps. Survey findings mirrored findings from other stages, which suggests that responses from stages 1 and 2 were minimally biased by our recruitment methods. Further, we administered the SUS verbally with usability testing participants over Zoom, which is different from the typical questionnaire-based approach. While we requested that participants give honest feedback in order to facilitate improvements to the app, it is possible that this approach resulted in response bias.

The results of these activities have led to the production of Amp, a mobile app that has the content, tools, community, and resources identified as important among YBGBM and HIV service providers in our HCD process. At the time of writing, the Amp mobile app is being developed with Phase II funding through a Small Business Innovation Research Grant provided by the National Institutes on Minority Health and Health Disparities. Amp will be evaluated in a nationwide randomized controlled trial in 2024, which will test the effectiveness of the application on knowledge, social support, linkage to care, medication adherence, retention in care, viral suppression, mental health, and HIV related stigma among a relatively large sample of YBGBM living with HIV. Our team plans to disseminate Amp to organizations nationwide, including HIV service organizations, public health departments, and community organizations. Ideally, Amp will be offered to people immediately after an HIV diagnosis as well as to existing clients who are living with HIV.

This study adds to the limited research on using HCD to develop HIV mHealth programs. The evaluation study will be among the first randomized controlled studies of an HIV intervention developed with HCD. Although we attribute the early signs that Amp will be feasible, acceptable, usable, and effective to the HCD process through which it was developed, further research is needed to understand the impact of HCD relative to other approaches. Future studies could also investigate using HCD to create mHealth self-management programs for other chronic illnesses.

We used HCD in a highly collaborative, multi-stage process to develop an innovative mobile app that holistically addresses multiple health and wellness needs identified by YBGBM and HIV service providers. Each stage of the process identified barriers and facilitators to program uptake, with findings used to make the app more useful and engaging. These results support prior studies finding that intensive involvement of end users and stakeholders using HCD can yield innovative, acceptable, and culturally-appropriate programs.