All study procedures were approved by Johns Hopkins School of Medicine institutional review board (IRB00211578).

Data were obtained from 12 focus groups and 1 in-depth interview among BSMM from Baltimore, MD, between October 2019 and May 2020 [38]. Participants were recruited using a combination of active (eg, contacting participants from previous studies who provided written consent to be called for future research) and passive (eg, advertising on Craigslist, obtaining participant referrals) strategies [35,37]. Participants were eligible based upon the following criteria: self-identifying as Black or African American, self-identifying as a man, aged 18 years or older, self-reporting as HIV negative, having oral/anal sex with at least 1 male partner in the previous 6 months, and residing in Baltimore, MD.

Most focus groups (n=9) were conducted in a private designated research space at Johns Hopkins School of Nursing. The protocol for the last 3 focus groups was updated to a virtual, synchronous format for safety due to COVID-19. Details regarding the protocol for virtual, synchronous focus group conduct have been published elsewhere [35]. The focus group guide was designed using LCT and HBM constructs, and groups were stratified by 3 age groups: 18 to 24 years, 25 to 34 years, and 35 years and older. Groups were limited to 5 or fewer for feasibility and to allow all participants to respond to each question [35,37]. An in-depth interview was conducted because only 1 participant attended a scheduled focus group and still wanted be a part of the study and share thoughts regarding the intervention [14,38]. Two facilitators led the focus groups (and interview). One moderated discussions and recorded notes; the other recorded notes, observed group dynamics, and conducted administrative activities [35,37]. In-person participants provided written informed consent; virtual participants provided oral informed consent that was documented by the study team prior to beginning the focus group [35].

Data collection included reflexive debriefing among the facilitators before and after every focus group to review research intentions, identify potential biases, explore personal challenges of the investigative team, and explore preliminary themes [35,39,40]. Data collection began with a written or online survey via Qualtrics assessing demographic and behavioral characteristics along with perceived HIV risk, assessed using the 8-item Perceived Risk of HIV scale [41], the scores for which ranged from 10 to 40, indicating low-to-high perceived HIV risk. Focus groups were conducted after the survey and lasted 50 to 75 minutes. Facilitators began by asking participants for details regarding their current lifestyles and social activities and how their lives and self-image would change if they received a positive HIV test result to learn more about their perceived severity of disease. Facilitators also targeted domains regarding participants’ thoughts about their behaviors and their risk relative to learning the high HIV prevalence within the city and likelihood of infection for BSMM nationally [6]. Facilitators then shared details regarding PrEPme using handouts and live demonstrations on a smartphone and asked questions regarding attitudes toward using the app to self-monitor sexual risk behaviors. After obtaining attitudes toward the app, facilitators asked questions regarding the usefulness of a PCA in the intervention, attitudes toward sharing their PrEPme diary with the PCA, ideal PCA characteristics, and potential barriers to working with a PCA. All participants were compensated $80 [35].

All focus groups were audiorecorded and transcribed verbatim by an IRB-approved company. The first author reviewed all transcriptions for fidelity to audio files and revised text files as needed prior to coding in ATLAS.ti (version 8.4, ATLAS.ti Scientific Software Development GmbH). The facilitators reviewed all focus group recordings, transcripts, and notes and then designed a codebook for descriptive thematic analysis using the questions in the focus group guide (Multimedia Appendix 1). The first author coded transcripts and systematically identified themes using an adapted pile-sorting approach whereby quotes associated with specific codes in ATLAS.ti were into an Excel (Microsoft Corp) spreadsheet and organized [15,40,42]. The facilitators then reviewed the quotes associated with codes and sorted them into piles for similarity within Excel to represent themes. Facilitators identified themes as patterns or novel responses associated with specific codes or focus group questions. To identify a range of themes, novel responses that at least 1 person in the group mentioned were also considered [40,43]. Between-group analysis was conducted to identify potential thematic differences by age.

Data resulted in 5 groups of BSMM aged 18 to 24 years, 4 groups aged 25 to 34 years, 3 groups aged 35 years and older, and 1 in-depth interview with a participant aged 18 to 24 years (n=39). All except 1 participant completed the survey (Table 1). Most who completed the survey identified as homosexual, gay, or same gender-loving (26/38, 68%), were employed (26/38, 69%), and single (25/38, 66%). A total of 34% (13/39) of participants had ever been diagnosed with an STI, of whom 38% (5/13) had repeated at least one STI. Perceived Risk of HIV scores ranged from 13 to 40 (x̄=22.4, s=5.3), and HIV Incidence Risk Index for Men Who Have Sex With Men scores ranged from 5 to 28 (x̄=16.2, s=5.7). Most (26/38, 68%) were interested in using a smartphone app to self-monitor their sexual behaviors. Half (19/39, 50%) of surveyed participants had ever used PrEP; 12 out of the 39 reported being current PrEP users.

Focus groups yielded important themes regarding perceived HIV risk and attitudes toward the proposed multicomponent intervention.

Across and within age groups, participants provided polarizing responses regarding perceived HIV risk. They mentioned that HIV infection would impact their mental health due to lifestyle changes and anticipated stigma from society and potential partners, but they expected to overcome those challenges since contemporary medications improved longevity and quality of life. Despite being presented with Centers for Disease Control and Prevention data regarding HIV prevalence and incidence among BSMM locally and nationally, participants across age groups still perceived their HIV risk as low—even if they disclosed drug use or condomless sex. Each group used positive terms such as goal-oriented and caring to describe themselves and reckless, impulsive, and drug addict to describe individuals they thought were likely to acquire HIV. However, they also mentioned they occasionally had the same characteristics as someone they perceived as high risk for HIV despite not equating their current sexual behaviors as high risk. BSMM in the 2 youngest groups specifically compared their current behaviors to their past and their friends who they believed had riskier behaviors or who had acquired HIV.

Across groups, participants reported barriers regarding using PrEPme such as remembering to use it and providing honest responses. Those aged 18 to 34 years generally agreed that having a smartphone app diary could be useful for individuals with higher risk behaviors (eg, individuals who use drugs, practice condomless sex with multiple partners), but men age 35 years and older did not think it would be useful or used among BSMM. Participants across all age groups reported that seeing their sexual activities summarized on the dashboard could trigger feelings of internalized stigma regarding their risk behaviors (see Textbox 1).

Participants agreed that having an in-app dashboard of sexual risk behaviors would decrease BSMM’s willingness to provide honest responses and use PrEPme for long. Some added that seeing their condomless sex and receptive anal sex activities in the app could exacerbate internalized stigma and shame that would cause them to discontinue use because society, gay men, and clinicians stigmatize those behaviors.

The youngest 2 age groups mentioned concerns that their sexual diary could be hacked by app developers or seen by community health workers who were in-app support systems for PrEP linkage. These concerns could be related to anticipated stigma and shame regarding the sexual behaviors documented in the app. However, participants did not mention this explicitly or associate these directly.

Overall, most participants across age groups affirmed that having another BSMM to discuss sexual health and PrEP before visiting a PrEP-prescribing clinician was acceptable if he was relatable and trustworthy. They specified that the PCA should be Black and a current PrEP user so they could share their experiences with side effects and maintaining adherence (see Textbox 2).

These and other characteristics such as aesthetics (eg, style of dress), language use, personal disclosures, and neighborhood familiarity served as indicators for BSMM to use to anticipate how stigmatizing, trustworthy, or relatable the PCA would be.

Despite agreeing that having another BSMM as a PCA was acceptable in a clinical intervention, there were age group differences regarding additional PCA characteristics and the practical role of the PCA. Specifically, those aged 18 to 24 years shared preferences for having a PCA who was older or a Black woman, partly due to anticipated within-group stigma from another BSMM who might judge their sexual behaviors as risky. Older participants shared that they would prefer not to engage with a PCA in practice because they did not want to relieve clinicians of the responsibility of asking detailed and personal questions to know them as patients. They also mentioned concerns that working with a PCA prior to visiting a clinician could be another barrier to PrEP initiation for interested BSMM. Still, the older men agreed that the PCA should be a BSMM who was professional yet relatable to discuss specific cultural issues and PrEP-related concerns such as side effects, efficacy, and relative risks for HIV infection by sexual positioning.

Across groups, participants acknowledged that a benefit of a PCA in this intervention included being able to discuss sexuality freely and trusting the PCA’s cultural competency more than that of a clinician who was not a BSMM (see Textbox 3).

They mentioned that a PCA could help circumvent potentially stigmatizing or unsatisfactory conversations with clinicians regarding sexual risks but were not interested in the PCA discussing health concerns with the clinician on their behalf. Their interest in sharing their PrEPme dashboard with a PCA or clinician was contingent on their perception of how judgmental and trustworthy either would be.

We designed a multicomponent strategy to increase perceived HIV risk and PrEP initiation among BSMM and obtained in-group perspectives regarding the intervention components for refinement. Overall, BSMM had low-to-moderate perceived HIV risk. Using a PCA as an interventionist could be feasible and acceptable among younger BSMM. However, findings are inconclusive regarding feasibility of using the diary in PrEPme to record and review sexual behaviors for this group. We found age-related differences in attitudes toward the intervention components, which could result from different exposures to HIV-related deaths by age group and changes in personal agency and perceived support needs due to older age [14,44]. Novel insights regarding stigma and shame also emerged that informed considerations for intervention refinement.

BSMM held dissonant attitudes between their sexual behaviors and HIV risk, perceiving their risk as low. Other studies among BSMM have found similar results [17,45]. Having lower perceived HIV risk could be an attempt to combat stereotype threat, which is a perceived risk of conforming to negative stereotypes about a group [46,47] that exacerbates poor patient-clinician communication, refusal of clinical recommendations, and medication nonadherence among Black patients [47,48]. For BSMM, the threat of being stereotypically high risk or sexually deviant could be an important, understudied phenomenon impacting HIV and PrEP initiation. Anticipated and internalized stigma and shame regarding self-monitoring sexual behaviors, discussing sexuality with professionals, and receiving an HIV diagnosis were salient. Additionally, practicing receptive anal sex, or being a bottom, can also carry specific shame and stigma among BSMM due to masculinity stereotypes [49,50]. However, stereotype threat decreases with race-concordant clinicians [43,51]. Therefore, a PCA must be a trained professional with substantial sexual health literacy who understands the issues concerning internalized stigma, shame, and stereotype threat and can help BSMM navigate their perceived HIV risk to make decisions about PrEP. A culturally responsive PCA could circumvent some of the multilevel barriers to perceived HIV risk and PrEP initiation among young BSMM with cultural familiarity, acceptable language, explanation of privacy protocols, and personal disclosures to build trust [35,43].

Findings regarding perceived susceptibility and severity of HIV among BSMM expanded our perspectives regarding the intervention framework. Based upon findings that BSMM had low perceived HIV risk, internalized stigma and shame regarding sexual risk behaviors, and preferences for a culturally congruent PCA, we incorporated possible selves theory [52,53] along with LCT and HBM to refine the intervention strategy. Possible selves represent individuals’ ideas of what they could and want to become in addition to what they are afraid of becoming and can incentivize future behavior [53,54]. Therefore, an intervention using a PCA should consider how congruent with and relatable the peer is with a future self for HIV-negative BSMM beyond race and sexuality concordance [14,35,43]. Factors such as cultural familiarity, professionalism, open-mindedness, sexual health literacy, and PrEP use could be key additional PCA characteristics that build trust, mitigate stereotype threat, and increase willingness to initiate PrEP among BSMM [35,37].

Future research should explore the relative impact of implementing this multicomponent strategy on perceived HIV risk and PrEP initiation among BSMM. Since a sexual health diary app could trigger internalized stigma and clinical research teams are not culturally congruent with BSMM, future research should implement interventions led by culturally congruent investigators familiar with these psychosocial barriers among this population. Future research could explore the potential impact of the dual role of an investigator or clinician as a PCA. Having PrEP clinical trials led by BSMM researchers and clinicians could be an important yet understudied structural, social, and psychological factor in HIV prevention among BSMM.

Limitations should be considered. This study included a convenience sample and is not representative of HIV-negative BSMM. HIV status was self-reported. Groups were intentionally conducted with a small number of participants to increase rapport, trust, and privacy and to ensure that everyone could respond to every question if desired [35,37]. Information regarding implications of the intervention regarding on-demand and injectable PrEP was limited.

Qualitative studies are crucial to designing culturally relevant interventions for BSMM because they identify how socioecological factors such as stigma impact HIV risk for this group [55]. Despite the proliferation of smartphone apps for health promotion and HIV prevention [56], we found that using PrEPme to record and review sexual risk behaviors may not be as innocuous as anticipated. Therefore, we plan to train the PCA to anticipate stereotype threat and manage insider-outsider dynamics to demonstrate relatability, trustworthiness, care, and professionalism. In the refined intervention, we plan to encourage participants to use the diary in PrEPme and ensure that the PCA is a nonjudgmental possible self who can help BSMM navigate feelings of stigma, shame, low perceived HIV risk, and PrEP hesitancy as participants reflect on their behaviors. Since providing objective HIV risk scores alone does not increase PrEP uptake among BSMM [7], the PCA will be trained in motivational interviewing to partner with BSMM to explore their health goals, perceived HIV risk, and relative interest in PrEP given their responses to the diary in PrEPme. We named the refined intervention strategy Possible based upon findings from this study.