Between October 2017 and March 2018, we conducted two focus groups and three in-depth interviews with MSM and TGP who use methamphetamine or who have worked with people who use methamphetamine (eg, peer educators). Focus groups were facilitated by two authors (VMM and JDS) and interviews were conducted by VMM. Focus groups and in-depth interviews followed semistructured interview guides (see Multimedia Appendices 1 and 2) that included questions regarding what information about PrEP and the study should be provided to participants, what types of recruitment materials should be used, and whether participants thought people would be interested in participating in the study as well as specific questions to develop the peer navigation and text messaging interventions. The questions about the text messaging intervention included reviewing messages used in an earlier study by Reback et al [23], which were aimed at reducing methamphetamine use and sexual behaviors associated with risk for HIV acquisition. These text messages were developed using the Social Support Theory, the Health Belief Model, and the Social Cognitive Theory. The Social Support Theory posits that instrumental, informational, and emotional support from one’s network can lead to positive changes in health behavior. The Health Belief Model describes how one’s beliefs about health behaviors and risks to one’s health are associated with engagement in protective behaviors. Finally, the Social Cognitive Theory describes how the interaction of the individual, their behavior, and the environment are related to health behavior. The study done by Reback et al demonstrated a reduction in methamphetamine use and unprotected anal intercourse with nonprimary partners among MSM who used methamphetamine [24]. Focus groups and interviews were anonymous and the audio recordings were transcribed.

The first focus group was done with 9 peer educators at Project Needle and Sex Education Outreach Network (Project NEON) of the Seattle Counseling Service. Project NEON is a harm reduction program that aims to reduce sexual and drug-related risks associated with the use of methamphetamine. In order to generate preliminary data for this project, the authors had collaborated with the Project NEON peer educators on earlier work to try to better understand PrEP use among MSM and TGP who use methamphetamine and to increase PrEP education in this community [13,16]. This work included developing PrEP education materials with the Project NEON peer educators and recommendations for evaluating text messages and peer navigation as potential supports for PrEP adherence among MSM and TGP who use methamphetamine.

The second focus group was recruited through Seattle Area Support Groups (SASG), now Peer Seattle, a nonprofit organization that provides peer emotional support and development services to lesbian, gay, bisexual, transgender, and queer or questioning individuals, plus other sexual and gender minorities (LGBTQ+). The second group consisted of 10 MSM who had used methamphetamine. For these two focus groups, the researchers gained access to participants through long-standing relationships with these two community organizations.

After conducting these two focus groups, we chose to conduct research activities with individuals who represented perspectives that had not already been shared, specifically LGBTQ+ youth and people who are not cisgender. We originally planned to conduct an additional two focus groups with up to 10 participants each representing these populations; however, we were advised that interviews would be more appealing to individuals who are not cisgender since there was less chance of being “outed” in a one-on-one activity. We recruited eligible interview participants through palm cards and word-of-mouth. Palm cards were brought to local agencies that provide services to LGBTQ+ youth and people who are not cisgender. The researchers also informally provided palm cards to, or discussed the study with, community members in these networks with whom they had up to two decades of collaboration. Despite these various approaches to recruitment, we only conducted three interviews out of the 20 planned.

At the end of each focus group or in-depth interview, participants were given a US $40 gift card to an online retailer. The University of Washington Institutional Review Board approved this study (#00004760). All participants were given an information sheet for participation; written consent was waived.

The protocol was drafted in parallel with formative activities. Therefore, after each focus group and interview, the facilitators (VMM and JDS) met to discuss the major themes that arose during the discussion to inform ongoing study development in an iterative process. The first version of the study protocol was finalized in April 2018 and submitted for ethics review and approval.

After completion of the formative work, two authors (VMM and NF) performed a retrospective content analysis [25] of the focus group and in-depth interview transcripts to describe which specific opinions and recommendations influenced protocol development and the study interventions. First, they reviewed transcripts individually and compared them to the final study protocol. Then they met and discussed which themes from the early formative work had most influenced the resulting interventions and protocol until they reached consensus.

Across focus groups and interviews, participants reported that they thought MSM and TGP who use methamphetamine would be interested in the study. Participants offered a variety of suggestions about the peer navigation and text messaging interventions. Their recommendations focused on ways to make these interventions relevant and accessible to MSM and TGP who use methamphetamine.

Focus group and interview participants thought that the peer should be someone who is patient, nonjudgmental, not transphobic, and has experience with the target population.

Participants thought that it would be important for the peer to help with appointment reminders and medication support, including refills. One peer recommended the option of holding on to medications for participants, as the peer had prior experience providing a few days’ supply of medications to someone he had sponsored and reported that it had helped with medication compliance. They also recommended that the peer be flexible in order to meet study participants outside of the “9-to-5” schedule and in a variety of settings.

They also discussed barriers the peer should anticipate, including difficulties reaching study participants.

Participants discussed how setting boundaries would be important for the peer, including appropriate times to expect them to answer the phone and what interactions would or would not be appropriate if the peer and participant were visible to each other on the internet (eg, on a hookup app).

The number of texts per day that participants thought would be appropriate ranged from two to six and there was general consensus that three per day was not excessive. Focus group and interview participants suggested that the study participants should be allowed to select the timing of the messages, since people who use methamphetamine may not be accessible during the daytime. There were variable responses to the example text messages that were shown to participants (see Table 1).

Some participants liked the content, as evidenced by the following quotes.

The text messages from the Social Support Category were particularly liked.

In addition, text messages that provided “straightforward” information, like “Needle exchange 2nite @ __________,” were preferred, although one participant said receiving these may make them feel guilty.

Some other examples of potential text messages were regarded less highly.

The variety of opinions on our example text messages resulted in us developing many text message options: 540 text messages across 11 content categories.

Focus group and interview participants consistently recommended that study participants should be able to choose their own text message content.

Across focus groups and interviews, there was consensus that straightforward PrEP reminders would be best (eg, “take your pill”) and that we should “change up” the messages so they were not the same each day. Daily PrEP reminders were discussed in the context of the effects of methamphetamine, including “losing days.” One participant recommended including the passage of days within the texts themselves to help participants be aware of what day it is and how much time is passing.

Multiple participants recommended using email in conjunction with text messages, because it is common to lose one’s phone.

Participants also emphasized that using emojis was very important and could help “with discretion.”

Finally, in our second focus group, one participant reported a technology phobia when high on methamphetamine and advised that receiving texts could cause paranoia for some.

Endorsement of potential paranoia in this population by focus group attendees led to a protocol exclusion criterion regarding discomfort with technology.

In the focus groups and interviews, we shared educational materials that had been previously developed during formative work done with Project NEON to increase PrEP education and uptake among MSM and TGP who use methamphetamine [19]; we then asked participants for feedback about the card images and content. Participants liked the nonjudgmental phrasing “we aren’t here to judge your drug use” but recommended replacing “your drug use” with “you.” Some, but not all, participants thought that the inclusion of the word “tweaker” was stigmatizing; however, there was consensus that the word “partying” would be a preferable word choice and widely understood among the LGBTQ+ community to refer to methamphetamine use. Using terms familiar to individuals who engage in party and play (ie, having sex while using methamphetamine) was liked and recommended (eg, “parTy” or “PnP” [party and play]). Participants recommended images that did not blatantly show drug use or sex but conveyed these activities discretely (eg, smoke in the background and an image suggestive of a sex act). Images that were gay friendly and transgender inclusive were also endorsed.

Participants were concerned that our recruitment approach and materials would not reach a subset of MSM and TGP who use methamphetamine and who may be at particularly high risk for HIV, including those who are homeless and engage in survival sex work.

Participants recommended that we have later hours for study visits (ie, after 5 PM) to make the study more accessible to potential participants.

Participants thought that retention may be challenging among MSM and TGP who use methamphetamine and, to retain participants, it would be important to be nonjudgmental.

There was general consensus that our planned reimbursement of a US $20 gift card to an online retailer for study surveys three times during the study period was sufficient. However, one participant thought it would be better to provide a gift card to a brick-and-mortar store due to the lack of internet access among the target population.

The initial version of our study protocol was implemented in June 2018 (see Multimedia Appendix 3). In addition to meeting the criteria for PrEP use at the clinic of enrollment, eligible participants have to be 18 years of age or older; must be MSM or TGP; must be able to understand, read, and speak English; must report methamphetamine use in the past 3 months; must have a cell phone that can send and receive text messages; and must intend to remain in the King County area for at least 6 months. Exclusion criteria include PrEP use in the prior month; circumstances that preclude provision of informed consent, make participation unsafe, or make it unlikely that they would be able to participate for 6 months; or discomfort or anxiety with regard to text messaging. The criterion of discomfort or anxiety with text messaging was a direct result from our formative work.

Study procedures are summarized in Table 2 and the specific intervention components and procedures that resulted from our formative work are listed in Table 3. Briefly, eligible participants who choose to enroll are provided a link to a baseline survey and, if they complete it within 3 days of their initial PrEP appointment, they are randomized to one of the four study interventions: standard of care, text messaging, peer navigation, or combined text messaging and peer navigation. Participants are followed at 1 month, in accordance with the clinic’s procedures; 3 months; and 6 months. Follow-up surveys are sent to participants at 3 and 6 months; for each completed survey, participants are sent a US $20 gift card to an online retailer by email. All participants are invited to participate in an in-depth interview when they complete or discontinue the study. Participants are provided an additional US $40 gift card to an online retailer for completing an in-depth interview. We chose not to include a gift card option to a brick-and-mortar store because we needed the ability to send the gift cards to individuals electronically, since online study surveys are administered remotely and end-of-study interviews can also be done via teleconference or phone.

We incorporated suggestions from our formative work into the peer position and training. We included prior experience working or networking with substance users, particularly methamphetamine users as a required criterion for the peer position and created a role with flexible work hours and locations. Upon hiring, study onboarding includes standard research training sessions as well as training sessions with specific modules related to peer support, substance use, and boundary setting. Peers also receive ongoing peer support. The peer attempts to meet with all participants randomized to the peer navigation intervention within one week. During this initial contact, the peer completes a harm reduction–based needs assessment (see Multimedia Appendix 4). We anticipated that a peer would be able to provide assistance to 7-10 participants based on a recommendation from leaders of a local peer counselor program for persons who use methamphetamine. However, because the peer navigation is individually tailored and participants have varying levels of need, the study team meets regularly to ensure that the number of participants the peer is assisting is appropriate, and additional peers can be hired as needed.

The peer checks in with participants receiving the peer support intervention at least once per week to help support PrEP persistence, adherence, and study retention. The peer provides tailored support and referrals (eg, to shelters, needle exchanges, etc) in response to participants’ needs, which may include appointment reminders, direct coordination with the pharmacy and PrEP clinic, and providing transportation.

Based on our focus groups and interviews, participants are asked to select the timing of text messages and content areas in the baseline survey. The survey asks participants to select two ranges of 2-hour time periods within 24 hours (eg, 12 AM-2 AM and 2 AM-4 AM) during which to receive the messages. Each participant receives one daily PrEP adherence reminder, one text containing general PrEP information, and a third text from the content areas of their choice. Participants can select as many content areas as they choose, including local social and health services; general health information; harm reduction information for safer methamphetamine use or injection drug use; bondage and discipline, dominance and submission, and sadism and masochism (BDSM) and toys; PnP; STIs; condoms; anal care; and communication. Participants are also asked in the baseline survey which word they prefer for the daily adherence reminder (ie, “PrEP,” “pill,” “medication,” or other); their preference is programmed into their daily text messages.

Since the messages that were reviewed during our formative work aimed to reduce methamphetamine use and sexual behaviors for HIV and were not specific to PrEP, we created additional categories of messages related to PrEP knowledge and adherence. In response to suggestions from the focus groups and interviews, our text messages include emojis and the PrEP reminder texts are straightforward. In addition, we created a variety of PrEP reminder messages so they can vary throughout study participation as recommended by participants in our formative work. Based on the baseline survey selections, text messages are randomly selected from our text library and programmed to be sent at the chosen times. We chose Telerivet [26] as our text messaging platform because it is affordable and has the functionality for two-way messaging, including emojis in text messages, and for uploading personalized text message schedules for participants.

Finally, our education and recruitment materials were modified based on the suggestions from focus group and interview participants, as shown in Figure 1.