High study participation but diverging adherence levels: qualitatively unpacking PrEP use among adolescent girls and young women over two years in Eastern Cape, South Africa

The Community PrEP Study (CPS) was a mixed methods study conducted among 603 AGYW aged 16–25 years in Eastern Cape Province, South Africa from 2018 to 2021 to evaluate the impact of community-based platforms for increasing access and adherence to daily oral PrEP for HIV prevention (Medina-Marino et al., 2021a, 2021b, Mudzingwa et al., 2022a, Mudimu et al., 2022b). AGYW were recruited from two communities (one peri-urban, one rural) in the Buffalo City Metro (BCM) Health District. In 2018, Eastern Cape had an estimated HIV prevalence of 12.1% and an incidence of 1.91% among AGYW 16–25 years (Johnson & Dorrington, 2022; Johnson et al., 2017). As of 2017, 15.6% of youth aged 15–24 in Eastern Cape Province reported early sexual debut (before age 15), and 14.4% of children aged 18 and younger had experienced orphanhood (Simbayi et al., 2017). Most residents speak isiXhosa, and approximately 25% of the population in BCM resides in informal housing (Council, 2017).

Details of the CPS protocol and baseline results have been previously published (Medina-Marino et al., 2021a, 2021b). Briefly, AGYW were recruited from community-based HIV testing service (HTS) platforms, including: (1) home-based HTS, and (2) mobile pop-up sites. Upon receiving negative HIV test results, AGYW were referred to study-established community-based PrEP initiation services. AGYW who met inclusion criteria and consented to be part of the study were randomly assigned to one of three study arms (PrEP medication pick-up only; one-on-one adherence counselling; and group adherence counselling), provided same-day PrEP and invited to return for monthly PrEP refills for 24 months. All participants completed a social-demographic and sexual risk behavior survey at study baseline using an ACASI platform and additional behavioral questionnaires at months 3, 12, and 24. PrEP adherence was measured at months 3, 6, 12, 18, and 24 via collection of dried blood spots (DBS) to detect intracellular tenofovir diphosphate (TFV-DP) levels; TFV-DP levels provide an estimate of cumulative dosing over the past four to six weeks (Anderson et al., 2018). Participants were given refreshments, sanitary pads, and offered HIV and pregnancy testing at monthly PrEP refill visits. Additionally, they received 30 ZAR (~ $2 USD) for each visit where a DBS sample was collected and 100 ZAR (~ $7 USD) for each interview. Other health services such as STI testing and Hepatitis B vaccination were provided at pre-specified study visits (Medina-Marino et al., 2021a). This study is reported in accordance with the consolidated criteria for reporting qualitative research (COREQ) guidelines (Supplemental Table 1) (Tong et al., 2007).

A total of 89 participants were purposively recruited to complete qualitative interviews to further explore their experiences initiating and adhering to PrEP throughout the study. This manuscript examines a subset of (n = 22) these participants who all had high rates of return for monthly PrEP refills, and attended at least three of five study visits in which DBS samples were collected. We then stratified these 22 study participants into two groups: (1) those (n = 7) with high levels of TDF-DP in DBS (≥ 700 fmol/punch) averaged across all their DBS measurements during the study (“High TFV-DP”), and (2) those (n = 15) with low to moderate levels of TDF-DP in DBS (< 700 fmol/punch) averaged across all their DBS measurements during the study (“Low TFV-DP”). This threshold (≥ 700 fmol/punch) correspond to an average of four doses per week in males in the USA (Anderson et al., 2018), the level associated with 100% protection among men who have sex with men (Grant et al., 2014), and has been previously been used as a threshold for high adherence among South African AGYW (Celum et al., 2020). Participants for this specific qualitative subsample were sampled across the entire study, including participants from the control arm (medication pick-up only) and intervention arms (individual or group adherence counselling); participants were also sampled across multiple interview categories (Supplemental Table 2).

The Information-Motivation-Behavioral skills model of behavior change (IMB) (Fisher et al., 2006, 2008) was used to conceptualize and develop questions for the interview guides (Table 1). IMB-informed questions were used to gather information on AGYW-specific phenomena and factors that influence uptake and adherence such as: (a) information, misinformation and concerns, (b) personal and social motivational factors and (c) behavioral skills, tools and strategies needed for both PrEP uptake and adherence. Interviews explored individual PrEP narratives and topics such as enablers and barriers to PrEP adherence, study visit experiences, social support for PrEP use, self-perceived risk over time, and what participants did with unused PrEP. Participants were telephonically approached and invited for an interview. Interviews were conducted in isiXhosa by trained research assistants highly knowledgeable of study community socio-cultural dynamics. Interviews lasted approximately 20–50 min in duration and were completed between September 2019 and April 2021. Interviews were audio-recorded, translated and transcribed in English, and reviewed for quality by qualitative research staff.

For qualitative analysis, transcripts were coded in Dedoose (2022) following an inductive approach using previously described methods (Mudzingwa et al., 2022a). Codebooks were developed using an iterative approach drawing from review of transcripts and research objectives (Medina-Marino et al., 2021a), considering critical constructs and domains of the IMB model. Code domains covered PrEP knowledge and misconceptions, personal and social motivators for PrEP uptake and adherence, PrEP routines and reminders, PrEP pick-up behaviors and study experiences. The coding team consisted of five researchers [blinded for review] who met weekly to discuss findings, resolve discrepancies, and establish consensus. Given the limited theoretical framing of AGYW PrEP use, we selected a thematic analytical approach. For those with low levels of TFV-DP, code reports containing relevant transcript excerpts were compiled for the following key codes: “Discarding PrEP”, “Sharing PrEP”, “Stockpiling PrEP”, “Selling PrEP”, “Non-PrEP related reasons for study visits”, and “PrEP Recommendations”. Research team members wrote summary memos highlighting themes that emerged from each code report. Given the relatively small sample of interviewed participants with high levels of TFV-DP that met our inclusion criteria, a case approach was followed in which each participant’s transcript was reviewed in depth and summarized in an analytical matrix. The matrix included columns to summarize each High TFV-DP participant’s motivation for using PrEP, social support, PrEP adherence enablers, PrEP adherence barriers, and recommendations for future support interventions. Results from these analyses were discussed by the research team during weekly calls and at a meeting with the larger protocol team (listed co-authors) for critical input.

Descriptive statistics for the subset of AGYW who completed qualitative interviews was analyzed using Stata 17 (StataCorp, 2021); given the purposive sampling, no additional statistical analysis was conducted. TFV-DP levels in DBS samples at M3, M6, M12, M18, and M24 were graphically displayed to visualize PrEP adherence over time in the study for each qualitative participant included in this subset (Fig. 1; Supplementary Figure S1). The limit of detection for TFV-DP in DBS samples was 16.6 fmol/punch; any measurement below this was counted as zero.

Ethics approval for this research was obtained from the ethics review committee at the University of Cape Town (HREC 289/2018). Separate approval was given by the Eastern Cape Provincial Department of Health to conduct the research in Eastern Cape Province. Written informed consent was obtained prior to participant initiation in the study and prior to each interview. A ministerial waiver of consent was obtained for participants 16–17 years old, with parental consent obtained upon participant request. To ensure confidentiality of study participants, participants’ quotes are represented with TFV-DP level in DBS (High or Low), study site (peri-urban or rural), and age.

Participants’ demographic and behavioral characteristics are summarized in Table 2; these should be interpreted with caution given the small sample size. Those with low levels of TFV-DP were younger (median age 17), were more likely to be attending school and recruited from a mobile pop-up HTS site. A similar percentage from both groups (~ 60%) reported being sexually active during the study. All AGYW with high levels of TFV-DP disclosed their PrEP use to either a family member, partner or friend. Only 67% of those with low levels of TFV-DP disclosed their PrEP use to a family member, partner or friend. Of note, those with high levels of TFV-DP also reported more total PrEP disclosure events and more supportive reactions when they disclosed their PrEP use. Both groups had few participants (n ≤ 2) who self-reported low adherence to PrEP. As measured by TFV-DP in DBS across the two-year study, the average adherence for those with high levels of TFV-DP was over tenfold higher than for those with low levels of TFV-DP (1046 fmol/punch vs. 95 fmol/punch; Fig. 1, Supplementary Figure S1).

AGYW with low levels of TFV-DP highlighted barriers to PrEP use including negative talk casting doubt on PrEP, misconceptions, and mistrust of PrEP by members of their social networks and community. For example, one woman said that she considered discarding PrEP because of “the way people were talking about it – it was like a disease pill” (Low, peri-urban, age 16). Another described reasons why others were not using PrEP: “not trusting this PrEP if it’s working or maybe being told bad things about PrEP by the friends but not wanting to come and cancel herself here at the study” (Low, peri-urban, age 23). Other Low TFV-DP participants described a lack of support from family or partners, feeling judged by others, unresolved side effects which were not discussed with study staff, and an overall lack of interest in PrEP. One described how her brother (her “support system” in life) accused her of being HIV positive because she was using PrEP: “when I started using PrEP he [brother] use to swear at me saying that I’m using ARVs and I’m hiding them” (Low, rural, age 16). Low TFV-DP participants described how their social networks were unfamiliar with PrEP, questioning its purpose and effects on the body.

Low TFV-DP participants from the rural study site described more examples of being in “desperate” situations, with greater lack of resources than those from the peri-urban site, as influencing their study visit attendance. Rural participants shared how some AGYW came “when they are desperate of money that’s when they decide to come and take PrEP” (Low, rural, age 16); “they are desperate for that R30 [$2 USD]” (Low, rural, age 23). Another said, “maybe there is nothing to eat at home … maybe will start thinking that I can come to PrEP to collect R30 [$2 USD]… or maybe she is craving for chips” (Low, rural, age 16). Peri-urban participants were also motivated to come due to money, snacks, and sanitary pads, but those participants did not elaborate on what difference that might make in their lives the way some rural participants did.

AGYW described receiving better health services at the study site (e.g., friendlier staff, shorter queues) compared with public clinics as a significant motivator. Others shared how the group or one-on-one PrEP adherence sessions motivated them to come for PrEP study visits. Notably, many shared that the most valuable part of the PrEP adherence sessions was not the PrEP-specific education they received, but rather the opportunity to learn career and goal-setting skills (e.g., interviewing, CVs) and to discuss interpersonal problems they were facing at home.

Some AGYW also described peer pressure from friends to visit the site, and social desirability of not wanting to disappoint study staff by missing study visits. For example, one participant described not feeling at risk for HIV, but still wanting to come to the study: “Because I did not have a boyfriend, so I did not see the reason to take it. But I was continuing coming […] I didn’t want to skip my date here. I wanted to continue with my study nicely” (Low, peri-urban, age 23).

AGYW with low levels of TFV-DP described what happened to unused PrEP, including stockpiling, discarding, selling, and sharing PrEP (see Table 3 for representative quotations for each pattern).

In contrast to those with low levels of TFV-DP in their DBS, participants who had high levels of TFV-DP described what motivated them to adhere to PrEP over a two-year time period (see Table 4 for representative quotations).

Both those with low and high levels of TFV-DP in their DBS shared multiple suggestions for how to improve PrEP uptake and adherence in their communities. Among AGYW with low TFV-DP, many recommended increasing education in schools and clinics to counteract PrEP misconceptions: “Go to schools and tell them more about PrEP and what kind of a pill it is. Because most people believe that PrEP is something else. Or when you go to the clinic there could be people there to explain to people about PrEP and how to take PrEP” (Low, rural, age 16). Some Low TFV-DP participants thought longer-acting PrEP methods like an injection or once-monthly pill would improve use of PrEP, such as this woman: “If they can be injected with PrEP they can come. They can’t say I took it while they did not when they are injected” (Low, rural, age 16).

Those with high levels of TFV-DP recommended improving relational support from family members and friends for PrEP use and broad campaigns to improve PrEP knowledge in communities. One young women with high levels of TFV-DP suggested that for someone to succeed in using PrEP, “a person must consult her parents so that her parents can understand and also her friend so that they may do the same thing” (High, rural, age 25). To increase knowledge about PrEP, other High TFV-DP participants recommended PrEP awareness campaigns to clearly explain what PrEP is, how it works, and who it is for. For example, one participant suggested using posters, community committees, and participants “that are accurate on PrEP” who can explain the benefits of PrEP to others and counteract “some people misinterpret PrEP” (High, peri-urban, age 23). While High TFV-DP participants said they preferred picking up PrEP at the study site due to friendly staff and quick pick-up, most said they would still pick up PrEP at the public clinic in the future.

Another High TFV-DP participant recommended that her peers need to overcome feelings of shame associated with taking PrEP: “A person must not have pride and do what is right for herself. [,,,] They should not be ashamed; they should put pride aside. A person doesn’t know what tomorrow brings. This is the way a person should be safe" (High, rural, age 25). For her personally, the one-on-one counseling sessions at the study site helped to raise her own awareness of peer pressure and “focus on what I do and use my time for the right things” (High, rural, age 25).