Trust in the provider and accurate self-reported PrEP adherence among adolescent girls and young women in South Africa and Zimbabwe: HPTN 082 study

Of 1.7 million new HIV infections globally in 2019, 25% occurred in adolescent girls and young women (AGYW) aged 15–25 years in sub-Saharan Africa [1]. Pre-exposure prophylaxis (PrEP) is highly effective in preventing HIV acquisition when taken consistently [2‐4]. Despite high initial uptake of PrEP by African adolescent girls and young women (AGYW) in demonstration studies, adherence declined within six months of starting PrEP [5‐8]. Poor adherence to PrEP can lead to HIV acquisition. Consequently, an AGYW with HIV faces lifelong antiretroviral therapy (ART) to prevent AIDS and risk of transmission to sex partners and children [3, 9, 10].

The World Health Organization (WHO) recommends quarterly PrEP adherence monitoring for patients on PrEP [11]. The cost of biomarkers of PrEP adherence using plasma, an indicator of PrEP taken within the last 2–3 days, was approximately USD $50–150 per test. Intracellular tenofovir-diphosphate (TFV-DP) in dried blood spots (DBS), an indicator of average PrEP doses per week taken over the last 4–6 weeks, is about USD $200 per test. Because of DBS cost and burden of collection, patient-reported adherence is advantageous, as it is simple to ascertain by questionnaire, is inexpensive, and is a low-burden measure. In studies among adolescents in Africa and the United States (U.S.), patient-reported ART was significantly associated with viral load, suggesting that self-reported PrEP adherence may be used for clinical monitoring in the absence of biomarker assessment [12, 13]. However, self-reported PrEP adherence was substantially higher than biomarker concentrations of PrEP in several placebo-controlled PrEP clinical trials [14‐16].

Trust is necessary for patients to believe that the health provider will act in their best interest in an inherently asymmetrical patient-provider relationship [17]. Hall et al. described that trust is at the core of the doctor-patient relationship, a defining characteristic that gives the relationship meaning, importance, and substance [18]. A literature review of qualitative studies on the role of trust in healthcare systems in sub-Saharan Africa found that trust between patients and providers was a critical element in fostering demand for health services, including HIV services [19]. Respect for individual patient needs and understanding of the patient’s values and beliefs are essential factors in providers acquiring trust and delivering quality healthcare, especially in the context of reproductive health and HIV-related health services in sub-Saharan Africa [20]. Cultural norms about adolescents’ sexual encounters outside of marriage contribute to provider bias and AGYW experiencing discriminatory care [21]. A youth-friendly, nonjudgmental environment that recognizes the dynamic nature of adolescents’ sexuality and self-autonomy will likely foster open communication between provider and patient to formulate patient-centric PrEP adherence strategies jointly [22, 23].

There is, however, limited research on the role of trust in the provider and PrEP adherence. Qualitative research conducted among women in sub-Saharan Africa in placebo-controlled efficacy trials and open-label PrEP demonstration projects revealed that community distrust of the study drug or the study team contributed to low adherence to PrEP [24‐26]. Poor clinician attitude was identified as one reason for PrEP discontinuation in a recent qualitative study that examined PrEP discontinuation in participants who initiated PrEP in HIV comprehensive care centers in Kenya, [27]. Similarly, qualitative studies on preventing perinatal transmission services in Africa described that negative attitudes of care providers and disrespectful treatment of patients by providers resulted in abandoning ART [28]. In VOICE, a placebo-controlled efficacy trial of oral PrEP and vaginal tenofovir gel for HIV prevention among African women, trust in the researchers was recognized as a facilitator by women with biomarker evidence of adherence [29].We did not find published research that examined whether patients are likely to be transparent about their PrEP adherence if they have a trusting relationship with their provider.

In this study, we quantitatively evaluated the association between trust in PrEP providers (research clinicians and counselors) and the concordance between self-reported and biomarker PrEP adherence among AGYW from the HIV Prevention Trials Network (HPTN 082) study. We hypothesized that having a trusting relationship with the PrEP provider was associated with concordant adherence (high self-reported and high biomarker) and concordant non-adherence (low self-reported and low biomarker) and that a low level of trust in the PrEP provider was associated with discordant non-adherence (high self-reported and low biomarker). In addition, we hypothesized that the perception of greater likelihood of HIV acquisition was associated with PrEP adherence [30]. Hence, we examined if perception of greater likelihood of HIV acquisition was associated concordant adherence.

This study provides empirical evidence that AGYW from Africa who had a trusting relationship with their PrEP providers were more likely to have congruence between patients’ self-reported adherence and concentrations of DBS TFV-DP than those without a trusting relationship. The finding is important because to increase PrEP adherence, providers must understand if patients are taking the medication. Since it is expensive and burdensome to collect biomarkers and self-report is easy, inexpensive, and presents little burden to patients, if self-reported adherence is accurate, then with this information, providers can deliver additional support or interventions to those who report low adherence.

We found about one-third of participants were discordant non-adherent, which is comparable to the SEARCH population-based universal PrEP access study in East Africa [41]. There are likely multiple reasons for discordant non-adherence. Even though AGYW who participated in the study were told that they could stop and start PrEP, they may have perceived that not adhering as prescribed would disappoint the study team and were inclined to report higher adherence than those not participating in a study [42]. The participants may also not have fully understood that PrEP providers respected their self-autonomy to choose how often to take PrEP pills, depending on changing ‘seasons’ of HIV acquisition likelihood (i.e., between relationships, the partner was away, or partner living with stable low viral load) [42‐47]. Moreover, AGYW may have taken PrEP pills on an event-driven basis during the time of possible HIV exposure and subsequently reported higher adherence. The World Health Organization has recommended an event-driven 2 + 1 + 1 oral PrEP strategy (two pills taken between 2 and 24 h before sex, then a third pill 24 h after the first two pills, and a fourth pill 48 h after the first two pills) for MSM [48]. Data about the efficacy of event-driven PrEP dosing are not available for women and were not recommended in HPTN 082, although women may still have taken PrEP episodically. Since DBS drug concentration is a cumulative measure of pill taking during the prior 1–2 months, it cannot differentiate how well AGYW adhered during the occasions of possible HIV exposure in the past month; this may at least partly explain discordant non-adherence. On the other hand, due partly to adolescents’ cognitive-developmental stage and the overall challenges in knowing whether one’s partner is likely to be living with HIV, AGYW may not have accurately perceived their HIV acquisition likelihood [49‐51]. However, this study signaled greater odds of concordant adherence among those who perceived a greater likelihood HIV acquisition (p = 0.07). This is consistent with findings of increased PrEP adherence among AGYW [52]. Given only four new HIV diagnoses were observed in the HPTN 082 study over a 1-year period [8], future research is needed to assess if AGYW are following an event-driven (i.e., on-demand) adherence strategy but do not feel empowered to communicate actual adherence patterns to their provider and the reasons for not adhering to taking daily PrEP pills as prescribed by the provider.

There are a few limitations to our study. The ‘trust’ indicator questions used in the paper were collected at month three and not from a scale validated for this population and we evaluated the three questions separately, which may limit validity. Therefore, developing developmentally appropriate trust questionnaire for AGYW will likely lead to more accurately capturing factors that affect having a trusting relationship with the provider. For example, the components of the Sandford Trust in Physician Scale includes questions about quality of care, maintaining personal information confidential, treating participants nonjudgmentally, and prioritizing participant needs over research goals [53].

The provider trust indicators collected in the HPTN 082 trial did not allow us to discern trust issues stemming from encounters with clinicians, other staff, or the healthcare system. Further, the study did not collect data on which aspects of interactions with providers and staff contributed to a low level of trust. However, all three trust indicators signaled high trust in the providers. Finally, adherence goal setting was not captured at each visit to reflect the dynamic nature of the AGYW’s sexual partnerships, which may be better aligned with patient-reported adherence. Goal setting recognizes AGYW’s self-autonomy, which can facilitate open communication with providers, including identifying barriers such as depression experienced by AGYW [35]. To build trust, PrEP providers must treat AGYW nonjudgmentally for their choices, including PrEP use and sexual behavior [21, 54]. A greater understanding of trust factors is essential to create effective tools that improve patient-provider interactions.

Our study confirms that trust in the PrEP provider increases concordance between patient-reported PrEP adherence and biomarker concentration of PrEP for AGYW suggesting that when a patient trusts their provider, patient-reported adherence is an adequate adherence measure. With accurate knowledge of PrEP adherence, providers can deliver additional support or interventions as needed to increase the daily medication. Therefore, education and training to build trusting relationships between providers and AGYW may lead to more accurate reporting of PrEP adherence, and in-turn greater participation of AGYW in formulating an HIV prevention plan.