Mechanisms of Action for mobile SMS PrEP adherence intervention

ABSTRACT Effectiveness of pre-exposure prophylaxis (PrEP) for HIV prevention methods currently available to cisgender women hinges on daily medication adherence, yet >50% of women discontinue within the first month of use. Digital technologies to improve adherence to HIV treatment and prevention interventions are rapidly expanding in the WHO African Region, despite mixed effectiveness findings. Systematic reviews note the near absence of articulated mechanisms of action (MoA) for HIV digital health strategies which limits translation into practice and thus population-level impact. Tools from behavioral health research and implementation science are useful to elucidate, articulate, and test MoA for digital strategies. There is substantial pragmatic value in understanding MoAs and moderators for how digital strategies work, including adapting intervention message content balance to emphasize the most potent mechanisms and tailoring content to specific populations. We propose an R21 to utilize data from the mWACh PrEP study (R01NR019220), a hybrid effectiveness-implementation trial that tests the impact of a two-way, interactive, SMS-based platform on PrEP adherence and maintenance (continuing PrEP refills and use) during pregnancy and postpartum in Kenya. The parent study includes 5 study sites based at public sector antenatal clinics in a setting with high HIV prevalence. To date N=600 (100% of planned enrollment) women are enrolled (1:1 randomization) to be followed until 9 months postpartum. The trial utilizes the Information-Motivation-Behavior (IMB) model, which specifies 1) knowledge acquisition, 2) motivation, and 3) behavioral skills as potential MoA. Within this R21, we propose to test hypothesized MoA for the mWACh PrEP digital strategy, assess specific context variables as moderators, and characterize experiences of MoA activation. In Aim 1, we will test whether knowledge acquisition, motivation, and behavior skills are MoA through which mWACh PrEP influences PrEP adherence and maintenance. We will consider both self-reported adherence, hair concentrations of PrEP, and PrEP continuation. We will employ deductive analysis using structural equation modeling and inductive analysis using coincidence analysis. In Aim 2, we will elucidate whether MoA are moderated by mental health, relationship climate, sexual behavior, access to care, and medical history, using structural equation modeling. Finally, in Aim 3, we will characterize how and why MoA are activated by assessing the experiences of women, SMS content, and health provider insights. We will conduct in-depth interviews with women, analyze two-way messages between women and providers, and engage facility and study staff through a workshop to understand activation of MoA. Our study is designed to move the field from theoretical to empirical knowledge of how and why women maintain and sustain PrEP use behaviors. Our findings will have pragmatic utility beyond HIV prevention in pregnancy for adapting digital message content to emphasize potent mechanisms and tailor message content to diverse populations.