Characterizing PrEP Need and Use in Eswatini: Individual and Programmatic Insights Through Data Triangulation and Novel Epidemiological Methods

About This Grant

PROJECT SUMMARY/ABSTRACT Background. In Eswatini, adult HIV prevalence is over 31%, and national adult HIV incidence is the highest in the world at 7.77 new infections per 1,000 persons annually. While the country has almost met UNAIDS’ 95-95- 95 goals for epidemic control, HIV incidence is declining at a glacial pace, pointing to the need for comprehensive HIV prevention programs. Eswatini’s scale-up of HIV pre-exposure prophylaxis (PrEP) began in 2018, with nearly 100,000 individuals initiating PrEP in-country since. The proposed study will develop better methods to determine who could most benefit from PrEP (i.e., PrEP need) in Eswatini, and whether that need is being met at programmatic and individual levels, to inform future implementation of HIV prevention programs. Aims. In line with the NIMH Division of AIDS Research’s strategic priorities, this study will 1) compare models for predicting who could most benefit from PrEP in Eswatini; 2) compare novel PrEP-to-Need Ratios (PnRs) to original PnRs at the facility level; and 3) examine individual-level prevention-effective persistence and trajectories over time. Approach. This research will use data from the 2021 Eswatini Population-based HIV Impact Assessment (N=11,199) and Eswatini’s HIV-1 Recent Infection Surveillance (N~32,780) to build machine learning models to predict both recent and long-term but newly diagnosed HIV infections, providing insight into vulnerable subpopulations most in need of PrEP. Using routinely collected health facility data (N=150 facilities), PnRs will be created that account for heterogeneity in infection timing (i.e., recent vs. long-term), as well as subpopulation variability. Facilities will be compared over time. Lastly, using data from the DYnamics of Contraception in Eswatini (DYCE) Study (N=321), changes in PrEP need over time, relative to changes in use of PrEP and other prevention methods, will be assessed using an operationalization of prevention-effective persistence. Latent class analysis will be used to detect actionable differences between classes of prevention-effective persistence trajectories. Each aim is designed to identify gaps in HIV prevention efforts and guide the continued scale-up of PrEP in Eswatini to ultimately reduce HIV incidence. Training. Ms. Wallach’s training plan leverages her quantitative analysis and HIV research experience to advance her skills and launch her career as an independent HIV investigator. Her training goals are to develop a deep, nuanced understanding of PrEP need, gain experience working with and triangulating various data sources, gain skills in machine learning methods and latent class analysis, gain experience with longitudinal data analysis, and build capacity for effective research communication and dissemination. Under the guidance of her Primary Sponsor, Ms. Wallach will receive tailored mentorship from a team of experienced HIV researchers who work with the included data sources, conduct PrEP-related research, and have expertise in the advanced epidemiologic methods herein. Training will occur at Columbia University, a high-caliber institution with specialized research programs, including in infectious disease epidemiology, and that houses ICAP, a global health and HIV implementation and research organization.