Mining Social Media for Barriers and Facilitators to PrEP Uptake and Persistence - PROJECT SUMMARY Pre-exposure prophylaxis (PrEP) coverage in the United States is low. Currently, only 36% of those eligible are using the medication. To increase uptake and persistence, we must address the widespread uncertainty regarding PrEP among the public (e.g., regarding costs, side effects, details of the drug regimen), particularly among groups at higher risk for HIV. This involves understanding people’s questions and concerns about PrEP and ensuring that guidelines and health communication directly address those uncertainties. Many people seek information about PrEP on the Internet; thus, their online posts/discussions offer insights into knowledge gaps regarding PrEP at the population level. To identify those gaps, the trainee proposes an innovative, community- driven natural language processing (NLP) analysis guided by the co-creating knowledge translation framework to achieve the following specific aims: (1) use NLP to extract user-generated online posts, including inquiries and concerns about initiating or continuing PrEP, from 10 years of Google and Reddit data. (2) Iteratively work with input from the Center for AIDS Research (CFAR) Community Liaison Council (CLC), a diverse group of community members living in the Southeast, the region with the highest HIV prevalence, to qualitatively refine NLP output and identify (A) terminology that the general public uses regarding PrEP, i.e., “PrEP-related lexicon, and (B) barriers and facilitators to uptake and persistence. (3) Compare the results with the CDC 2021 PrEP Clinical Practice Guidelines to identify consistencies and discrepancies between the language people use when discussing PrEP and the information they seek, compared to what is provided in the guidelines. The impact of this study is significant, as it will result in (1) a community-built PrEP-related lexicon, a compilation of terms that the public uses to refer to PrEP and common search queries; (2) an understanding of PrEP-related knowledge gaps and the barriers and facilitators to PrEP uptake and persistence at the population level; and (3) inform community-driven PrEP information dissemination. Additionally, this work will support the training of Ms. Carlson, who is committed to becoming a high-caliber, NIH-funded researcher focused on using community-driven NLP to guide interventions that improve access to HIV care and prevention services. Ms. Carlson’s 3-year training plan includes: (1) increasing capacity in qualitative methods, (2) developing skills in NLP, and (3) cultivating skills in community-based participatory research (CBPR) and co- design methods. The mentor team, Dr. Sales (Primary Sponsor) and Drs. Sarker and Chandler (Co-Sponsors) will provide oversight and guidance to Ms. Carlson during her training in HIV prevention services, qualitative, NLP, CBPR, co-design methods, and subsequent research activities. Ms. Carlson will leverage resources within the doctoral program, Rollins School of Public Health, and Emory University. The candidate, mentorship team, and environment are extremely well-positioned to achieve the proposed research and training aims.
