An AI-Based Chatbot for Doxycycline as Post-Exposure Prophylaxis Among Men Who Have Sex with Men

Miriam Hospital, Providence RI

Investigators

Abstract

Bacterial sexually transmitted infections (STIs)—including syphilis, gonorrhea, and chlamydia—are increasing in the United States, with disproportionately high rates among men who have sex with men (MSM). These infections can lead to serious health complications such as neurological and cardiovascular syphilis, infertility, pelvic inflammatory disease, and increased HIV transmission. Doxycycline, a widely available and safe antibiotic, has shown promise as post-exposure prophylaxis (doxy-PEP) to prevent bacterial STIs. Motivational interviewing (MI) is effective in promoting HIV prevention behaviors, especially when paired with the Transtheoretical Model (TTM) to assess readiness for change and tailor messaging. However, traditional MI requires multiple sessions and is resource-intensive. This pilot study aims to overcome those limitations by developing and testing SAI (STI Prevention AI-Chatbot), an AI-based chatbot powered by the latest ChatGPT model. In Specific Aim 1, SAI will be tailored to assess STI risk and TTM stage, deliver stage-appropriate education, and use MI-based counseling to promote doxy-PEP uptake among MSM. In Specific Aim 2, we will conduct an open pilot trial with 25 adult MSM recruited via social media to evaluate SAI’s accuracy, usability, appropriateness, and user satisfaction, followed by iterative refinement. If successful, this study will demonstrate the feasibility of using AI to scale behavioral interventions for STI prevention and support broader implementation of doxy-PEP strategies among high-risk populations. Modified Project Title An AI-Based Chatbot for Doxycycline as Post-Exposure Prophylaxis Among Men Who Have Sex with Men