Feasibility and Acceptability of Sarhachat™: An Ai-Powered Chatbot for Sexual and Reproductive Health Education and Decision Support

Abstract

Objective:This study examines the feasibility and acceptability of SARHAchat™, an AI-powered chatbot designed to improve access to evidenced-based sexual and reproductive health (SRH) education, counseling, and contraceptive decision-making in underserved U.S. regions.Methods:SARHAchat™ was developed following the Ottawa Decision Support Framework and International Patient Decision Aid Standards. An advisory panel provided iterative feedback, and the chatbot was beta-tested. In Phase One, the chatbot was pilot-tested with 44 participants, assessing user experience and SRH decision support. A mixed methods post-intervention survey, incorporating the Person-Centered Contraceptive Counseling (PCCC) measure, evaluated acceptability and practicality.Results:32 participants completed the survey, with SARHAchat™ receiving a 47% "excellent" score on the PCCC measure, with respect for user preferences rated highly and lower scores for decision support comprehensiveness . Thematic analysis indicated high user satisfaction, privacy appreciation, and perception of empathy, with some limitations in personalization.Conclusion:The findings support SARHAchat™ as a feasible and acceptable tool for SRH counseling and decision support. Privacy and patient-centered interactions enhance its utility, though further refinement for personalized responses is necessary.Innovation:SARHAchat™ leverages generative-AI to deliver unbiased, empathetic, accurate and accessible SRH care, addressing misinformation and healthcare inequities by meeting informational needs within underserved populations.

Note:
Funding Information: No funding or grant was received for this study.

Declaration of Interests: Kandyce Brennan, Tianlong Chen, and Maureen Baker declare not conflicts of interest.

Ethics Approval Statement: The University of North Carolina at Chapel Hill Institutional Review Board approval IRB #24-0503.

Keywords: Artificial intelligence, patient engagement, decision aid, patient education, sexual and reproductive health, misinformation