JMIR Publications

ABSTRACT

Background:

Young adults aged 18–24 in rural communities face compounding barriers to sexual and reproductive health (SRH) information and care, including provider shortages, clinic closures, and privacy concerns inherent to close-knit communities. In the absence of clinical access, many turn to digital platforms where up to 40% of reproductive health content has been identified as inaccurate or misleading. Although AI-enabled conversational health tools (chatbots) offer a scalable and private strategy for SRH information delivery, their real-world impact depends on overcoming critical translational barriers in rural communities, including low institutional trust, data privacy concerns, and systematic underrepresentation of rural users in AI development. SARHAchat is an AI-enabled SRH conversational health prototype developed through prior work; this proposal describes a funded research program to co-design and evaluate it for use with rural young adults.

Objective:

The proposed research pursues three interconnected aims: (1) identify multilevel determinants of trust and acceptability of AI-enabled conversational health tools among rural pregnant-capable individuals aged 18–24 in North and South Carolina; (2) co-design an enhanced SARHAchat prototype with rural community members and stakeholders using human-centered design (HCD); and (3) evaluate the feasibility and acceptability of the enhanced prototype through a 6-month community-based pilot. The overarching goal is to generate generalizable methods and knowledge to guide responsible development and deployment of conversational AI health tools in rural and underserved communities.

Methods:

The proposed research employs a sequential three-phase mixed-methods design. In Aim 1a, we propose conducting 14–32 semi-structured interviews with pregnant-capable individuals (ages 18–24) in Rural-Urban Commuting Area (RUCA) 4–10 counties in NC and SC, followed by rapid thematic analysis to generate a conceptual framework of multilevel trust determinants. In Aim 1b, two structured iterative community feedback sessions with 10–12 stakeholders will guide co-design and adaptation of an enhanced SARHAchat prototype using HCD principles. In Aim 2, a 6-month non-randomized feasibility pilot will recruit 75 pregnant-capable individuals aged 18–24 from RUCA 4–10 zip codes; a sequential explanatory mixed-methods design (quantitative surveys followed by qualitative exit interviews with a purposive subsample of 20) will assess feasibility, acceptability, usability, and implementation outcomes guided by the RE-AIM framework.

Results:

This proposal was funded in February, 2026. Ethical review board approval is pending from the University of North Carolina at Chapel Hill Institutional Review Board. Aim 1a recruitment is planned to commence in June, 2026. Aim 1b community co-design sessions are scheduled for August–December, 2026. The Aim 2 feasibility pilot is projected to open in February, 2027, with all data collection concluding by May, 2027. Full study results are expected to be submitted for publication in Spring 2028.

Conclusions:

This funded proposal addresses a critical equity gap in AI-enabled digital health design by embedding trust as a foundational design principle, rather than a post-hoc outcome, in an AI-enabled SRH tool developed with and for rural young adults. The resulting conceptual framework, community-informed prototype, and feasibility data will support a subsequent R01-scale effectiveness trial and contribute generalizable methods applicable to responsible conversational AI development and implementation across sensitive health domains and underserved populations.