Designing for Supportive Accountability: Using Conversational Agents to Sustain Patient Engagement in PTSD

Researcher
Saeed Abdullah

Sponsoring Agency
National Science Foundation

Summary

Approximately 8 million adults in the United States suffer from Post-traumatic stress disorder (PTSD) in a given year. It causes staggering individual and societal costs including higher mortality risk and significant national health care expenditure. While evidence-based PTSD treatments exist, their dissemination is limited due to logistical issues (e.g., cost and lack of trained professionals). To address these issues, recent studies have used mobile technologies for intervention and treatment delivery. However, ensuring engagement and adherence to these technologies remains a serious challenge. This project aims to address these engagement and adherence issues by developing a finite-state conversational agent (CA) for therapeutic content delivery and day-to-day illness management. Findings from this project can lead to more effective methods for keeping people engaged.

The proposed conversational agent will provide interactive psychoeducation, assessment tools, and personalized illness management strategies to patients with PTSD. It will support turn-taking and interactivity, which will result in active information flow in both ways rather than just passive delivery of information content. The implementation of the CA will involve dialogue management, handling user input for branching, and delivering appropriate message for a given context. For sustaining longitudinal user engagement, it will leverage the Supportive Accountability model with a specific focus on maintaining social presence and process accountability through personalization, goal setting, and contextualized support. The CA will be developed as a stand-alone module that can be incorporated into a mobile app. This will streamline the deployment of the system to patients. The project will collect real-world data to determine domain specific usability issues and information needs. It will also perform a systematic evaluation to understand the causal pathway between a CA and patient engagement. That is, it will assess what components of a CA work for patients, in what contexts, and for how long. Publications and source code resulting from the project will be publicly available.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

Research Area
Artificial Intelligence and Big Data
Health and Bioinformatics
Human-Computer Interaction

Term
 -