TY  - JOUR
AU  - Trivedi, Ritu
AU  - Shaw, Tim
AU  - Sheahen, Brodie
AU  - Chow, Clara K
AU  - Laranjo, Liliana
PY  - 2025
DA  - 2025/3/12
TI  - Patient Perspectives on Conversational Artificial Intelligence for Atrial Fibrillation Self-Management: Qualitative Analysis
JO  - J Med Internet Res
SP  - e64325
VL  - 27
KW  - atrial fibrillation
KW  - conversational agents
KW  - qualitative research
KW  - self-management
KW  - digital health
KW  - patient perspective
KW  - conversational artificial intelligence
KW  - speech recognition
AB  - Background: Conversational artificial intelligence (AI) allows for engaging interactions, however, its acceptability, barriers, and enablers to support patients with atrial fibrillation (AF) are unknown. Objective: This work stems from the Coordinating Health care with AI–supported Technology for patients with AF (CHAT-AF) trial and aims to explore patient perspectives on receiving support from a conversational AI support program. Methods: Patients with AF recruited for a randomized controlled trial who received the intervention were approached for semistructured interviews using purposive sampling. The 6-month intervention consisted of fully automated conversational AI phone calls (with speech recognition and natural language processing) that assessed patient health and provided self-management support and education. Interviews were recorded, transcribed, and thematically analyzed. Results: We conducted 30 interviews (mean age 65.4, SD 11.9 years; 21/30, 70% male). Four themes were identified: (1) interaction with a voice-based conversational AI program (human-like interactions, restriction to prespecified responses, trustworthiness of hospital-delivered conversational AI); (2) engagement is influenced by the personalization of content, delivery mode, and frequency (tailoring to own health context, interest in novel information regarding health, overwhelmed with large volumes of information, flexibility provided by multichannel delivery); (3) improving access to AF care and information (continuity in support, enhancing access to health-related information); (4) empowering patients to better self-manage their AF (encouraging healthy habits through frequent reminders, reassurance from rhythm-monitoring devices). Conclusions: Although conversational AI was described as an engaging way to receive education and self-management support, improvements such as enhanced dialogue flexibility to allow for more naturally flowing conversations and tailoring to patient health context were also mentioned. Trial Registration: Australian New Zealand Clinical Trials Registry ACTRN12621000174886; https://tinyurl.com/3nn7tk72 International Registered Report Identifier (IRRID): RR2-10.2196/34470 
SN  - 1438-8871
UR  - https://www.jmir.org/2025/1/e64325
UR  - https://doi.org/10.2196/64325
DO  - 10.2196/64325
ID  - info:doi/10.2196/64325
ER  -