Designing Deterministic Content for an AI Chatbot to Support Patients in Cardiovascular Self-Management: Insights from Qualitative Interviews.
7 affiliations
- 1University of Colorado School of Medicine, Aurora, CO, USA.
- 2Department of Community and Behavioral Health, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora, CO, USA.
- 3Adult and Child Center for Outcomes Research and Delivery Science, University of Colorado School of Medicine, Aurora, CO, USA.
- 4Department of General Internal Medicine, Denver Health and Hospital Authority, Denver, CO, USA.
- 5MHealth Impact Lab Community, Colorado School of Public Health, Aurora, CO, USA.
- 6Clinic Chat, LLC, Denver, CO, USA.
- 7Institute for Health Research, Kaiser Permanente Colorado, Aurora, CO, USA.
Abstract
Text message-based chatbots powered by artificial intelligence (AI) show promise for supporting self-management, a critical component of chronic cardiovascular disease care. However, limited research has examined how these chatbots should be designed to resonate with patients from different backgrounds. This study explored how text-based, AI-driven chatbot systems can be optimized to encourage self-management behaviors among Spanish and English-speaking patients. Participants were recruited from two clinical sites including an integrated safety-net health system and a Federally Qualified Health Center: Denver Health and Salud Family Health Centers. Semi-structured interviews were conducted, and qualitative coding continued until thematic saturation was reached. Interview topics included awareness and perceptions of chatbots, reactions to a demonstrated prototype, usability, trust and credibility, privacy concerns, preferences for personalization and messaging logistics, goal setting, and anticipated engagement with the program. We interviewed 10 patients of whom 5 were Spanish-speaking and 5 were English-speaking. Three major themes emerged: (1) individualized and personable support, (2) a motivating and simple self-management tool, and (3) an integrated informational gateway. Participants emphasized the importance of tailoring messages to personal circumstances such as age, income, retirement status, and preferred language. A conversational, human-like tone without medical jargon was viewed favorably. Participants valued goal setting with consistent follow-up and preferred a user-friendly, engaging design. Integration with existing healthcare systems and providers was identified as essential for fostering trust and credibility. Tailoring chatbot design to users' preferences and life contexts, combined with ease of use, may enhance engagement and support behavior change. Integrating chatbots within established healthcare systems may further build trust and credibility, representing an important and underexplored design consideration.
Related
- Cancer-related loneliness management strategies: qualitative interviews of British adults
- Surviving cancer, managing the heart: Awareness of potential cardiovascular toxicity and barriers to cardiovascular self-management among women with breast cancer: A qualitative descriptive study
- TVET teachers’ perceptions of ICT integration in resource-constrained teaching contexts: a qualitative thematic analysis of interviews from Hebei, China
- “It’s not necessarily science based”: understanding organ utilisation decision-making in cardiothoracic transplantation in the UK through qualitative interviews
- Caregivers’ experiences of PM2.5 prevention for preschool children in Thailand: a qualitative study with focus group interviews
BibTeX
@article{fouts2026designing,
title = {{Designing Deterministic Content for an AI Chatbot to Support Patients in Cardiovascular Self-Management: Insights from Qualitative Interviews.}},
author = {Delaney Fouts and Makala D Carrington and Catia Chavez and Emily Kosirog and Amy D Lu and Pamela N Peterson and Cathryn Perreira and Lisa M C Sandy and Joy Waughtal and Sheana Bull and P Michael Ho},
journal = {Patient preference and adherence},
year = {2026},
month = nov,
doi = {10.2147/ppa.s628813},
url = {https://europepmc.org/article/MED/42999827},
note = {CC BY-NC},
}
RIS
TY - JOUR
TI - Designing Deterministic Content for an AI Chatbot to Support Patients in Cardiovascular Self-Management: Insights from Qualitative Interviews.
AU - Delaney Fouts
AU - Makala D Carrington
AU - Catia Chavez
AU - Emily Kosirog
AU - Amy D Lu
AU - Pamela N Peterson
AU - Cathryn Perreira
AU - Lisa M C Sandy
AU - Joy Waughtal
AU - Sheana Bull
AU - P Michael Ho
PY - 2026
DA - 2026/11/03
JO - Patient preference and adherence
DO - 10.2147/ppa.s628813
UR - https://europepmc.org/article/MED/42999827
AB - Text message-based chatbots powered by artificial intelligence (AI) show promise for supporting self-management, a critical component of chronic cardiovascular disease care. However, limited research has examined how these chatbots should be designed to resonate with patients from different backgrounds. This study explored how text-based, AI-driven chatbot systems can be optimized to encourage self-management behaviors among Spanish and English-speaking patients. Participants were recruited from two clinical sites including an integrated safety-net health system and a Federally Qualified Health Center: Denver Health and Salud Family Health Centers. Semi-structured interviews were conducted, and qualitative coding continued until thematic saturation was reached. Interview topics included awareness and perceptions of chatbots, reactions to a demonstrated prototype, usability, trust and credibility, privacy concerns, preferences for personalization and messaging logistics, goal setting, and anticipated engagement with the program. We interviewed 10 patients of whom 5 were Spanish-speaking and 5 were English-speaking. Three major themes emerged: (1) individualized and personable support, (2) a motivating and simple self-management tool, and (3) an integrated informational gateway. Participants emphasized the importance of tailoring messages to personal circumstances such as age, income, retirement status, and preferred language. A conversational, human-like tone without medical jargon was viewed favorably. Participants valued goal setting with consistent follow-up and preferred a user-friendly, engaging design. Integration with existing healthcare systems and providers was identified as essential for fostering trust and credibility. Tailoring chatbot design to users' preferences and life contexts, combined with ease of use, may enhance engagement and support behavior change. Integrating chatbots within established healthcare systems may further build trust and credibility, representing an important and underexplored design consideration.
SN - 1177-889X
ER -