Accepted for/Published in: Journal of Medical Internet Research
Date Submitted: Aug 21, 2025
Date Accepted: Jul 31, 2026
Development and Usability Assessment of a Health Education Conversational Agent for Gastric Cancer patients: An Action Research Approach
ABSTRACT
Background:
Health education of gastric cancer patients has been delivered face-to-face which might cause challenges in understanding the information. Conversational agents have increasingly been adopted for healthcare use to provide innovative solutions for patient education.
Objective:
The objective of this study was to develop a health education conversational agent and test its performance and usability in promoting gastric cancer patients’ health literacy.
Methods:
The artificial intelligence-guided conversational agent was developed based on the OpenMEDLab2.0 foundation model and the Retrieval-Augmented Generation (RAG) architecture. An action research approach was adopted to implement this intervention. A four-step cycle including diagnosis and planning, act and implementation, evaluation, and reflection were undertaken in a gastric cancer center in China. Quantitative and qualitative methods were used to understand the barriers and enablers of the novel health education intervention.
Results:
Totally, 44 patients and 20 stakeholders (13 clinical nurses, 3 nursing managers, 2 doctors and 2 health experts) participated. Favorable outcomes in terms of accuracy, safety and usability were achieved. Overall accuracy rate of 94% and knowledge hit rate of 95% were reported. The mean chatbot usability questionnaire score was 91.9 (SD 3.6). Three themes of the real-world user experiences were identified: perceived usefulness, ease of use and intention to use.
Conclusions:
Our health education conversational agent is a feasible innovation for gastric cancer patients. The intervention might need to be optimized and validated by larger group of patients. Further testing on patient outcomes might confirm the value of artificial intelligence in healthcare for the future.
Citation
Per the author's request the PDF is not available.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.