Integrating AI-assisted world café discussion into clinical reasoning training: an exploratory study in medical education

Background: This study aimed to investigate the implementation pathway and educational effectiveness of an AI-assisted World Café model within the Clinical Comprehensive Reasoning and Ability Training course. Specifically, the study evaluated the model’s potential to enhance students’ acceptance of AI tools, optimize the clinical reasoning learning process, and inform the development of novel teaching assessment practices.
Methods: A single-group pretest-posttest quasi experimental design was employed. Participants comprised 55 third-year clinical medicine students at Kunming Medical University. The instructional design integrated the World Café discussion method, Case-Based Learning, and AI-assisted tools. Students received explicit training in standardized prompt engineering to facilitate structured and critically reflective interactions with AI. Multi source data were collected using perception questionnaires, clinical knowledge tests, and a double scoring protocol involving both AI analysis and blinded human instructor evaluation.
Results: Following the course intervention, students’ acceptance of AI-assisted learning demonstrated a statistically significant increase (p < 0.05), with 81.81% of participants reporting active use of AI for learning purposes. In the evaluation of case discussion records and PPT, AI generated scores exhibited consistency with human instructor scores across multiple assessment dimensions, supporting its reliability as a supplementary formative assessment tool. Although the improvement in pretest to posttest accuracy rates for clinical knowledge did not reach statistical significance, a consistent positive trend was observed. Standardized prompt training substantially enhanced students’ favorable perceptions of AI output quality, with 81.48% of students expressing a positive attitude toward AI generated content.
Conclusion: The AI-assisted World Café model demonstrates effectiveness in enhancing medical students’ acceptance and informed utilization of AI technologies and holds practical value in optimizing the clinical reasoning training process and supporting formative assessment strategies. This model offers a feasible and replicable reference point for the integrative innovation of medical education pedagogies in the era of AI augmented learning.