Related Experiment Video
Updated: Mar 25, 2026
05:04
Author Spotlight: Evaluating Clinicians' Adoption of Ultrasound-Guided Vascular Cannulation Through Simulation Training
Published on: August 9, 2024
1.8K
Scaling Multimodal Agentic AI in Medical Education: Multisite Cross-Sectional Study of Simulation Effectiveness in
Chris Jacobs1,2, Hans Johnson3, Kirsty Brownlie3
1Department of Psychology, University of Bath, Bath, England, United Kingdom.
JMIR Formative Research
|March 23, 2026
Summary
Conversational artificial intelligence (AI) shows promise for medical training, excelling in clinical accuracy but needing improvement in realistic dialogue. It is best suited as a supplementary tool for skills development.
Area of Science:
- Medical Education Technology
- Artificial Intelligence in Healthcare
- Clinical Skills Training
Background:
- Conversational AI offers solutions to traditional medical consultation training challenges like cost and scheduling.
- Limited evidence exists on medical professionals' perceptions of AI patient interactions and their educational value.
Purpose of the Study:
- Evaluate perceptions of conversational AI patient simulations in primary care consultation training.
- Assess functional fidelity, conversational realism, educational value, and implementation readiness of AI simulations.
Main Methods:
- Cross-sectional evaluation at a UK medical school with students and general practitioners.
- Participants used the SimFlow conversational AI system for standardized scenarios.
- Multidomain questionnaire assessed AI realism, content, educational value, feedback, and usability; data analyzed statistically.
Main Results:
- High ratings for medical content (97.8% plausible); positive educational value (median 4.0).
- Moderate scores for AI realism (median 3.0), with higher ratings from experienced AI users.
- Moderate-to-strong agreement between individual and group rankings; qualitative themes included authenticity, limitations, potential, and implementation.
Conclusions:
- Conversational AI exhibits strong functional fidelity (clinical accuracy) but has limitations in conversational realism.
- AI shows potential as a supplementary tool for clinical skills training, not high-stakes assessment.
- Future development should focus on dialogue naturalness and feedback mechanisms.