Related Experiment Video
Updated: Jun 12, 2025

05:04
Author Spotlight: Evaluating Clinicians' Adoption of Ultrasound-Guided Vascular Cannulation Through Simulation Training
Published on: August 9, 2024
876
Enhancing Medical Interview Skills Through AI-Simulated Patient Interactions: Nonrandomized Controlled Trial
Akira Yamamoto1, Masahide Koda2, Hiroko Ogawa3,4
1Department of Hematology and Oncology, Okayama University Hospital, Okayama, Japan, Okayama, Japan.
JMIR Medical Education
|September 23, 2024
Summary
Medical students improved interview skills using AI-simulated patients. This AI training showed educational effectiveness and safety, supplementing traditional methods for enhanced clinical practice.
Area of Science:
- Medical Education
- Artificial Intelligence in Healthcare
- Clinical Skills Training
Background:
- Medical interviewing is a crucial clinical skill with limited practical training opportunities in Japanese medical schools.
- Artificial intelligence (AI) applications are expanding in medicine, but its use in medical interview education is underreported.
- There is a need for innovative training methods to enhance medical students' interviewing capabilities.
Purpose of the Study:
- To evaluate the effectiveness of AI-simulated patient interviews in improving medical students' interview skills.
- To assess the impact of large language model (LLM)-based AI patient simulations with feedback on clinical performance.
- To investigate the safety and educational utility of AI in medical interview training.
Main Methods:
- A nonrandomized controlled trial involving fourth-year medical students in Japan.
- An intervention group (35 students) used an LLM-based AI simulation program, while a control group (110 students) did not.
- Primary outcome: Pre-Clinical Clerkship Objective Structured Clinical Examination (pre-CC OSCE) scores for medical interviewing; Secondary outcomes: Simulation-Based Training Quality Assurance Tool (SBT-QA10) surveys.
Main Results:
- The AI intervention group achieved significantly higher medical interview scores on the pre-CC OSCE compared to the control group (p=.01).
- A trend of inverse correlation was observed between the SBT-QA10 scores and pre-CC OSCE performance.
- The AI simulation program demonstrated no significant safety concerns during the study.
Conclusions:
- AI-simulated patient interviews offer a safe and effective supplementary educational tool for medical students.
- The current AI platform shows educational benefits for medical interviewing skills but has limitations in improving nonverbal communication.
- AI-driven simulation should complement, not replace, traditional medical education methods for comprehensive skill development.

