Can ChatGPT Teach Tendon Repair Suturing? A Comparison With Video Instruction in a Simulation-Based Environment: A...

Area of Science:

• Medical Education
• Surgical Simulation
• Artificial Intelligence in Healthcare

Background:

• Novice surgical trainees require effective methods for learning complex techniques.
• Traditional video-based instruction is a common tool in surgical education.
• Conversational AI, like ChatGPT, presents a novel approach to guided learning.

Purpose of the Study:

• To compare the efficacy of ChatGPT-guided versus video-based instruction for the Modified Kessler tendon repair technique.
• To assess technical performance, satisfaction, and self-confidence in novice medical trainees.
• To evaluate the potential of AI in surgical skills training.

Main Methods:

• A quasi-experimental study involving 33 novice medical students.
• Two self-directed learning groups: ChatGPT-Tutored and Video-Tutored.
• Training involved 5 hours of practice on a tendon simulator with respective instructional methods.

Main Results:

• The Video-Tutored group achieved significantly higher Objective Structured Assessment of Technical Skills (OSATS) scores (p=0.0004).
• Competency rates were higher in the video group (52.6%) compared to the ChatGPT group (10.5%) (p=0.002).
• Video instruction led to significantly greater satisfaction (p<0.0001) and self-confidence (p<0.0001).

Conclusions:

• Video-based instruction is more effective than ChatGPT for teaching tendon repair to novice trainees.
• ChatGPT guidance resulted in lower technical performance, satisfaction, and confidence.
• Conversational AI shows promise as an adjunct tool, with potential for blended learning approaches in surgical education.