JMIR Publications

ABSTRACT

Background:

Peer teaching is an established pedagogical approach in medical education, but traditional peer teaching methods face challenges in providing consistent, evidence-based support. Artificial intelligence (AI) tools offer potential to enhance peer teaching by providing students with on-demand access to medical knowledge and diagnostic reasoning support.

Objective:

To evaluate the effectiveness of AI-assisted peer teaching compared to traditional peer teaching in orthopedic clinical education, focusing on knowledge acquisition, clinical skills development, student engagement, and long-term knowledge retention.

Methods:

This historical cohort study compared two consecutive cohorts of medical students at Zhejiang University School of Medicine Affiliated Second Hospital. The control group (taught in 2024, n=96) received traditional peer teaching, while the intervention group (taught in 2025, n=94) received AI-assisted peer teaching with access to Deepseek AI throughout the course. Primary outcomes included post-intervention knowledge scores, knowledge gain, and Objective Structured Clinical Examination (OSCE) performance. Secondary outcomes included student engagement, satisfaction, and 3-month follow-up knowledge retention. Statistical analyses employed independent samples t tests and chi-square tests, with effect sizes calculated using Cohen d.

Results:

The AI-assisted group demonstrated significantly higher post-intervention knowledge scores (79.69±8.41 vs 75.33±9.26, P=.0008, Cohen d=0.492) and greater knowledge gain (11.77±8.80 vs 8.56±9.19, P=.015, d=0.356) compared to the control group. OSCE total scores were significantly higher in the AI-assisted group (80.95±7.57 vs 76.24±9.23, P=.0002, d=0.558). Notably, clinical reasoning skills showed the largest improvement (20.28±3.12 vs 18.06±4.11, P<.0001, d=0.607). The AI-assisted group also demonstrated higher discussion participation (3.43±0.74 vs 3.09±0.85, P=.005, d=0.417), peer teaching satisfaction (3.87±0.69 vs 3.44±0.86, P=.0002, d=0.559), and perceived method effectiveness (3.76±0.68 vs 3.29±0.92, P=.0001, d=0.574). At 3-month follow-up, the AI-assisted group maintained significantly higher knowledge scores (77.36±8.60 vs 72.84±10.42, P=.002, d=0.473). Students in the AI-assisted group used the AI tool an average of 12.30±4.01 times per week, rating its usability at 3.87±0.79, utility at 3.66±0.87, and accuracy at 3.27±0.83 on a 5-point scale.

Conclusions:

AI-assisted peer teaching significantly enhanced knowledge acquisition, clinical skills development, and student engagement in orthopedic education, with effects persisting at 3-month follow-up. The intervention was particularly effective in developing clinical reasoning skills, a critical competency for medical practice. These findings support the integration of AI tools to augment peer teaching in medical education, though further randomized controlled trials are needed to establish causality and optimal implementation strategies.