Investigators at Peking Union Medical College Hospital conduct a national cross-sectional survey among 456 Chinese critical care academic physicians to evaluate Generative AI adoption in standardized residency training. They report that 64.7% use AI clinically and 33.1% in teaching, primarily for content generation and simulation, while highlighting ethical concerns such as over-reliance and data privacy. Respondents endorse enhanced AI training and ethics education to support effective, equitable integration in medical education.

Key points

• Nationwide survey of 456 critical care academic physicians using a Wenjuanxing questionnaire to assess Generative AI usage.
• Logistic regression shows clinical AI use predicts teaching adoption (adjusted OR 1.59, 95% CI 1.06–2.38, p=0.0239).
• Top teaching applications: querying pedagogical content (79.5%) and generating instructional materials (65.6%).

Why it matters: This study clarifies critical care educators’ AI adoption behaviors and ethics priorities, guiding responsible integration of AI in medical training.

Researchers at Sultan Qaboos University's College of Medicine and Health Sciences use the MAIRS-MS questionnaire to evaluate medical students' AI readiness following preclinical exposure, revealing moderate preparedness overall yet significant gaps in cognition, particularly in AI terminology and data science.

Key points

• Students scored lowest in the cognition domain (mean=3.52), reflecting gaps in AI terminology and data-science knowledge.
• Vision domain achieved the highest score (mean=3.90), indicating strong ability to anticipate AI’s applications, risks, and limitations.
• No statistically significant differences in overall AI readiness were found based on gender or prior exposure to AI topics.

Why it matters: Assessing and improving AI readiness among medical students highlights crucial training gaps and guides curriculum enhancements for future healthcare innovations.

A cross-sectional study led by Zagazig University and collaborators conducted a structured online survey of 423 medical students from ten Egyptian universities, assessing their understanding, attitudes, and practices regarding generative artificial intelligence. Findings indicate 61.5% satisfactory knowledge levels, higher scores among males and clinical-phase students, and widespread use of Chat-GPT tools for academic tasks.

Key points

• An 8-question knowledge score, 13-item attitude Likert scale, and 7-item practice frequency scale evaluated generative AI competencies among 423 Egyptian medical students.
• Binary logistic regression revealed male gender (OR=1.87), 6th October University affiliation (OR=3.55), and clinical-phase status (OR=0.54) as significant predictors of satisfactory AI knowledge (p<0.05).
• Students primarily employed Chat-GPT 3.5 (37.1%) and 4 (35.2%) for grammar correction, assignment preparation, research, and idea generation, correlating with knowledge scores (r=0.303, p<0.001).

Why it matters: Understanding medical students’ readiness for generative AI informs curriculum design for future healthcare education and practice.

A recent scoping review by BMC Medical Education illustrates how generative AI, particularly ChatGPT, transforms psychiatric education. By creating case vignettes, simulations, and refined assessments, the study showcases AI’s ability to mirror clinical reasoning challenges. An example includes AI-built illness scripts that supplement traditional teaching, providing nuanced insights for evolving medical training.