October 13, 2025CureusOpen Access

Performance of ChatGPT and Large Language Models on Medical Licensing Exams Worldwide: A Systematic Review and Network Meta-Analysis With Meta-Regression

View Full Paper

Ask AI

Bookmark

Authors

AKAlousious KasaggaPeking University ASAmy R. SapkotaUniversity of Maryland, College Park GCGichin ChangaramkumarathChongqing Medical University

Discussion

Member takes

Implication

Systematic review found multiple large language models performed well on medical licensing exams, indicating potential in education and decision support.

Key Points

Most large language models achieved over 60% accuracy on medical licensing exams, with GPT-o1 leading at 95.4%.
The study analyzed 120 evaluations across 10 exam systems and identified significant performance variations by language and model.
Meta-regression revealed model type and language significantly impacted exam performance, with lower scores in Chinese and Japanese exams.
Heterogeneity among results was largely explained by model version and exam system, highlighting the importance of standardized assessments.

Cite This Study

Kasagga et al. (2025) studied this question.

synapsesocial.com/papers/68ec51df42911f61ef8b201b https://doi.org/10.7759/cureus.94300

View Full Paper

Ask AI

Bookmark

Implication

Systematic review found multiple large language models performed well on medical licensing exams, indicating potential in education and decision support.

Performance of ChatGPT and Large Language Models on Medical Licensing Exams Worldwide: A Systematic Review and Network Meta-Analysis With Meta-Regression

Discussion

Member takes

Implication

Key Points

Cite This Study

Performance of ChatGPT and Large Language Models on Medical Licensing Exams Worldwide: A Systematic Review and Network Meta-Analysis With Meta-Regression

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider