Back

Artificial intelligence in health professions education: A state-of-the-art meta-review

Hossain, M. M.; Hossain, P.; Roy, T. J.; Das, J.; Tasnim, S.; Ma, P.; Liaw, W.

2025-10-21 medical education
10.1101/2025.10.20.25338371 medRxiv
Show abstract

The growing adoption of artificial intelligence (AI) technologies in healthcare is transforming modern healthcare systems, necessitating current and future healthcare providers to be educated on the meaningful use of AI in their academic and professional activities. Despite an emerging body of literature emphasizing the use of AI in health professions education (HPE) and the availability of multiple reviews on this topic, there is a lack of meta-research evidence that can provide a broader overview of the evidence landscape reported across the existing systematically conducted literature reviews. This meta-review aimed to synthesize evidence on the applications of different AI technologies in HPE, multi-level factors influencing the applications of AI in HPE, and associated outcomes from existing systematically conducted literature reviews (SCLRs). A total of 48 eligible SCLRs were identified from six databases and additional sources, and the synthesized findings suggest emerging use cases of multiple AI technologies among HPE users and institutions, including AI-assisted instructional delivery, augmenting learning sessions, content optimization, and providing feedback. While most reviews reported positive HPE-related outcomes, there are critical challenges at the user and institutional levels, which should be considered for effective AI implementation in HPE. Building AI capacities among HPE users and facilitating AI resources development are critical for AI adoption. This meta-review may inform HPE and broader healthcare communities to advance knowledge and practice on evidence-based AI in HPE settings.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.