PedMedQA: Comparing Large Language Model Accuracy in Pediatric and Adult Medicine
INTRODUCTION. Large language models (LLMs) have the potential to revolutionize healthcare, including aiding in clinical decision-making. However, recent work suggests that LLM performance in pediatric cases may be weaker than adult cases. A key limitation in evaluating these differences is the lack of pediatric-specific benchmarks, making it difficult to systematically assess how well LLMs generalize to pediatric scenarios.
Authors
- Bertrand Lebouché (ORCID: https://orcid.org/0000-0002-1273-9393)
- Nikhil Jaiswal (ORCID: https://orcid.org/0000-0002-6213-0316)
- Esli Osmanlliu (ORCID: https://orcid.org/0000-0001-5590-8866)
- Dan Poenaru (ORCID: https://orcid.org/0000-0002-6267-6140)
- Yuanchao Ma (ORCID: https://orcid.org/0000-0002-4048-1705)
Institutions
- Montreal Children's Hospital (CA)
- McGill University Health Centre (CA)
- Centre for Advancing Health Outcomes (CA)
- Mila - Quebec Artificial Intelligence Institute (CA)
- Polytechnique Montréal (CA)
- McGill University (CA)
Publication Details
- Journal
- Pediatrics Open Science
- Published
- 2025-05-19
- DOI
- https://doi.org/10.1542/pedsos.2025-000485
- Citations
- 4
- Primary Topic
- Machine Learning in Healthcare
- Type
- article
- Field-Weighted Citation Impact
- 9.92
Funders
- Canadian Institutes of Health Research
- Fonds de Recherche du Québec - Santé