Large language models as medical code selectors: a benchmark using the International Classification of Primary Care
Large language models show strong potential for automating ICPC-2 coding, even without fine-tuning. This work offers a benchmark and highlights challenges, but findings are limited by dataset scope and setup. Broader, multilingual, end-to-end evaluations are needed for clinical validation.
Authors
- Kees van Boven (ORCID: https://orcid.org/0000-0002-4422-9443)
- Luis Fernández López (ORCID: https://orcid.org/0000-0003-4097-6887)
- Vinicius Anjos de Almeida
- Marcelo Finger
- Vinicius de Camargo
- Raquel Gómez-Bravo
- Egbert van der Haring
Institutions
- Radboud University Nijmegen (NL)
- Universidade de São Paulo (BR)
- Institute of Mathematics and Computer Science (MD)
- Nation University (TH)
- Centre Hospitalier Henri Laborit (FR)
Publication Details
- Journal
- JAMIA Open
- Published
- 2026-01-06
- DOI
- https://doi.org/10.1093/jamiaopen/ooag017
- Primary Topic
- Machine Learning in Healthcare
- Type
- article
- Field-Weighted Citation Impact
- 0.00