Exploring the potential of large language models for assessing medication adherence to the ESC heart failure guidelines
Task-specific fine-tuning of LLMs is necessary for optimal performance, and selecting the appropriate LLM for this is important. Without fine-tuning, both general-domain and medical-specific LLMs performed close to random guessing, revealing key limitations in their adaptability to specialized tasks. Medical-specific LLMs showed no clear advantage over general-domain LLMs.
Authors
- Ameen Abu Hanna
- Noman Dormosh (ORCID: https://orcid.org/0000-0003-3152-6200)
- Iacer Calixto
- Machteld Boonstra (ORCID: https://orcid.org/0000-0001-7550-0489)
- Folkert W Asselbergs
Institutions
- National Institute for Health and Care Research (GB)
- Amsterdam Neuroscience (NL)
- Farr Institute (GB)
- GGD Amsterdam (NL)
- Public Health Service of Amsterdam (NL)
- Amsterdam University Medical Centers (NL)
- University College London (GB)
- University of Amsterdam (NL)
Publication Details
- Journal
- JAMIA Open
- Published
- 2025-11-03
- DOI
- https://doi.org/10.1093/jamiaopen/ooaf155
- Primary Topic
- Medication Adherence and Compliance
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Institute for Health and Care Research
- European Commission
- Nederlandse Organisatie voor Wetenschappelijk Onderzoek
- HORIZON EUROPE Framework Programme