Using Large Language Models to Detect and Understand Drug Discontinuation Events in Web-Based Forums: Development and Validation Study
This study demonstrated the effectiveness of open- and closed-source LLMs, such as GPT-4o and BART, for detecting DDEs and their root causes from publicly accessible data through zero-shot classification. The robust and scalable framework we propose can aid researchers in addressing data-sparse clinical research questions. The launch of open-access DDE datasets has the potential to stimulate further research and novel discoveries in this field.
Authors
- Michelle Alvarado (ORCID: https://orcid.org/0000-0001-9649-214X)
- Xiang Zhong (ORCID: https://orcid.org/0000-0002-6214-5876)
- William Trevena (ORCID: https://orcid.org/0000-0001-7011-6867)
- Alexander Semenov (ORCID: https://orcid.org/0000-0003-2691-4575)
- Devin Devlin (ORCID: https://orcid.org/0009-0005-2207-673X)
- Alp Oktay (ORCID: https://orcid.org/0009-0007-2075-4896)
- Aarya Yogesh Gohil (ORCID: https://orcid.org/0009-0008-3603-9480)
- Sai Harsha Chittimouju (ORCID: https://orcid.org/0009-0002-4420-9220)
Institutions
- Microsoft (United States) (US)
- University of San Diego (US)
- University of Florida (US)
Publication Details
- Journal
- Journal of Medical Internet Research
- Published
- 2024-11-07
- DOI
- https://doi.org/10.2196/54601
- Citations
- 3
- Primary Topic
- Topic Modeling
- Type
- article
- Field-Weighted Citation Impact
- 1.04