Developing and Evaluating a Natural Language Processing Algorithm to Detect Insulin Pump and Continuous Glucose Monitor Use in Electronic Health Records of Patients with Type 1 Diabetes
We showed that NLP can accurately identify insulin pump and CGM users from unstructured EHR notes, substantially outperforming billing code-based methods. This scalable approach can support system- and population-level evaluations of diabetes technologies.
Authors
- Eli Ipp (ORCID: https://orcid.org/0000-0003-1588-0146)
- Nestoras Mathioudakis (ORCID: https://orcid.org/0000-0002-0210-655X)
- Jeffrey Feng (ORCID: https://orcid.org/0000-0002-6927-5200)
- Alex Bui (ORCID: https://orcid.org/0000-0002-4702-1373)
- Bing Zhu (ORCID: https://orcid.org/0000-0002-5700-1253)
- Tannaz Moin (ORCID: https://orcid.org/0000-0002-5035-6641)
- J. C. W. Graham
- Estelle Marla Everett (ORCID: https://orcid.org/0000-0002-7468-296X)
- NICHOLAS J. JACKSON (ORCID: https://orcid.org/0000-0002-9515-5050)
- Ryan Alyson-Yao Tiu (ORCID: https://orcid.org/0000-0003-1189-2836)
Institutions
- Johns Hopkins University (US)
- University of California, Los Angeles (US)
- Johns Hopkins Medicine (US)
- UCLA Medical Center (US)
- University of California System (US)
- Los Angeles Medical Center (US)
- VA Greater Los Angeles Healthcare System (US)
- Harbor–UCLA Medical Center (US)
Publication Details
- Journal
- Diabetes Technology & Therapeutics
- Published
- 2025-09-29
- DOI
- https://doi.org/10.1177/15209156251383828
- Primary Topic
- Diabetes Management and Research
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Institutes of Health
- Office of Naval Research
- National Institute of Diabetes and Digestive and Kidney Diseases
- Office of Behavioral and Social Sciences Research