Identifying Diabetes Related-Complications in a Real-World Free-Text Electronic Medical Records in Hebrew Using Natural Language Processing Techniques
Authors' conclusion
NLP algorithms and machine learning processes may enable more accurate identification of diabetes complications in EMR data.
Study context
- Population
- Patients with diabetes aged ≥35 years, Leumit Health Services members, 2012–2017; 2.3 million records2.3 million records
- Participants
- 41,469 patients
- Comparison
- Natural language processing (NLP) algorithm vs International Classification of Diseases (ICD) codes, each against reviewers
- Outcome
- Precision, recall and F-score for identification of diabetes-related complications (CVD, neuropathy, nephropathy, retinopathy, diabetic foot, cognitive impairments, mood disorders, hypoglycemia)
Methodological detailsrule-based and machine learning NLP, validation against independent nurse reviewers (gold standard)
- Rule-based and machine learning NLP
- Validation against independent nurse reviewers (gold standard)
Performance
As reported in the abstract
Mean F-score vs reviewers (gold standard)
| Outcome | NLP algorithm | ICD codes |
|---|---|---|
| Mean F-score | 86% | 51% |
Original publication
Identifying Diabetes Related-Complications in a Real-World Free-Text Electronic Medical Records in Hebrew Using Natural Language Processing Techniques. J Diabetes Sci Technol. Published 30 January 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →