Extracting and Classifying Drug Discontinuations From Estonian Electronic Health Records: Development and Validation Study - Summary - MDSpire

Extracting and Classifying Drug Discontinuations From Estonian Electronic Health Records: Development and Validation Study

  • By

  • Hendrik Šuvalov

  • Nikita Umov

  • Maria Malk

  • Markus Haug

  • Sven Laur

  • Marek Oja

  • Sirli Tamm

  • Sulev Reisberg

  • Jaak Vilo

  • Raivo Kolde

  • June 17, 2026

  • 0 min

Share

Objective:

To examine the discontinuation of antidiabetics and statins among Estonian patients using large language models (LLMs) to identify reasons for discontinuation and categorize them.

Approach:
    Key Findings:
    • Identified 24,040 unique anamneses for antidiabetics and 27,290 for statins.
    • Classified negative experiences as discontinuation reasons, even when there was no explicit mention of stopping.
    Interpretation:

    The study demonstrates the potential of LLMs to systematically analyze unstructured clinical narratives for insights into medication discontinuation.

    Limitations:
    • additional
    Conclusion:

    This research highlights the utility of LLMs in pharmacoepidemiological studies, particularly in low-resource languages, for understanding medication discontinuation.

Original Source(s)

Extracting and Classifying Drug Discontinuations From Estonian Electronic Health Records: Development and Validation Study

  • by Hendrik Šuvalov, Nikita Umov, Maria Malk, Markus Haug, Sven Laur, Marek Oja, Sirli Tamm, Sulev Reisberg, Jaak Vilo, Raivo Kolde

  • June 17, 2026

Related Content

Cardiac magnetic resonance biomarkers for cardiovascular complications in diabetes: a systematic review and meta-analysis

Vitamin D: The Answer May Be Genetic

A post hoc D2d analysis found vitamin D3 supplementation was linked to lower diabetes risk only among patients with ApaI AC or CC genotypes.

Machine Learning–Based risk stratification for in-hospital mortality in ICU patients with cardiovascular diseases and diabetes