EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

LLM-Assisted Clinical Data Harmonization: Combining Automated ETL Generation with Semantic Vocabulary Mapping for OMOP CDM

Falk Meyer-Eschenbach, Martin Vogel, Philipp Jacob, Severin Kohler, Louis Agha-Mir-Salim, Loreen Ruhm, Christof von Kalle, Elias Grünewald, Oliver Kumpf, Felix Balzer

Studies in Health Technology and Informatics · 2026

DOI: 10.3233/shti260883 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Studies in Health Technology and Informatics, 2026. Verfasst von Falk Meyer-Eschenbach und 9 weiteren Autor:innen. Beteiligte Einrichtung: Institute of Medical Informatics, Charité - Universitätsmedizin Berlin, Berlin, Germany. Hauptthemen laut MEDLINE: Large Language Models, Semantics, Vocabulary, Controlled, Natural Language Processing. Die Arbeit wurde bisher 1-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Transforming clinical data into the OMOP Common Data Model requires structural schema mapping and semantic vocabulary harmonization, demanding both technical and clinical expertise. We investigated Large Language Models for both tasks using the eICU Collaborative Research Database. For structural transformation, context-enriched prompting with White Rabbit profiling reports guided Gemini 2.5 Pro to generate PostgreSQL Extract-Transform-Load (ETL) scripts, populating eight OMOP core tables after five debugging iterations. For semantic mapping, a hybrid approach combining vector similarity search with LLM refinement targeted Logical Observation Identifiers Names and Codes (LOINC) and Anatomical Therapeutic Chemical (ATC) codes, validated by clinical experts from Charité – Universitätsmedizin Berlin. Our approach achieved 93.9% precision for medications (n=148) and 78.5-96.8% for laboratory terms (n=158) depending on consensus criteria. LLM refinement raised medication precision from 60.8% to 93.9% and laboratory precision from 62.0–88.6% to 78.5–96.8%. Together, these phases reduce manual effort in OMOP transformation across three clinical domains, while expert supervision remains essential.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Large Language ModelsSemanticsVocabulary, ControlledNatural Language ProcessingElectronic Health RecordsHumans

Stichwörter der Autor:innen

Clinical Data HarmonizationETLLarge Language ModelsOMOP Common Data ModelVocabulary Mapping

Autor:innen und Institutionen

  • Institute of Medical Informatics, Charité - Universitätsmedizin Berlin, Berlin, Germany.
  • Clinical Study Center, Berlin Institute of Health, Charité, Germany.
  • Institute of Informatics, Humboldt-Universität zu Berlin, Berlin, Germany.
  • Institute of Informatics, Freie Universität Berlin, Berlin, Germany.
  • Department of Anesthesiology and Intensive Care Medicine, Charité, Germany.

Zitieren

APA

Meyer-Eschenbach, F., Vogel, M., Jacob, P., Kohler, S., Agha-Mir-Salim, L., Ruhm, L., von Kalle, C., Grünewald, E., Kumpf, O., & Balzer, F. (2026). LLM-Assisted Clinical Data Harmonization: Combining Automated ETL Generation with Semantic Vocabulary Mapping for OMOP CDM. Studies in Health Technology and Informatics. https://doi.org/10.3233/shti260883

Vancouver

Meyer-Eschenbach F, Vogel M, Jacob P, Kohler S, Agha-Mir-Salim L, Ruhm L, et al. LLM-Assisted Clinical Data Harmonization: Combining Automated ETL Generation with Semantic Vocabulary Mapping for OMOP CDM. Studies in Health Technology and Informatics. 2026. doi:10.3233/shti260883

BibTeX
@article{MeyerEschenbach2026,
  author = {Meyer-Eschenbach, Falk and Vogel, Martin and Jacob, Philipp and Kohler, Severin and Agha-Mir-Salim, Louis and Ruhm, Loreen and von Kalle, Christof and Grünewald, Elias and Kumpf, Oliver and Balzer, Felix},
  title = {LLM-Assisted Clinical Data Harmonization: Combining Automated ETL Generation with Semantic Vocabulary Mapping for OMOP CDM},
  journal = {Studies in Health Technology and Informatics},
  year = {2026},
  doi = {10.3233/shti260883}
}
Kostenlos starten →