EUVIMEDEUVIMED entdecken
Open Access

Advancing delirium detection through the Open Health Natural Language Processing Consortium and the Evolve to Next-Gen Accrual to Clinical Trials Network

Förderung der Delir-Erkennung durch das Open Health Natural Language Processing Consortium und das Evolve to Next-Gen Accrual to Clinical Trials Network

Sunyang Fu iD, Min Ji Kwak, Jaerong Ahn, Zhiyi Yue, Shreyas Ranganath, Joseph R Applegate, Andrew Wen, Liwei Wang, Chenyu Li iD, Michele Morris, Kelly M Toth, Timothy D Girard und 20 weitere

The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences · 06.11.2025 · 80 (12)

DOI: 10.1093/gerona/glaf207 ↗

Auf einen Blick

Die Arbeit beschreibt erste Schritte zur Standardisierung und Evaluation eines NLP-basierten Delir-Erkennungsmodells für elektronische Gesundheitsakten über mehrere Institutionen hinweg. Unterschiede in Screening und Dokumentation beeinflussten die Leistung.

Fragestellung

Wie lässt sich ein NLP-basiertes Delir-Erkennungsmodell für EHRs institutionenübergreifend standardisieren, evaluieren und in die i2b2/ENACT-Plattform übertragen?

Methode

Multisite-Kontextanalyse und multisite NLP-Evaluation im ENACT NLP Working Group-Kontext; Fokus auf Standardisierung, Evaluation und Translation eines Delir-Erkennungsmodells.

Ergebnisse

Die Kontextanalyse zeigte Herausforderungen durch unterschiedliche lokale Screeningpraktiken, harmonisierungsbedürftige Definitionen und komplexe zeitliche Logik. Die NLP-Leistung variierte je nach Screening, Dokumentation, Semantik und Syntax der Notizen.

Einschränkungen

Das Abstract nennt keine formalen Limitationen.

Schlussfolgerung der Autor:innen

Die Arbeit ist ein wichtiger erster Schritt zu skalierbarer und standardisierter NLP-basierter Delir-Erkennung über Institutionen hinweg. Der kollaborative Ansatz lieferte ein robusteres Rahmenwerk für heterogene EHR-Systeme.

Kernaussagen

  • Unterschiedliche Screeningpraktiken erschwerten die Delir-Erkennung
  • Dokumentationsmuster und Notizstruktur beeinflussten die Modellleistung
  • Kollaboration zwischen Institutionen stärkte das Rahmenwerk

Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.

Kurz gefasst

Fachartikel in der Zeitschrift The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences, online erschienen am 29.09.2025. Verfasst von Sunyang Fu und 31 weiteren Autor:innen. Beteiligte Einrichtung: McWilliams School of Biomedical Informatics, University of Texas Health Science Center at Houston, Houston, Texas, United States. Hauptthemen laut MEDLINE: Natural Language Processing, Delirium, Electronic Health Records. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 34 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NCATS NIH HHS, the Mayo Clinic Robert and Arlene Kogod Center on Aging, NIH NCATS CTSA.

Auf einen BlickAbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Delirium is often underdiagnosed in clinical practice and is not routinely coded for billing. While manual chart review can identify delirium, it is labor-intensive and impractical for large-scale studies. Natural language processing (NLP) can analyze unstructured text in electronic health records (EHRs) to extract meaningful clinical information.

Methods

To support national integration of NLP for EHR-based delirium identification across different institutions, we launched the Delirium Interest Group within the national Evolve to Next-Gen Accrual to Clinical Trials (ENACT) NLP Working Group. This paper outlines our initial efforts to standardize, evaluate, and translate an NLP-based delirium detection model into the i2b2/ENACT platform.

Results

Multisite contextual inquiry identified several key challenges, including variations in local screening practices (eg, tools used, documentation frequency, and quality control), the need for harmonized definitions in the context of EHRs, and the complexity of modeling temporal logic. Multisite NLP evaluation revealed variable performance degradation driven by differences in delirium screening practices, clinical documentation patterns and semantics, and note syntactic structures.

Conclusion

Our work represents an important first step toward enabling scalable and standardized NLP-based delirium detection across institutions. By engaging diverse institutions through the ENACT NLP Working Group, we identified shared challenges and site-specific variations that impact model implementation and performance. Our collaborative approach enabled the development of a more robust framework for delirium identification across heterogeneous EHR systems. Future efforts will build on this foundation to enhance the validity, usability, and translational impact of delirium detection.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Natural Language ProcessingDeliriumElectronic Health RecordsHumansClinical Trials as Topic

Stichwörter der Autor:innen

Artificial IntelligenceDeliriumElectronic Health RecordsNatural Language ProcessingTranslational AI

Autor:innen und Institutionen

  • McWilliams School of Biomedical Informatics, University of Texas Health Science Center at Houston, Houston, Texas, United States.
  • McGovern Medical School, University of Texas Health Science Center at Houston, Houston, Texas, United States.
  • School of Public Health, University of Texas Health Science Center at Houston, Houston, Texas, United States.
  • Department of Health Information Management, University of Pittsburgh, Pittsburgh, Pennsylvania, United States.
  • Center for Research, Investigation, and Systems Modeling of Acute Illness (CRISMA) in the Department of Critical Care Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, United States.
  • Department of Biomedical Informatics and Data Science, University of Alabama at Birmingham, Birmingham, Alabama, United States.

ORCID-Profile: Sunyang Fu, Chenyu Li, Justin F Rousseau, Chao Yan, Tyler J Murphy, Chan Mi Park, Elmer V Bernstam, Shyam Visweswaran, Hongfang Liu

Förderung

  • NCATS NIH HHS
  • the Mayo Clinic Robert and Arlene Kogod Center on Aging
  • NIH NCATS CTSA
  • NIA NIH HHS
  • NIGMS NIH HHS
  • RRD VA
  • VA
  • Ingram Chair in Surgical Sciences, Cancer Prevention and Research Institute of Texas

Zitieren

APA

Fu, S., Kwak, M. J., Ahn, J., Yue, Z., Ranganath, S., Applegate, J. R., Wen, A., Wang, L., Li, C., Morris, M., Toth, K. M., Girard, T. D., Osborne, J. D., Kennedy, R. E., Garduno-Rapp, N. E., Reeder, P., Rousseau, J. F., Yan, C., Chen, Y., …, & Liu, H. (2025). Advancing delirium detection through the Open Health Natural Language Processing Consortium and the Evolve to Next-Gen Accrual to Clinical Trials Network. The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences, 80(12). https://doi.org/10.1093/gerona/glaf207

Vancouver

Fu S, Kwak MJ, Ahn J, Yue Z, Ranganath S, Applegate JR, et al. Advancing delirium detection through the Open Health Natural Language Processing Consortium and the Evolve to Next-Gen Accrual to Clinical Trials Network. The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences. 2025;80(12). doi:10.1093/gerona/glaf207

BibTeX
@article{Fu2025,
  author = {Fu, Sunyang and Kwak, Min Ji and Ahn, Jaerong and Yue, Zhiyi and Ranganath, Shreyas and Applegate, Joseph R and Wen, Andrew and Wang, Liwei and Li, Chenyu and Morris, Michele and Toth, Kelly M and Girard, Timothy D and Osborne, John D and Kennedy, Richard E and Garduno-Rapp, Nelly-Estefanie and Reeder, Phillip and Rousseau, Justin F and Yan, Chao and Chen, You and Patel, Mayur B and Murphy, Tyler J and Malin, Bradley A and Park, Chan Mi and Jia, Heling and Pagali, Sandeep and Palmer, Allyson K and Sauver, Jennifer St and Sohn, Sunghwan and Bernstam, Elmer V and Visweswaran, Shyam and Wang, Yanshan and Liu, Hongfang},
  title = {Advancing delirium detection through the Open Health Natural Language Processing Consortium and the Evolve to Next-Gen Accrual to Clinical Trials Network},
  journal = {The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences},
  year = {2025},
  volume = {80},
  number = {12},
  doi = {10.1093/gerona/glaf207}
}
Kostenlos starten →