EUVIMEDEUVIMED entdecken
FallberichtOpen Access · CC BY 4.0

Improving medication error classification using a reasoning large language model

Anders Krifors iD, Theodor Beskow, Magnus Jonsson, Karl-Johan Lindner, Jenny Calås, Veronica Arwsbo, Kristian Sandström, Christer Norström

JAMIA Open · 06.01.2026 · 9 (1)

DOI: 10.1093/jamiaopen/ooag004 ↗

Kurz gefasst

Fallbericht in der Zeitschrift JAMIA Open, online erschienen am 24.01.2026. Verfasst von Anders Krifors und 7 weiteren Autor:innen. Beteiligte Einrichtung: Centre for Clinical Research Västmanland, Uppsala University, Västerås Hospital, 721 89 Västerås, Sweden. Die Arbeit wurde bisher 9-mal zitiert (Crossref) und nennt 18 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Region Västmanland and Mälardalen University.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objectives

To assess the performance of a reasoning large language model (LLM) in identifying medication errors in medical incident reports. Materials and Methods OpenAI’s O4-mini LLM was adapted using prompt engineering on 75 000 anonymized incident reports from the Västmanland region of Sweden (2019-2024). To guide the prompt design, we used a subset of 2434 reports, which were manually reclassified by pharmacists as medication-related or not. For validation, 200 reports (January 2024-March 2024) were independently classified by 2 pharmacists to establish a reference classification. Moreover, the LLM performed binary classification, with concordance rates measured against the expert consensus.

Results

The LLM achieved a concordance rate of 96.0% (192/200; 95% CI, 92.3-98.3) with expert classification. Eight cases (4.0%) showed disagreements, primarily due to linguistic ambiguity or context-dependent interpretation. Five cases involved pharmacists classifying reports as non-medication-related, while the LLM classified them as medication-related, with the reverse in 3 cases. Subcategorization accuracy was 76.5%.

Discussion

The LLM showed expert-level performance, outperforming existing automated methods. Thus, its integration into incident reporting systems might improve the efficiency, accuracy, and consistency of patient safety monitoring.

Conclusion

This validated AI-driven method can be integrated directly into clinical informatics workflows, enabling healthcare organizations to rapidly and consistently identify medication errors, ultimately enhancing patient safety outcomes.

Schlagwörter

Stichwörter der Autor:innen

artificial intelligenceincident reportingmedication errorsnatural language processingpatient safety

Autor:innen und Institutionen

  • Centre for Clinical Research Västmanland, Uppsala University, Västerås Hospital, 721 89 Västerås, Sweden.
  • Unit of Infectious Diseases, Department of Medicine Huddinge, Karolinska Institutet, 171 77 Stockholm, Sweden.
  • Norrvy Intelligence AB, 117 59 Stockholm, Sweden.
  • Department of Pharmacy, Region Västmanland, 721 89 Västerås, Västmanland County, Sweden.
  • School of Innovation, Design and Engineering, Mälardalen University, 721 23 Västerås, Sweden.

ORCID-Profile: Anders Krifors

Förderung

  • Region Västmanland and Mälardalen University

Zitieren

APA

Krifors, A., Beskow, T., Jonsson, M., Lindner, K. J., Calås, J., Arwsbo, V., Sandström, K., & Norström, C. (2026). Improving medication error classification using a reasoning large language model. JAMIA Open, 9(1). https://doi.org/10.1093/jamiaopen/ooag004

Vancouver

Krifors A, Beskow T, Jonsson M, Lindner KJ, Calås J, Arwsbo V, et al. Improving medication error classification using a reasoning large language model. JAMIA Open. 2026;9(1). doi:10.1093/jamiaopen/ooag004

BibTeX
@article{Krifors2026,
  author = {Krifors, Anders and Beskow, Theodor and Jonsson, Magnus and Lindner, Karl-Johan and Calås, Jenny and Arwsbo, Veronica and Sandström, Kristian and Norström, Christer},
  title = {Improving medication error classification using a reasoning large language model},
  journal = {JAMIA Open},
  year = {2026},
  volume = {9},
  number = {1},
  doi = {10.1093/jamiaopen/ooag004}
}
Kostenlos starten →