EUVIMEDEUVIMED entdecken
Vergleichsstudie

Routine laboratory panels classify internal medicine ICD-10 code groups: comparison with frontier large language models and laboratory-only specialist assessment

Yusuf Yesil iD, Alpay Medetalibeyoglu iD, Evin Ademoglu iD

Laboratory Medicine · 05.10.2026 · 57 (6)

DOI: 10.1093/labmed/lmag074 ↗

Kurz gefasst

Vergleichsstudie in der Zeitschrift Laboratory Medicine, online erschienen am 06.10.2026. Verfasst von Yusuf Yesil und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Biochemistry, Istanbul Faculty of Medicine, Istanbul University, Istanbul, Türkiye. Hauptthemen laut MEDLINE: Internal Medicine, International Classification of Diseases. Die Arbeit nennt 21 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

Routine laboratory panels are nearly universal, but the panels’ joint information is underused. We evaluated contemporaneous classification of International Statistical Classification of Diseases, Tenth Revision (ICD-10) code groups from same-encounter laboratory results.

Methods

We developed 17 eXtreme Gradient Boosting (XGBoost) classifiers in 242 648 adult internal medicine encounters using age, sex, and results from 34 assays, with ICD-10 codes used only as outcomes. Stratified 10-fold cross-validation assessed discrimination, calibration, and decision curve performance. Leakage-free recalibration and fold-specific thresholds targeted 95% specificity. Using an 80-patient temporal holdout cohort, we compared the model with 3 large language models accessed through an application programming interface and findings from 1 specialist physician.

Results

The mean cross-validated area under the curve (AUC) was 0.893. With identical preprocessing, the mean AUC was 0.892 for XGBoost and 0.809 for logistic regression. Recalibration changed the mean intercept and slope to −0.002 and 0.999, respectively, and the Brier score from 0.099 to 0.044. The holdout micro-averaged AUC was 0.880 vs 0.729 to 0.739 for language models; paired differences were 0.141 to 0.151 (all P < .005). Matched-specificity sensitivity differences favored the model but were imprecise.

Discussion

Routine panels contain substantial information about contemporaneous ICD-10 coding, supporting prospective multicenter evaluation rather than a diagnostic claim.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Internal MedicineInternational Classification of DiseasesHumansLarge Language ModelsBoosting Machine Learning AlgorithmsFemaleClassification AlgorithmsMaleMiddle Aged

Stichwörter der Autor:innen

clinical decision supportdecision curve analysisgradient boostinglaboratory medicinelarge languagemodelsrisk calibration

Autor:innen und Institutionen

  • Department of Biochemistry, Istanbul Faculty of Medicine, Istanbul University, Istanbul, Türkiye.
  • Department of Internal Medicine, Istanbul Faculty of Medicine, Istanbul University, Istanbul, Türkiye.

ORCID-Profile: Yusuf Yesil, Alpay Medetalibeyoglu, Evin Ademoglu

Zitieren

APA

Yesil, Y., Medetalibeyoglu, A., & Ademoglu, E. (2026). Routine laboratory panels classify internal medicine ICD-10 code groups: comparison with frontier large language models and laboratory-only specialist assessment. Laboratory Medicine, 57(6). https://doi.org/10.1093/labmed/lmag074

Vancouver

Yesil Y, Medetalibeyoglu A, Ademoglu E. Routine laboratory panels classify internal medicine ICD-10 code groups: comparison with frontier large language models and laboratory-only specialist assessment. Laboratory Medicine. 2026;57(6). doi:10.1093/labmed/lmag074

BibTeX
@article{Yesil2026,
  author = {Yesil, Yusuf and Medetalibeyoglu, Alpay and Ademoglu, Evin},
  title = {Routine laboratory panels classify internal medicine ICD-10 code groups: comparison with frontier large language models and laboratory-only specialist assessment},
  journal = {Laboratory Medicine},
  year = {2026},
  volume = {57},
  number = {6},
  doi = {10.1093/labmed/lmag074}
}
Kostenlos starten →