EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Beyond Traditional Covariates: An Interpretable Machine Learning Workflow for Improved Hybrid Pharmacometric Modeling

Freek J A Relouw iD, Jort J M T Lokers iD, Tim Preijers iD, Sebastiaan D T Sassen iD, Natal A W van Riel iD, Birgit C P Koch iD

CPT: Pharmacometrics & Systems Pharmacology · 2026-08 · 15 (8)

DOI: 10.1002/psp4.70313 ↗

Kurz gefasst

Fachartikel in der Zeitschrift CPT: Pharmacometrics & Systems Pharmacology, online erschienen am 04.08.2026. Verfasst von Freek J A Relouw und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, the Netherlands. Hauptthemen laut MEDLINE: Vancomycin, Machine Learning, Anti-Bacterial Agents, Models, Biological. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 51 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Model‐informed precision dosing is often constrained by the limited generalizability of traditional population pharmacokinetic models, especially in critically ill patients. A hybrid machine learning‐population pharmacokinetic framework is proposed to improve a priori pharmacokinetic predictions by integrating real‐world clinical data. This approach was applied to vancomycin trough concentration prediction. Two widely used two‐compartment population pharmacokinetic models provided individual pharmacokinetic parameter estimates. Maximum a posteriori Bayesian estimation was used to adjust population parameters for individual patients based on drug administration records, therapeutic drug monitoring values, and patient‐specific covariates from the MIMIC‐IV database. The resulting clearance and central volume of distribution estimates served as training targets for XGBoost and symbolic regression models. Machine learning‐predicted parameters were reinserted into the original pharmacokinetic equations to generate a priori vancomycin trough concentrations without reliance on therapeutic drug monitoring input. The hybrid models demonstrated improved prediction accuracy over traditional population pharmacokinetic covariate models and reduced vancomycin trough concentration prediction error by up to ~20%. XGBoost generally provided the highest predictive performance, while symbolic regression produced interpretable mathematical expressions revealing associations between non‐traditional clinical predictors and pharmacokinetic parameters, highlighting a trade‐off between accuracy and interpretability. This framework illustrates the potential of combining machine learning with population pharmacokinetic modeling to refine pharmacokinetic parameter estimation and support more precise, individualized dosing. The workflow is adaptable to other drugs and patient populations, offering a generalizable methodological strategy to identify non‐traditional predictors and enhance existing pharmacometric model performance in real‐world clinical settings.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

VancomycinMachine LearningAnti-Bacterial AgentsModels, BiologicalHumansBayes TheoremBoosting Machine Learning AlgorithmsPredictive Learning ModelsDrug MonitoringWorkflowFemaleMale

Stichwörter der Autor:innen

XGBoostempirical Bayesian estimatesintensive care unitmachine learningmodel‐informed precision dosingpopulation pharmacokineticssymbolic regressionvancomycin

Substanzen

VancomycinAnti-Bacterial Agents

Autor:innen und Institutionen

  • Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, the Netherlands.
  • Department of Hospital Pharmacy, Erasmus University Medical Center, Rotterdam, the Netherlands.
  • Rotterdam Clinical Pharmacometrics Group, Rotterdam, the Netherlands.
  • Division of Neonatology, Department of Neonatal & Pediatric ICU, Erasmus MC, Rotterdam, the Netherlands.
  • Department of Intensive Care Medicine, Radboud University Medical Center, Nijmegen, the Netherlands.

ORCID-Profile: Freek J A Relouw, Jort J M T Lokers, Tim Preijers, Sebastiaan D T Sassen, Natal A W van Riel, Birgit C P Koch

Zitieren

APA

Relouw, F. J. A., Lokers, J. J. M. T., Preijers, T., Sassen, S. D. T., van Riel, N. A. W., & Koch, B. C. P. (2026). Beyond Traditional Covariates: An Interpretable Machine Learning Workflow for Improved Hybrid Pharmacometric Modeling. CPT: Pharmacometrics & Systems Pharmacology, 15(8). https://doi.org/10.1002/psp4.70313

Vancouver

Relouw FJA, Lokers JJMT, Preijers T, Sassen SDT, van Riel NAW, Koch BCP. Beyond Traditional Covariates: An Interpretable Machine Learning Workflow for Improved Hybrid Pharmacometric Modeling. CPT: Pharmacometrics & Systems Pharmacology. 2026;15(8). doi:10.1002/psp4.70313

BibTeX
@article{Relouw2026,
  author = {Relouw, Freek J A and Lokers, Jort J M T and Preijers, Tim and Sassen, Sebastiaan D T and van Riel, Natal A W and Koch, Birgit C P},
  title = {Beyond Traditional Covariates: An Interpretable Machine Learning Workflow for Improved Hybrid Pharmacometric Modeling},
  journal = {CPT: Pharmacometrics & Systems Pharmacology},
  year = {2026},
  volume = {15},
  number = {8},
  doi = {10.1002/psp4.70313}
}
Kostenlos starten →