Research progress in machine learning-assisted individualized vancomycin medication
- VernacularTitle:机器学习辅助万古霉素个体化用药的研究进展
- Author:
Mingyue CHEN
1
;
Yifei LEI
1
;
Xuewu SONG
2
;
Jinqi LI
2
Author Information
1. Dept. of Pharmacy，Sichuan Academy of Medical Sciences & Sichuan Provincial People’s Hospital （Affiliated Hospital of UESTC）/Personalized Drug Research and Therapy Key Laboratory of Sichuan Province，Chengdu 610072，China;School of Medicine UESTC，Chengdu 610054，China
2. Dept. of Pharmacy，Sichuan Academy of Medical Sciences & Sichuan Provincial People’s Hospital （Affiliated Hospital of UESTC）/Personalized Drug Research and Therapy Key Laboratory of Sichuan Province，Chengdu 610072，China
- Publication Type:Journal Article
- Keywords:
machine learning;
individualized medication;
dose;
prediction model
- From:
China Pharmacy
2026;37(17):2310-2314
- CountryChina
- Language:Chinese
-
Abstract:
Vancomycin is a glycopeptide antibiotic widely used for severe infections caused by gram-positive bacteria. Nevertheless， its clinical application is limited by a narrow therapeutic window and inter-individual differences in pharmacokinetics. Conventional dosing regimens rely on clinical experience and population pharmacokinetic models， which make precise therapy difficult. The emergence of machine learning provides new insights into tackling the above challenges. By integrating multi-dimensional data including demographic characteristics， clinical indicators and combined medications， machine learning can assist individualized medication for patients and improve the capacity for vancomycin dose prediction， in vivo exposure assessment and renal toxicity risk warning. However， the application of machine learning still faces multiple challenges such as poor data quality， weak model generalization ability and insufficient interpretability. Its clinical value needs to be further confirmed by prospective studies and external validation. Future research may consider deep integration of machine learning and population pharmacokinetic models to better support vancomycin individualized medication.
