BackgroundVancomycin trough concentration is closely associated with clinical e cacy and toxicity. Predicting vancomycin trough concentrations in pediatric patients is challenging due to signi cant interindividual variability and rapid physiological changes during maturation.
AimThis study aimed to develop a machine-learning (ML) model to predict vancomycin trough concentrations and determine optimal dosing regimens in pediatric patients using various machinelearning (ML) algorithms.
MethodA single-center retrospective observational study was conducted from January 2017 to March 2020. Pediatric patients who received intravenous vancomycin and underwent therapeutic drug monitoring (TDM) were enrolled. Seven ML models [linear regression (LR), gradient boosted decision trees (GDBT), support vector machine (SVM), decision tree (DT), random forest (RF), Bagging, and extreme gradient boosting (XGBoost)] were developed using 31 variables. Performance metrics including R-squared (R 2 ), mean square error (MSE), root mean square error (RMSE), and mean absolute error (MAE) were compared, and important features were ranked.
ResultsThe study included 120 eligible trough concentration measurements from 112 patients. Of these, 84 measurements were used for training and 36 for testing. Among the seven algorithms tested, XGBoost showed the best performance, with low prediction error and high goodness of t (MAE = 2.55, RMSE = 4.13, MSE = 17.12, and R 2 = 0.59). Blood urea nitrogen, serum creatinine, and creatinine clearance rate were identi ed as the most important predictors of vancomycin trough concentration.
ConclusionAn XGBoost ML model was developed to predict vancomycin trough concentrations and aid in drug treatment predictions as a decision-support decision-support technology.
