An Ensemble Model for Prediction of Vancomycin Trough Concentrations in Pediatric Patients

Abstract

Purpose This study aimed to establish an optimal model to predict vancomycin trough concentrations by using machine learning. Patients and Methods We enrolled 407 pediatric patients (age < 18 years) who received vancomycin intravenously and underwent therapeutic drug monitoring from June 2013 to April 2020 at Xinhua Hospital affiliated to Shanghai Jiaotong University School of Medicine. The median (interquartile range) age and weight of the patients were 2 (0.63–5) years and 12 (7.8–19) kg. Vancomycin trough concentrations were considered as the target variable, and eight different algorithms were used for predictive performance comparison. The whole dataset (407 cases) was divided into training group and testing group at the ratio of 80%: 20%, which were 325 and 82 cases, respectively. Results Ultimately, five algorithms (XGBoost, GBRT, Bagging, ExtraTree and decision tree) with high R 2 (0.657, 0.514, 0.468, 0.425 and 0.450, respectively) were selected and further ensembled to establish the final model and achieve an optimal result. For missing data, through filling the missing values and model ensemble, we obtained R 2 =0.614, MAE=3.32, MSE=24.39, RMSE=4.94 and a prediction accuracy of 51.22% (predicted trough concentration within ±30% of the actual trough concentration). In comparison with the pharmacokinetic models ( R 2 =0.3), the machine learning model works better in model fitting and has better prediction accuracy. Conclusion Therefore, the ensemble model is useful for the vancomycin concentration prediction, especially in the population of children with great individual variation. As machine learning methods evolve, the clinical value of the ensemble model will be demonstrated in the clinical practice.

Construction and Interpretation of Prediction Model of Teicoplanin Trough Concentration via Machine Learning

“…Its main ability is to gather and interpret any relevant data even on a large scale and thus, transforms medicine to a data-driven approach. Precision treatment is one of the top applications of machine learning, where a patient receives tailored medical care, such as personalized dose adjustment, plasma concentration prediction, and adverse drug events prediction (19)(20)(21)(22). Ensemble learning, one of the key features of machine learning, comes from a combination of various models that is capable of producing a final prediction.…”

“…It enables computers to access previous data without being explicitly programmed. Many researches have reported that the predicting accuracy of machine learning approach exceeded the PPK method ( 20 , 25 ). Huang et al constructed an ensemble prediction model of vancomycin trough concentrations, and compared with PPK model.…”

“…Huang et al constructed an ensemble prediction model of vancomycin trough concentrations, and compared with PPK model. Their findings showed that machine learning model works better with higher accuracy of prediction ( 20 ). The evaluation parameters ( R 2 and accuracy) of our ensemble predicting model have surpassed its vancomycin counterpart, suggesting that our model has a good prediction effect and prospect of clinical application.…”

A Machine Learning Model to Predict Risperidone Active Moiety Concentration Based on Initial Therapeutic Drug Monitoring

“…Hence, the selected variables were of great significance to the next risperidone TDM value. There was another study of 407 patients predicting vancomycin trough concentration using an ensemble machine learning model with five algorithms (XGBoost, GBRT, Bagging, ExtraTree, and decision tree), achieving a prediction accuracy (predicted trough concentration within ±30% of the actual trough concentration) of 51.22% in a test cohort ( 26 ). In comparison, our study had better performance (accuracy of the predicted TDM within ±30% of the actual TDM is 54.82%) in the test cohort and also larger sample size, which led to a more mature model for application and reference.…”

Case Report: Monitoring Vancomycin Concentrations and Pharmacokinetic Parameters in Continuous Veno-Venous Hemofiltration Patients to Guide Individualized Dosage Regimens: A Case Analysis

“…Vancomycin concentrations were determined by high-performance liquid chromatography, as described previously ( Huang et al, 2021 ). The linear range for the assay was 2.0–100 mg/L, the lower detection limit was 1 mg/L, and intra- and inter-day precision and accuracy values were both within 15%.…”