Towards precision dosing of vancomycin: a systematic evaluation of pharmacometric models for Bayesian forecasting

Abstract

Objectives: Vancomycin is a vital treatment option for patients suffering from critical infections, and therapeutic drug monitoring is recommended. Bayesian forecasting is reported to improve trough concentration monitoring for dose adjustment. However, the predictive performance of pharmacokinetic models that are utilized for Bayesian forecasting has not been systematically evaluated. Method: Thirty-one published population pharmacokinetic models for vancomycin were encoded in NONMEM ® 7.4. Data from 292 hospitalized patients were used to evaluate the predictive performance (forecasting bias and precision, visual predictive checks) of the models to forecast vancomycin concentrations and area under the curve (AUC) by (a) a priori prediction, i.e., solely by patient characteristics, and (b) also including measured vancomycin concentrations from previous dosing occasions using Bayesian forecasting. Results: A priori prediction varied substantiallydrelative bias (rBias): e122.7e67.96%, relative root mean squared error (rRMSE) 44.3e136.8%, respectivelydand was best for models which included body weight and creatinine clearance as covariates. The model by Goti et al. displayed the best predictive performance with an rBias of e4.41% and an rRMSE of 44.3%, as well as the most accurate visual predictive checks and AUC predictions. Models with less accurate predictive performance provided distorted AUC predictions which may lead to inappropriate dosing decisions. Conclusion: There is a diverse landscape of population pharmacokinetic models for vancomycin with varied predictive performance in Bayesian forecasting. Our study revealed the Goti model as suitable for improving precision dosing in hospitalized patients. Therefore, it should be used to drive vancomycin dosing decisions, and studies to link this finding to clinical outcomes are warranted.

Optimizing Predictive Performance of Bayesian Forecasting for Vancomycin Concentration in Intensive Care Patients

“…Thus, TDM samples that are drawn more closely to the ones to be forecasted in terms of time, are more likely to have better predictive abilities. Similar findings were reported in a previous study that if only a single sample was utilized, Bayesian predicted concentrations were less accurate when obtained using the first ('oldest') observed concentration compared with the most recent observed vancomycin concentration (14). By contrast, in the general patient population, it was reported that taking more TDM data into account did improve the performance of Bayesian forecasting for vancomycin (15).…”

“…The time distance to the last historical data is thus usually fixed. Yet, it is our expectation that a greater time distance to the last historical data comes with higher PE, as was also shown by Broeker et al in a mixed patient population receiving vancomycin (14). Last, although the time-varying covariate creatinine clearance was taken into account in the analysis, further inclusion of time-varying characteristics is expected to improve the performance of Bayesian forecasting using the standard MAP method.…”

Prospective validation of a model‐informed precision dosing tool for vancomycin in intensive care patients

“…Guo et al 23 21 Furthermore, a prospective fit-forpurpose evaluation 13 was lacking in the analysis by Guo et al 23 to evaluate how well vancomycin exposure could be predicted from a previous PK assessment. A fit-for-purpose evaluation of various population PK models for vancomycin in adult patients was recently also performed by Broeker et al 24 This study performed a similar evalua- (open circles) also be a limitation, as it is known that some populations, such as critically ill 25 or neutropenic 26 patients, exhibit altered vancomycin PK. We therefore postulate that our prospective fit-for-purpose evaluation in the critically ill population is of added value.…”

Antimicrobial therapeutic drug monitoring in critically ill adult patients: a Position Paper#

“…(a) the first dose can initially be individualised for patients by using Monte Carlo simulations; (b) the sources of PK variability can be separated into intra-and inter-individual variability; (c) PK sampling before reaching steady state is possible; and (d) through Bayesian estimation, the entire antimicrobial PK profile can be estimated from a single PK sample. Bayesian estimation is most accurate when an optimal sampling time is chosen to provide the most information about the drug's PK behaviour in the patient and a suitable population PK model matching the target population to serve as the Bayesian prior model [32]. After initial TDM, it is recommended to repeat TDM (within 1-2 days for most drugs) to confirm therapeutic exposures have been achieved and again thereafter if there are concerns of significant changes to PK (e.g.…”