Prospective validation of a model-informed precision dosing tool for vancomycin treatment in neonates.

We recruited 48 neonates (50 vancomycin treatment episodes) in a prospective study to validate a model-informed precision dosing (MIPD) software. The initial vancomycin dose was based on a population pharmacokinetic model and adjusted every 36-48 h. Compared with a historical control group of 53 neonates (65 episodes), the achievement of a target trough concentration of 10-15 mg/L improved from 37% in the study to 62% in the MIPD group (P = 0.01), with no difference in side effects.

Subgroup identification-based model selection to improve the predictive performance of individualized dosing.

Currently, model-informed precision dosing uses one population pharmacokinetic model that best fits the target population. We aimed to develop a subgroup identification-based model selection approach to improve the predictive performance of individualized dosing, using vancomycin in neonates/infants as a test case. Data from neonates/infants with at least one vancomycin concentration was randomly divided into training and test dataset. Population predictions from published vancomycin population pharmacokinetic models were calculated. The single best-performing model based on various performance metrics, including median absolute percentage error (APE) and percentage of predictions within 20% (P20) or 60% (P60) of measurement, were determined. Clustering based on median APEs or clinical and demographic characteristics and model selection by genetic algorithm was used to group neonates/infants according to their best-performing model. Subsequently, classification trees to predict the best-performing model using clinical and demographic characteristics were developed. A total of 208 vancomycin treatment episodes in training and 88 in test dataset was included. Of 30 identified models from the literature, the single best-performing model for training dataset had P20 26.2-42.6% in test dataset. The best-performing clustering approach based on median APEs or clinical and demographic characteristics and model selection by genetic algorithm had P20 44.1-45.5% in test dataset, whereas P60 was comparable. Our proof-of-concept study shows that the prediction of the best-performing model for each patient according to the proposed model selection approaches has the potential to improve the predictive performance of model-informed precision dosing compared with the single best-performing model approach.

DosOpt: A Tool for Personalized Bayesian Dose Adjustment of Vancomycin in Neonates.

BACKGROUND: Our main aim has been to design a framework to improve vancomycin dosing in neonates. This required the development and verification of a computerized dose adjustment application, DosOpt, to guide the selection. METHODS: Model fitting in DosOpt uses Bayesian methods for deriving individual pharmacokinetic (PK) estimates from population priors and patient therapeutic drug monitoring measurements. These are used to simulate concentration-time curves and target-constrained dose optimization. DosOpt was verified by assessing bias and precision through several error metrics and normalized prediction distribution errors on samples simulated from the Anderson et al PK model. The performance of DosOpt was also evaluated using retrospective clinical data. Achieved probabilities of target concentration attainment were benchmarked against corresponding attainments in our clinical retrospective data set. RESULTS: Simulations showed no systemic forecast biases. Normalized prediction distribution error values of the base model were distributed by standardized Gaussian (P = 0.1), showing good model suitability. A retrospective test data set included 149 treatment episodes with 1-10 vancomycin concentration measurements per patient (median 2). Individual concentrations in PK estimation improved probability of target attainment and decreased the variance of the estimation. Including 3 individual concentrations in the kinetics estimation increased the probability of Ctrough attainment within 10-15 mg/L from 16% obtained with no individual data (95% confidence interval, 11%-24%) to 43% (21%-47%). CONCLUSIONS: DosOpt uses individual concentration data to estimate kinetics and find optimal doses that increase the probability of achieving desired trough concentrations. Its performance started to exceed target levels attained in retrospective clinical data sets with the inclusion of a single individual input concentration. This tool is freely available at http://www.biit.cs.ut.ee/DosOpt.