| dc.description.abstract | Introduction and aims: International guidelines recommend area under the curve (AUC)-guided dosing, preferably by using model-informed precision dosing (MIPD), as the most accurate and optimal way to monitor vancomycin dosing. The current practice of trough guided dosing is not only suboptimal, but also considered complex to manage by clinicians in Helse Vest. Therefore, it is necessary to evaluate the current routines and, preferably, identify an adequate pharmacokinetic (PK) model for individual, AUC guided vancomycin dosing for the patient population.
Method: In a prospective cohort study, data from patients treated with vancomycin was collected from three different hospital wards in Helse Vest. The data consisted of a set of variables, including doses and serum concentrations of vancomycin, creatinine, sex, age, and weight. AUC estimation was performed in the PK software DoseMeRx®. The non-linear mixed effects modeling software Monolix® was used to compare a selection of published and commercially available pharmacokinetic models from PK software and assess the goodness of fit to our population data.
Results: A total of 65 patients were included in the prospective study, with a total of 165 trough concentrations considered for further analysis. The majority of patients had trough concentrations outside the recommended range of 15-20 mg/L on the fourth day of treatment, while the estimated AUC in the same time period showed contrasting results with 60 % of patients within the therapeutic AUC range. The one-compartment PK model from DoseMeRx® published by Buelga et al was identified as an adequate model for AUC guided vancomycin dosing in our population.
Conclusion: Considerable interindividual variability was found in the AUC level estimated from the same trough level, indicating that trough concentrations are a poor indicator for reaching a therapeutic AUC level and strengthening the argument for implementing AUC guided vancomycin dosing in Helse Vest. With the current knowledge on model evaluation, the Buelga model implemented in DoseMeRx® seems adequate for clinical use in our population, indicating that establishment of PK models adapted to the local population is not required. | |
