Vancomycin therapeutics and monitoring: a contemporary approach

• Published in: Internal medicine journal 2013-02, Vol.43 (2), p.110-119
• Main Authors: Avent, M. L., Vaska, V. L., Rogers, B. A., Cheng, A. C., van Hal, S. J., Holmes, N. E., Howden, B. P., Paterson, D. L.
• Format: Article
• Language: English
• Subjects: Animals; Bayes Theorem; clinical pharmacology; computerized models; drug monitoring; Drug Monitoring - methods; Drug Monitoring - trends; Humans; Linear Models; Methicillin-Resistant Staphylococcus aureus - drug effects; Methicillin-Resistant Staphylococcus aureus - physiology; Pharmacokinetics; Staphylococcal Infections - blood; Staphylococcal Infections - drug therapy; Vancomycin - pharmacokinetics; Vancomycin - therapeutic use; Vancomycin Resistance - drug effects; Vancomycin Resistance - physiology
• ISSN: 1444-0903; 1445-5994
• DOI: 10.1111/imj.12036

Vancomycin remains a clinically useful antibiotic despite the advent of several alternative drugs. Optimising vancomycin therapy with therapeutic drug monitoring is widely recommended. The aim of therapeutic drug monitoring is to help the clinician to achieve target pharmacodynamic parameters in the case of vancomycin, an area under the concentration time curve/minimum inhibitory concentration ratio of ≥400. Vancomycin monitoring methods can be categorised into four categories: empiric trough concentrations; linear regression analysis (one‐compartment model), population methods and Bayesian estimation procedures. Although the empiric trough concentrations and population methods are easy to use and require minimal resources, there are large differences in the published vancomycin model parameters. This demonstrates that there is great variance in pharmacokinetic parameters between the models and a single vancomycin model cannot be applied to all patient populations. The linear regression and Bayesian methods recommended more accurate dosage regimens; however, they require additional resources such as information technology and healthcare personnel with background training in pharmacokinetics. The Bayesian methods offered additional advantages such as calculation of doses based on a single‐serum concentration and optimisation of the patient's previous pharmacokinetic data to determine subsequent dosage regimens. Computerised programs, utilising the Bayesian estimation procedures, are able to achieve target concentrations in a greater percentage of patients earlier in the course of therapy than the empiric trough concentrations and population methods. We recommend the use of these programs providing there is appropriate expertise available to make appropriate recommendations.