Poster Presenter

Therapeutic Drug Monitoring of Vancomycin: evaluation of initial dosing methods at AL-Amiri Hospital in Kuwait
Yousef Mohammed Allanqawi
Kuwait

Population-based pharmacokinetic methods are commonly employed to predict volume of distribution (Vd) and clearance (Clvanco) in order to calculate doses required to achieve desired levels. Substantial variability exists in Vd and CLvanco, therefore the objective of this study was to compare three different methods (Conventional kinetic, Ambrose and Matzke methods) for their predictive ability in estimating Vd and Clvanco., and (b) the frequency with which the dosage from each method would yield simulated steady-state peak and trough levels within a specified range. Methods: Based on measured steady state vancomycin concentrations (n=63 patients), Vd and Clvanco were calculated using Sawchuk-Zaske and three population methods. The predictive ability of the estimated parameters from each method was assessed using predictive performance analysis (mean prediction error, ME, mean squared error MSE, and root mean squared error RMSE). In addition, the individual parameters were used to simulate peak and trough levels, using first order kinetic equations from doses derived from the population equations. The frequency with which each equation would have achieved target peak and trough concentrations was determined. Results: The Ambrose method had the best combination of the least bias (ME=0.2, p<0.05 for Vd ; ME= -5.0 p<0.05 for Clvanco) and best precision. (MSE=175 and RMSE=10 for Vd; MSE=665 and RMSE=21 for Clvanco). This method also produced a higher frequency of simulated steady-state levels within the target ranges specified. Conclusion: Ambrose method proved the most reliable of those evaluated and could be applied in Kuwaiti hospital. However, in all methods, a large number of patients were outside the target ranges. Therefore, individualization of vancomycin dose regimen based on measured drug level is of great need.