Aris Dokoumetzidis and Leon Aarons
School of Pharmacy and Pharmaceutical Sciences, University of Manchester, UK.
Objectives: To investigate the propagation of population pharmacokinetic information across clinical studies by applying Bayesian techniques. The aim is to summarize the population pharmacokinetic estimates of a study in appropriate statistical distributions in order to use them as Bayesian priors in consequent population pharmacokinetic analyses.
Methods: Various data sets of simulated and real clinical data were fitted with WinBUGS, with and without informative priors. The posterior estimates of fittings with non-informative priors were used to build informative priors and the whole procedure was carried on in a consecutive manner. The estimates of fittings with informative priors where compared with meta-analysis fittings of the respective combinations of data sets. Also, approximate estimates of the population parameters were calculated by applying the Bayes theorem directly on the posterior distributions of the population parameters obtained by fittings with non-informative priors.
Results: Good agreement was found between the fittings with informative priors and the respective meta-analysis fittings, for the simulated datasets. Agreement was found not only on the population parameter estimates but also on the respective precisions. Also, reasonable agreement was observed in the clinical data, for most model parameters. Further, the computational times were much smaller for the prior method compared to the meta-analysis, due to the large datasets used with the latter.
Conclusions: The results of a population pharmacokinetic analysis may be summarized in Bayesian prior distributions that can be used consecutively with other analyses. The procedure is an alternative to the meta-analysis and gives comparable results. It has the advantage that is much faster than the meta-analysis, and can be performed when the summarized data are not actually available.
Reference: PAGE 13 () Abstr 484 [www.page-meeting.org/?abstract=484]
Poster: poster