Advances and Applications in Statistics
Volume 47, Issue 3, Pages 247 - 263
(December 2015) http://dx.doi.org/10.17654/ADASDec2015_247_263 |
|
RECYCLED ESTIMATION OF POPULATION PHARMACOKINETICS MODELS
Shaul K. Bar-Lev and Benzion Boukai
|
Abstract: This paper proposes a specialized re-sampling scheme appropriate for estimation of the population parameters in the mixed effects nonlinear regression models often used in clinical pharmacokinetics. The approach taken here is based on the observation that the random weighting method for re-sampling discussed in [19], [11] and in [3], can effectively be generalized for hierarchical situations involving mixed-effects and a ‘sample’ of observed drug-concentration curves from a population of interest. This approach enables us to capture, via the repeated recycling (with appropriately chosen random weights) of the model’s parameter estimates, the respective sampling distributions and the various components of the intra-individual as well as the inter-individual variability. Thereby, this methodology allows for a data-based assessment of the model’s adequacy as well as of the accuracy of the parameters estimates. We illustrate this proposed recycling methodology with an application to a real data set from clinical pharmacokinetic study as well as by simulations. |
Keywords and phrases: nonlinear regression, hierarchical models, mixed effects, re-sampling, recycled, bootstrapping, population pharmacokinetics. |
|
Number of Downloads: 397 | Number of Views: 929 |
|