Predicting Shrinkage of Individual Parameters in More Complex NLME Models Using Bayesian Fisher Information Matrix

Authors: Gaurav M, Castleman A, Owen JS
Conference: AAPS
Keywords: Bayesian methodology, BPFIM, Complex NLME Models, Shrinkage
Division: Clinical Pharmacology and Pharmacometrics (CPP)

PURPOSE

When data are sparse, parameters derived from a non-linear mixed effects model analysis can shrink to the mean and can be misleading. The objective of this project was to predict the shrinkage on parameters using Bayesian methodology and test whether the results of a published 1 compartment model example by Combes et al., are applicable to more complex models.

Presented at: American Association of Pharmaceutical Scientists (AAPS) PharmSci 360; November 3-6, 2019; San Antonio, TX

By, Mitali Gaurav, Andrew Castleman, Joel S Owen