|
Activity Number:
|
486
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 10, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #304934 |
|
Title:
|
Locally D-Optimal Designs Based on Models Combining Emax and One-Compartment Models
|
|
Author(s):
|
Sam Hedayat*+
|
|
Companies:
|
University of Illinois at Chicago
|
|
Address:
|
851 S. Morgan Street, Chicago, IL, 60607-7045,
|
|
Keywords:
|
pharmacokinetic ; pharmacodynamic ; compartmental model ; non-linear model ; Emax model ; d-optimal design
|
|
Abstract:
|
Four- and five-parameter models combining a pharmacokinetic compartmental model and a pharmacodynamic Emax model are studied here. Locally D-optimal (LD) designs and two classes of robust designs are investigated based on these two models. The required support size of a locally D-optimal design is found to be between four and six, based on the four-parameter combined model. Further, a sufficient condition of when the support requires the minimum number of design points is given based on the four-parameter model. Examples of the LD designs are presented for a certain set of nominal parameter values. Simulation results show LD designs and robust LD designs cannot be traded in terms of coefficient of variation and point estimation if the nominal values of the unknown parameter aren't close to the true values. It's demonstrated that LD design should be employed instead of robust LD designs.
|
