Back to the Top
Hello, everybody.
I want to optimize my sample designs badly, However the software AdaptII needs the users to offer the parameters sigma inter(SDinter) and sigma slope(SDslope) themselves. Unfortunately, the authors only afforded the variance model description : var{v(t)} =(sigma inter + sigma slopey(t))2. As an example assume sigma inter =0.0and sigma slope =0.1 in Eq. (2.9). Then the variance model represents an error process with a standard deviation of 10% of the measured quantity (i.e.,10% coe\0xFB03cient of variation). While there is another different explanations for the two parameters of the same model: where sigma inter is closely related to the sensitivity of the analytical method and sigma slope is closely related to the coefficient of variation(abbreviated from Thesis of Pakawadee Sermsappasuk). Especially for the meaning of sigma slope, there exist contradictions. So my questions are listed as follows: firstly, which is the purpose of sample design, optimizing subsequent formal large sample size exp
Back to the Top
The following message was posted to: PharmPK
Dear Dr. Qi Feng,
From what I understand from ADAPT II, you do not necessary need to have a variance model equal to (SDslope*Y + SDinter)^2; it is an error model suggested in the documentation of ADAPT II. Other type of error model could be used. The purpose of optimal sampling design is to get the most information from you data with the minimum of samples based on the a priori the model and variance, at least this how I interpret it.
If you already have data, fit them (with ADAPT or your favorite software) and then you can use these parameters in the optimal sampling of ADAPT. Perhaps it would be of interest to you to know that ADAPT 5 is now available at
http://bmsr.usc.edu/Software/ADAPT/ADAPT.html
I would use either ITS or MLEM module of ADAPT 5 to get population estimates.
I hope this help!
Jean
--
Jean R. Lavigne M.Sc.
Associate Director, Clinical Pharmacology
MDS Pharma Services
4705 Dobrin
St-Laurent, Quebec, Canada
H4R 2P7
|
Want to post a follow-up message on this topic?
If this link does not work with your browser send a follow-up message to PharmPK@boomer.org with "Sample design" as the subject | Support PharmPK by using the |
Copyright 1995-2011 David W. A. Bourne (david@boomer.org)