Modelling Parallel bears [🇷 for BE/BA]

posted by yjlee168 Homepage – Kaohsiung, Taiwan, 2010-04-23 23:14 (5088 d 09:06 ago) – Posting: # 5207
Views: 55,669

Dear d_labes,

Thank you for your message.

❝ IMHO definitely yes.


O.k.

❝ To ask it again: What is the idea behind your model, where does it come from? Do you have a reference for me? lm() or lme() is only a question of considering some effects as random or not.


No reference for this. Simply try to translate the codes of Power to know into R. When we were developing bear for parallel BE study, we tried to get exactly the same results as Power to know or WinNonlin. The codes of Power to know for parallel BE study looks like
PROC GLM DATA=EXAMPLE;
CLASS SUBJ TRT;
MODEL LAUCT LAUCI LCMAX=TRT;
ESTIMATE ‘A vs. B’ TRT 1–1;
LSMEAN TRT;
RUN;

So we used lme() (not lm()) to meet this requirement for validation purpose.
lmeCmax_ss<-lme(Cmax_ss ~ drug, random=~1|subj, data=Data, method="REML" )
Just like analyzing a replicate crossover. I can not find the results obtained from analyzing with WinNonlin using a linear mixed effects model. As far as I could remember, we got pretty similar results as with WinNonlin at that moment. Then when search the answer for your question int his thread, I find that we probably can simply use lm() as what we do in non-replicate 2x2x2 crossover BE study, such as
lmCmax_ss<- lm(Cmax_ss ~ drug , data=TotalData)
Which model is more appropriate for analyzing parallel BE study with R?

All the best,
-- Yung-jin Lee
bear v2.9.1:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan https://www.pkpd168.com/bear
Download link (updated) -> here

Complete thread:

UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,636 registered users;
104 visitors (0 registered, 104 guests [including 9 identified bots]).
Forum time: 07:20 CET (Europe/Vienna)

With four parameters I can fit an elephant,
and with five I can make him wiggle his trunk.    John von Neumann

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5