vijay
☆    

2008-04-11 18:36
(5829 d 17:54 ago)

Posting: # 1769
Views: 3,593
 

 Three Way Cross over Dose Proportionality [🇷 for BE/BA]

Hello All,

I am analyzing data from a study with the following design.
3 treatments (1-5mg IV; 2-5mg Extravascular (EV); 3-10mg EV) of a well known drug with highly variable PK assigned randomly to 8 subjects in a three-way cross over, 3 period study. All subjects receive the three treatments by the end of 3 periods of the study. (Total number of sequences =5; sequences not distributed evenly among subjects) The aim of the study was to asses absolute BA and dose proportionality of this EV formulation administered at two different doses (5 and 10 mg).

Intuitively, looking at the plots and the ratio of the PK parameters we can conclude that the two EV doses are approximately dose proportional. (Normally if there are more than 2 doses I could have performed a regression analysis)
In order to report statistical significance I carried out an ANOVA to see if the 90 % CI on the ratio of log transformed, dose-normalised PK parameters (Cmax, AUC0-60 and AUC0-inf) are within an acceptable range of 80-125 %.
I use R for my analysis. My 2 questions are as follows:
1. While doing ANOVA should I be analyzing the data including sequence as a Fixed effect along with Treatment and Period? I ask this because, the main aim is to compare the two EV formulations, but as the IV formulation is random within the sequence for different subjects, should I include the IV data while analyzing the study?
2. Can I do a straight forward head on pairwise comparison between the two EV doses if I know from the ANOVA model above (in 1) that there are no period, treatment and sequence effect to get my CI's.

Thanks
Vijay
UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,639 registered users;
110 visitors (0 registered, 110 guests [including 6 identified bots]).
Forum time: 11:30 CET (Europe/Vienna)

Nothing shows a lack of mathematical education more
than an overly precise calculation.    Carl Friedrich Gauß

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