Back to the Top
The following message was posted to: PharmPK
Dear All,
If we conduct the BE study using 2 way crossoverdesign and from ANOVA
results some say that we could calculate the intra-subject variation.
Could you please tell me thw way to calculate it which MS error
should we
use (from Sub(seq)???). Another point is how to calculate the 90%
Confidence interval from Geometric mean, can we use Excel software?
Thanks
Wichittra
--
Dr. Wichittra Tassaneeyakul
Department of Pharmacology
Faculty of Medicine
Khon Kaen University
Khon Kaen 40002
Thailand
Back to the Top
The following message was posted to: PharmPK
Dear Wichittra,
while calculating ANOVA for two-way cross over design,PK Parameters
Cmax,AUC's are calculated using
mixed effect ANOVA model using Type III sum of squares and taking in
to consideration different factors
responsible for variation. These factors are tested at different
level of significance depending upon
levarge of factors.
Intra Subject variability is nothing but it is considered as residual
variability after accounting for variabilty
between treatment,period and subject and it is a major determinant to
calculate estimated sample size.
for calculating 90% CI following procedure is adopted
90% confidence interval for the ratio of the test and reference
product averages (least square means) are
calculated using the 90% confidence interval for the differences in
the averages (arithmetic means) of the
log-transformed data and then taking the antilog of the obtained
confidence limits
we can use Excel for calculating 90% CI, But performing ANOVA in
Excel,does not provide more detailed summary
outputs as compared to SAS and other statistical softwares.
Best Regards,
Nand Kishore Rawat
Bio-Statistician (CPP)
Ranbaxy Research Laboratories
Gurgaon-122001 India
Back to the Top
To calculate intra-subject variation you can use the formula below:
100*sqrt(exp(MSE)-1)
Bang Qian Xu
Back to the Top
Which error term to use depends on which SAS proc you use. If you
use GLM then use subject nested within sequence as the error term.
If you use the Mixed procedure, you use the true error term from the
model.
The geometric mean can be calculated as first taking tha natural log
of the data then calculating the arithmetic mean of the natural log
transformed data. The arithmetic mean is then exponentiated. This
is the geometric mean.
Good luck,
pete bonate
Back to the Top
The following message was posted to: PharmPK
Dear Wichittra!
You wrote:
>If we conduct the BE study using 2 way crossoverdesign and from ANOVA
>results some say that we could calculate the intra-subject variation.
>Could you please tell me thw way to calculate it which MS error
>should we use (from Sub(seq)???).
>
Yes.
CV-intra = sqrt(exp(MSE)-1)
>Another point is how to calculate the 90%
>Confidence interval from Geometric mean, can we use Excel software?
>
Funny, we had this today ;-)
Have a look at this thread:
http://forum.bebac.at/mix_entry.php?id=559
Good look,
Helmut
--
Helmut Schuetz
BEBAC
Consultancy Services for Bioequivalence and Bioavailability Studies
Neubaugasse 36/11
1070 Vienna/Austria
tel/fax +43 1 2311746
Web http://BEBAC.at
BE/BA Forum http://forum.bebac.at
Back to the Top
The following message was posted to: PharmPK
Dear All,
we are conducting Bio-equivalence study on Parallel design. Can
someone Pls let me know
the formula to Calculate InterCV in Parallel design using SAS
Software.
Best Regards,
Nand Kishore Rawat
Bio-Statistician (CPP)
Ranbaxy Research Laboratories
Gurgaon-122001 India
Back to the Top
Dear Wichittra,
You wrote:
>If we conduct the BE study using 2 way crossoverdesign and from ANOVA
>results some say that we could calculate the intra-subject variation.
>Could you please tell me thw way to calculate it which MS error
>should we use (from Sub(seq)???).
My Reply:
MS error from Sub(seq)is used to calculate inter subject variation.
To calculate intra-subject variation, MS error obtained from entire
ANOVA
model is used.
Best Regards,
harshada
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 "Intra-subject variation calculation" as the subject | Support PharmPK by using the |
Copyright 1995-2011 David W. A. Bourne (david@boomer.org)