- On 16 Dec 2006 at 13:20:46, "V.G.S.Raghavendra Mamidi" (mamidi.ganesh.aaa.gmail.com) sent the message

Back to the Top

Dear All,

I have got a pivotal parallel study (two treatment) for statistical

analysis using SAS software. Since I have not dealt with this type of

design in SAS soft ware, could anybody suggest me how to proceed with

the statistical analysis.

Or else please tell me is it ANOVA one-way procedure used for a

parallel study (two treatment) in SAS soft ware,In ANOVA one way

could anybody tell me which test to be applied (Bartlette,

Duncan.....and so on).

And also please tell me what factors/effects (treatment or subject or

both) should be taken for statistical analysis in ANOVA one-way

procedure

Please tell me any guideline or books or any websites which gives a

clear picture regarding the statistical analysis of a parallel study.

Note: Kindly treat this as high priority and reply me at the earliest

as it is very urgent

--

Regards,

MGR

Statistical Analyst - On 16 Dec 2006 at 10:44:46, "Ed O'Connor" (Ed.OConnor.at.tandemlabs.com) sent the message

Back to the Top

You have got at least 2 treatments, multiple doses? multiple time

points, repeated measures. it would seem that you would be looking

at the differences of equivalence of the two drugs, each dose, each

time point so it should be drugXdoseXtime, that would be 3 way-main

effects drug then interactions drugXdose, drugXtime and doseXtime

with DRUGXDoseXtime, unless I am missing something. - On 17 Dec 2006 at 02:41:37, jfinsi (jamil.finsi.at.gmail.com) sent the message

Back to the Top

Hi,

Well the design path analysis would be oneway Anova and a bartlett's

test

Good luck - On 18 Dec 2006 at 10:37:22, "Nadeem Irfan Bukhari" (nadeemirfan_bukhari.-at-.imu.edu.my) sent the message

Back to the Top

The following message was posted to: PharmPK

Dear MGR

To me, for two treatments in parallel design, parameteric (unpaired t

test) or nonparameteric (MannWhitney test).

Nadeem - On 18 Dec 2006 at 10:23:01, Navdeep Coelho (Navdeep.Coelho.-a-.biovail.com) sent the message

Back to the Top

The objective of this analysis is not clear but assuming that it is a

bioequivalence type study and you have parametric data, then you can

analyze the data using PROC GLM. For a parallel design, you only need

to consider treatment as a factor in the model. If applicable, then

you might also want to log transform the parameters accordingly. And

since there are only two treatments, you don't need Duncan's test for

multiple comparisons. Bartlett's test is for testing non-normality of

the data.

You can also consult the FDA guidelines for guidance on the

statistical analysis.

Nav Coelho

Biostatistician, Biovail Contract Research

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 "Statistical Analysis of a Pivotal Parallel study by using SAS" as the subject

PharmPK Discussion List Archive Index page

Copyright 1995-2010 David W. A. Bourne (david@boomer.org)