LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (January 2008, week 1)Back to main SAS-L pageJoin or leave SAS-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:         Mon, 7 Jan 2008 08:10:24 -0800
Reply-To:     RolandRB <rolandberry@HOTMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         RolandRB <rolandberry@HOTMAIL.COM>
Subject:      Re: %unistats (and %unipvals) updated for adjustment for centre
Comments: To:
Content-Type: text/plain; charset=ISO-8859-1

On Jan 6, 1:41 pm, RolandRB <> wrote: > I have updated my %unistats macro (which calls the %unipvals macro) so > that you can adjust for centre effects for the PROC GLM call. Only one > centre variable can be specified and the model statement will be > generated either in this form for the default modelform=short > parameter setting: > > model response=trtcd centre /ss1 ss2 ss3 ss4; > > or for modelform=long as: > > model response=trtcd centre centre*trtcd /ss1 ss2 ss3 ss4; > > By setting dsmodelanova= you can save the ModelANOVA dataset and so > post-process the p-values. You can change errortype=3 to another value > if you need to for selecting on p-values. > > New versions can be browsed on the "updates" page or downloaded from > the download page. > >

If people wonder why I have updated the macros for adjustment for centre effects then it is because of the ICH guidelines. See ICH E3 and ICH E9 3.2

But then I am sure everybody follows these guidelines with respect to multicentre trials and they do all the recommended analysis. And I guess everybody's standard software for analysing clinical trials can handle this. The answer is, naturally, "of course!". Have a Happy 2008, everyone!


Back to: Top of message | Previous page | Main SAS-L page