[R] Metafor: Moderator variables for each study arm

[Viechtbauer Wolfgang (STAT)]

Dear Faiz,

in principle, you can include whatever moderators you want in a meta-regression model. A technical issue here is that the two follow-up variables are probably strongly correlated, so this may create issues with the model fitting and it also complicates the interpretation of the results (the notion of: what happens to y when increasing x1 while holding x2 constant is then questionable). Since the dependent variable in meta-analyses is (typically) a measure of a group contrast, it may also be reasonable to use the contrast (i.e., difference) of the follow-up times as a moderator. It all depends on what your question is: Do you believe that the *difference between the two groups* depends on the absolute follow-up time or rather on the difference in follow-up duration between the two groups?

Best,

--   
Wolfgang Viechtbauer, Ph.D., Statistician   
Department of Psychiatry and Psychology   
School for Mental Health and Neuroscience   
Faculty of Health, Medicine, and Life Sciences   
Maastricht University, P.O. Box 616 (VIJV1)   
6200 MD Maastricht, The Netherlands   
+31 (43) 368-5248 | http://www.wvbauer.com   


> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org]
> On Behalf Of Faiz Ahmad Khan
> Sent: Monday, February 20, 2012 16:00
> To: r-help at r-project.org
> Subject: [R] Metafor: Moderator variables for each study arm
> 
> Hello,
> 
> I am trying to do a meta-analysis where each study has two arms, similar
> to the BCG data set. However, follow-up duration was different for each
> study arm, so I would like to fit a model that uses the length of follow-
> up in each arm as a moderator. Is this possible?
> 
> For example, if this was the BCG data, each study would have 2 follow up
> variables: 1 for the treatment group, and 1 for the control group.
> 
> thanks for the help,
> Faiz
> 
> .
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-
> guide.html
> and provide commented, minimal, self-contained, reproducible code.