I hope this isn't a stupid question ... when combining 2X2 cross-over studies and parallel group studies in an IPD (individual patient data) meta-analysis, is it normal to discard the 2nd period of the cross-over studies? It seems terribly wasteful. I've read the section of Whitehead's book titled 'inclusion of trials with different study designs' and it seems to recommend that we obtain estimates of the treatment effect for each study and then include these in a meta-analysis. Is this equivalent to the pooled IPD meta-analysis I want to run?
Cheers
Mike
_________________________________________________________________
The next generation of MSN Hotmail has arrived - Windows Live Hotmail
http://www.newhotmail.co.uk
|