Multivariate Random Effects Meta-regression

[Dearbhla Kelly]

Hello!

Does anyone have an appropriate syntax for this in a pooled analysis of RR/95% CIs with various covariates?

Any help would be much appreciated!

D

[Marcos Almeida]

You may wish to read this text.

[Dearbhla Kelly]

Thanks Marcos. I've been using this syntax:

metareg lnrr Variable1 Variable2 Variable3, wsse(se) eform

Is this correct for a multivariate meta-regression? The problem I encountered when I tried to put all the covariates in is that it says "insufficient observations" which I don't understand...

[Marcos Almeida]

The best approach for your query is to follow the FAQ advice. Sharing command, results and (real or not) data (for this, use CODE delimiters or - dataex - command) may underline the problem.