[R-meta] Using meta package when SMDs and SEs/95% CIs are already known - R-SIG-meta-analysis

Wed, Jun 21, 2023 4:14 AM #

Hello,

I am currently conducting a meta-analysis for which I have already calculated SMDs with corresponding standard error/95% CIs. Rather than redoing the calculations, I was wondering if the meta package for R allows me to use SMDs and CIs as input to run a meta-analysis. I have gone through the documentation and it seems like I cannot, but I would love to know about any other possibilities.
The goal is to run a limit meta-analysis using the limitmeta function from the metasens package. Since this function takes an object of class 'meta' as an argument, I need to use a function that actually outputs such an object. Any help is very much appreciated. Thank you in advance!

Best,

Ben Glaser

Tobias Saueressig

Wed, Jun 21, 2023 4:42 AM #

An HTML attachment was scrubbed...
URL: <https://stat.ethz.ch/pipermail/r-sig-meta-analysis/attachments/20230621/c0bc81bb/attachment.html>

Guido Schwarzer

Wed, Jun 21, 2023 7:15 AM #

I should have read this reply first before posting my answer. ;-)

It is also possible to provide study data with information on SEs or confidence limits, i.e.,

metagen(mySMD, myseSMD, lower = lowerSMD, upper = upperSMD, sm = "SMD")

The standard error is only calculated from the lower and upper limits if the standard error of a study is missing.

Best,
Guido