Meta-analysis: calculating adjusted effects based on meta-regression result

[Clemens Zimmermann]

Hi!

I am working on a random-effects meta-analysis and meta-regression has identified a highly significant continuous covariate (time). I would now like to calculate an adjusted effect estimate for the random-effects model - ideally also for a subgroup analysis where the subgroups have different means for the variable time, which means I need to be able to manually select a specific time point for adjustment.

Does someone know which command can be used for that? I use metan but am happy to try something else if that will give me an adjusted estimate, maybe even on a graph (forest plot). I found a solution for R (https://www.metafor-project.org/doku...sted_effects?s[]=effect&s[]=estimate) but not for Stata. Can someone help?

Cheers,
Clemens

[Bruce Weaver]

Hello Clemens Zimmermann. I see that was your first post to the forum, so welcome to Statalist. I don't have time right now to study your problem in detail, but I just wanted to make sure you know that you can use the -margins- command after -meta regress-. I suspect you can get it to produce the output you're looking for. Good luck.

[Clemens Zimmermann]

Thanks Bruce Weaver, both for the welcoming words and the hint with the margins command. I will play around with it and see how it goes!