Meta analysis for subgroups
Hi everyone! I need some help. Is there anyone that use JASP for meta analysis? I have some subgroups including different drug treatment and the results include their effects on blood glucose levels. Some of them increase but some of them decrease the blood glucose levels. I think I need to perform subgroup analysis evaluating the different subgroup effects in the meta analysis.How could I do it? Thank you in advance
Comments
Hi, you have two options: you can use a meta-regression where you use the variable as a factor predictor. This allows you to estimate the treatment levels at the different levels of subgroups. You can use the "Estimated marginal means" to obtain the subgroup-specific treatment estimates. The advantage of this is that you use all of your data to estimate the between-study heterogeneity.
Alternatively, you can use the "Filter" feature in the data editor to select only some levels of the subgroup variable and perform the meta-analysis on a subset of the data set.
Cheers,
Frantisek