[R-meta] robust variance estimator in meta-analyses of rare events (proportions)

Guido Schwarzer @c @ending from imbi@uni-freiburg@de
Thu Oct 4 14:32:52 CEST 2018


You only provided R code to conduct a classic meta-analysis using 
inverse variance weights.

Instead of using the inverse variance method, Stijnen et al. (2010) - 
part of your reference list - suggest to use generalized linear mixed 
models for meta-analysis of proportions; see section 3.1.

This method is available in R packages metafor (R function rma.glmm) and 
meta (R function metaprop calling internally rma.glmm from metafor).

Best wishes,

More information about the R-sig-meta-analysis mailing list