I am attempting to conduct a secondary random-effects meta-analysis (umbrella analysis) of pooled prevalence of mental health problems during COVID19 using the metaprop() function from the meta package in R. The vast majority of studies in the datafile report only the pooled prevalence and CIs and total sample size, rather than n events & total sample size.
I am aware that I can create a proxy n events variable using the pooled prev & sample size and I can run a metaprop model using this input, but I have a feeling this is not necessarily the most correct approach, as it is not taking into account the reported error/variance of the pooled prev. So far, I have not been able to find a way to conduct this analysis using the reported pooled prevalence and CI estimates from the primary meta analyses. Is this possible within the meta or metafor packages in r?
Note., I looked into the metaumbrella package, which would be perfect except it does not yet have functionality for analyses with pooled prevalence.
Thanks in advance for any guidance!