[R-meta] Question regarding metarate calling the Poisson model for meta-analysis

[Guido Schwarzer]

Hi all,

I wanted to implement exact Poisson confidence limits for individual 
study results in metarate() for some time. This thread motivated me to 
finally do it.

The development version of metarate() on GitHub has a new argument 
'method.ci' to specify the type of CI limits: "Poisson" or "NAsm" for 
normal approximation

install.packages("remotes")
remotes::install_github("guido-s/meta")

library(meta)

data(dat.nielweise2008, package = "metafor")

m <- metarate(x1i, t1i, data = dat.nielweise2008,
   studlab = paste0(authors, ", ", year),
   method.ci = "Poisson", irscale = 1000)

print(summary(m), digits = 2)

forest(m)

Best, Guido