[R] Meta-Analysis: Metaplot Labels, SE, and Summary

[Michael Dewey]

At 02:37 29/06/2009, Ishwar Bridgelal wrote:
>Hello,
>I'm ashamed to admit that I'm an R and meta-analysis neophyte just when my
>colleagues and I are attempting to run R for that latter purpose. We would
>like to construct a graph of the 95% confidence intervals of odds ratios
>(inverse odds ratios, actually). We are using the metaplot() function in the
>"rmeta" package. While we can construct a basic plot, we cannot configure it
>as desired.

library(rmeta)
?forestplot


>One, we would like to label each stratum with the study author(s) and date,
>but when we include alphabetical characters in a column in our matrix
>(labels=id, where id=m1[,5] , m1 is a matrix and [,5] is the column
>containing authors and dates), we obtain the following error message:
>"expected 'a real' got 'Kovacs'. (Kovacs is the author of the first study
>listed in the matrix.) How can we label the plot with authors and dates if
>this method of inputting data does not work?
>
>Finally, we are unclear about the arguments "summn", "sumse", and "sumnn".
>To what should we set these arguments in order to create a summary line that
>illustrates the confidence interval of the mean of all effect sizes in our
>meta-analysis?
>
>Thank you very much. I apologize for my proven inability to resolve these
>issues without assistance.
>
>Best,
>
>Ishwar Bridgelal
>University of Southern California
>
>         [[alternative HTML version deleted]]

Michael Dewey
med at aghmed.fsnet.co.uk
http://www.aghmed.fsnet.co.uk/home.html