Add prediction interval forest plot (package "meta")

Dear all,

I am struggling to add a prediction interval to a forest plot that was
created with forest.meta(), package "meta".

I checked the source of forest.meta() and realized that it is heavily
relying on grid. I am lacking any experience with grid graphics. So, I
am having difficulties to find out where the random effects estimate is
actually plotted in order to add prediction intervals.

Any help is much appreciated!

Bernd


R version 3.0.1 Patched (2013-05-28 r62825)
Platform: x86_64-w64-mingw32/x64 (64-bit)

locale:
[1] LC_COLLATE=German_Germany.1252  LC_CTYPE=German_Germany.1252
LC_MONETARY=German_Germany.1252
[4] LC_NUMERIC=C                    LC_TIME=German_Germany.1252

attached base packages:
[1] grid      stats     graphics  grDevices utils     datasets  methods
  base

other attached packages:
[1] meta_2.3-0

loaded via a namespace (and not attached):
[1] tools_3.0.1

At 19:34 06/06/2013, Bernd Weiss wrote:

Dear all,

I am struggling to add a prediction interval to a forest plot that was
created with forest.meta(), package "meta".

I checked the source of forest.meta() and realized that it is heavily
relying on grid. I am lacking any experience with grid graphics. So, I
am having difficulties to find out where the random effects estimate is
actually plotted in order to add prediction intervals.

I appreciate this is not a direct answer to your question but the
equivalent function in metafor (available from CRAN) does this and it
might be easier to use that.

Am 07.06.2013 13:17, schrieb Michael Dewey:

At 19:34 06/06/2013, Bernd Weiss wrote:

Dear all,

I am struggling to add a prediction interval to a forest plot that was
created with forest.meta(), package "meta".

I checked the source of forest.meta() and realized that it is heavily
relying on grid. I am lacking any experience with grid graphics. So, I
am having difficulties to find out where the random effects estimate is
actually plotted in order to add prediction intervals.

I appreciate this is not a direct answer to your question but the
equivalent function in metafor (available from CRAN) does this and it
might be easier to use that.

Thanks, Michael!

Yes, I vaguely remember that the metafor package 
can calculate the PIs but most of our analyses 
rely on the meta package. On the other hand, 
this might be a good reason to finally switch to metafor.

You probably refer to the function addpoly() 
which can be used for that. I will check it out.

Again, thanks for your help,

Bernd