0

I've found there is a way to create a forest plot of summary estimates with the 'metafor' package, which can be found here: Meta-analysis: Forest plot of summary estimates using metafor package

Is there a solution for the meta package, too? With over 30 studies, the byvar function produces a forest plot that does not fit the window.

Community
  • 1
  • 1
user3875022
  • 116
  • 4
  • Thanks to Guido Schwarzer, the maintainer of the 'meta' package, I've found a solution. A good solution is to save the plot in .pdf. You have to play around with the 'width' and 'height' parameters. – user3875022 Dec 08 '14 at 18:31

1 Answers1

0

Here an example: 

### the data

d <- structure(list(study = structure(c(1L, 2L, 3L, 4L, 5L, 6L, 7L, 
7L, 8L, 8L, 8L, 10L, 9L, 11L, 11L, 12L, 13L, 14L, 14L, 15L, 16L, 
17L, 18L), .Label = c("Bort et al. 2012                ", "Boyl et al. 2004                ", 
"Cart et al. 2007                ", "Coryet al. 2009                 ", 
"Cosoff 1998                     ", "Dell'a 2011                     ", 
"Dilan 2003                      ", "Dilton et al. 1997              ", 
"Mac et al. 2001                 ", "Man et al. 2006                 ", 
"Okan 2011                       ", "Orol et al. 2006                ", 
"Pinto et al. 2003               ", "Simone et al. 2004              ", 
"Strahowski et al. 1992          ", "Tamara 2002                     ", 
"Viera et al. 2001               ", "Zucchi et al. 2006              "
), class = "factor"), xi = c(60, 40, 13, 107, 3, 32, 1, 16, 33, 
1, 20, 46, 27, 30, 22, 78, 35, 33, 5, 2, 4, 3, 4), ni = c(200, 
140, 56, 427, 20, 508, 25, 19, 53, 32, 44, 191, 288, 50, 46, 
918, 151, 360, 115, 41, 70, 129, 80), group = structure(c(3L, 
3L, 1L, 3L, 3L, 3L, 1L, 1L, 2L, 1L, 1L, 3L, 3L, 1L, 2L, 3L, 3L, 
1L, 2L, 1L, 1L, 4L, 4L), .Label = c("anxiety     ", "depression  ", 
"mixed       ", "remission   "), class = "factor")), .Names = c("study", 
"xi", "ni", "group"), row.names = c(NA, -23L), class = "data.frame", codepage = 1252L)

attach(d)

### the code

### load the library

library(meta)

library(metafor)

#### the model 
#### Freeman-Tukey Double arcsine transformation
#### Empirical Bayes estimator  and Hartung and Knapp adjustment

model <- metaprop(xi,ni,sm="PFT",hakn=TRUE, method.tau="EB")

########################
#
# subgroup analyses
#
########################


modelsub <- update(model, byvar=group)

summary(modelsub)

forest(modelsub,studlab=paste(study), print.byvar=FALSE)

#################################
#
# solution to a poor forest plot
#
#################################

### a good solution is to save the plot in .pdf 
### you have to play around with the 'width' and 'height' parameters 
### I used  width=12,height=13 (as higher the values, as more space you will get)

pdf(file="good forest plot.pdf", width=12,height=13)

forest(modelsub,studlab=paste(study), print.byvar=FALSE)

dev.off()
Werner
  • 14,324
  • 7
  • 55
  • 77
user3875022
  • 116
  • 4
  • sessionInfo() R version 3.0.2 (2013-09-25) Platform: x86_64-w64-mingw32/x64 (64-bit) locale: [1] LC_COLLATE=Italian_Italy.1252 LC_CTYPE=Italian_Italy.1252 [3] LC_MONETARY=Italian_Italy.1252 LC_NUMERIC=C [5] LC_TIME=Italian_Italy.1252 attached base packages: [1] grid stats graphics grDevices utils datasets methods [8] base other attached packages: [1] metafor_1.9-3 Matrix_1.1-0 Formula_1.1-1 meta_3.7-1 psych_1.4.5 [6] foreign_0.8-57 loaded via a namespace (and not attached): [1] lattice_0.20-24 tools_3.0.2 – user3875022 Dec 08 '14 at 18:32