[R-meta] Zero seTE in network MA

[Guido Schwarzer]

Dear Carla,

I discussed the check for zero treatment arm variances with Gerta. We 
realized that we have been overcautious with respect to zero treatment 
arm variances which can happen in real live (as your example nicely 
shows). We changed the behaviour of netmeta() such that a warning 
instead of an error is triggered for zero treatment arm variances 
(similar to the warning for re-ordering of treatment arms).

This change is implemented in the development version of *netmeta* 
available on GitHub:

install.packages("devtools")
devtools::install_github("guido-s/netmeta", ref = "develop")


In your second example, treatment effects are inconsistent as you missed 
a minus sign for the comparison exotic.forest vs forest.harvested (see 
attached R code). ;-)

Best wishes,
Guido

-- 
Dr. Guido Schwarzer (sc at imbi.uni-freiburg.de)
Institute for Medical Biometry and Statistics
Stefan-Meier-Str. 26, D-79104 Freiburg | Phone: +49 (0)761 203 6668
http://www.imbi.uni-freiburg.de        | Fax:   +49 (0)761 203 6680

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://stat.ethz.ch/pipermail/r-sig-meta-analysis/attachments/20170922/d243d42f/attachment.html>
-------------- next part --------------
## Read dataset with wrong comparison for 'Exotic forest' vs 'Harvested forest'
##
example <- read.csv("example.csv")


## Correct dataset
##
example.corrected <- example
example.corrected$TE[4] <- -example.corrected$TE[4]


library(netmeta)
options(width = 120)


## Use corrected dataset
##
netmeta(TE,seTE,treat1, treat2, studlab, data = example.corrected,
        comb.random = TRUE, sm="Log response ratio", details.chkmultiarm=TRUE,
        ref = "forest.harvested")


## Just 1 study
##
netmeta(TE,seTE,treat1, treat2, studlab, data = example.corrected[-1, ],
        comb.random = TRUE, sm="Log response ratio", details.chkmultiarm=TRUE,
        ref = "forest.harvested")


## Error ...
##
netmeta(TE,seTE,treat1, treat2, studlab, data = example,
        comb.random = TRUE, sm="Log response ratio", details.chkmultiarm=TRUE)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: example.csv
Type: text/csv
Size: 380 bytes
Desc: not available
URL: <https://stat.ethz.ch/pipermail/r-sig-meta-analysis/attachments/20170922/d243d42f/attachment.csv>