[R-meta] help with meta-analysis

[Viechtbauer Wolfgang (SP)]

Dear Nadine,

I took a quick look at your code. I think you are not specifying the right model. Please take a look at:

http://www.metafor-project.org/doku.php/analyses:konstantopoulos2011#a_common_mistake_in_the_three-level_model

This aside, this doesn't fully account for use of the same control group in multitreatment studies. To do so, you need to compute the covariance among the sampling errors as described here:

http://www.metafor-project.org/doku.php/analyses:gleser2009

This, plus a multilevel model (with random effects for studies and estimates within studies) gets you a model that I would consider appropriate.

As for studies with 0 events in both arms -- some would argue that they should be dropped completely (which would imply using drop00=TRUE). This is a matter of debate. You can always do a sensitivity analysis.

Best,
Wolfgang

-----Original Message-----
From: R-sig-meta-analysis [mailto:r-sig-meta-analysis-bounces at r-project.org] On Behalf Of Nadine Vogt
Sent: Sunday, 22 October, 2017 15:02
To: r-sig-meta-analysis at r-project.org
Subject: [R-meta] help with meta-analysis

ATTACHMENT(S) REMOVED: Challenge trials for R with clustering.txt | Metaanalysis in R.R 

Hi everyone,

I am new to meta-analysis and relatively new to R, and am hoping for some advice regarding a meta-analysis! We are interested in the ability of a vaccine to reduce clinical signs in dogs, and used a dichotomous outcome to characterize the presence/absence of clinical signs in vaccinated and non-vaccinated dogs. 

I am currently using a rma.mv model to account for clustering by trials which use the same control group for multiple treatment groups. I recently stumbled across the paper by Ruecker et al. 2017 (http://onlinelibrary.wiley.com/doi/10.1002/jrsm.1259/abstract) and am now wondering if I should be making an adjustment for the use of the same control group using one of those methods as opposed to using clustering methods?

I've attached my dataset and code.

If the rma.mv model is appropriate, I just want to confirm that the studies with 0 events in both arms should be included. I am also wondering how the weights for the studies are calculated in rma.mv, and whether I need to adjust them for the studies with 0 events (although it seems to already do this, by assigning relatively low weights)?

Thanks in advance!

Nadine Vogt