[R-meta] Question about netmeta

[Gerta Ruecker]

Dear Antonio,

Thank you for your question. netmeta is based on a consistency model. In 
addition, there are some features to investigate (in)consistency, see 
functions decomp.design() (to look at consistency under the assumption 
of a full design-by-treatment interaction random effects model) or 
netheat() (for visualization) or netsplit() (for separating direct and 
indirect evidence). However, there is no modelling of inconsistency 
providing parameter estimates under an inconsistency model. Insofar, the 
sentence in the cited reference ("The modeling process provides flexible 
options for the incorporation of heterogeneity and inconsistency in the 
estimation") is misleading.

To briefly answer your question below: If you run netmeta, your results 
(parameter estimates) are always based on the consistency model.

Best,

Gerta


Am 08.09.2017 um 11:13 schrieb antonio.facciorusso at virgilio.it:
> Dear
>        Prof Schwarzer and Prof. Rucker,
>        i am Dr Antonio Facciorusso from the University of Foggia
>          (Italy).
>        I have recently submitted a network meta-analysis (in the
>          field of gastrointestinal endoscopy) performed with netmeta,
>          and among the points raised by the reviewers, i was asked
>          whether i used a consistency model or not.
>        My question is: does netmeta allows to perform both
>          models (consistency and inconsistency random effect network
>          meta-analysis)? If so, which is the code to adopt a consistency
>          random effect model (as suggested by the reviewer)?
>        In this paper i read that netmeta allows both models
>          but i did not understand which one i obtained
>        http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0115065#pone-0115065-t001
>
>        
>        
>
>        
>        I used the following codes:
>
>   p1 <- pairwise(list(Treat1, Treat2),list(ADR1,ADR2),time=list(N1,N2),studlab=Study,data=x)
>   where ADR is the outcome of interest (Adenoma detection rate)
> net1=netmeta(p1,sm="RR",comb.fixed=FALSE, comb.random=TRUE)
>
>
> Thank you in advance for your kind response
>        
>
>        
>        Antonio
>      
>      
>
>      
>    
>
>
>
>
>
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-meta-analysis mailing list
> R-sig-meta-analysis at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis