Non-comparative studies in a meta-analysis

#1
I am currently busy with the statitics of a meta-analysis using non-comparative studies. Therefor I used the RevMan tool by Cochrane Initiative. But because the studies van non-comparative it does not work. I tried it in different ways, but I can't make it work. Can someone help me pool this data in any way and with any program and compare them too each other?
 
#3
As shown in the attached file, the data I gathered for the treatment modality volar plating are from other studies than the data I gathered for the treatment modality plaster cast. So in the studies that I included in my meta analysis non compare both treatments directly.
 

hlsmith

Omega Contributor
#4
If I remember correctly we had a conversation like this one before. I have only used RevMan once, so I can't help with particulars. Though, if exposure A was compared to exposure B, while exposure B was also compared to exposure C, you can work towards finding differences in A compared to C using Network Meta-analysis, which has been a developing and popular area of research over the past 5 years. Though, I am unsure if RevMan has functionality related to these comparisons.
 
#5
Thank you once again. But I still am not able to do the analysis. I have decided to compare all treatments to the Golden Standard of plaster casting. So no network meta analysis. I have read the book Introduction of Meta-Analysis by Borenstein which gives a good inside into meta-analysis. I have tried different programs like RevMan, RStudio with the Metafor package, SPSS with Wilson Macros. RevMan is very easy but doesn't work for my analysis because the data from the different treatments are from different studies and not directly compared in the studies. For weeks/months I have been working on the statistics.

If I know the heterogeneity and z score I know to use either the fixed effects model or the random effects model and get the RR's for the treatment modalities I want to compare. I have 5 treatment modalities in total which I want to compare. I want them all to be compared with the Golden Standard (plaster casting).

I just am not able to pool and compare the data I found and shown in the attached file above.
 

hlsmith

Omega Contributor
#6
I would like to become more familiar with meta-analyses etc., but I am fairly busy right now and do not have any topics that I would like to investigate. The following is a link to R's netmeta package. I realize it is not easy to just jump right into R and these analyses. Have you run basic meta-analyses yet. Perhaps you need a progression of steps to get you to where you need to be.


If I go to do one of these in the future, I will let you know and update this thread.


Thanks.


https://cran.r-project.org/web/packages/netmeta/netmeta.pdf
 
#7
Your help is very much appreciated! I really mean it. Thanks for the time you have spend answering my questions. Do you perhaps know someone who can help me or a website that can help me besides the netmeta page of R?
 
Last edited: