CRSU-Apps/MetaInsight

Single arm studies should be flagged during validation

JanionNevill opened this issue · 7 comments

JanionNevill commented

When data includes single arm studies, the analysis functions cannot be run, and produce errors. This is a data issue which should be flagged at data upload, like with other data issues.

bischi999 commented

Hi, how can I include single arm studies? is there a workaround option?

JanionNevill commented

@bischi999 Unfortunately this is a limitation of the packages that we use to run the analysis, so there is no known workaround.

bischi999 commented

@JanionNevill thanks a lot for your answer!
Do you know a software or code where this is possible?

JanionNevill commented

@bischi999 I'll ask my colleagues if they have any suggestions.

BalancedmonkeY commented

Hi @bischi999!
Including single-arm studies into a network meta-analysis is not a straight-forward process. There are a few papers in the literature for how to adapt current methods to include single-arm studies, but I wouldn't say any of these methods are 'standard practice', or that one method has been universally agreed.
As such, I am not aware of any software/code that does this.

My advice would be to consult a statistician in your institution that can help you decide whether including the single-arm studies is necessary, and if so, running an advanced model/method to incorporate them.

bischi999 commented
BalancedmonkeY commented

I have not come across that before and sounds very specific to your project. Unfortunately, I cannot advise any further. Furthermore, we try to keep user feedback/queries to just the tools themselves and how to use them/improve usability/functionality.

I hope you find the expertise you need :)