[R-meta] meta-regression for incidence rate

Michael Dewey ||@t@ @end|ng |rom dewey@myzen@co@uk
Mon Jun 1 17:05:09 CEST 2020


Dear Giorgio

This may be related to a long discussion on this list recently. You can 
see the original post 
https://stat.ethz.ch/pipermail/r-sig-meta-analysis/2020-May/002082.html 
and then follow the rest of a quite extensive thread. I am not 
sufficiently expert in time-to-event studies to say whether it does 
answer your question but it might be informative.

Michael

On 31/05/2020 19:27, Giorgio Maria Agazzi wrote:
> Dear all,
> I am planning to do a meta-regression with the intent to compare trial
> patient characteristics to the incidence of an outcome in the time period
> between t=0 and t=2 years. Every study reports the number at risk at t=0
> and t=2 years and the kaplan-meier curve, so I'm able to digitize the curve
> and extract the percentages of events.
> I was thinking of doing a meta-regression and using as effect size the
> freeman-tukey transformed incidence rate. I'm not sure this is the best way
> and I have also another question, how can I calculate the total number at
> risk as requested by metafor? for the number of events I can calculate it
> by digitizing the plot and calculating the percentage of people that had
> the outcome.
> 
> Thank you very much!
> 
> gma
> 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> R-sig-meta-analysis mailing list
> R-sig-meta-analysis using r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis
> 

-- 
Michael
http://www.dewey.myzen.co.uk/home.html



More information about the R-sig-meta-analysis mailing list