[R-meta] Meta Analysis with HR values

[Guido Schwarzer]

Am 17.02.19 um 20:32 schrieb Awais Wahab:
> Dear Sir/Madam!
> I am doing meta analysis in R, and have HR and CIs values of different
> studies. There are couple of studies which have reported Disease specific
> survival (DSS) and Disease free survival (DFS) and overall survival
> analysis is presented in percentages in different studies. .
>
> So, is there any method to convert these percentages into Hazard Ratios to
> do meta analysis by combining them with studies having similar results like
> (DSS, DFS and OSS).

Methods exist to approximate a hazard ratio from Kaplan-Meier curves, 
e.g. Tierney et al. (2007), which provide several survival probabilities 
/ percentages.

However, I am not aware of any reliable method to approximate a hazard 
ratio from a single time point, e.g., 5 year survival probability in 
experimental and control group.

Best wishes, Guido

Reference:
Tierney JF, Stewart LA, Ghersi D, Burdett S, Sydes MR. Practical methods 
for incorporating summary time-to-event data into meta-analysis. Trials. 
2007;8:16.

	[[alternative HTML version deleted]]