2
$\begingroup$

Is it possible to a meta-analysis of hazard ratios, where the hazard ratios are from a continuous covariate with different ranges in the different studies? The covariate is a log base 2 transformed microarray gene expression from different cohorts.

$\endgroup$
0

1 Answer 1

3
$\begingroup$

It is certainly possible to do this assuming you have the standard errors for the (log) hazard ratios. One thing you might want to explore is using some form of average or other typical value of your log transformed covariate as a moderator variable in a meta-regression. Such a moderator would be study-level so would tell you whether the hazard differed according to whether you were in a cohort with high or low average values of the covariate. You would not need to do this if you had some other covariate which was closely related to the average.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.