Brms alternative to frequentist Cox regression (Cox PH assumption violated)

Interfaces brms
1

Does brms has an alternative to frequentist Cox regression, that I can not use, since the PH assumption is violated (HR varies between different time points). As much as I know, AFT is also not a good option as time ratios also change during time?

Then I thought that maybe PH assumption does not matter if i report median survival during follow up with CIs and then I started think about Bayesian ways.

Does brms has an alternative for the aforementioned analyses? If yes, could you please share tutorial?

2

Yes, Cox regression is unofficially supported. You can get the details in GitHub issue #230. I’ve also worked through a couple examples here.

3 Likes