Joint Modeling in rstanarm: predicting survival times then fitting for them a survival model to estimate hazard ratio

Interfaces rstanarm
1

Dear rstanarm developers

I am planning to use rstanarm to fit a joint model.

Ideally, what I want to do is to fit a joint model and make Individual-specific predictions for out-of-sample individuals to get the predicted survival times (1st question: is that possible? I saw in the vignette the predictions are for survival probabilities not for times). Then for those survival times, I want to fit a cox model to estimate hazard ratio (2nd question: which function can be used to run only a survival analysis). Is there any way to do that in a single model, separately or in whatever way the package support?

Thank you for your help

Best,
Tasnim

2

If you are using a model that has a constant hazard ratio or a parametric function of time that is the log hazard ratio, use that model for hazard ratio inference directly. But you didn’t give us any information about why you have a joint model or how you set it up.