Collinearity in cox model with time dependent variables

[aimeehayes0]

Is there a way to check for collinearity in a cox regression model with time dependent variables?

Can PROG REG and VIF be used??

Grateful for any assistance.

[PaigeMiller]

So, yes, you can indeed run the data through PROC REG to compute VIFs of your many time dependent variables.

It's not clear to me what you would do if there was significant collinearity between two time dependent variables -- would you remove one (or more) of the times? What exactly are the time dependent variables, anyway?

[aimeehayes0]

Thank you Paige.

It is a survival analysis in patients with a particular type of cancer. The variables are baseline patient characteristics, tumour characteristics, scan characteristics etc. I think I would have to remove collinear variables from the model.