[R] Stratified Cox proportional Hazard Model

Terry Therneau therneau at mayo.edu
Tue May 22 15:26:57 CEST 2007


 You will have to be more clear about what exactly you are trying to do.
I, at least, am quite unsure about what you want.


   fit <- coxph(Surv(time, status) ~ x1 + strata(x2))
 
The hazard ratio(s) wrt x1 have the same meaning for a stratified or an 
unstratified model.   If you are asking about hazard ratios for x2 --- there are 
none.  That is the point of a strata; you are declaring a variable to NOT be 
proportional hazards, and thus there is no single "hazard ratio" that describes 
it.

	Terry Therneau



More information about the R-help mailing list