[R] survival in R

[Troels Ring]

Dear friends - I hope it is OK to ask a question on the principles of 
survival analysis. I'm a medical person in this project, but wonders 
about the statistical methods proposed. We have a huge dataset and 
assumption of proportional hazard is said to be fulfilled if survival 
time is fractionated into short, intermediate and long time survival. I 
wonder if it should not be better to have a more generalized method 
encompassing the entirety of the observation period.

Best wishes

Troels Ring, MD
Department of nephrology
Aalborg, Denmark