[R] nested case-control study

Terry Therneau therneau at mayo.edu
Tue Mar 1 14:13:04 CET 2011


1. Using offset(logweight) in coxph is the same as using an "offset
logweight;" statement in SAS, and neither is the same as case weights.

2. For a nested case control, which is what you said you have, the
"strata" controls who is in what risk set.  No trickery with start,stop
times is needed.  It does no harm, but is not needed.
  In a case-cohort design the start= stop-epsilon trick is one way to
set risk sets up properly.  But that's not the design you gave for your
study.  

Terry T.



On Mon, 2011-02-28 at 17:02 -0800, array chip wrote:
> Terry, thanks very much! 
> 
> Professor Langholz used a SAS software trick to estimate absolute risk
> by creating a fake variable "entry_time" that is 0.001 less than the
> variable "exit_time" (i.e. time to event), and then use both variables
> in Phreg. Is this equivalent to your creating a dummy survival with
> time=1?
> 
> Another question is, is using offset(logweight) inside the formula of
> coxph() the same as using weight=logweight argument in coxph(),
> because my understanding of Professor Langholz's approach for nested
> case-control study is weighted regression?
> 
> Thank you very much for the help.
> 
> John



More information about the R-help mailing list