| To: | A.Perperoglou@lumc.nl |
|---|---|
| Subject: | Re: baseline survival in time-dependent model |
| From: | Terry Therneau <therneau@mayo.edu> |
| Date: | Thu, 24 Oct 2002 10:00:17 -0500 (CDT) |
| Cc: | s-news@lists.biostat.wustl.edu |
| Reply-to: | Terry Therneau <therneau@mayo.edu> |
A Perperoglou asked:
"after fitting a Cox time-dependent model I plot the survival function. I
know that S-plus estimates the Survival
at the mean value of the covariates, but what happens whith the time
dependent model?
What is the value of time that S-plus uses to estimate the survival
function?"
The computation is easy, but there are several pitfalls with respect to
the interpretation or use of baseline survival curves after a Cox model
with time-dependent covariates. See section 10.2.4 of the book by
Therneau and Grambsch for an extended discussion and examples.
Terry Therneau
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: Gaussian kernel, ripley |
|---|---|
| Next by Date: | plotting points on a time series graph, Burbidge, Robert |
| Previous by Thread: | baseline survival in time-dependent model, Perperoglou, A. (MStat) |
| Next by Thread: | [Fwd: RE: find row&col], Catalina Mesina |
| Indexes: | [Date] [Thread] [Top] [All Lists] |