st: STCOX survival estimates and SEs with shared Frailty

["Cameron L. Aldridge" <aldridge@ualberta.ca>]

For All those Survival Analyses Experts out there.......

I am estimating a cox proportional hazards model, to look at survival of 
young birds (chicks).  In many cases, I have multiple individual chicks 
with a brood marked and followed, but they are not independent of each 
other.  Thus, I estimating a shared frailty model as follows:

stcox X1 X2 X3, shared (Brood_num)

Where the X's are my covariates, and Brood_num is a unique identifier 
for the specific brood of chicks.

However, I can not figure out how (if I can) generate an appropriate 
survival or hazard function, that correctly estimates the SE (& 
Confidence Intervals) associated with the cumulative function (i.e. 
based on the latent effect of chicks within broods).

I can only seem to revert back to the basic K-M function, and possibly 
use a robust cluster for Brood_num to inflate my SE on my 
survival/hazard function.

Is there any way to appropriately account for the covariance of multiple 
chicks within a brood, and correctly estimate the survival function and 
SE's based on the shared frailty within Broods?

I have searched the lsitserv archives, but had any luck!

Thanks
Cam

--
_______________________________________
Cameron L. Aldridge	
Department of Biological Sciences,			
University of Alberta
Edmonton, AB   T6G 2E9
Voice: (780) 492-6267 Fax: (780) 492-9234

Field Address: 
Box 112 Manyberries, AB  T0K 1L0
Voice/Fax: (403) 868-2025
http://www.ualberta.ca/~aldridge/


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/