Dear Carl,
One approach is to rearrange the dataset in long format. That may be
difficult, but working with time-dependent covariates will be mu-u-u-uch
easier. Allow me to recommed the book "Introduction to survival analysis
with Stata" (I'm not sure I get the title right) which has helped me a lot.
Sincerely,
Peter Jepsen.
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Cunlin Wang
Sent: 16. november 2003 18:24
To: [email protected]
Subject: st: Time-dependent covariates in Survival analysis.
Looking for how to conduct the time-dependent survival analysis using Stata.
The question is trying to identify the association of HCV and overdose
death. Each subject has both constant variables and repeated measurements on
interested covariates (time-dependent). Also, some subjects had HCV
seroconversion during the follow-up therefore have two observations: one for
HCV negative and one for HCV positive (staggered entry and therefore
correlated each other). The dataset has been established into "wide" format
with each observation representing one subject with hundreds of variables. I
know how to fit the time-dependent Cox model in SAS, but don't know how to
adjust the correlation between observations (if anyone can advise here will
be appreciated as well).
I know the Stata can easially use id() option in "stset" to adjust the
correlation, but don't know how to deal with the time-dependent covariates.
Any advice? Appreciate that .....
SIncerely.
Carl. JHU.
*
* 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/
*
* 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/
