Hi Luis,
Multinomial logits are no good? Jenkins only notes that it's "rather complex" in the case where the data is interval-censored, and [this is implied] the intervals are long relative to the timescale of variation in the hazard rates. From your description, it's not obvious that this is the problem.
Best,
Marc Pelath
------------------------------
Date: Wed, 12 Mar 2008 12:52:02 +0100
From: "Luis Ortiz" <[email protected]>
Subject: st: Competing risks models with discrete time data
Dear Statalisters,
I am running an event-history (survival) analysis with discrete time (panel)
data. My risk period may potentially end in different, concurring events.
They might not be independent. I am almost sure they are not.
It seems that modelling competing risks with discrete time data is "rather
complex". Stephen Jenkins says so in one of its lessons, generously provided
on-line.
Do you happen to know if STATA10 provides any improvement in this regard;
that is, if there is any improvement that allows for modelling competing
risks with discrete time panel data?
Many thanks for your attention
Luis Ortiz
*
* 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/
