Re: st: competing hazards models with stata?

[Jesper Sorensen <sorensen@mit.edu>]

Assuming that your setup meets all of the standard assumptions for 
modeling competing risks (eg that they're not interdependent), then 
you can do this by simply estimating rate models for each of the 
events separately -- ie first model event A treating event B as a 
censoring event (along with other forms of censoring), and then model 
event B, treating event A as a censoring event.

Alternatively you can use TDA (transition data analysis) 
http://steinhaus.stat.ruhr-uni-bochum.de/tda.html


//Jesper


> I need to estimate a survival-type model (either a Cox model or a
> parametric model will do) with two competing hazards. Is there any way to
> do this with stata so that I would get a two coefficients for each
> independent variable, ie, one for each of the two hazards? (what I need
> could be viewed as a survival equivalent of a multinomial logit model).
>
> If yes, I would then like to even do something a little more complicated,
> because one of the hazards can repeat while the other can happen only once
> (like the medical studies where someone can contract an illness many times
> but only die once).
>
> Any insight would be appreciated. Also, if not stata, might there be any
> other packages that could do this?
>
> cheers, Arturs
>
>
> _________________________________________________________________________
>
> Arturs Kalnins
> Assistant Professor
> Department of Management and Organization
> Marshall School of Business
> Bridge Hall 307H
> University of Southern California
> Los Angeles, CA 90089-0808
> kalnins@usc.edu
>
> *
> *   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/