Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: Proportional hazard assumption test for stcrreg
From
Adam Olszewski <[email protected]>
To
[email protected]
Subject
Re: st: Proportional hazard assumption test for stcrreg
Date
Mon, 5 Aug 2013 16:10:36 -0400
Hi Christel,
For competing risk regression, the options are:
- "graphical method" predict Schoenfeld residuals and plot them the
same way that -estat phtest- does (see -help stcrreg
postestimation##predict- ); I suppose one could write a test to obtain
a p-value similar for the -estat phtest- but I do not know if such
tests have been validated;
- "time interaction method" - study each regressor's interaction with
time using the tvc() option of stcrreg; this may be time consuming in
your dataset, since stcrreg is very slow with tvc();
I hope this helps,
Adam Olszewski
On Mon, Aug 5, 2013 at 9:53 AM, Christel Häggström
<[email protected]> wrote:
> Dear Statausers,
>
> I am using stcrreg for competing risk analyses, which assumes proportionality of hazards over time, but I haven't found any way to test the proportionality assumption. The Cox model (stcox) which also build upon the same assumption have a post estimation command (estat phtest) which is easy to use, but this option is missing in stcrreg. I have a dataset with approx. 300 000 individuals and 5 covariates which I want to include in the competing risk regression. Other statausers that have used stcrreg, how did you check for this? Any suggestions?
>
> Thank you for the consideration.
> Christel
>
>
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/faqs/resources/statalist-faq/
> * http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/