Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: cox regression survival curves


From   Maarten buis <maartenbuis@yahoo.co.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: cox regression survival curves
Date   Sat, 13 Sep 2008 16:28:26 +0100 (BST)

--- moleps islon <moleps2@gmail.com> wrote:
> Is there a way to display confidence intervals using stcurve after
> running a cox regression (I need to adjust for the covariates-so sts
> graph is not as easy) ?

That has not been implemented, and I am not surprised. The very purpose
of Cox regression is not having to make assumptions about the baseline
hazard curve. This is done by letting the baseline hazard drop out of
the likelihood. The price you pay for that is that it now becomes
harder to make inference about the hazard curve (and thus the survival
curve), Cox regression just isn't designed for that purpose. -sts
graph- does have an -adjust()- option, but you can't combine that with
the -ci- option. You might be able to implement the procedure discussed
in (Hutson 2001).

Hutson, Alan D. (2001) Bootstrap-type confidence intervals for
quantiles of the survival distribution. Statistics in Medicine, 20(11):
1693--1702.

-- Maarten

-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room N515

+31 20 5986715

http://home.fsw.vu.nl/m.buis/
-----------------------------------------


      
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index