[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Re: multiple imputation and survival analysis

From   Maarten buis <>
Subject   Re: st: Re: multiple imputation and survival analysis
Date   Fri, 1 May 2009 20:37:09 +0000 (GMT)

--- From: "Stephanie L Kent" wrote:
> > I'd like to run a cox regression in data declared
> > survival after using multiple imputation.
> > 
> > I ran the "ice" command then using the imputed data,
> > the micombine command followed by the stcox model. 
> > However, this produced models with coefficients rather
> > than hazard rates.  

--- On Fri, 1/5/09, Martin Weiss wrote:
> The example from the help file produces hazard ratios as
> default, and you have to request coefficients. I am having a
> hard time to believe that it behaves differently because of
> the imputation..

This is actually normal behavior, as -stcox- is not run separately
as Stephanie wrote, but within -micombine-. Stephanie can specify
the -eform- option within -micombine- to get hazard ratios. 

As a matter of terminology it is good to know that -stcox- never 
gives you hazard rates, but always hazard ratios. The whole point
of -stcox- is that you do not have to specify a functional form 
for the baseline hazard rate.
-- Maarten

Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen


*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index