Date: Thu, 10 Oct 1996 21:56:58 GMT
Reply-To: David Nichols <nichols@SPSS.COM>
Sender: "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From: David Nichols <nichols@SPSS.COM>
Organization: SPSS, Inc.
Subject: Re: random effects proportional hazards
In article <v01510103ae7fff4e64ab@[164.111.164.4]>,
Jerry McGwin <mcgwin@EYES.UAB.EDU> wrote:
>we are conducting a pooled analysis of six different studies. each study
>collects a similar set of independent variables and determines events
>(outcomes) using a standard protocol. the primary outcome of interest is
>measured as time to event. therefore, we would like to use proportional
>hazards regression for the analysis. however, a priori we consider that
>there will be heterogeneity across studies that will need to be accounted
>for. the literature on meta- and pooled-analysis suggests the use of
>"random effects models" in this situation. any advice regarding the
>incorporation of random effects using the cox model?
>
>regards.
>
>gerald mcgwin, jr, ms
>
>=====================================================
>department of epidemiology department of ophthalmology
>school of public health school of medicine
>tidwell hall eye foundation hospital
>u n i v e r s i t y of a l a b a m a at b i r m i n g h a m
>720 south 20th street 1720 university blvd.
>birmingham, al 35294 birmingham, al 35294
>205-975-7369 205-325-8614
>205-xxx-xxxx (fax) 205-325-8692 (fax)
>gmcgwin@epi.soph.uab.edu mcgwin@eyes.uab.edu
>=====================================================
You might want to check with the folks at the Research Triangle Institute
to see if there's any reason why SUDAAN can't be profitably used on such
data.
--
-----------------------------------------------------------------------------
David Nichols Senior Support Statistician SPSS, Inc.
Phone: (312) 329-3684 Internet: nichols@spss.com Fax: (312) 329-3668
-----------------------------------------------------------------------------
| 
