Since you've decided to do an analysis match on clinic, I assume you are
worry about a potential confounding effect of clinic. You can control by
the clinic effect using conditional logistic (clogit) or just by
including clinic as a variable in an ordinary logistic regression model.
In fact, because the number of clinics is very small and it seems that
you've matched for logistic reasons (not to increase efficiency), I'd be
inclined to use clinic as a potential confounder in an ordinary logistic
regression model. That way I'd be able to evaluate a "clinic effect",
something that can not be done in a conditional model with clinic as the
matching factor. Now, if you think that clinic is interacting with race,
then you should not adjust for clinic. Instead you should test for the
interaction in an ordinary logistic regression model including a (race x
clinic) term in your model. If the interaction is not significant, then
you should assess whether clinic should remain in the model as a
confounder. You could also test for the interaction by including the
(race x clinic) term in a conditional model (clogit), but I don't see
any advantage over the previous approach.
I hope this helps,
Leonelo E. Bautista, MD, DrPH
Assistant Professor
University of Wisconsin Medical School
Population Health Sciences
610 Walnut Street, 703 WARF
Madison, WI 53726-2397
Phone: (608)265-6176
Fax: (608)263-2820
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Ricardo
Ovaldia
Sent: Monday, November 24, 2003 7:46 AM
To: [email protected]
Subject: st: conditional logistic regression
Dear Stata users,
I posted a message last week and did not receive any
replies. So let me ask it more simply:
How can I best analyze (estimate and adjust for
covariates) a dichotomous outcome measured in a 100%
sample of patients at 5 different clinics when there
is significant clinic to clinic variation in the
measured outcome?
Thank you,
Ricardo.
--- Ricardo Ovaldia <[email protected]> wrote:
> Dear Stata users,
>
> I have data from women seen at 5 outpatient clinics.
> One of the outcome variables of interest is whether
> or
> not women are routinely performing self-breast
> exams.
> We are interested in determining factors associated
> with this outcome. In the entire dataset African
> American women perform self-breast exams less often
> than Caucasian women (OR=1.96). However in one of
> the
> clinics the rate among African American is higher
> than
> in Caucasian women (an interaction between clinic
> and
> race?). I had decided to estimate all odds ratios
> using -clogit- to account for the clinic effect.
> When
> I do this for race I get an even more significant
> OR=
> 2.5678 (same direction). Is it correct to use
> -clogit-
> as I did or do I need to something different such as
> a
> stratified analysis?
>
> Thank you,
> Ricardo.
>
>
>
> =====
> Ricardo Ovaldia, MS
> Statistician
> Oklahoma City, OK
>
> __________________________________
> Do you Yahoo!?
> Free Pop-Up Blocker - Get it now
> http://companion.yahoo.com/
> *
> * 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/
__________________________________
Do you Yahoo!?
Free Pop-Up Blocker - Get it now
http://companion.yahoo.com/
*
* 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/
