Re: st: SVY medians and Elixhauser

[richgold@ix.netcom.com]

the question about Elixhauser is not clear to me; I am currently away from my office and don't have access to any literature, but I have used her code (downloadable from the AHRQ web site); her (I am referring to Ann Elixhauser when I say "her") code sets up 29 indicator variables, one for each comorbidity (yes, I know that the original article has 30, but her code dropped the second one several years ago)

so, what do you mean by "index"?

Rich

-----Original Message-----
>From: Steve Samuels <sjsamuels@gmail.com>
>Sent: Apr 7, 2013 7:41 PM
>To: statalist@hsphsun2.harvard.edu
>Subject: Re: st: SVY medians and Elixhauser
>
>
>
>To get medians and other quantiles with survey data, use Roger Newson's
>-sccendif- command, part of his -scsomersd- package, available from
>SSC. It will compute medians and CIs that account for the weighting and
>first stage clusters, but not the survey strata. You'll have to
>download some additional commands. Details are in the -help-.
>
>Steve
>sjsamuels@gmail.com
>
>On Apr 7, 2013, at 4:01 PM, Mike Butterfield wrote:
>
>Hi All,
>
>I am using the Nationwide Inpatient Sample 2008 and have been having
>trouble with two things:
>
>1.  I would like to find median values for some of my continuous variables,
>but have not yet found a way to do this while using SVY.
>
>2. I want to use the Elixhauser comorbidity index in my logistic regression
>model and have downloaded the code, but do not know how to actually use
>it.
>
>I would really appreciate anyone could refer me to some
>annotated/output or other resources so
>I could figure these things out.
>
>Best,
>Mike B., UC-Berkeley SPH
*
*   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/