Date:         Thu, 4 Dec 2008 12:28:23 -0800
Reply-To:     Sassy <AugustinaO@GMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Sassy <AugustinaO@GMAIL.COM>
Organization: http://groups.google.com
Subject:      Proc Genmod using a Possion distribution
Comments: To: sas-l@uga.edu
Content-Type: text/plain; charset=ISO-8859-1

Hey all,

I am doing a study were I want to compared the actual number of
hospital visits patients had to the expected number of hosptial
visits
they are supposed to have based on the medication they are taking.
we are lookin at visits within one year only.  I think this analysis
requires a poisson distribution using proc genmod, which i have not
used before.  I also want to control for other factors that might
affect the frequency of visits.  does anyone  have an idea on how to
carry out this analysis?  I also included a sample of what my data
looks like and the syntax i want to use.


example of data:

variable name:
med= type of medication
e_visit= expected visit
a_visit= actual visit
difference = difference between actual and expected visit

med       a_visit e_visit difference
type 1       10     12         -2
type 1       20     12          8
type 1       8      12         -4
type 2       6       6            0
type 2      10      6            4
type 3       2       3           -1
type 3      15      3           12
type 3      3        3            0


proc genmod data=test;
where medication_type='type 1';
model difference= age  race sex comorbid_conditions /
dist=poisson
link=log;
run;