Odds ratio tertile plots

[catch18]

Hi,

I have several tables and I'm trying to present some graphically

I have just run adjusted logistic regressions for 3 different response variables. Each response variable was run with 5 different main effects. For instance I did:

PROC logistic data= work plots=oddsratio (logbase=10);*final model;

class smoker(ref = "Yes") race2(ref = "1") Mixed_tert (ref='2') bsex (ref='0')/param=ref;

MODEL AGAc (event='1') = Mixed_tert smoker race2 bsex;

RUN;

PROC logistic data= work plots=oddsratio (logbase=10); *final model

class smoker(ref = "Yes")  race2(ref = "1")  Ovalo_tert (ref='2')/param=ref;

MODEL AGAc (event='1') = Ovalo_tert GA38 smoker race2;

RUN;

For each of the 3 response variables is it possible to have 5 columns of odds ratio plots, horizontally, (one for each main effect variable) and to include adjusted and unadjusted rather than the other variables in each model? Is it also possible to include the p values for trend?

I would really  appreciate any help!

Many thanks.

[SteveDenham]

I think this is going to require a combination of ODS output and PROC SGPLOT or SGPANEL. You may want to look at some of the examples in the documentation. The SGPANEL gives a nice example of a 2x2 panel.

SteveDenham

[catch18]

Many thanks for the suggestions, Steve. I used ggplot2 to create them.

Thanks.