Programming the statistical procedures from SAS

odds ratio interpretation

New Contributor
Posts: 2

odds ratio interpretation

[ Edited ]

Hi all,


Please help me with the following model: (DRUG:aspirin, paracetamol,ibuprofen CHANGE:increased, decreased, unchanged AGE:continous variable BASE_VALUE:Continous variable NO_INJECT:continous COVARIATE:0,1)


proc logistic data=new descending;
CLASS drug(Ref='Aspirin') / Param=REF;
MODEL change= Age base_value No_inject drug Age*drug covariate/expb;


When I used the above code, I didnot get ODDS Ratio for variables in Interaction, so I used


ods graphics on;
proc logistic data=PrepAMDONLY21 plots(only)=(oddsratio(range=clip))descending;
class Txtype /param=ref;
model Change_VA1YR_Outcome = Age drug Base_value No_inject Age*drug covariate;
oddsratio drug;
oddsratio Age;
oddsratio Base_value;
oddsratio No_inject;
oddsratio Covariate;
contrast 'Pairwise Paracetamol vs Aspirin' drug 1 0 / estimate=exp;
contrast 'Pairwise Ibuprofen vs Aspirin' drug 0 1 / estimate=exp;
contrast 'Pairwise Paracetamol vs Ibuprofen ' drug 1 -1 / estimate=exp;
ods graphics off;


Please let me know if this is correct!? Attached is the interpretation. Please help me out with the intrepretation if the model is correct.


Thansk in Advance,


Super User
Posts: 10,210

Re: odds ratio interpretation

Can you use LSMEAN ?

Check the example 17 in proc logistic's documentation.


proc logistic data=Neuralgia;
class Treatment Sex / param=glm;
model Pain= Treatment|Sex Age;
oddsratio Treatment;
lsmeans Treatment / e diff oddsratio cl adjust=bon;
SAS Employee
Posts: 308

Re: odds ratio interpretation

The CONTRAST statements are meaningless since Drug is involved in interaction and so, by definition, its effect depends on the level of the interacting variable, Age.  Your ODDSRATIO statement for Drug gives you estimates of the odds ratios comparing Drug levels at the mean of Age.  If you want to make the Drug comparisons at a different level of Age, use the AT option in the ODDSRATIO statement.

Ask a Question
Discussion stats
  • 2 replies
  • 3 in conversation