interpreting multinomial logistic regression

[Kyra]

Hi,

I need help in interpreting multinomial logistic regression. Please find attached my SAS output.

Exposure  pills is number of pills prescribed which is continuous.

Outcome  pillsconsumed is pills consumed in categories. It is divided in three categories:

                         0: no pills consumed,

                         1: 1-10 pills consumed

                         2:  10+ pills consumed

My code is

proc logistic data=new;
class pillsconsumed (ref = "0") /param=reference;
model pillsconsumed = pills / link = glogit;
run;

Thanks a lot in advance.

[cminard · Accepted solution]

An increase of 1 pill prescribed is associated with a 1.033 fold increased odds of consuming 1-10 pills compared with no pills.

An increase of 1 pill prescribed is associated with a 1.053 fold increased odds of consuming 10+ pills compared with no pills.

But this analysis seems a bit odd practically speaking.  Maybe agreement (ie, kappa)?

[Kyra]

Thank you very much! Our dataset has pills consumed in categories and not continuous that is why we had to analyze this way.

thanks