Help with Interrupted time series analysis: Code Optimization

tina_ting — Mon, 09 Jun 2025 08:16:10 GMT

Hello SAS experts,

I used the SAS code recommended in this link to perform an ITS analysis.

https://support.sas.com/kb/70/498.html

If I want to take the ITS code for a one-group, one-stage design with a binary response, and modify it to a one-group, two-stage and two-group, two-stage design, how should I change it?

If I also want to include covariates (categorical variables), where in the syntax should I place those variables?

Thank you very much for your help!

/*one-group, one-stage, Binary Response*/
original month1-6
iv1=0/1 (at month3)
iv2=0/1 (at month5)

original month1= month1 
original month2= month2
original month3= month3 
original month4= month1
original month5= month2
original month6= month3

proc gee data=a; 
        class id iv1;
        model y(event="1") = iv1 month iv1*month / dist=bin noint;
        repeated subject=id;
        run;
proc gee data=a; 
        class id iv1;
        model y(event="1") = iv1 iv1*month / dist=bin noint;
        repeated subject=id;
        effectplot fit(x=month plotby(pack)=iv1) / clm;
        effectplot fit(x=month plotby(pack)=iv1) / link clm;
        estimate 'm3 log odds' iv1 1 0 iv1*month 3 0 / ilink;
        estimate 'm4 log odds' iv1 0 1 iv1*month 0 1 / ilink;
        estimate 'm3 to m4 odds ratio' iv1 -1 1 iv1*month -3 1 / exp cl;
        store gee;
        run;

topic Help with Interrupted time series analysis: Code Optimization in Statistical Procedures

Help with Interrupted time series analysis: Code Optimization