contrast.sas

ods output SolutionF = fixfx_02 FitStatistics=fitstats_02 LRT=lrt_02  LSMeans=lsm_02  Estimates=estdiffs_02;
proc mixed data=work.tmp method = reml ; * asycov asycorr covtest ic ; 
    class id group_blindint(ref = 'X') time(ref = '1'); 
    model stepcount = baselinescu group_blindint | time  / solution outpm = work.outpm_02 ; *influence(effect=id iter=5); 
    repeated time / subject=id type = ar(1) R RCORR;
    lsmeans group_blindint * time / at means e cl ; * tdiff pdiff;

    estimate 'mean Y at t3' baselinescu 4310.844 
                            intercept 1
                            group_blindint 1 0 0
                            time 0 1 0
                            group_blindint * time 0 1 0 0 0 0 0 0 0 ;
    estimate 'mean Z at t3' baselinescu 4310.844 
                            intercept 1
                            group_blindint 0 1 0
                            time 0 1 0
                            group_blindint * time 0 0 0 0 1 0 0 0 0 ;
                            
* To get the difference between the two just subtract the terms listed for the above estimates ;
    estimate 'Y-Z at t3'   group_blindint 1 -1 0
                           group_blindint * time 0 1 0 0 -1 0 0 0 0 / cl;

run;