04-19-2018 03:03 PM - edited 04-19-2018 03:23 PM
I'm do have a lot of experience working with mixed method models, so I hope someone can assist.
In the data, the placebo is where avisitn=4. I need to get output by all treatment groups trt01pn * placebo (or where trt01pn=4).
Can someone help me modify this code or do I need to create some other variable for placebo? Thanks!
proc mixed data = adeff method = reml;
where trtn ne 9 and avisit ne "Baseline";
class usubjid trt01pn avisitn ;
model chg = trt01pn avisitn trt01pn*avisitn / solution ddfm=kr cl;
repeated avisitn / sub=usubjid type=sp(pow)(week);
lsmeans trt01pn * avsitn / cl;
lsmeans avisitn / cl;
ods output lsmeans = lsm_chg;
I've seen it done like this with 6 treatment groups but, not sure what this means. Is there another way?
/* Day 7 comparisons */
estimate 'treat1-placebo7' trt01pn 1 0 0 -1 0 0
trt01pn*avisitn 1 0
0 0 / divisor=1 e cl;