I am trying to compare concordance using SAS 9.4 on different models utlizing
proc phreg data=MyData concordance=uno(diff se seed=1234 iter=100);
model Time*Status(0)=V1 V2 V3;
roc 'V1 + V2' V1 + V2;
roc 'V2 + V3' V2 + V3;
run;
but would like to go the other way around it. I wanna add variables to my base model. I can't seem to figure out how to compare models by adding instead of removing variables from a base model. Say I want to see if adding a variable to traditional risk factors for stroke changes concordance and p-values. I would like to do something like this:
proc phreg data=MyData concordance=uno(diff se seed=1234 iter=100);
model Time*Status(0)=V1 V2;
roc 'V1 V2 + V3' V1 V2 V3;
roc 'V1 V2 + V3 +V4' V1 V2 V3 V4;
run;
And so on. Trying this SAS hits me with "can't use variable that isn't in the original model"
Using the upper way leaves me comparing submodels to the largest model, instead of comparing the additive effect of new variables.
Is there a way around this, or am I trying to do something that can't be done or is misunderstood in the first place?
