Hi,
I used proc phreg to conduct survival analysis but failed to get the c statistic from the model. I kept receiving an error message from the log: 'Neither concordance nor ROC analysis is performed with the START/STOP specification' .
ods graphics on;
PROC PHREG DATA=temp.POP_long2 concordance plots=roc rocoptions(at=365 to 2920 by 365);
class c_age (ref='1') c_education (ref='1') c_gender (ref='2') c_race (ref='1')
c_married(ref='0') c_access(ref='0') c_financialDifficulty(ref='0') c_depression(ref='0')
c_Anxiety(ref='0') c_demclas(ref='3') c_carecompanion(ref='0')
c_ParticipationRestriction(ref='0')/param=reference;
MODEL (start,stop)*flag_visitFallout_t(0)=c_age c_gender c_race c_education c_married c_access c_financialDifficulty
c_ParticipationRestriction c_depression c_Anxiety Comorbidities_f
IADLs c_demclas c_carecompanion /TIES=EFRON;
RUN
I also tried something like:
plots=auc rocoptions(method=ipcw(cl seed=1234));
but that didn't work neither.
I wonder if there is any way to work around this. Thanks for your inputs!
