P-value from Proc life test

[aboli]

Hi ,

Does proc life test, give the correct p-value if data does not have an event for one of the strata   group?

I am getting a below warning message as well as my data zero event in one treatment group.

The likelihood ratio test for strata homogeneity is questionable since some strata have no events.

   ods output censoredsummary=a1 productlimitestimates=a3  Quartiles=a4 HomTests=a5;
    proc lifetest data=ind timelist=1 to 6 by 1 outsurv=outsurv reduceout alpha=0.05 method=km conftype=loglog;
        time aval*cnsr(1);
        strata trtpn;
    run;

[sbxkoenk · Accepted solution]

Try this Google search ...

"The likelihood ratio test for strata homogeneity is questionable since some strata have no events." site:communities.sas.com

(don't forget the quotes)

BR, Koen