Does anyone know how to suppress the following warning from proc lifetest?
"WARNING: The likelihood ratio test for strata homogeneity is questionable since some strata have
no events."
thanks!
RC
Best would be to ensure that all of the strata have events.
Why do you want to disable a warning that you may be using an inappropriate method for the given data?
I agree not ideal. I am making a number of reports in an automated process. In the long runI could make the lifetest code conditional on how many strata and if they have events, but in the short term I'm trying to get my code to run without any warnings.
Remove that Strata which have no event .
A WHERE clause would probably handle this.
Steve Denham
what if the strata is treatment arm?
If the specification you received request summary statistics for time-to-even data such as median survival
time, confidence intervals, and 25th – 75th percentiles, you still need to use the LIFETEST procedure and
you can use the NOTEST option to skip the inappropriate statistical test. For example, the following
codes will not issue the warning message because it does not perform the inappropriate statistical test,
but it will still give you the summary statistics.
ods output Quartiles=quartiles;
proc lifetest data=adtte method=KM alpha=0.05;
by subgroup;
time AVAL*CNSR(1);
strata trt01pn/notest;
run;
SAS Innovate 2025 is scheduled for May 6-9 in Orlando, FL. Sign up to be first to learn about the agenda and registration!
Learn the difference between classical and Bayesian statistical approaches and see a few PROC examples to perform Bayesian analysis in this video.
Find more tutorials on the SAS Users YouTube channel.
Ready to level-up your skills? Choose your own adventure.