I am looking to do a Kaplan Meier analysis to look at survival in two groups. I have 21 people in group A and 17 people in group B. 5 people in group B died at various times, 0 people in group A died, ie all my data is censored for Group A.

I am getting a log-rank p-value of 0.028 when I do the KM analysis. My question is can I still use the KM for survival analysis given the absence of events in one group and if so, is the p-value valid.

Thanks in advance


Yes, you can use the log rank test in that case. The log rank test is something called a "score test" (I'm getting technical here), which remains valid under those conditions; however under ideal circumstances, score tests are often less powerful than likelihood ratio tests.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.