It is well acknowledged that we cannot extrapolate the lifespan T of the censored observation.

In the Kaplan-Meier estimation, the estimated $s(t)$ for $t∈[t_j,t_{j+1}]$ is estimated using:

enter image description here

I am confused by the fact that the censored observation is excluded from the denominator n at the next event point. Since we are not sure if their lifespan is more than the observed lifespan, how could we treat them as dead during the past interval?


Your Answer

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

Browse other questions tagged or ask your own question.