Suppose we could estimate the value of the survival function S.t / for every possible survival time from t = 0 onward. Clearly S(t) → 0 as t becomes large. It can be shown that the mean survival time...

Suppose we could estimate the value of the survival function S.t / for every possible survival time from t = 0 onward. Clearly S(t) → 0 as t becomes large. It can be shown that the mean survival time is equal to the area under this “complete” survival curve. Why are we unable to estimate mean survival from the Kaplan–Meier result when the largest follow-up time is censored? To gain insight, contrast the survival curves for the 6-MP and placebo groups in Fig. 3.2.