“…We focused on phase 3 clinical trials with a survival outcome measure, a frequent setting in hemato-oncology, to illustrate based on real data, two main interesting uses of the Bayesian approach, as a simple tool for early stopping decisions and in borrowing of external data. Indeed, though these developments are not new, they are poorly used in practice, and Bayesian analyses of phase 3 trials, even recently, mostly use posterior densities of outcomes [11] , or only historical controls data [12] . More specifically, the aim of this paper was two-fold: (i) to assess how sequential Bayes analyses may allow early decisions (termination) of the trial; (ii) to assess whether the borrowing of external data (for both the control and the experimental groups) in the analysis would allow optimizing the current trial design.…”