Choice Of The Prior In The Bayesian Design For The Clinical Trials.

Bayesian statistics have the advantage of being easily established and derived. Thus, we will use this approach to find the distribution of the predictive probabilities of the data not observed yet in the conception of clinical trials phase II. Such perspectives are necessary in cases or in reason treat ethical preoccupations or ether to achieve trials which are particularly toxic or expensive. We propose, in reason of neutrality, to calculate the predictive probabilities within a Bayesian case, where the prior is non informative in different models apply in clinical trials.