Gravar-mail: Ethics of clinical trials from bayesian perspective: Medical decision making should use posteriors, not priors