Fig. 5

Mean Sample Size and Trial Duration for Bayesian Adaptive Designs under the Harmful Scenario. a and (b) are Bayesian adaptive design 1 that only allow early stopping for success based on skeptical prior; (c) and (d) are Bayesian adaptive design 2 that only allow early stopping for futility based on enthusiastic prior; (e) and (f) are Bayesian adaptive design 3 that allow early stopping for success based on skeptical prior, or early stopping for futility based on enthusiastic prior; (g) and (h) are Bayesian adaptive design 4 that allow early stopping for either success or futility both based on non-informative prior