Decision-Making in Drug Development: Application of a Model Based Framework for Assessing Trial Performance

摘要: This chapter proposes a general framework for the formal integration of model-based predictions and their uncertainty in planning prospective trials quantitative decision-making. Standard operating characteristics such as statistical power, which are conditional on chosen effect size, quantify performance design. Optimising based solely power does not fully address needs drug development teams interested understanding compound well proposed study Many Phase 3 fail due to lack significant efficacy despite being adequately powered. Power take into consideration likelihood achieving assumed treatment effect. Metrics probability correct decision, Go reaching target value evaluate trial. A conceptual clinical trial simulation (CTS) approach is outlined calculating these metrics ‘false positive’ negative’ error rates metrics. An example presented illustrate CTS procedure show how different choices design, analytic technique metric influence making decisions.