In the last month Andrea Bassi presented his work at the Adaptive Designs and Multiple Testing Procedures Workshop 2017 and at the annual conference of the International Society for Clinical Biostatistics (ISCB). The purpose of the workshop was to present state of the art research on adaptive designs and multiple testing procedures, while the ISCB conference focused on the methodologies used in the design and analysis of clinical research.
Andrea discussed the latest development of his BIOMARKER project in a talk titled “A novel decision-theoretic framework for adaptive multi-arm trials”. In this presentation, he introduced a decision-theoretic multi-arm multi-stage (MAMS) trial framework and evaluated the performance of the design. The proposed multi-arm adaptive design outperforms standard non-adaptive designs by selecting the best drug available more often (percentage of selection 0.8 to 14.3 percent higher, with the same expected number of patients enrolled). This means that by adopting this adaptive design it is possible to plan clinical trials that require a lower amount of patients to reach the same frequentist operating characteristics, when comparing multiple treatment arms.