Statistics Colloquium, Dr. Samiran Ghosh

Title: Non-inferiority Intervention/Clinical Trial Design: Challenges, Solutions and Some Open Problems

Abstract: Non-inferiority trials are becoming very popular in comparative effectiveness research. These trials are required to show that an experimental drug/intervention is not inferior to a known reference by a small pre-specified amount. Since existing treatment option is available for many disease areas, simple placebo control trials are becoming less ethical. Also with the recent slowing down of blockbuster drug discovery, many pharmaceutical companies are focusing on non-inferiority design where the new experimental treatment may have additional advantages beyond simple efficacy. Hence, non-inferiority trials are of great importance to the pharmaceutical companies. In this talk I will present different statistical challenges one can face in designing a non-inferiority trial. We will cover issues related to construction of non-inferiority margin and hypothesis testing, both from frequentist and Bayesian points of view. Many open research questions will be presented which needs further attention. No clinical trials background is  assumed for this talk.