Academy of Mathematics and Systems Science, CAS Colloquia & Seminars | Speaker: | Dr.Ming T. Tan,Department of Biostatistics, Bioinformatics and Biomathematics, Georgetown University, Washington DC | Inviter: | 孙六全 研究员 | Title: | Causal Inference and Statistical Challenges in RWE Impacted Clinical Trials | Time & Venue: | 2019.09.09 16:00 N620 | Abstract: | The increasing availability of real world data (RWD) and the subsequent real world evidence has continued to impact statistical theory and methods in drug development and translational research. RWD can influence trial design and augment randomized controlled trial (RCT). Although RCT continues to be the gold standard to show evidence of efficacy and safety of an experimental treatment, it is not always feasible and a historical control clinical trial has to be used. RWD allows better selection and utilization of external controls and thus innovative trial designs. However the application of RWD is far more complex depending on the diseases and nature of treatments. Can the efficacy of a treatment be really supported by RWE? Is the statistical inference causal? We focus on one motivated by immunotherapy trials on rare cancer. We will introduce an external or historical control adaptive clinical trial design; and designs based on landmark survival to increase statistical power as applied to a recent immunology therapy trial on a rare disease, ocular melanoma, which compares relapse-free survival (RFS) rate at 3 years of patients with locally treated high-risk patients versus a matched contemporaneous control population. Many problems are still unresolved. (Work in collaboration with Yuan A and Yin A). | | | |