1081 Burrard Street
Dr. Hongbin Zhang
Postdoctoral Fellow, CHÉOS
Abstract: Adaptive randomized clinical trials (ARCT) have the potential to make research more efficient, more informative, and more likely to demonstrate the potential effects of treatment. However, a good ARCT needs to address a number of issues associated in the study design. Motivated by a grant proposal on the prevention of dementia/Alzheimer’s disease, Dr. Zhang will explore ways to address the statistical estimation bias in the adaptation process. In this seminar, he will focus on bias correction when estimating the efficacy of combination therapy in a factorial design setting. After evaluating a few estimators’ bias magnitude, mean square error, and percentage coverage rate, a selection strategy is obtained. Although originating from ARCT, this finding is useful even in non-adaptive clinical trials, especially when combination therapy is being assessed.
Biography: Before joining CHÉOS in 2004, Dr. Hongbin Zhang worked in the Department of Statistics at UBC and MDA, a tech company. As a statistician at CHÉOS, he is involved with a number of studies, including The Cedar Project. In February 2015, Dr. Zhang completed his PhD degree in Statistics, and is currently completing a postdoctoral fellowship under the supervision of Dr. Hubert Wong.