主 题: Estimating the Causal Effect of Flu Shot for Influenza in an Encouragement Design Study
报告人: Prof. Xiao-Hua Zhou (University of Washington)
时 间: 2005-06-29 下午 2:00 - 3:00
地 点: 理科一号楼 1364
A traditional randomized clinical trial is the established gold
standard for estimating the causal effects of treatments. However,
for some treatments, it is impossible to perform such a randomized
trial due to ethical and other reasons. For example, in our flu shot
reminder study, it would be unethical to randomize high-risk adult
patients to receive or not to receive
flu shots in order to evaluate causal effects of having flu on
patient outcomes. A better way to learn about the treatment causal
effects is to perform a randomized encouragement design study (EDS).
A randomized EDS randomly assigns subjects to receive or not to
receive an encouragement for the use of a treatment. In recent
years, there has been a rapid growth in utilizing encouragement
designs to study causal treatment effects. Since the randomization
to encouragement leads to a natural instrumental variable under some
plausible assumptions, the randomized EDS provides a tool for
estimating causal effects of the treatment on patient outcomes.
In this talk, we focus our attention on estimating causal effects on
binary outcomes and generalize the moment method proposed in Frangakis
and Rubin (1999) to a randomized clinical trial with crossover
non-compliance and missing data, as in our flu shot reminder study.
We also develop a maximum likelihood (ML) approach for this type
