B0707
Title: Group sequential testing under instrumented difference-in-differences approach
Authors: Samrat Roy - University of Pennsylvania (United States) [presenting]
Abstract: Unmeasured confounding is a major obstacle to reliable causal inference based on observational studies. Instrumented difference-in-differences (iDiD), a novel idea connecting instrumental variables and standard DiD, ameliorates the above issue by explicitly leveraging exogenous randomness in an exposure trend. The above idea of iDiD is utilized, and a novel group sequential testing method is proposed that provides valid inference even in unmeasured confounders. At each time point, the average or conditional average treatment effect under the iDiD setting is estimated using the data accumulated up to that point and testing the significance of the treatment effect. The joint distribution of the test statistics is derived under the null using the asymptotic properties of M-estimation, and the group sequential boundaries are obtained using the pending functions. The performance of the proposed approach is evaluated on both synthetic data and Clinformatics Data Mart database (OptumInsight, Eden Prairie, MN) to examine the association between rofecoxib and acute myocardial infarction (AMI), and the method detects significant adverse effects of rofecoxib much earlier than the time when it was finally withdrawn from the market.