|
Abstract:
|
Conventional controlled clinical trials employ randomization and blinding to ensure the balance of baseline covariates between study arms. The effect of an experimental treatment is formally tested via a pre-specified hypothesis reflecting the study’s primary objective, which directly links to the primary efficacy endpoint, such as, is experimental treatment effective in reducing cognitive decline at a pre-specified landmark time? To address such clinical question in an observational study, the targeted maximum likelihood estimation (TMLE) method has been suggested. For a longitudinal control-matched study, assessing treatment effect in the presence of missingness can be very challenging, which depends also on trial duration and pattern of missingness between arms. In this paper, we investigate a prospectively control-matched longitudinal design using longitudinal TMLE via simulation studies. We compare the performance of LTMLE with a few commonly used methods. We discuss the results and their implication to conducting a feasible real-world observational study.
*This abstract reflects the views of the authors and should not be construed to represent the views or policies of US FDA
|
