Design of RCTs to Identify Safe and Efficacious Treatments Among Responders, Tolerators, and Compliers
*Scott Emerson, University of Washington 

Keywords: Missing Data, RCT design

The National Academy of Science Oversight Committee convened to address Missing Data in Clinical Trials held a number of workshops to gain the perspectives of academic and industry researchers, as well as government regulators. A common theme that was presented was the issues surrounding treatments designed for symptomatic control in chronic disease. In such a setting, it is not uncommon to have missing data arise due to treatment dropout for the specified reasons of adverse events or lack of response. Nevertheless, there often seem to be a subset of patients who can both tolerate the treatment and who seem to respond. In this talk I discuss some RCT design strategies that might be used to allow rigorous, per randomization analyses to identify treatments that are safe and effective in only a subset of the patients who might meet initial eligibility criteria.