Highly efficient clinical trial designs for reliable screening of under-performing treatments: Application to the COVID-19 Pandemic

Abstract

Background: The COVID-19 pandemic presents challenges for clinical trials including urgency, disrupted infrastructure, numerous therapeutic candidates, and the need for highly efficient trial and development designs. This paper presents design components and rationale for constructing highly efficient trials to screen potential COVID-19 treatments. Methods: Key trial design elements useful in this circumstance include futility hypotheses, treatment pooling, reciprocal controls, ranking and selection, and platform administration. Assuming most of the many candidates for COVID-19 treatment are likely to be ineffective, these components can be combined to facilitate very efficient comparisons of treatments. Results: Simulations indicate such designs can reliably discard underperforming treatments using sample size to treatment ratios under 30. Conclusions: Methods to create very efficient clinical trial comparisons of treatments for COVID-19 are available. Such designs might be helpful in the pandemic and should be considered for similar needs in the future.