Statistical Challenges of Launching a Randomized Controlled Trial of a COVID-19 Therapeutic at the Start of the Pandemic

• Activity Number: 91 - High Dimensional Data, Causal Inference, Biostats Education, and More
• Type: Contributed
• Date/Time: Monday, August 9, 2021 : 10:00 AM to 11:50 AM
• Sponsor: ENAR
• Author(s): Pamela Shaw* and Grace Choi and Haidaliza Soto-Calderon and Lizette Grajales and Katharine Bar
• Companies: Kaiser Permanente Washington Health Research Institute and University of Pennsylvania and University of Pennsylvania and University of Pennsylvania and University of Pennsylvania
• Keywords: Clinical Trials; COVID19; convalescent plasma; study design; composite endpoints

Abstract

Since the start of the SARS-CoV2 pandemic in the United States in March 2020, over 5000 clinical trials have been registered on ClincialTrials.gov. Many of these trials were launched with great uncertainty about incidence, disease outcomes, and in a setting of a rapidly changing standard of care. We will describe the experience of designing and implementing a randomized controlled trial of the efficacy of convalescent plasma for the treatment of patients hospitalized with COVID19, which randomized its first patient within 2 months of the first diagnosed case of COVID19 in the local area. We will present the rationale for the study design and endpoints and briefly describe the study outcomes. We will focus discussion on aspects of the design and planned study analyses that were particularly robust to the dynamic clinical setting and other aspects that proved more challenging than anticipated. We will conclude with discussions of some lessons learned that more broadly apply to robust clinical trial design, particularly in a rapidly changing clinical setting.