Bayesian Methodologies To Address Clinical Development Challenges For COVID-19 Drugs, Devices And Biologics

By Natalia Muhlemann, MD, MBA, Rajat Mukherjee, PhD, and Caroline Claire Morgan-Bouniol, DPhil (authors affiliation: Cytel, Cambridge, MA, USA)

The COVID-19 pandemic has elevated the challenge of designing and executing clinical trials within a substantially shortened time frame, and with limited data on the course of a newly emerged disease. According to the COVID-19 trials tracker, there are 1,476 ongoing trials as of June 10, 2020.

As explained in the recent article by Natalia Muhlemann, Rajat Mukherjee, and Frank Harrell (https://www.fharrell.com/post/bayes-covid/), there are numerous challenges when designing COVID-19 trials. Challenges include lack of prior data for candidate interventions / vaccines due to the novelty of the disease and the evolving standard of care as knowledge accumulates on the COVID-19 disease, and emerging evidence from completed trials.

The sense of urgency incites clinical researchers to invoke innovative trial design approaches to expedite the identification of efficacious interventions without compromising patient safety and scientific rigor. Bayesian statistical methods are very well suited to address these challenges due to their ability to adapt to knowledge that is gained during a trial.