Rescuing Randomized Trials: How to Detect and Correct Unlucky Randomizations as They Occur
Robert Greevy, PhD VUMC
A fundamental misconception of randomized trials is that randomization prevents selection bias. However, it is not uncommon to see multimillion dollar randomized trials with clinically meaningful differences in baseline covariates between treatment arms. Trials that have had an “unlucky” randomization provide a lower quality of evidence. As we monitor trials for adverse events and deviations from protocol, we should monitor trials for unlucky randomizations. This talk will present our team’s experience of monitoring randomization quality in a randomized trial using REDCap, encountering an unlucky randomization, and the remedy adopted. The majority of this talk will be accessible to a broad clinical audience.