Episode 67, “A Statistician Reads JAMA: A Futile Issue”
On Ep. 67 of "In the Interim…", Dr. Scott Berry dissects two recent JAMA trials that both stopped early for futility, revealing critical lessons about trial design.
The TRACK trial (rivaroxaban in advanced kidney disease) had no pre-specified futility rules, leaving the DSMB to make tough post-hoc decisions.
In contrast, the VICTORY trial (vitamin C in severe burns) applied explicit interim futility boundaries with operating characteristics established by simulation.
Scott advocates defining these rules up front and employing Bayesian predictive probability for task-appropriate futility analysis.
For practical, empirical insight for trial design teams listen to the full episode today:
https://www.berryconsultants.com/resource/67-a-statistician-reads-jama-a-futile-issue