Originally broadcast on Oct 29, 2019 at 02:00 PM Central Time (US and Canada)

Historical Borrowing in Clinical Trials

Presented by Dr. Scott Berry, Berry Consultants

Clinical trials never occur in a vacuum. Historical data such as previous randomized trials, observational data, or real-world evidence is ubiquitous. Incorporating this information into a clinical trial has the potential to decrease sample size and/or improve decision accuracy, but only if the external evidence is similar to the current trial. When this similarity is absent, external borrowing can result in large biases, decreased power, or inflated type I error depending on the discrepancy. In this webinar we will discuss how to quantify the pros and cons of borrowing, how to quantify the “sweet spot” under which historical borrowing is beneficial and discuss the factors that might make historical data likely to fulfill those conditions. This information allows investigators to assess whether historical borrowing is promising for their own trials.

Bio: Scott Berry is President and a Senior Statistical Scientist at Berry Consultants, LLC. He earned his MS and PhD in statistics from Carnegie Mellon University and was an Assistant Professor at Texas A&M University before co-founding Berry Consultants in 2000. He has led Berry Consultants to be widely regarded as the premier Bayesian consulting company in the world. Since 2000, he has been involved in the design of hundreds of Bayesian adaptive clinical trials of pharmaceuticals and medical devices and has become an opinion leader in the field of Bayesian adaptive clinical trials.