All Times EDT
Keywords: Platform Trials, Non-Concurrent Controls
Somewhat surprisingly the value of non-concurrent controls in master protocols has been underutilized. When presented with the problem it is usually done as a single arm enrolling in a trial with previous non-concurrently enrolled controls in a "disconnected: time period. The value of platform trials though, is that there are many arms in the trial with overlapping connected time periods. So, these non-concurrent control patients are connected through time to the arm of interest and are incredibly valuable for making inferences. In this talk we will present the comparisons of methods for utilizing (or not) non-concurrent controls and demonstrate the inefficiency of ignoring data.