|
Abstract:
|
In addition to increased recent usage of DMCs to monitor accruing data in clinical trials, the scope of potential DMC responsibilities has expanded noticeably. Trial aspects on which DMCs may consider action include, but are not limited to stopping for demonstrated efficacy (usually based upon a group sequential scheme); stopping for futility if results are poor; stopping for ethical reasons based on safety concerns; modification based on inaccurate design assumptions or emerging external information. A relatively recent development currently eliciting much interest involves implementation of an adaptive design scheme; changes during a trial could potentially involve any of a number of aspects (sample size, treatment arms, population, etc.). This expanded scope raises new challenges for effectively addressing the motivations for interim analyses, while avoiding compromising validity of trial results. Different operational models have been proposed, e.g., separate boards for different decision types. But proper analytic thresholds for different actions potentially under DMC responsibility may not be independent. We discuss challenges and possible remedies, and present illustrations.
|
