Robust Inference in Clinical Trials

• Activity Number: 447
• Type: Topic Contributed
• Date/Time: Wednesday, August 1, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Jitendra Ganju*+
• Companies: Gilead Sciences
• Keywords: robust inference ; multiple comparisons ; randomized trials ; permutations

Abstract

Trials are planned with multiple objectives and multiple endpoints. Usually, one objective and one endpoint is labeled "primary." It can be challenging to robustly interpret results from such trials. For example, analysis of the primary objective usually depends on a single statistic. But reliance on a single statistic for formal inference is a risky strategy, particularly for complex and expensive trials. The statistic may underperform because the underlying assumptions required for optimal performance may not be met. A different statistic may give quite a different answer. For another example, a trial may have 3 endpoints, one of which is labeled primary, and the other two secondary but the alpha is split in three. Do the labels "primary" and "secondary" cause unnecessary confusion? The speakers will share real life examples and offer solutions with the goal of making inferences more robust.