|
Activity Number:
|
251
- SPEED: Biopharmaceutical Methods and Application I, Part 2
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 29, 2019 : 2:00 PM to 2:45 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
|
Abstract #307600
|
|
|
Title:
|
Design of Clinical Trials for Bivariate Endpoints
|
|
Author(s):
|
Junxiao Hu* and Patrick Blatchford and John Kittelson
|
|
Companies:
|
University of Colorado and University of Colorado and University of Colorado
|
|
Keywords:
|
Clinical Trial;
Bivariate outcome;
Phase II-III clinical trial
|
|
Abstract:
|
Clinical trials commonly measure treatment effects on multiple endpoints. Mapping parameter space to regions of benefit and lack of benefit is fundamental to any design but capturing clinical meaning with multiple endpoints is not straightforward. This paper proposes a statistical design framework for clinical trials in a bivariate outcome space. A rectangular hyperbola is used to define a bivariate null curve that allows transitive (non-constant) tradeoff between the two endpoints. The outcome space is ordered using the distance from the null curve that defines a statistic for bivariate inference. The asymptotic distribution of the outcome statistic is derived and two approaches for inference are described. Asymptotic properties of these approaches are examined with simulation studies. Both approaches preserve operating characteristics when the statistic is asymptotical Gaussian. In addition, one approach is robust for model misspecification and its interpretation corresponds with the commonly reported confidence rectangle. The methods are illustrated with two major trials of new LDL-cholesterol lowering treatments to reduce the risk of cardiovascular events.
|
Authors who are presenting talks have a * after their name.
