American Statistical Association
The win ratio method of Pocock et al. (2012) has become an increasingly popular choice for analyzing composite endpoints that incorporate failure events of varying importance. The current work adapts the win ratio approach from two sample comparisons to regression. We relate the composite outcomes to the covariates via a novel semiparametric proportional win model, under which the regression parameters can be interpreted as certain transformations of the conditional win ratio given a pair of covariates. For inference, we propose a class of weighted estimating functions in form of one-sample U-statistics. We derive the asymptotic distributions for the resulting estimators and construct corresponding robust variance estimators. In addition, we develop a set of graphical and statistical tools for assessing the various model assumptions. Simulation studies demonstrate satisfactory finite-sample performances of the proposed methods. A dataset from a recent large-scale cardiovascular clinical trial is analyzed as an illustration.
|Date:||Wednesday, October 4, 2017|
|Time:||4:00 - 5:00 P.M.|
Memorial Sloan Kettering Cancer Center
Department of Epidemiology and Biostatistics
485 Lexington Avenue
(Between 46th & 47th Streets)
2nd Floor, Conference Room B
New York, New York
**Outside visitors please email firstname.lastname@example.org for building access.
You must be on the security list to enter the floor.