Theory, Methods & Applications
Statistical Methods for Composite Endpoints
Win Ratio and Beyond
How do we combine survival and nonfatal events into a meaningful measure of treatment effect? Explore the statistical ideas, estimands, and methods behind composite endpoints.
From the Course to the Book
These materials grew out of a short course taught at the 2024 Annual Meeting of the Society for Clinical Trials (SCT). The five chapters form the starting structure for this book, connecting statistical theory with applications in R.
A Pairwise Comparison
What counts as a win?
Compare survival first, then time to first hospitalization. Move the time horizon to see the comparison change.
Illustrative pair, not trial data. Complete histories assumed.
Adapted from the time-horizon question in Chapter 1.
The Chapters
Introduction
From patients' event histories to clinically interpretable treatment comparisons.
Hypothesis Testing
Win ratio statistics, recurrent events, and the design of adequately powered trials.
Nonparametric Estimation
Define treatment effects through restricted win ratios, average win time, and while-alive loss.
Semiparametric Regression
Regression methods for relating composite outcomes to treatment and other covariates.
Discussions
Connect the methods and examine questions in adjustment, monitoring, and evidence synthesis.