DagStat 2025 tutorial: Generalized pairwise comparisons

Instructors: Johan Verbeeck (University of Hasselt, Belgium), Brice Ozenne (University of Copenhagen, Denmark)

Location: Berlin 10099, Humboldt-Universität zu Berlin, Unter den Linden 6, room D24-1.406.

Date: 23/03/2025

Content: you will find here the slides (doc directory), codes (.R files) and data (data directory) used for the tutorial.

• BuyseTest_intro.R: general introduction to the BuyseTest package
• revisit_CHARM.R: analysis of composite endpoint (death, hospitalization) via GPC
• revisit_prodige.R: assessing benefit-risk (death, quality of life) via GPC
• revisit_EB.R: analysis of a small sample trial by GPC

Installation of the software

First you should install:

• the R software (https://www.r-project.org/)
• an integrated development environment (IDE) such as R studio (https://posit.co/download/rstudio-desktop/)

Then open your IDE and download the BuyseTest package from CRAN:

install.packages("BuyseTest")

or from GitHub:

library(remotes)
install_github("bozenne/BuyseTest")

The CRAN version is the latest stable version and the Github version is the latest development version. Either way you should obtain version 3.1.0 (as of March 2025):

packageVersion("BuyseTest")

[1] '3.1.0'

To run one of the R scripts you will also need the packages dplyr and survminer.

How to run the code

You will need to use your own computer to run the R code provided in the tutorial. After having installed the R software and the BuyseTest package (see previous paragraph), you can access the R code by either:

• clicking on top right green button ‘Code’ and then on ‘Download ZIP’. Once the download is completed you just have to unzip “tutorial-DagStat2025-GPC-main.zip” and then you can access the .R files.
• clicking on a specific .R file and either copy paste the code or click on the download button ↓ (top right, next to ‘Raw’)