R Code repository for the manuscript 'Validation of prediction models in presence of competing risks: a guide through modern methods' published in BMJ.
The repository contains the following code:
-
Prediction_CSC_minimal.R : the companion R script for the manuscript. This script reproduces all main tables and figures of the manuscript. The file evaluates the performance of a cause specific hazards prediction model. Data are available here.
-
Prediction_CSC.md : a markdown document containing a more in-depth version of the script, with details on model development, descriptive tables and supplementary plots. This document requires installing several add-on packages that are not needed to run the minimal script. The RMarkdown source code (.Rmd) is here. Additional functions useful for this script are available here.
-
Additional code to alternatively develop a competing risk prediction model using the subdistribution hazard approach (Fine & Gray) is here. The Rmarkdown source code (.Rmd) is here. A more concise R source code (.R) is here.
-
sharing_CSC_model.R : example/template of how to share a cause-specific hazards prediction model for external validation, without having to share the original development data.
If you are git user, you can clone the directory by using
git clone https://github.com/survival-lumc/ValidationCompRisks.gitOtherwise, you can simply download a zip file containing the directory by clicking Code -> Download ZIP at the top-right of this Github page. Extract the zipped files to a directory of your choice.
Afterwards, you can double-click the ValidationCompRisks.Rproj file to open an Rstudio session in the directory you have just downloaded. This will ensure all file-paths called in the files are maintained. The minimal script and the .Rmd files can now be executed.
| Name | Affiliation | Role |
|---|---|---|
| Daniele Giardiello | University of Milan-Bicocca (IT) | Author .Rmd files/maintainer |
| Edouard Bonneville | Leiden University Medical Center (NL) | Author minimal .R and CSC sharing scripts/review of .Rmd scripts |
| Nan van Geloven | Leiden University Medical Center (NL) | Review of both .R and .Rmd scripts |
| Maarten van Smeden | University Medical Centre Utrecht (NL) | Review of .Rmd script |