Welcome to successR

successR is the portal that provides various tools for computation of success probabilities of clinical trials.

Scope

The purpose of successR is to provide a comprehensive resource for training, methodology, and computational tools around success probabilities. The main focus is on data driven conditional probability (DDCP). However, the broader context of PTS assessments within Roche is also considered.

The idea is that successR will be continuously updated to become a comprehensive resource.

Subject matter experts

If you have questions, please reach out to our SMEs

successR was created and maintained previously by Kaspar Rufibach. We also thank the many colleagues who provided input and/or served as SMEs: Simona Rossomanno, Bernadette Surujbally, Jianmei Wang, Yifan Wang, Iris Yan, Imran Hassan. Beki Finch, Molly He, Nina Qi, and Marcel Wolbers.

Obtaining and installing relevant packages

  • rpact is pre-installed on OCEAN.
  • The current version of bpp can be installed from CRAN.
  • Computations on this page assume bpp version \(\geq 1.0.4\) and rpact \(\geq 3.1.1\) is installed.
# check installed package versions
packageVersion("bpp") >= "1.0.4"
[1] TRUE
packageVersion("rpact") >= "3.1.1"
[1] TRUE

Validation status

The bpp package is not formally validated. Usage is at your own risk, so please plausibilize all the numbers that you get out of the package.

Updates to this page

  • 15.07.26: Updated About page following handover and Training Materials to cover 2025 roll-out of Providentia v2.0.
  • 25.03.24: Added new example, illustrating use of DDCP to derive PivGo gating criteria.
  • 10.11.22: Some of the material on back-engineering a prior from a generic PTS has been moved from the three cases studies on the Gazyva program, JACOB trial, and MIRROS trial trial and centralized in a new Tutorial.
  • 05.12.21:
    • Based on feedback from the user community, bpp Version \(\geq 1.0.4\), which is available from CRAN here, now contains wrapper functions for each endpoint type. All tutorials and the exercises have been re-written using these wrapper functions. See here for documentation of additional changes to bpp.
    • Previously, the methodology tab had a section Connection to other quantities. This has now been made into a separate tab under Tutorials and extended.
  • 24.11.21: After all sites had their hands-on tutorials, solutions to exercise questions are now available and linked.
  • 15.11.21: Added Introductory video by Greg Spaniolo and Kaspar Rufibach on R&D academy nucleus to training page.
  • 28.10.21: Added some more description to the three “updating after interim” tutorials.
  • 13.10.21: Implemented input given by Jianmei Wang - thanks!
  • 05.10.21: Added case study Applying the UNICYCLE framework: RO7172508 and ADAs.
  • 22.09.21: Updated exercises after first roll-out: added context around the first case study.