Related Experiment Video
Updated: May 21, 2026
05:37
An R-Based Landscape Validation of a Competing Risk Model
Published on: September 16, 2022
Qualified prediction system for allograft failure in real world settings: extended validation study
Marc Raynaud1, Agathe Truchot1, Sofia Naser1
1Université de Paris Cité, INSERM, PARCC U970, Paris Institute for Transplantation and Organ Regeneration (PITOR), Paris, France.
BMJ Medicine
|May 20, 2026
Summary
The integrative Box (iBox) system accurately predicts long-term kidney transplant failure risk across diverse populations and clinical scenarios. This validated model supports its use in clinical practice and trials for kidney allograft failure risk stratification.
Area of Science:
- Nephrology and Transplant Medicine
- Biostatistics and Predictive Modeling
- Clinical Trial Design
Background:
- Kidney allograft failure poses a significant long-term risk for transplant recipients.
- Accurate prediction models are crucial for risk stratification and clinical trial endpoints.
- The integrative Box (iBox) system is a predictive model for long-term kidney allograft failure.
Purpose of the Study:
- To comprehensively validate the integrative Box (iBox) system for predicting long-term kidney allograft failure.
- To extend the context of use for the iBox system in clinical trials.
- To support the wider implementation of the iBox system in clinical practice.
Main Methods:
- An extended validation study involving 12,683 kidney transplant recipients from 21 academic centers across Europe, North America, and South America.
- Utilized derivation (n=4000) and validation (n=8683) cohorts with follow-up until November 2024.
- Assessed iBox performance using discrimination, calibration, overall fit, and clinical utility, including flexible versions (e.g., race-free eGFR) and various clinical contexts.
Main Results:
- All iBox algorithm versions demonstrated robust discrimination and overall fit (C-index 0.79–0.87) across derivation and validation cohorts.
- Calibration varied across cohorts; decision curve analysis showed positive net benefit for iBox algorithms.
- The iBox system outperformed eGFR slope and anti-HLA donor-specific antibodies in predictive ability, maintaining performance across diverse scenarios and extended follow-up periods.
Conclusions:
- The iBox system exhibits robust and versatile predictive performance across diverse real-world settings and clinical scenarios.
- These findings support the reliability of the iBox system for risk stratification in routine clinical practice.
- The iBox system is suitable for use as a surrogate endpoint in clinical trials for kidney allograft failure.