Predictive Score for Posttransplantation Outcomes.

Area of Science:

• Nephrology
• Transplantation Medicine
• Biostatistics

Background:

• Current kidney transplant scoring tools primarily rely on posttransplant data.
• Predicting patient and allograft survival is crucial for successful kidney transplantation.
• There is a need for predictive models utilizing pretransplant information.

Purpose of the Study:

• To develop and validate a novel scoring tool for predicting posttransplant outcomes.
• To utilize pretransplant laboratory data for predicting kidney transplant success.
• To compare the performance of the new model against existing prediction tools.

Main Methods:

• A cohort of 15,125 hemodialysis patients undergoing first deceased donor kidney transplants was analyzed.
• Cox models were employed to predict mortality, allograft loss, and combined failure.
• The cohort was split into development (N=10,083) and validation (N=5,042) sets; C-statistics and bootstrap methods assessed performance.

Main Results:

• The developed model incorporated 10 pretransplant predictors, including recipient and donor characteristics.
• The new model demonstrated strong predictive discrimination (C-statistics: 0.70 for mortality, 0.63 for graft failure).
• The model's performance was validated, showing good calibration and discrimination.

Conclusions:

• A novel prediction tool utilizing pretransplant data has been developed.
• This tool accurately predicts kidney transplant outcomes, including graft survival.
• The TransplantScore.com model shows potential to outperform current prediction methods.