*Purpose: The purpose of this study was to examine the difference between the estimated and actual values of graft survival and patient survival in kidney transplant recipients.

*Methods: The predictive accuracy of EPTS and KDPI were assessed using the area under the receiver operating characteristic curve (AUC) for the overall cohort and according to age categories (<55, 55-60, 61-65 and 66-70). We also calculated the difference between estimated failure from the actual failure probability (assessed using the cox regression model) along with absolute difference between KDPI, EPTS and the actual graft and patient survival values which was subsequently used to assess the difference across age groups using one-way analysis of variance.

*Results: The final analysis included 2,032 patients from UNOS with either a graft failure and/or death between 1/2015 and 12/2016. The KDPI was not significantly predictive of graft failure (AUC =0.52; 95% CI 0.49 to 0.54; p=0.18). The EPTS was moderately predictive of overall survival (AUC=0.67; 95% CI 0.65 to 0.70; p<0.001). The overall predictive ability of KDPI for graft failure across age categories ranged between AUC of 0.52 and 0.62 and for EPTS between 0.50 and 0.65. The difference in actual versus predicted graft failure probabilities was statistically significant across all age group comparisons and ranged between 9% to 25%. The same was true for EPTS as well with a difference ranging from 26% to 55%.

*Conclusions: Our analysis suggests kidney donor profile index and the estimated post-transplant survival scores are not predictive of graft failure and overall survival, respectively and further research is needed to identify and improve the predictive utility of these measures. Until then, quality metrics to measure individual program outcomes in addition to the predictive values for potential donor and recipient combinations should be interpreted with caution.

Based on OPTN data as of September 8, 2017

« Back to 2019 American Transplant Congress