The prognostic Kidney Donor Risk Index (KDRI)-developed and internally validated in the United States-is a widely used tool to predict transplant outcome of a deceased donor kidney. The KDRI is currently used for longevity matching between donors and recipients in the United States. We aimed to externally validate the KDRIdonor-only and KDRIfull as proposed by Rao et al (2009). KDRIdonor-only consist of 10 donor factors, and KDRIfull with an additional 4 transplant factors. We used the Dutch Organ Transplantation Registry to include 3201 adult recipients transplanted from 2002 to 2012. The median Dutch KDRI was 1.21 and comparable with the year 2012 in the United States (median of 1.24). The calibration-slope was 0.98 and 0.96 for the KDRIfull and KDRIdonor-only, respectively, indicating that predictions of graft failure were on average similar. The discriminative ability (Harrell C) of the KDRIfull and the KDRIdonor-only at 5 years was 0.63 (95% confidence interval [CI], 0.62-0.64), and 0.62 (95% CI, 0.61-0.63), respectively. We found misspecification of 3 KDRI factors: age (P = 0.002), weight (P = 0.017), and cold ischemia time (P < 0.001). Adding the use of inotropic drugs before donation (P = 0.040) and the interaction between circulatory-death donor kidneys and prolonged cold ischemic time (>24 hours vs 12 hours; P = 0.059) could improve predictive ability. The KDRI performs equal in the Dutch population. Discriminative ability of the KDRI indicates limited clinical use for adequate individualized decisions. An updated KDRI may contribute to a standardized policy meeting the growing demand of donor kidneys in the Eurotransplant region.
Read full abstract