*Purpose: The donor risk index (DRI) is one of the most established risk indices in liver transplant (LT). DRI, which was reported in 2006, was developed by obtaining patient information who received LT before the MELD era (1998-2002). The improvement of donor/recipient management and changes of allocation system over the decades may have influenced the predictive power of DRI. The aim of this study was to investigate chronological changes of predictive power of DRI and elucidate which factors gained/lost power of prediction.

*Methods: The study includes 80,998 patients who received LT from adult deceased donors and were registered in the scientific registry of transplant recipients between 1998-2015. The prognostic power of DRI was evaluated by Harrell’s C-index of 3-years graft survival (GS) and area under curve (AUC) using ROC analysis of 1-year GS. The contribution of prognostic power was evaluated by Wald test calculated by Cox regression analysis.

*Results: The C-index and AUC of the DRI in original cohort was 0.59 and 0.59, respectively. Both indicators declined over time regardless of the mean DRI score was almost constant over time (Figure 1). As for the factors of DRI, age of donor significantly lost prognostic power during study period. Also, race, sharing, and partial/split lost their prognostic influence over time. On the other hand, donor after circulation death, cold time, cause of death, and height of donor kept their prognostic power similar to the original DRI (Figure 2).

*Conclusions: Half of the DRI component factors decreased their prognostic power. Recalibration of the DRI is needed.

« Back to 2019 American Transplant Congress