Date of Conferral







Branford McAllister


The purpose of this experimental research was to determine whether using delta-MELD as a criterion for the liver transplant patient selection process could improve the U.S. liver allocation system. This research closed a gap in current literature on the utility of delta-MELD for liver transplant patient selection. The frameworks of systems theory, the analytic hierarchy process, and the Kalman filter contributed to the development of 2 simulation models of the liver allocation system: one that used delta-MELD and one that did not use delta-MELD. The research question examined whether using delta-MELD could improve the liver allocation system by reducing the number of patients dropping off the wait list and lowering the average MELD score. Statistical t tests of 2 independent scenarios (allocation with and without delta-MELD), each with 70 runs of 180 simulated days on the liver allocation wait list, did not indicate a significant improvement to the liver allocation system by using delta-MELD for liver allocation. However, observations made from the simulation experiment, such as the median patient wait time being 11 months and delta-MELD being more variable at the end-stage of liver diseases, provided insights into how to improve the model of the liver allocation process. In addition, observations made from the status 1 patient subgroup (patients in ICU with about 7 days to live), which were excluded from this research, suggested including status 1 patients and expanding the simulation timespan from 180 to 360 days to better capture the delta-MELD variability from patients at the end-stage of liver disease. This research provides empirical evidence on the applicability of the delta-MELD criterion for non-status 1 patients, and yields recommendations to include status 1 patients in an improved simulation of the donor liver system while using delta-MELD as criterion.