MELD and PELD: Application of survival models to liver allocation

In 1998, the Department of Health and Human Services (DHHS) issued the final rule,1 in which the principles of organ allocation were defined, to govern the operation of the Organ Procurement and Transplant Network (OPTN). This rule included the following guidelines for organ allocation: (1) organs should be allocated to transplant candidates in the order of medical urgency; (2) the role of waiting times should be minimized, and (3) attempts should be made to avoid futile transplants and to promote efficient use of our scarce donor organs. The consensus opinion to minimize waiting time was based on 2 recent reports that analyzed the impact of waiting time on survival of liver patients on the United Network for Organ Sharing (UNOS) waiting list: one from the Institute of Medicine2 and a second report from Freeman et al.3 Both studies concluded that waiting time did not correlate with death on the waiting list and therefore should be de-emphasized in developing a new allocation algorithm. The challenge put forth by this conclusion was to create an allocation policy that made the most effective use of organs especially by making them available, whenever feasible, to the most medically urgent patients who are appropriate candidates for transplantation. This challenge was accepted by the UNOS Liver and Intestinal Committee, whose task it was to (1) make an assessment of the current allocation policy including the Child-Turcotte-Pugh (CTP) score, (2) evaluate a number of previously published survival models that were developed to estimate survival of patients with end-stage liver disease, and (3) develop a new disease severity index to be utilized to allocate liver donor organs in the future. After careful deliberation and extensive input from transplant hepatologists and surgeons, a number of guidelines were established for creating an index of disease severity to estimate survival in patients with chronic liver disease. By consensus, it was determined that such a disease severity index should rely on a few, readily available, objective variables that would be generally applicable to a heterogeneous group of patients with end-stage liver disease, to determine the risk of dying. Finally, the severity index should be clinically and statistically validated and be able to predict the probability of death in groups of patients with chronic liver disease who are demographically diverse and of varying etiology and disease severity. There was agreement among the committee that such a new index should not be introduced without careful prospective evaluation of the potential impact that such a model may have on the gravely ill patients awaiting liver transplant.