Congress dictated in 1998 that neither place of residence nor place of listing should be a major determinant of access to a transplant, but geographic disparities in access to kidney and liver transplantation are a major problem that has persistently worsened. Kidneys are allocated by waiting time, but there are 8-fold differences in median waiting times across donation service areas (DSAs). Livers are intended to go to the candidate with the highest MELD (model of end-stage liver disease) score, but instead median MELD varies by 10 points across DSAs, and 90-day probability of LT for candidates with high MELD varies from 18%-86%. The Organ Procurement and Transplantation Network (OPTN) unanimously resolved in 2012 that the existing geographic disparity in allocation? is unacceptably high, calling for studies of optimized systems. A few ad hoc fixes have been proposed, but optimizing policy has rarely been explored, and no comprehensive solution has ever been implemented. Even the two recent major policy changes, Share-35 and the new kidney allocation system (KAS), did not explicitly address geographic disparities. Clamorous scientific controversy over whether the geographic disparity in organ availability is truly structural or merely reflects heterogeneity in clinical practice and organ procurement organization performance contributes to the policy stalemate. We have started working closely with the OPTN Liver Committee to design optimized sharing districts that would significantly reduce geographic variation in access to deceased donor livers. However, this preliminary work is currently limited by poor metrics of geographic variation and uncorrected measures of supply and demand. Furthermore, other promising concepts, like redrawing DSA boundaries, or dynamic prioritization of candidates without pre-determined maps, have not yet been explored or evaluated. Finally, no work has been done to adapt optimization models designed for liver allocation to the different priorities of kidney allocation. We will develop new approaches to making transplantation more equitable, considering both clinical practice and allocation systems. Using an innovative adaptation of mathematical optimization and statistical inference to transplant policy, we will isolate the fundamental drivers of geographic disparity and outline promising policies to make deceased donor organs equally available to candidates across the country. We will address the following aims: (1) To construct metrics of geographic variation in access to liver and kidney transplantation that adjust for heterogeneity inherent to clinical practice; (2) To measure fundamental demand for and supply of organs; (3) To understand the impact of recent allocation changes on clinical decision-making and geographic disparities; and (4) To design and test optimized sharing systems that reduce geographic disparity while explicitly accounting for uncertainty in supply and demand. The proposed research directly addresses a Congressional mandate and uses the approach favored by the OPTN. More equitable organ distribution will save lives and improve the care of over 100,000 patients currently at risk of dying on the waiting list.
Candidates waiting for kidney and liver transplants in different parts of the country face very different odds of receiving an organ before they die on the waiting list, and these disparities in allocation are large, troubling, and violate federal law. The Organ Procurement and Transplantation Network has stated that the existing geographic disparity in allocation is unacceptably high, and has called for studies of optimized systems to replace the ad hoc fixes that have so far failed to resolve the problem. The goal of this research is to carefully determine how to measure geographic disparity in access to transplants, to model the supply of organs and the demand for transplants, and then to use mathematical optimization techniques to reduce geographic disparity by designing the best possible organ allocation systems.
