A New Empirical Model of Deceased Donor Kidney Allocation

Kidney transplantation is the treatment of choice for End-Stage Renal Disease (ESRD). A kidney transplant improves life expectancy and quality of life and is cost-effective through savings in dialysis costs. More than 100,000 patients are currently on the national kidney waiting list, but, in a typical year, under 10,000 patients are transplanted with a deceased donor kidney. The benefits of transplantation are specific to the recipient and donated kidney, and depend on the risk of graft failure, sex, age/weight differences and the degree of antigen mismatch. Improving matches between patients and donors can therefore increase benefits from the donor pool.
A recent change in the allocation system used to match deceased donor kidneys with patients prioritizes expected graft survival by offering kidneys from healthier donors to patients who are expected to live longer. This change also prioritizes highly sensitized patients for compatible deceased donor kidneys. The new allocation system replaced a coarse point system based on few donor and patient characteristics and the patient’s waiting time. Broadly speaking, the change seeks to offer the relatively scarce resource to those who will benefit from it the most. While the organ offer system can affect patient priority, the ultimate decision of whether or not to accept an offer remains with the patient and the surgeon. This decision is likely to be influenced by three factors: 1) whether waiting for a more suitable donor is preferable to accepting the current offer, 2) the patient’s risk tolerance given his/her current dialysis experience, and 3) center-specific incentives created by competition among centers. Preliminary evidence suggests that KDRI thresholds for acceptance vary by medical factors as well as the patient’s position on the match list. Anecdotal evidence also suggests that surgeons often advise patients to wait for a more preferable organ. These aspects of patient-surgeon choice and its dependence on the organ offer system has been previously ignored during the allocation system design. Investigating the determinants of this choice can suggest improvements in two ways. First, offering kidneys to patients who are likely to accept the offer can reduce the lag time between organ procurement and transplantation (cold-time), which can directly increase the life-years supported by an organ. Second, any offer system will have patients who are commonly prioritized and others who have low priority. While sometimes justified due to specific ethical or fairness considerations, such differences can cause unintended harms. For example, patients with low priority may accept organs that would provide more benefit to others. The goal of the proposed research is to establish and estimate a new empirical model of agent choices for analyzing the impact of the recent changes to the offer system and to evaluate alternative designs. The novelty in our approach is the direct consideration of decision rules in the design process. The recent change presents a unique opportunity for this exercise. An improved understanding of agents’ choices promises avenues for leveraging the match-specific benefits of particular patient-donor pairs.