Unpaired Kidney Exchange

Research Seminars

Overcoming the Double Coincidence of Wants Without Medium of Exchange

For an incompatible patient-donor pair, kidney exchanges often forbid receipt-before-donation (the patient receives a kidney before the donor donates) and donation-before-receipt, causing adouble-coincidence-of-wants problem. The proposal, the Unpaired kidney exchange algorithm, uses “memory” as a medium of exchange to eliminate these timing constraints. In a dynamic matching model, they prove that Unpaired delivers a waiting time of patients close to optimal and substantially shorter than currently utilized state-of-the-art algorithms. Using a rich adminis-trative dataset from France, the authors show that Unpaired achieves a match rate of 57 percent and an average waiting time of 440 days. The (infeasible) optimal algorithm is only slightly better (58percent and 425 days); state-of-the-art algorithms deliver less than 34 percent and more than 695 days. they draw similar conclusions from the simulations of two large U.S. platforms. Lastly,the authors propose a range of solutions that can address the potential practical concerns of Unpaired.