A Look-Ahead Strategy for Non-directed Donors in Kidney Paired Donation

Abstract

While there is a growing need for kidney transplants to treat end stage kidney disease, the supply of transplantable kidneys is in serious shortage. Kidney paired donation (KPD) programs serve as platforms for candidates with willing but incompatible donors to assess the possibility of exchanging donors, thus opening up new transplant opportunities for these candidates. In recent years, non-directed (or altruistic) donors (NDDs) have been incorporated into KPD programs beginning chains of transplants that benefit many candidates. In such programs, making optimal decisions in transplant exchange selection is of critical importance. With the aim of improving the selection of chains beginning with an NDD, this paper introduces a look-ahead multiple decision strategy to select chains, that are easy to extend in the future. Simulation studies are adopted to assess performance of this strategy. Taking into account the extensibility of chains increases the number of realized transplants.

Kidney Exchange: An Operations Perspective

“…47 Simulations that consider this tension can be found in Santos et al (2017) and Ashlagi et al (2018b). 48 Some studies consider the problem of maximizing the expected number of matches (Klimentova et al 2016, Wang et al 2017, Bidkhori et al 2020 or using robust optimization (McElfresh et al 2019). Several papers consider identifying exchanges that are not necessarily disjoint, after which more than one crossmatch would be possible, while having "backup" options (Blum et al 2013, Bray et al 2015, Carvalho et al 2020.…”

An efficient algorithm to enumerate sets with fallbacks in a kidney paired donation program

“…A kidney paired donation (KPD) program provides a partial yet clinically important solution to such incompatibility [1][2][3]. A KPD program consists of (1,3,2), (3,4) and (3,6,4). (Note that (1,3,2) refers to a cycle in which the donor in pair 1 gives to the candidate in pair 3, donor 3 gives to candidate 2, and donor 2 gives to candidate 1 to complete the cycle.)…”

“…(Note that (1,3,2) refers to a cycle in which the donor in pair 1 gives to the candidate in pair 3, donor 3 gives to candidate 2, and donor 2 gives to candidate 1 to complete the cycle.) If we use the number of potential transplants to measure the value of each cycle and conduct a match run, then the solution would involve choosing either the cycle (1,3,2) or the cycle (3,6,4). With either choice we would realize three transplants if all vertices and edges are viable and otherwise, no transplant is realized.…”

“…With either choice we would realize three transplants if all vertices and edges are viable and otherwise, no transplant is realized. It is easy to see, however, that there is an advantage to the choice (3,6,4) if we consider the possibility of falling back to (3,4) as follows: if (3,6,4) succeeds, then 3 transplants are realized; if (3,6,4) fails but (3,4) succeeds, then the 2 transplants in (3,4) are realized; if both (3,6,4) and (3,4) fail, then no transplant is realized. Here (3,4) serves as a fallback.…”

Statistical Methods in Organ Failure and Transplantation

“…Using data obtained from the United States Renal Data System (USRDS), the proposed methods are used to compare mortality on hemodialysis versus peritoneal dialysis, a long-standing yet unresolved issue in ESRD treatment. The increasingly important topic of kidney paired donation (KPD) is studied by Wang et al (2018), who develop a novel multiple-look-ahead decision strategy to select transplant chains.…”