Effect of match-run frequencies on the number of transplants and waiting times in kidney exchange

Abstract

Numerous kidney exchange (kidney paired donation [KPD]) registries in the United States have gradually shifted to high-frequency match-runs, raising the question of whether this harms the number of transplants. We conducted simulations using clinical data from 2 KPD registries-the Alliance for Paired Donation, which runs multihospital exchanges, and Methodist San Antonio, which runs single-center exchanges-to study how the frequency of match-runs impacts the number of transplants and the average waiting times. We simulate the options facing each of the 2 registries by repeated resampling from their historical pools of patient-donor pairs and nondirected donors, with arrival and departure rates corresponding to the historical data. We find that longer intervals between match-runs do not increase the total number of transplants, and that prioritizing highly sensitized patients is more effective than waiting longer between match-runs for transplanting highly sensitized patients. While we do not find that frequent match-runs result in fewer transplanted pairs, we do find that increasing arrival rates of new pairs improves both the fraction of transplanted pairs and waiting times.

On Matching and Thickness in Heterogeneous Dynamic Markets

“…20 There is some concern that this behavior is inefficient (and arguably a result of competition). However, numerical simulations by Ashlagi et al (2017) suggest that in steady-state there is essentially no harm from frequent matching (though having multiple small platforms does harm efficiency). Moreover, MSA is not facing any competition.…”

“…11 Akbarpour et al (2014) consider a homogeneous model with departures and finds that the optimality gap of the policy that matches without waiting remains constant as the match probability decreases. Moreover, using data-driven simulations, Ashlagi et al (2017) study the impact of match-run frequency, and show that among polices that match periodically (e.g., every week or every day), high matching frequencies perform best. 12 This paper builds on these findings, and only analyzes myopic policies that search for a match upon arrival of a new agent.…”

The 6‐year clinical outcomes for patients registered in a multiregional United States Kidney Paired Donation program ‐ a retrospective study

“…Despite this change APD-KPD patients had only 3% of transplants with delayed graft function (DGF), which was similar to 4% in OPTN-KPD and to 4.5% in OPTN-LD ( Figure S1B). Additional review of donor's age in APD-KPD showed that despite 7-h CIT grafts from 50to 64-year-old donors had similar 3-year graft survival as from young [18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35] year-old donors, which was confirmed by a parallel analysis of APD-LD, OPTN-KPD and OPTN-LD ( Figure S2A-E). Thus, we found no disadvantage in the use of older donors for the KPD program.…”

“…In contrast, one failed transplant in a cycle scrapped all planned transplants. Even with an average CIT of 7-h 50-64 year old donors produced good quality transplants comparable to [19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35] year old donors.…”

“…Recent reports described individual KPD programs with their own complexity and uniqueness [33,35,[49][50][51][52]. While demographics of the National Kidney Registry (NKR) were similar to other KPD and UNOS live donor programs, the NKR had 22.7% HS/VHS patients (cPRA > 80%) in comparison to 4.3% among LD in UNOS [51].…”