COMPUTATIONAL RESEARCH in BOSTON and BEYOND (CRIBB)

Kidney transplantation is the best option for many end-stage-renal-disease patients, but about one-third of patients who have a willing live donor will find that they have an incompatibility that prevents the direct donation. Paired kidney exchange is an innovative but embryonic approach that allows an incompatible donor/recipient pair to exchange kidneys with another donor and recipient in a similar situation. To choose the best method of selecting patients and donors to exchange kidneys before any national paired donation program begins, we examine different algorithms for arranging paired kidney exchanges on simulated patient populations. Optimal matchings over a graph representing the incompatible pairs and possible trades can guarantee that the patient-donor pairs are matched to the greatest possible benefit. Maximum edge weight matching affords both customizing each patient's match to his preferences and protecting disadvantaged groups. Including triple exchanges presents a computational challenge, as does projecting wait-list growth over more than three years.

This is work with Dr. Dorry Segev and Dr. Robert Montgomery, both transplant surgeons at the Johns Hopkins School of Medicine.