Domino paired kidney donation: a strategy to make best use of live non-directed donation.

Current models for allocation of kidneys from living non-directed donors Living non-directed (LND) donors, also known as altruistic, good Samaritan, anonymous, or benevolent community donors, are a new and rapidly growing source of solid organs for transplantation. The willingness of individuals to donate organs without a designated recipient has been unexpected, but has probably developed as a societal response to the growing crisis in organ availability. In the context of this shortage, health professionals have attempted to make the best use of kidneys from LND donors. We present a novel application of paired donation that has the potential to multiply the number of recipients who can benefi t from each LND donation. At present, there is no universally accepted system for allocation of organs from LND donors. Selection of recipients has been at the discretion of the transplant centres where LND donors have presented and has generally been guided by one of three models: donorcentric, recipient-centric, or sociocentric allocation. Each of these models is supported by valid ethical arguments. The main goal of donor-centric allocation is to ensure a successful outcome for the recipient. A good outcome provides justifi cation for medical professionals to assist a person who is not ill to put themselves in harm’s way to aid another. A positive result also gives an LND donor a sense that their eff ort was fruitful and worthwhile. However, this model dictates allocation to the healthiest patient on the transplant waiting list. These recipients are the most likely to have good outcomes on dialysis or with organs from deceased donors, and therefore are arguably the least in need. Recipient-centric allocation is based on the belief that society has a responsibility to protect its most vulnerable and disadvantaged members. Under this model, organs from LND donors are given to those patients in the greatest need, those for whom a kidney transplant might be truly life saving, or those disadvantaged under the existing system for allocation of kidneys from deceased donors. This model mainly benefi ts children, patients who have no vascular access, highly sensitised patients, and those with life-threatening medical illnesses related to dialysis. However, because the recipient-centric model accords priority to such patients, it tends to yield unacceptably poor transplant outcomes, and could lead to a negative public perception of LND donation. Under the third model, of sociocentric allocation, the LND donated organ is treated as a public resource that should be allocated in the fairest and most equitable way, irrespective of outcome or need. This rationale dictates that the recipient should be the patient at the top of the transplant waiting list administered by the United Network of Organ Sharing (UNOS). UNOS oversees the allocation of deceased donor organs in the USA, using a so-called match run algorithm that ranks potential recipients according to agreed criteria. The limitations of this model are that a patient at the top of the list will probably receive a kidney from a deceased donor in the near future, and that they will have already incurred the costs, and exposure to comorbidity, that result from a long period on dialysis. The waiting list for deceased donor kidneys can be circumvented by patients who fi nd a willing live donor. But direct donation might be complicated by diff erences in blood type and by HLA sensitivity. Some incompatible donor-recipient pairs enter into programmes that facilitate paired donation, also known as kidney paired donation. A donor and recipient who have incompatible blood groups or HLA sensitivity can be matched with another incompatible pair, to result in two compatible transplants (fi gure). Although there are many ways to match up a pool of incompatible pairs, the mathematical technique of optimisation helps to fi nd out which matches will yield the best results. Nevertheless, even in paired-donation programmes in which mathematical optimisation is applied, more than 50% of the incompatible pairs in the pool remain unmatched. In many cases, pools of incompatible donor-recipient pairs have a high proportion of patients with blood types that are hard to match and those with HLA sensitisation.