Hybrid Kidney Allocation: Improving Equity by Fair Integration of Deceased Donor-Initiated Chains in Kidney Exchange

Problem definition: Living donor kidney exchange programs (KEPs) expand access to transplants by enabling exchanges among patient-donor pairs. However, highly sensitized patients are disproportionately hard to match and face long waits. Deceased donor–initiated chains (DDICs) have been proposed to enable more transplants by diverting an occasional deceased donor kidney from the waitlist to start a KEP chain, which ends by returning a kidney to the waitlist. Yet, DDICs have seen virtually no implementation. In the U.S., this is largely due to concerns that the kidneys returned may differ in blood-type distribution from those diverted, disadvantaging waitlist patients of certain blood types. We study how DDICs can be integrated into KEPs while explicitly protecting blood-type-specific waitlist access. Methodology/results: We develop two forward-looking optimization approaches: one that solves a stochastic model using Sample Average Approximation (SAA) and one that solves a static model with forward-looking constraints. Through simulation, we show that forward-looking models are needed to preserve the blood-type distribution of donations to the waitlist. Under both models, even limited use of DDICs significantly increases transplant rates for highly sensitized patients. While the static approach requires only modest extensions to models as currently used in practice, the more complex and time consuming SAA-based approach yields larger equity gains. Managerial implications: DDICs can improve access to life-saving transplants for the most disadvantaged KEP patients without reducing access for any blood type on the waitlist. By explicitly controlling the blood-type distribution of waitlist donations, matching policies can address objections that have hindered DDIC adoption, calling for renewed policy evaluation.