Chains over cycles are selected by computer software optimization in the kidney paired donation program

Aim Failure to convert computer-identified possible exchanges into transplants limits Kidney Paired Donation (KPD) program. We have analyzed the efficiency of simultaneous closed loop exchanges (cycles) vs. non-simultaneous extended altruistic donor (NEAD) chains in generating KPD transplants. Methods/results The Alliance for Paired Donation (APD) utilizes an optimization algorithm selecting cycles or chains. All chains are initiated by a non-directed donor (NDD) and end with a bridge donor, who enters back into the pool. This bridge donor can then initiate another NEAD chain. Computer-generated potential cycles and NEAD chains are presented to transplant centers for review as formal offers. We tracked the progress in moving from offers to completed transplants to calculate the success rate. Between 2007 and 2014 the APD performed 225 transplants: 194 within the APD and 31 in collaboration with other KPD programs. Of the 194 APD transplants, 61 (31.4%) were performed through cycles (Fig. 1a) and 133 through NEAD chains (68.6%; Fig. 1b), showing that chains were two times more successful than cycles. The APD made 447 offers to transplant centers, of which 173 (38.7%) were cycles and 274 (61.3%) were chains. Consequently, the success rate of cycles was 14.5%, chains was 26.6%, while the overall success rate of 21.9%. Conclusions Although chains were not actively sought over cycles, the results showed that chains were preferentially found by the optimization algorithm and more successfully progressed to completed transplants than cycles. Thus, the concept of a bridge-donor introduces significant flexibility in generating transplants. We propose that NEAD chains are the most effective tool to increase the efficiency in KPD programs. Download full-size image