The Best‐Laid Schemes of Mice and Men Often Go Awry; How Should We Repair Them?

S. Gentry,D. Segev

DOI: https://doi.org/10.1111/ajt.13414

IF: 9.369

2015-10-01

American Journal of Transplantation

Abstract:Optimization algorithms have played a central role in matching for kidney paired donation to achieve the greatest benefit (1). In practice in the United States, matches selected using optimization are frequently scuttled, either because one or more of the pairs has already matched in a different registry, or because a donor turns out to be unacceptable to, or has a positive crossmatch with, a matched recipient (2). These failed matches pose a challenge for matching systems: what should then be done with the pairs involved? An intuitive method for handling this, which achieves transplant rates almost as high as the ideal world (where all proposed matches would succeed), is to move forward with all of the matches that were successful, add the failed pairs back into the pool, and re-optimize (1).