Our Lifesaving Algorithm.

In the U.S., more than 100,000 people are waiting for a kidney transplant, but only 17,000 of them will receive a kidney this year. Of those, only 6,000 come from living donors. For the kidney patients who are fortunate enough to have a friend or family member step forward to donate, most are incompatible.

For patients and their families, the odds are overwhelming. The Alliance for Paired Kidney Donation is rewriting that story.

By including those incompatible pairs in the paired donation registry, the APKD is seeing more than a 50% rise in matches. That means 50% more patients who are finding the gift of life. With more pairs in the system, the rise in matches will continue to climb, and that’s exactly what the APKD is working toward.

The goal is to get the transplant center personnel to refer patients to the alliance for paired kidney donation so we can match more people and save more lives. The Scoring Rubric below explains what that process is and is meant for the audience of clinical people at the transplant center.

How does it work?

The APKD uses a Nobel prize-winning algorithm, developed by Alvin Roth, to match one incompatible pair with another. A scoring rubric, developed by a team of healthcare professionals, is used help find the best possible match. 

Background on what the algorithm is doing:

For example, suppose Tom needed a kidney and his brother Stan was willing to donate one of his kidneys to Thom, but he was not a suitable tissue “match”.

Similarly, Nancy needs a kidney and her high school teacher, Randy is willing to donate a kidney, but he does not match Nancy.

The APKD algorithm finds suitable pairings, Stan could safely donate his kidney to Nancy (and Randy to Tom), such that everyone benefitted. The APKD software creates chains like this with Transplant centers around the country.

Link to helpful resource: https://www.today.com/video/how-a-kidney-chain-is-saving-lives-an-awesome-gift-130256453794

Scoring Rubric

The APKD software awards points according to the following criteria:

DR-locus mismatches

0 mismatches

2.0

1 mismatch

1.0

Any other mismatches

0.0

Zero-antigen mismatch

Yes

6.0

No

0.0

High PRA

>=80

10.0

>=50

6.0

Any other PRA

0.0

Travel distance

Same center or same city

3.0

Any other distance

0.0

Pediatric recipient

Age <=5

4.0

Age <= 17

2.0

Any other age

0.0

Recipient once was donor

Yes

6.0

No

0.0

Known negative cross match and PRA >=80

Yes

6.0

No

0.0

 

“The way the algorithm works in kidney transplants begins with data that we speak of as a compatible graph. You can think of a compatible graph as the participants in the kidney exchange, where patients and donors are pairs. Sometimes there are no direct donors, and sometimes there are people who are waiting for a donor kidney, but don’t have a living donor. What the algorithm does is try to recommend a set of transplants as it tries to find the maximum gauged set of matches. So, it tries to find as many of those as possible, and strings them together in a chain, so the surgeons can go ahead and accomplish those transplants.”
Alvin Roth
2012 Nobel Prize in Economics Sciences