Phase II of the offer filters pilot will enable further improvements to the kidney donation process and increase organ utilization across the country.
United Network for Organ Sharing will soon start the second phase of a pilot project aimed at improving the kidney donation process and increasing organ utilization across the country. Emerging from a collaborative effort among the donation and transplantation community, the offer filters tool will help organ procurement organizations find accepting kidney offer candidates more quickly by allowing transplant program staff to preemptively filter the donor offers that they do not want to receive. Additionally, the project will use big data analytics to identify consistent, multi-variable refusal patterns at kidney programs. “We expect that by reducing unwanted kidney offers, we’ll be able to reduce cold ischemic time and ultimately increase organ acceptance,” said UNOS business architect Rob McTier. Phase II of the offer filters pilot project is scheduled to start in the summer.
Participating programs will be able to activate offer filters, review effectiveness
During phase I, which began in June 2019, pilot project participants representing 29 kidney programs from across the country used the Offer Filters Explorer tool to identify the organ offers that they did not want to receive. Though the filters did not actually screen offers during this phase, UNOS information technology team members used the information provided by participants to collect data on whether each offer would have been screened if the filter was turned on. The filter information was then displayed when a program received a kidney offer. “We reviewed data from the pilot participant’s feedback after phase I ended in late 2019,” McTier said, adding that the team referred to the data to inform phase II of the project.
In phase II, participating transplant programs will be able to activate the offer filters. When a filter is on, the filter criteria will be compared with donor data before organ offers are sent. When an OPO initiates an organ offer, a bypass code will be applied to all candidate offers in which the kidney donor data meets the filter criteria, the filter is turned on and no response has already been entered.
Participating programs will be able to exclude certain transplant candidates from being bypassed by making adjustments to the filter’s candidate exclusion criteria, but otherwise the filters will apply to all candidates in a participating program’s waitlist. Participating programs will also be able to generate real-time reports to see the effectiveness of their filters.
Recently, McTier answered questions about the offer filters tool and talked about what phase II means for participating OPOs and organ transplant programs.
Can you tell me about the genesis of the project?
The idea for this project first came to my attention when I had a conversation with Dr. Ken Andreoni. He’s the director of kidney and pancreas transplant at UF Health in Gainesville, Florida and a past president of UNOS Board of Directors. He was showing me a spreadsheet he had developed at his transplant program to evaluate organ offers. In the course of the conversation, we started talking about whether there was a way to program the type of spreadsheet he had developed into UNetSM in order to increase the efficiency of the organ offer process.
As we originally envisioned it, a transplant program would describe the types of organ offers that they would accept. But as we collaborated with the working group and further examined this idea, we wondered if participants were able to describe all the different types of organ offers that they would accept. Based on that interaction, we changed the idea so that participants would describe the offers they didn’t want and we decided to call it offer filters.
After phase I was complete, what kind of feedback did you get? How did you use that information to inform phase II?
We conducted the phase I pilot from June until September of 2019 with participants from 29 kidney programs. During the phase I pilot, we weren’t actually bypassing organ offers. We instead created modeling that looked at a program’s past two years of organ offers and came up with recommended filters for them to use. The participating programs could modify these recommend filters or add new filters of their own. Then we observed the offers that the program received and compared them to their filters for three months, and surveyed the pilot participants about their individual filters. We asked them if they would have turned on the filters we provided to them. In those surveys, many responded that they would bypass at least on one of the filters, so we felt like that was a pretty good green light for proceeding with phase II.
In phase I, pilot participants set filters for their programs and got reports at the end. In phase II, we’ll actually have real-time reporting in the pilot so participants can see what’s happening with their filters, whether they choose to bypass or not.
What should participating transplant program members know about setting up candidate exclusion criteria for each filter?
In addition to the real-time reporting, we also are going to provide candidate exclusion criteria for each filter as an additional safeguard. This will allow transplant programs to be able to exclude certain candidates based upon their CPRA, whether there is zero-antigen mismatch, and the candidate’s age. For example, you could come up with a filter that would normally filter out kidney offers if the donor is more than 60 years old and the distance exceeds 500 nautical miles, unless the candidate’s CPRA exceeds 80 percent, or the candidate is a zero-antigen mismatch.
The algorithm will determine if the donor meets that criteria, but also if the candidate meets the exclusion criteria. So for the example I just provided of the 60-year-old donors that are from far away, the high CPRA candidates and the zero-antigen mismatch candidates would continue to get offers, but the other candidates on the list would get bypassed.
How are those criteria used during the filters process in DonorNet®?
In UNet right now, the transplant programs have donor acceptance criteria on the waitlist, but those criteria are only applied when an OPO runs a match at the beginning of the process, usually before they’ve gone to the operating room, before cross-clamp, and before there’s cold ischemic time on the donor or the organs. A big difference with the offer filters project is that we’re applying those filtering criteria at the time the OPO makes the offer. That way, the offer filters will have the most up-to-date information, and we can include additional criteria like cold ischemic time.
How will transplant programs and OPOs benefit from the offer filters tool?
In the long run, we’re trying to increase the number of transplants by getting to organ offer acceptance faster. In the current process, there are people who are getting offers way down on the match run saying, “I would have taken this if you had called me sooner, but now this organ has so much cold ischemic time on it so I’m going to have to turn it down.”
What we’re attempting to do is to make the process more efficient so that OPOs don’t need to make as many unwanted organ offers, and transplant programs don’t have to respond to those offers, which will ultimately decrease the number of organs that are deemed not suitable for transplant and increase organ acceptance, particularly with hard-to-place organs that currently go way down the match.
How will patients benefit?
Our ultimate goal is to improve outcomes for the candidates on the waiting list. This means getting organs to candidates who could benefit from them as quickly and efficiently as possible, and minimizing the amount of cold ischemic time that the organ accumulates. Getting the offer to the right candidate in a more targeted manner can lead to an increase in organ utilization as well as in the number of candidates receiving a transplant. Reducing the amount of cold ischemic time can also help improve outcomes for those candidates after transplant.
