1. Removal of DSA and Region from Kidney and Pancreas Distribution
Kidney and Pancreas Transplantation Committees
Introduce yourself, your committee and what organization you’re from.
Note that this is a joint project between the kidney and pancreas committees, and we are looking for feedback on both distribution systems so please keep that in mind when you listen to the proposal. A pancreas committee member will be available to answer questions for this presentation as well.

2. What problem does the paper address?
Ensure compliance with the Final Rule by proposing potential solutions to remove DSA and region from allocation policies
Concept paper does not change policy
Purpose to inform and gather feedback on options for:
Removing DSA and region in kidney and pancreas distribution
Aligning allocation policies with the Final Rule
Moving kidney and pancreas policies toward continuous distribution framework
This proposal seeks to ensure compliance with the Final Rule by proposing potential solutions to remove DSA and region from allocation policies.
The concept paper does not propose a change to policy. It serves to gather feedback and more evidence on policy solutions before such a solution is issued for public comment.
The purpose is to inform and gather feedback on options for:
Removing DSA and region in kidney and pancreas distribution
Aligning allocation policies with the Final Rule
Moving kidney and pancreas policy towards single allocation framework
The workgroup wishes to move allocation towards a unified framework of continuous distribution, as directed by the OPTN Board of Directors at the December 2018 meeting. This work represents step-wise progress towards that goal.

3. Important Considerations
Proposed solutions represent a list of policy options the KP workgroup chose to model
KP Workgroup continues to consider framework variations not included in this concept paper
No decisions made about potential policy solutions
Concept paper is being pursued to solicit community feedback on Workgroup progress and findings as well as variation preferences
Kidney committee continues to deliberate further prioritization of pediatrics and prior living donor candidates in classification tables
The proposed solutions we will discuss represent a list of policy options the kidney pancreas (KP) workgroup chose to model to gather as much information about different options given the shortened time frame and limited modelling options. We are not looking for you to choose one particular model, rather we are looking for feedback on the type of frameworks selected, the distances selected, the use of proximity curves and if there are any other policy solutions to be considered. No decisions have made about potential policy solutions.
Finally, it is important to note that the Kidney Committee continues to deliberate further prioritization of pediatrics and prior living donor candidates in classification tables. This may be part of the ultimate policy solution submitted to the community for public comment and could be included in a second round of modeling following the receipt of community feedback.

4. What are the Possible Frameworks?
Fixed Concentric Circles
Variations in circle sizes
Hybrid Framework
Single fixed distance circle (500NM)
Variation in linear proximity point slope
The committee considered two possible solutions – fixed concentric circles and the hybrid framework.
Concentric Circles has been chosen by other organ types that have used this solution in replacing DSA and region with a framework more compatible with the final Rule. Fixed concentric circles were one of the frameworks approved by the Board in June 2018 as acceptable replacement options for DSA and region.
Ultimately the committees want to move kidney and pancreas allocation in the direction of continuous distribution. The hybrid framework would move distribution closer to the continuous distribution framework adopted by the board in December. The “hybrid framework” uses 1 circle from the donor hospital; beyond the circle is national distribution. In addition it includes proximity points using a linear curve both inside and outside of the circle. Two linear curve variations were modelled. A shallow curve favors broader distribution and a steep curve favors distribution closer to the donor hospital. I will describe this in more detail in the coming slides.
For each potential solution, allocation for kidney and pancreas policies would still determine the order in which potential transplant recipients would appear on the match run.

5. Fixed Concentric Circles Framework
First we’ll talk about the fixed concentric circle framework variations.

6. 150 NM/300 NM Fixed Concentric Circles
One potential solution considered by the committees replaces DSA and region with two fixed distance circles from the donor hospital. This solution would replace DSA with 150 nautical miles and region with 300 nautical miles.
This figure shows all kidney transplant hospitals with 150 NM and 300 NM circles around the top volume donor hospitals in each OPTN/UNOS region (in the contiguous United States) in order to illustrate how distribution would appear in different regions across the country for the 150/300 fixed circle variation. Darker circles represent a 150 NM radius and lighter circles represent a 300 NM radius.
Kidney and pancreas allocation would not be modified except to change geographic distribution: “DSA” and “region” would be removed and replaced with a 150 NM and 300 NM circle, respectively.

7. 250 NM/500 NM Fixed Concentric Circles
This figure shows all kidney transplant hospitals with 250 NM and 500 NM circles around the top volume donor hospitals in each OPTN/UNOS region (in the contiguous United States) in order to illustrate how distribution would appear in different regions across the country for a 250/500 fixed concentric circle variation.

8. Hybrid Framework
Next we’ll talk about the hybrid framework, which incorporates elements of the fixed circle framework and continuous distribution.

9. Hybrid Model Single 500NM Circle
This figure shows all kidney transplant hospitals with 500 NM circles around the top volume donor hospitals in each OPTN/UNOS region (in the contiguous United States) in order to illustrate how distribution would appear in different regions across the country

10. 500 NM Circle, No Proximity Points
We modelled a single 500nm circle with no proximity points as a control for the next two models in which we used proximity points. We chose 500NM since we did not have a 500 NM circle in our concentric circles model.
This figure shows how a single 500 NM circle from the donor hospital would replace DSA. Regional classifications from allocation policies would be eliminated.
Organ offers would go to candidates within the 500 NM circle before the offers would go to candidates outside the 500 NM circle (except for those exceptions noted in policy, such as mandatory national shares for the most highly-sensitized candidates).

11. 500 NM Circle, Shallow Proximity Points
This figure shows how a single 500 NM circle with shallow proximity points would work. As with the previous variation, DSA would be replaced with 500 nm from the donor hospital, and regional classifications would be eliminated.
Shallow proximity points favors broader distribution. For modelling purposes we chose shallow for both inside and outside of the circle. We did the same for steep points. It is possible to consider for example shallow inside and steep outside of the circle as an example. Of note the point values we modelled were conservative as will be seen in the travel distance slides.
Proximity points would be awarded to candidates closer to the donor hospital. Candidates would receive up to one point inside the circle, and up to two points when allocation moves outside of the circle.
A candidate registered at a transplant hospital approximately 500 NM away from the donor hospital would receive no proximity points. Proximity points are awarded linearly between those two points, so a candidate registered at a hospital 250 NM from the donor hospital would be awarded 0.5 points.
If no candidates within the 500 NM circle accept the organ offer, the organ moves out to national distribution. A candidate registered just outside the 500 NM circle would be awarded two proximity points and a candidate registered 2500 NM and beyond from the donor hospital would be awarded no proximity points.

12. 500 NM Circle, Steep Proximity Points
This figure shows how a single 500 NM circle with steep proximity points. As with the previous variation, DSA would be replaced with 500 nm from the donor hospital, and regional classifications would be eliminated. However, proximity points would be awarded to candidates closer to the donor hospital. Candidates would receive up to 2 points inside the circle, and up to 4 points when allocation moves outside of the circle.
Steep proximity points favors distribution closer to the donor hospital yet still broadens distribution.
A candidate registered at a transplant hospital approximately 500 NM away from the donor hospital would receive no proximity points. Proximity points are awarded linearly between those two points, so a candidate registered at a hospital 250 NM from the donor hospital would be awarded 1 point.
If no candidates within the 500 NM circle accept the organ offer, the organ moves out to national distribution. A candidate registered just outside the 500 NM circle would be awarded four proximity points and a candidate registered 2500 NM and beyond from the donor hospital would be awarded no proximity points.
Distance (NM) from Donor Hospital to Waiting List Candidate (Listing Center)

13. SRTR Modeling of Variations
Updated cohort of kidney, kidney-pancreas, pancreas candidates from Jan – Dec
Limitations:
Modeling does not account for changes in behavior
Acceptance behavior likely to change in response to organ availability at a center
Previous experience with the SAMs suggests that they under- predict the number of transplants that would occur in reality if a given policy scenario were adopted
Transplant counts and rates unlikely to decline in reality
The SRTR performed modeling based on an updated cohort of kidney, kidney-pancreas, pancreas candidates from 2017.
It is important to note some limitations of the KPSAM modeling
Modeling does not account for changes in behavior
Acceptance behavior likely to change in response to organ availability at a center
Previous experience with the SAMs suggests that they under-predict the number of transplants that would occur in reality if a given policy scenario were adopted
SRTR consistently stated to the KP workgroup that the KPSAM modeling is limited to current application practices. According to the SRTR, “the KPSAM was fit on acceptance occurring within a local (DSA), regional, and national framework, wherein there’s a strong preference for local offers. Acceptance behavior will likely change in response to changes in organ availability at a center, and transplant counts and rates may not decline in reality.”

14. Results of SRTR Modeling
Pediatric kidney transplant rates increased compared to adults
High-cPRA kidney, kidney-pancreas and pancreas transplant rates increased
Relatively more kidney and pancreas transplants occurred in African-American recipients compared to white recipients
Relatively more kidney transplant counts for:
Recipients with > 10 years of dialysis time
Recipients with 0-DR mismatches
This slide shows the impact on vulnerable populations and subgroups shown in the SRTR KPSAM modeling. Overall, positive impact was seen on several subgroups that can be considered vulnerable:
Pediatric kidney transplant rates increased compared to adults
High-cPRA kidney, kidney-pancreas and pancreas transplant rates increased
Relatively more kidney and pancreas transplants occurred in African-American recipients compared to white recipients
Relatively more kidney transplant counts for:
Recipients with > 10 years of dialysis time compared to recipients with less time
Recipients with 0-DR mismatches

15. Results of SRTR Modeling
No impact on waitlist mortality across organs
Overall, modeling projected decline in transplant rates and counts for kidney and pancreas alone
Increased kidney-pancreas transplant rates compared to baseline
No impact on waitlist mortality across organs
Lower kidney, pancreas transplant rates compared to the baseline for all variations
Increased kidney-pancreas transplant rates compared to baseline
Again, keep in mind the limitations of KPSAM modelling when interpreting transplant counts and transplant rates.

16. Projected Changes in Travel Distance
One of the key takeaways from the SRTR modeling requested by the Workgroup are the projected changes in the shape of distribution by distance. At their October 15th, 2018 meeting, the majority of the OPTN/UNOS Kidney Committee agreed that broader distribution of kidneys is a value that they would like to see strengthened in whichever framework variation is selected.  
In this figure, the top-left square represents kidney allocation, the top-right square represents kidney-pancreas allocation, and the bottom-left square represents pancreas allocation. From left to right, the distribution shapes represent the baseline policy (orange), the 500NM no points model, 500 NM shallow points model, and 500 NM steep points model (yellow), and the 150 NM/300 NM circles model and 250 NM /500 NM circles model (blue).
Each of the five proposed models would broaden kidney and kidney-pancreas distribution compared to the baseline, which represents distribution under the current policy. By contrast, pancreas alone distribution is projected to be less broad compared to the baseline. As noted before, pancreas alone represents a minority of all pancreas transplants, with a majority being comprised of KP transplants.
Orange = Baseline (Current policy)
Yellow = Hybrid framework variations
Blue = Fixed Concentric Circle Variations

17. Specific Feedback Requested
Pancreas Committee presenter

18. Specific Feedback Requested
Which framework do you prefer?
Within the framework you selected, which circle size(s) do you prefer?
Which points variation do you prefer for the hybrid framework?
Should there be different distribution systems for kidney and pancreas organs?
PANCREAS COMMITTEE MEMBER:
Hi everyone, I’m [name], a regional representative of the pancreas committee. As [name of kidney representative, or ‘my colleague’] emphasized, this is a joint concept paper put forward by both the kidney and pancreas committees. We would like to gather feedback from the community that can inform a policy solution to modify kidney and pancreas distribution. This policy solution will be presented in fall 2019 public comment.
Specifically, we would like feedback on the frameworks and themes presented, and whether there should be different distribution systems for kidney and pancreas. We want to hear both pancreas and kidney perspectives on the relative merits of the potential solutions discussed in the concept paper and in this presentation.
Feedback should be grounded in evidence tied to the Final Rule, such as the impact on efficiency in organ placement or on achieving the best use of donated organs.
After the presentation, you’ll be asked to key in your answers concerning specific preferences.

19. Specific Feedback Requested
How do you think replacing DSA and Region with either fixed concentric circles or the hybrid framework outlined in this presentation would affect organ acceptance behavior
Example: recovery practices of an OPO; evaluation of offers for a transplant program
Modeling is limited in predicting changes in transplant center and OPO behavior as a result of allocation policy changes
Do you have suggestions of what to include in a subsequent SRTR modeling request by the KP Work Group?
We also would like feedback on how your organization’s acceptance practices may change if DSA and region were removed from policy and replaced with one of the two frameworks outlined in this presentation – fixed distance circles, and the hybrid approach.
Because modeling is limited in predicting changes in transplant center and OPO behavior, feedback would be helpful in indicating how your organization’s behavior would change if the policies were modified.
Finally, please give us feedback any suggestions about what to include in a subsequent SRTR modeling request by the KP work group.
For these questions, we encourage feedback and discussion, and are not polling the room, because the feedback we’re looking for is more qualitative in nature.

20. Questions?

21. Extra Slides

22. Main metrics - Kidney
Modeling conducted by the SRTR based on a data request submitted by the Workgroup produced the results in this figure. This figure shows no change in waitlist mortality across the modeling options. As previously discussed, the projected decline in transplant rate/count is due to limitations of SRTR modeling, so caution should be taken in drawing policy conclusions from those particular metrics.
From top to bottom, the distribution scenario labels represent the baseline policy (BL), the 500NM no points model, 500 NM shallow points model, 500 NM steep points model, the 150 NM/300 NM circles model, and 250 NM /500 NM circles model.
Each model was run with 10 iterations to provide a measure of variability. The average results, along with the minimum and maximum, are provided.
Although the models show a reduction in transplants, this is likely because modeling is based on current acceptance behaviors that reflect allocation based on DSAs and OPTN Regions.

23. Main metrics – Kidney-pancreas
Modeling conducted by the SRTR based on a data request submitted by the Workgroup produced the results in this figure. For kidney-pancreas, an increase in transplant rates and counts was projected, but similar to kidney, waitlist mortality rate held steady across modeling options.
Although the models show a reduction in transplants, this is likely because modeling is based on current acceptance behaviors that reflect allocation based on DSAs and OPTN Regions.

24. Collaboration between kidney and pancreas committees
Committees worked together to develop SRTR modeling for both allocation systems
Why for both?
Most pancreata transplanted as kidney-pancreas (KPs)
Could be logistically challenging to have different systems
Not clear why the distribution for pancreas and kidney would need to be different
However…
Through Committee and Workgroup deliberation, members recognized the distribution systems may need to have separate solutions

25. Collaboration between kidney and pancreas committees
Concerns raised about same distribution distances for kidney and pancreas:
Pancreata have different acceptable ischemic time from kidneys
Distribution of pancreas programs fewer in number and more spread out
Procurement methods may be different
Issues of organ scarcity and discard rate may apply differently for kidney and pancreas
Different challenges to applying hybrid solution to pancreas allocation than KAS, that has points
Conclusion- gather more feedback from community and gather more evidence to make informed decision based on Final Rule:
Demonstrate sound medical judgment with data driven evidence
Achieve best use of donated organs
Promote the efficient management of organ placement measured in travel time and costs

26. Organ Center Travel Data (Kidney)
The figure represents transportation type and distance for deceased donor kidney transplant related organ placements facilitated by the UNOS Organ Center. The KP Workgroup determined that this data illustrates how the 150NM and 300NM circles could represent points of interest in changes in transportation, from driving to flying.
The median driving distance for deceased donor kidney related transplants facilitated by the UNOS Organ Center is close to 150 NM. Some outliers lie beyond the typical range for driving these deceased donor kidneys; however, a 300 NM circle would encompass a large majority of the kidneys driven but does not limit the majority of kidneys that were flown (those kidneys saw a median travel distance of ~1100 NM). Thus the smaller 150 NM circle would reflect a distance that would promote efficiency of organ placement.

27. Kidney CIT vs. Travel Distance
This figure illustrates how deceased donor kidneys were transplanted between January 2017 and June 2018 were allocated by distance, with local, Regional, and national distribution being depicted in blue, green and orange, respectively.
A large majority of kidneys are distributed locally and regionally, with 50 percent of kidneys being distributed within 75 NM, 75 percent of kidneys being distributed within 222 NM, and approximately 80 percent of kidneys being distributed within 500 NM. The majority of pancreata are also distributed locally and regionally, with 50 percent of pancreata distributed within 54 NM and 75 percent of pancreata being distributed within 164 NM.
The Workgroup determined that a 500 NM circle would be a rational boundary to achieve the best use of organs and promote an efficient organ placement system by limiting travel distance for kidney transplantation based on this data collected under current allocation polices, because these data may support the notion that organs are more likely to be transplanted when allocated closer to the donor hospital, thus reducing travel costs unless based on other principles of the Final Rule, such as sound medical judgement or best use of organs. The Workgroup acknowledges that the data is limited by reflecting current allocation practices that are shaped by distribution using DSA and region. Therefore, the Workgroup supplemented its review of current data with SRTR modeling to demonstrate the potential impact of a distribution system, understanding that these data possess their own limitations, as previously stated.