Increased logistical burden in circle-based kidney allocation

A U.S. Organ Allocation Revolution

In the United States, organ allocation policy is undergoing major revision to reduce geographic disparity that previously disadvantaged candidates living in some parts of the country¹. Fixed geographic donor service areas (DSAs) have been replaced with circles that offer organs first to candidates within a certain distance of the donor hospital. Further changes are mandated to eliminate all boundaries, even circles, in favor of continuously varying geographic allocation priority (continuous allocation). The change to circles has been disruptive, increasing the number of transplant centers and candidates required to place a kidney and increasing the logistical burden on transplant centers and local organ procurement organizations (OPOs) that handle deceased donors. Further, circles increased cold ischemia time (CIT) and distribution time. We believe these burdens should be accounted for as the U.S. moves toward continuous allocation.

On March 15^th 2021, deceased donor kidney allocation transitioned to an allocation system which uses a 250 NM circle around the donor hospital². Previously, each OPO worked primarily with the small set of transplant centers in one DSA. Replacing DSAs with circles was predicted to increase the complexity of the allocation system by increasing the number of necessary relationships between transplant centers and OPOs³.

Circles were implemented to align kidney allocation with the Final Rule, by reducing geographic variation in time on dialysis for transplanted patients². The Final Rule requires that allocation “shall not be based on the candidate’s place of residence or listing”, except to the extent required by other competing interests, among which is to “promote the efficient management of organ placement”⁴. Our findings suggest circles have decreased the efficiency of organ placement.

Increased Burden Analysis Methods

Using Scientific Registry of Transplant Recipient (SRTR) match run and transplant data for deceased donor kidneys from 10/05/2020 to 10/05/2021, we quantified the change in logistical burden due to circle-based kidney allocation. We compared the median number of transplant centers and candidates required to place a kidney pre- and post-circles. We further stratified this analysis by donor Organ Procurement and Transplantation Network (OPTN) region to determine how the change in logistical burden varied geographically. Finally, we compared mean CIT and mean distribution time pre- and post-circles both nationally and by donor OPTN region.

We defined center number at offer $50$ as the number of unique transplant centers that had at least one candidate on the match run with offer number $\le 50$. We defined distribution time as the duration from the match run submission to organ reperfusion, where organ reperfusion was estimated as the time at cross clamp plus ischemia time. Ischemia time was defined as cold ischemia time, plus warm ischemia time when reported.

Results: increased logistical burden of circles

Compared to the pre-circles era, kidneys were offered to more candidates (5 vs 10, p<0.0001) and centers (3 vs 5, p<0.0001) before being accepted; more centers were involved in the match run by offer number 50 (5 vs 11, p<0.0001); CIT increased by 1.7 hours (18.0 hours vs 19.7 hours, p<0.0001); and distribution time increased by 2.2 hours (42.0 hours vs 44.2 hours, p<0.0001) (Table 1).

Table 1.

Increased burden associated with circle-based kidney allocation.

	Pre-Circles	Post-Circles	P
Median Offer Number at Acceptance*	5	10	<0.0001
Median Center Number at Acceptance*	3	5	<0.0001
Median Center Number at Offer 50*	5	11	<0.0001
Mean Cold Ischemia Time (Hrs)	18.0	19.7	<0.0001
Mean Distribution Time (Hrs)**	42.0	44.2	<0.0001

Increased logistical burden is geographically heterogeneous, with a higher burden on OPOs and transplant centers in the Northeast and lesser burden for kidneys recovered in the less densely populated West (Figures 1 and 2). Kidneys recovered in densely populated Region 9, which includes New York State and New York City, were offered to many more candidates (10 vs 25, p=0.0006) and centers (5 vs 12, p<0.0001) before being accepted. Median center number at offer number 50 on the match run increased in every region, with the greatest increases being in Region 2 containing Maryland, New Jersey, Delaware, Pennsylvania, and West Virginia (6 vs 20, p < 0.0001), and Region 10 containing Michigan, Indiana, and Ohio (5 vs 17, p < 0.0001).

Figure 1. Geographic Variation of Changes in Logistical Complexity by Donor Region After Circle-Based Kidney Allocation*.

(a) Median offer number at acceptance; (b) Median center number at acceptance; (c) Median center number at offer number 50 on the match run.

* For each match run with at least one bypassed offer, we removed all bypassed offers and renumbered the offer and center number throughout the match run, unless all offers were bypassed until an accepted offer, in which case that match run was removed. When considering offer and center number at acceptance, we only considered accepted offers that ultimately resulted in transplant

Figure 2. Geographic Variation of Changes in Allocation Efficiency by Donor Region After Circle-Based Kidney Allocation.

(a) Mean cold ischemia time; (b) Mean distribution time**.

** Distribution time is defined as the duration from the match run submit date to reperfusion, where reperfusion was estimated as the time at cross clamp plus ischemia time. Ischemia time was defined as cold ischemia time, plus warm ischemia time when reported.

Continuous allocation

Circle-based allocation, like DSA-based allocation, uses hard geographic boundaries that determine which transplant centers will initially receive offers. Continuous allocation, which assigns a numerical score to each candidate according to their medical priority and proximity to the donor hospital, uses no hard geographic boundaries⁵. Under continuous allocation, a candidate at any transplant center could, in principle, be ranked highly on a match run regardless of location of listing or donor recovery. Consequently, continuous allocation might further disrupt the relationships between OPOs and transplant centers.

Circle sizes, like previous OPTN policies, were chosen ad hoc from a set of plausibly reasonable policies. In contrast, an optimization approach^6–8 would apply computational tools to design circles or continuous allocation scores that maximize transplant benefits while enforcing constraints, say, that limit logistical complexity while distributing organs equitably. Since increased burden varies greatly by geography, we could design a geographically heterogeneous continuous allocation score^9,10. Before implementation of any continuous allocation system, policymakers could examine the median center number by offer number 50, as we have done, to estimate the potential increase in logistical burden. We recommend approaches like these to carefully design continuous allocation scores to avoid further increased logistical burden in kidney allocation.

Conclusion

Circle-based kidney allocation increased logistical burden by increasing the number of transplant centers and candidates responding to offers before acceptance. Increased burden is likely greater than we estimated, because some transplant centers regularly evaluate offers before becoming primary, and would have evaluated offers beyond those accounted for here. Circle-based kidney allocation is additionally associated with an increase of 1.7 hours in mean CIT and an increase of 2.2 hours in mean distribution time. These delays are plausibly caused by the increased number of centers required to place an organ. We encourage the OPTN to attend to the logistical complexity of match runs in moving toward continuous allocation systems, perhaps by using optimization and design approaches to limit complexity while simultaneously reducing geographic disparity.

Abbreviations:

CIT

Cold Ischemia Time

DSA

Donor Service Area

HRSA

Health Resources and Services Administration

OPO

Organ Procurement Organization

OPTN

Organ Procurement and Transplantation Network

SRTR

Scientific Registry of Transplant Recipients