As we await implementation of the UNOS allocation policy entitled “The elimination of donor service area (DSA) from kidney allocation,” it’s important to reflect on the rationale, the process of policy development, potential consequences, and the path forward. The scope of the article includes the most common elements of the proposed policy that were discussed and debated during public comment. Cost, OPO performance, logistics, individual center impact, and different solutions for different geographic regions are among other topics that were raised during public comment and warrant further discussion in a different forum. In 2014, the new kidney allocation system (KAS) was a much-needed change that took over 10 y to develop and implement.1 Since implementation, many of the modeled projections have been realized; however, it was expected during development that postimplementation adjustments would be needed.2 While refining KAS, the Organ Procurement and Transplantation (OPTN) Kidney Committee received a directive in November 2017 from Health Resources and Services Administration (HRSA) that the use of DSAs and Regions in organ allocation policies could no longer be justified under the OPTN Final Rule. The Final Rule not only addresses that allocation policies “shall not be based on the candidate’s place of residence or place of listing, except to the extent required” but also includes the use of sound medical judgment to achieve the best use of donated organs, preserve the ability for transplant programs to decline organ offers, avoid wasting organs, avoid futile transplants, promote patient access to transplantation and efficient management of organ placement.3 It is well known that DSA’s have significant variations in size and shape, resulting in disparities in access to transplant.4 Many in the transplant community have expressed concern that the geography principle is being overemphasized in comparison to other principles, specifically that the presence of DSAs and regions is not the only determining factor in a candidate’s access to transplant. Rather, the performance and the practices of the Organ Procurement Organization (OPO) facilitating their allocation may contribute to a greater extent. OPO performance metrics were not included within the scope of policy development. OPTN data demonstrated that the adjusted estimated median waiting time to deceased donor kidney transplant varies greatly across the nation with median waiting times as high as 10.52 y and as low as 1.28 y.5 Further, Stewart et al developed the access to transplant score using pre-KAS and post-KAS data to measure relative differences in candidates’ access to transplant associated with patient characteristics such as blood type, calculated panel reactive antibody (cPRA), DSA of listing, age, and ethnicity. The score measures how each factor affects variability in transplant access. After KA, the DSA where a candidate is listed was found to have the strongest association with disparities (or highest variability) in access to transplantation, with blood type and cPRA close behind.6 Before the directive from HRSA, the OPTN Kidney Committee had been working on improvements to the KAS and recognized the disparity in waiting times and the need for broader sharing. The OPTN Kidney Committee embraced the opportunity to develop distribution policies allowing compliance with the Final Rule. More importantly, the goal is to ultimately transition allocation to a system of continuous distribution to address waiting time disparities and other ongoing challenges in kidney allocation. Continuous distribution will not only eliminate hard boundaries, but it will prioritize waiting list candidates based on a combination of attributes, including, but not limited to time on the waitlist, the efficient management of organ placement, expected posttransplant outcome and geography.7,8 The OPTN Kidney Committee engaged in robust discussions on equity, utility, efficiency, minimizing organ discard, and maximizing lives saved as they developed a framework for the community to consider in a concept paper. A hybrid framework was selected for further modeling, which utilized a single fixed-distance circle based on the distance from the donor hospital to the candidate’s place of listing, thus removing regional distribution. In this way, any organ that moves beyond the single fixed-distance circle would be considered a “national” organ offer. Transplant rates, the number of transplants, waitlist mortality rates, early posttransplant graft, patient survival rates, and many other metrics were used to assess the impact of a proposed change. After the Kidney-Pancreas Simulated Allocation Model (KPSAM)9 was obtained, the committee initially selected a 500 NM circle to achieve the goals of broader sharing. Potential consequences and criticisms cited during public comment included a projected decline in the number of transplants and transplant rates as well as exacerbation of racial disparities and increasing loss of life.10,11 As with any modeling, KPSAM is a tool for use in combination with clinical expertise and cannot predict changes in program behaviors. In addition, challenges exist with our current metrics due to wide variations in listing practices, waitlist management, center-level offer acceptance rates, geographic differences in end-stage renal disease patient referral behavior, as well as variation in OPO performance. Specifically, transplant rates have been shown to vary 10- to 20-fold in different areas of the country, corresponding with different average waiting times for patients after listing and cannot be explained by differences in candidate characteristics.6 The OPTN Kidney Committee has been charged with adjusting the model to fit the desired outcome of not decreasing the number of transplants. In KPSAM, a local indicator is used to represent local allocation within the DSA. After local allocation, organs are distributed with the region, of which there are 11 regions in the United States. National allocation occurs after regional allocation is exhausted or in certain categories where priority is given (cPRA > 100%). Since the proposed new unit of distribution is a 250 NM circle, local and regional allocation are no longer units of allocation. As a result, the use of the local indicator did not reflect the proposed changes and resulted in decreased number of transplants since kidneys were no longer being allocated locally. When the Scientific Registry of Transplant Recipients realized that a local indicator was used in KPSAM and was the cause of a projected, significant reduction in the expected number of transplants, they notified the kidney committee. With sound reasoning, the committee unanimously determined that the local indicator was no longer relevant and therefore should not be used in the proposed hybrid framework modeling (Meeting Summaries, OPTN Kidney Transplantation Committee 2018, 2019). Additional KPSAM limitations include offer acceptance and waitlist mortality measures. Offer acceptance uses a simple model of organ discard. For example, if an organ is offered to 200 candidates without an acceptance, it is marked as discarded, thus affecting the total number of transplants since offers accepted beyond 200 candidates are not included in the model.12 Waitlist mortality rates are censored at removal from the waitlist, so they only reflect the risk of death while waiting. They are not a measure of pretransplant mortality or postlisting survival since they do not include deaths that may occur after removal from the waitlist. The community also expressed concern about the projected increase in the percentage of kidneys flying, logistical complications secondary to broader distribution, and the effect on CITs, DGF, and organ discards. While the OPTN Kidney Committee initially believed that significant gains in equity would outweigh longer travel distances, it was acknowledged that from a community perspective, a compromise to a 250 NM circle was a more acceptable first step toward continuous distribution. Pediatric and prior living donor candidates were further prioritized by moving them up the KAS Kidney Donor Profile Index sequences in which they currently receive priority. All limitations were transparently discussed among the OPTN Kidney Committee and the transplant community. Based on the limitations of KPSAM modeling and sound judgment, it is unlikely that the number of transplants will decline as a result of broader sharing. Transplant rates may change in both positive and negative directions; nonetheless, it will be difficult to discern an exact cause given the heterogeneity in practice influencing this metric and the inability of modeling to predict behavior. Discard rates, offer rates, and mortality rates that also depend on center-level and OPO-level practices are far more complicated and will be monitored closely as a result of increased travel. With respect to travel and logistics, the debate has overstated the negative impact of increasing cold ischemia time (CIT) and delayed graft function (DGF), resulting in poorer outcomes. CIT and DGF are multifactorial and while both may result from broader sharing, it has been shown that prolonged CIT accumulates more frequently for reasons other than travel.13 Although not the norm or expected, there are multiple examples of transplant centers exceeding 24–36 h of CIT to utilize kidneys. In KAS, there are several examples of broader sharing including high cPRA candidates, who receive national priority. In paired kidney exchange where the stakes are high, living donor kidneys are frequently shipped from 1 coast to the other with minimal to no DGF and no difference in long-term outcomes.14-16 The kidney committee had extensive discussion on the impact of logistics of organ offers, specimen sharing, and allocation for effective crossmatching. There is no doubt there are opportunities for improvement in efficiencies throughout organ allocation to effectively achieve broader sharing. That said, we are already sharing many kidneys broadly; why are we so focused on limiting broader sharing on a larger scale to achieve equity? Solutions exist for inefficiencies. Our community can work on better alignment in its goals to serve our patients ethically. Days before implementation, a critical comment was filed on February 12, 2020, followed by a lawsuit filed on September 12, 2020. As a result, HRSA directed the OPTN to place implementation on hold until February 13, 2021. The policy was first presented to the community as a concept paper and then presented as a formal policy proposal. The community was given ample opportunity through Public Comment, regional meetings, and webinars, to express alternative opinions, which were discussed comprehensively. It is unfortunate that legal challenges have become a solution to solve policy differences. COVID-19 does complicate analysis of implementation outcomes since travel is affected and there are pandemic-related variations in transplants performed. However, claims regarding process, modeling, cold ischemia, discard, waitlist mortality were all extensively discussed and addressed. Creative energy is needed to support meaningful system enhancements, including universal virtual CXM, GPS units, financial models, cost assessments, uniform organ acquisition fees, etc.17 There were some important lessons learned from KAS not only with the policy itself but how we came together as a community to develop a solution. With its many imperfections, KAS taught us the value of debate and compromise. The elimination of DSA from kidney allocation policy was never intended to be a permanent solution, rather a step in the direction of continuous distribution. All of the groundwork laid by KAS places us in a favorable position to dynamically impact kidney allocation and address many of the concerns raised by the community. However, policy development over 10 y is far too long, and while the iterative work resulted in consensus and much-needed change, advances in technology, HLA, health services research, to name a few, are outpacing policy development needs. A more nimble approach to policy development is needed to meet the needs and expectations of our patients.
Search from 250 M+ research papersSearch from 250 M+ research papers
Feb 1, 2024
Jonathan M. Miller + 6
Open Access