Heavy LYFTing: KASting Pearls Before Swine
Article first published online: 10 JUN 2009
© 2009 The Authors Journal compilation © 2009 The American Society of Transplantation and the American Society of Transplant Surgeons
American Journal of Transplantation
Volume 9, Issue 7, pages 1489–1490, July 2009
How to Cite
Bromberg, J. and Gill, J. (2009), Heavy LYFTing: KASting Pearls Before Swine. American Journal of Transplantation, 9: 1489–1490. doi: 10.1111/j.1600-6143.2009.02688.x
- Issue published online: 29 JUN 2009
- Article first published online: 10 JUN 2009
- Received 31 March 2009, revised 31 March 2009 and accepted for publication 02 April 2009
In 2004, UNOS/OPTN instructed its Kidney Transplant Committee to review allocation policies with the aims to improve allocation and outcomes. The Committee identified a number of areas ripe for consideration, including trying to avoid extremes of donor/recipient matching, or placing young donor kidneys with presumed long functioning times, into very old recipients with limited life spans in the best of circumstances. Over the next 4 to 5 years, the Kidney Allocation Review Subcommittee (KARS) examined proposals, considered data, drew up plans for a new allocation system and solicited feedback from the transplant community and other stakeholders. This process resulted in the development of a new algorithm, the kidney allocation score (KAS) based on components of recipient dialysis time (DT), the donor profile index (DPI, a continuous variable rather than the current dichotomous standard criteria donor/expanded criteria donor [SCD/ECD] designations), and life years from transplant (LYFT, a calculation of the number of years of added life of a particular kidney transplanted to a particular recipient). Prior to adopting the new KAS, a public forum in St. Louis, MO, on January 26, 2009 revealed significant opposition to the plan, where many concerns were expressed by large segments of the transplant community as well as patient groups. In response to this, the adoption of KAS in its current form has been shelved, and a new approach will now have to be undertaken. Despite the disappointment that this has engendered in some, a great deal was learned by the analyses of the allocation schemes, and from the critiques of those schemes. In this issue of the American Journal of Transplantation are a series of seven viewpoints and perspectives, solicited by the editors, concerning these analyses and critiques. We feel there is a great deal that is very important and enlightening about these opinions, and far from being an outlet for unrestrained criticism, point to a number of areas of commonality of purpose on which most agree. Further, there is much to be learned by studying the history of the process that has allowed us to arrive at our current place.
Stock, who chaired the KARS, presents the historical context of the subcommittee and the process of its workings. He notes the conflicting goals of HHS, NOTA and the UNOS final rule; and how those goals influenced the specific plans that the subcommittee considered. He also notes that the overwhelming majority of controversy surrounded LYFT, while there was agreement that both DT and DPI are broadly supported by the transplant community and should likely remain, perhaps in altered forms, in subsequent allocation schemes. He provides a roadmap for future deliberations that includes DT, objective outcomes measures and amelioration of geographic disparities.
Murphy et al. provide the American Society of Transplantation (AST) perspective. There is strong support for DT. They note the requirements that any system be rational, transparent and easy to explain to patients, an issue noted by many others. They also suggest that a change to the organ allocation scheme that does not alter who gets transplanted but rather only alters the type of organ patients receive may be most acceptable.
Freeman et al. provide the American Society of Transplant Surgeons (ASTS) perspective. They note that LYFT is extremely complicated and when applied may compromise individual justice for patients. They further question the accuracy of the models and the data for both LYFT and DPI. They propose several alternative approaches and schemes, and note there are several areas not yet addressed by KARS, which may be more easily addressed than LYFT (their Table 1). Finally, they introduced the concept of incorporating patient choice into future organ allocation schemes and put forward very practical considerations such as the importance of a predictable allocation scheme that would enable maintenance of health status in waitlisted patients.
Hippen provides an eloquent personal viewpoint on LYFT, categorizing methodological concerns, moral concerns and unintended consequences including the erosion of public trust in the system. He notes, as did some of the other authors, that the current pediatric allocation scheme has decreased living donation and thus done harm to some, while benefiting a small subset. He has done an admirable job of encapsulating many of the concerns that others expressed at the meeting or in written comments to the subcommittee.
Leichtman and Wolfe provide a counterpoint to the many criticisms of LYFT, explaining its statistical underpinnings, particularly the C-statistic, and in terms that all of us nonmathematicians can readily appreciate. In particular, it should be noted that LYFT in fact performs slightly better than other commonly accepted statistical models including model for end-stage liver disease (MELD), lung allocation score and HLA weighting for kidney transplantation.
Segev provides another personal perspective on the mathematical basis of LYFT and other utility-based allocation systems such as the Eurotransplant Seniors Program. He considers the intended and unintended consequences of these systems and incorporates them into a framework for measurable risks and benefits (his Figure 2). This framework has explicit values assigned to utility, efficiency, equity, predictability, autonomy, controversy, trust and live donation.
Stegall provides a final personal perspective, tabulating areas of agreement, discussing alternatives for the relative importance assigned to LYFT and waiting time, and suggesting a compromise proposal for allocation. His suggestions provide a practical roadmap for getting the discussion back on track by focusing on principals and areas of commonality.
Far from being a dead end, the debate within the transplant community has revealed several important principles and findings. First, the community can work together, even on controversial topics, to seek goals that put patients first. Second, while some may see the process of the KARS as flawed because a new allocation scheme is not yet in place, the organizations were able to resist bureaucratic inertia and adopt something no matter how flawed. Instead the process will begin afresh with new ideas. Third, a consistent notion that came from virtually all the authors is that a system that is so complicated that it cannot be explained easily to patients, their families or the medical community at large is a disservice to patients and may erode trust in the system. This is a refreshing viewpoint that may show us the way out of so many other regulatory conundrums that utilize resources without helping patients. Lastly, there also seems to be agreement on the use of DT and DPI, and explicit examination of the consequences of allocation scheme. The challenge now is to discern what are the pearls of wisdom we have gained, and how best to communicate and promulgate those to the transplant community and our patients.