Answers are grounded in this paper and cite their sources.
As we approach the end of this century's first decade, it is appropriate to analyze the current state of the US donor pool and postulate mechanisms to stimulate its expansion. The Organ Donation Breakthrough Collaborative (ODBC) is a high-profile initiative sponsored by the US Department of Health and Human Services to increase organ donation.1 Since the program's launch in April 2003, the ODBC has identified and disseminated best practices throughout the donation, transplantation, and hospital communities via a series of local, regional, and national meetings. The Joint Commission on Hospital Accreditation and the Institute of Medicine have also prioritized organ donation within this decade.2, 3 These efforts to enhance donor identification, recruitment, and management have undoubtedly contributed to the 33% increase in the annual number of deceased donations that have yielded at least 1 recovered organ since 2000.4 Unfortunately, data on the deceased donation of liver allografts during this decade are not as robust and actually indicate a contracting donor pool.5 Concerning data vis-à-vis the US deceased-donation liver allograft pool include: DBD, donation after brain death; DCD, donation after cardiac death; DRI, donor risk index; ECD, extended criteria donor; ODBC, Organ Donation Breakthrough Collaborative; PSR, program-specific report; SRTR, Scientific Registry of Transplant Recipients; UNOS, United Network for Organ Sharing. The total number of deceased donation liver allografts transplanted has decreased for the past 2 years. From a peak of 6309 deceased donor liver allografts transplanted in 2006, the total number of deceased donor liver allografts transplanted decreased 4.9% through 2008. Reductions in the annual number of deceased donation liver allografts transplanted have occurred in donation after brain death (DBD) allografts, which have decreased 5% from a peak in 2006, and in donation after cardiac death (DCD) allografts, which have decreased 9.8% from a peak in 2007. Since a record-high number of consented DBD liver allograft deceased donors in 2006, the number of consented DBD donors declined 1% in 2007 and more than 3% in 2008. Although these decreases may not appear large, they are in stark contrast to the growth observed in the donor pool prior to 2007. The number of consented DBD liver allograft deceased donors increased 26% between 2000 and 2006 with an average annual increase of 4.1%. Since the inception of the ODBC, the average annual increase was 6.2% through 2006. The marked reductions in the consented DBD component of the liver allograft donor pool have been buffered by increases in the number of consented DCD donors; however, increases in the number of consented DCD donors have not yielded equivalent increases in the number of transplanted liver allografts. Throughout this decade, annual increases in the number of transplanted deceased donor liver allografts have not matched increases in the number of consented liver allograft deceased donors. In 2008, the percentage of liver allografts transplanted from the consented deceased donor pool was a decade low: 76.2%. The percentage of DBD allografts transplanted was 80.9%, which was the lowest since 2001, whereas the percentage of DCD liver allografts transplanted from consented DCD donors was a decade low: 34.5%. The number of consented donor liver allografts that are not recovered or are recovered and discarded is at decade highs for DCD and DBD donors. Unfortunately, 21.6% of the total number of consented donor liver allografts in 2008 were not recovered or were discarded. For consented DBD liver allograft donors, 17.9% of liver allografts were not recovered or were discarded in 2008; this is notably greater than the decade low of 14.4% observed in 2003 and equal to the not recovered/discard rate observed in 2001. For consented DCD liver allograft donors, the majority of liver allografts, a decade high of 54.6%, were not recovered or were discarded in 2008. The data are equally grim if one segregates the data sets into not recovered and recovered and discarded categories. Areas of sustained improvement in the US donor pool this decade include a marked decline in the number of DCD and DBD deceased donor liver allografts for which consent is not requested or obtained and increased allocation of liver allografts to research. During 2008, more than 10% of recovered DCD liver allografts were allocated to research. Although the underlying causes of these observations are multifactorial and require detailed statistical analysis, I believe that several factors are contributing to the contraction within the US donor pool. These include the following. The US donor pool is not a slice of Americana. Misinterpretation of the US donor pool as a random sampling of mainstream America impairs our recognition of the donor pool's natural limits. The traditional or reference population, which is assumed to consist of Caucasian males, 18 to 49 years of age, who have died as a result of trauma, actually represents a consistently decreasing minority of the donor population. In 2008, this group represented 20% of the total donor pool4; unfortunately, this small minority remains the allocation standard in public perception and statistical analysis. An allocation scheme that defines a standard or traditional allocation unit should incorporate that unit as the likely result of a wait-list priority allocation. This is a false expectation from our current donor pool. The robust increases in consented deceased donations in this decade have originated from increasing recruitment of elderly donors, DCD donors, donors with positive viral serologies, and behaviorally high-risk individuals.6, 7 This means a donor pool in which the incidence of hepatitis C, hepatitis B, diabetes, hypertension, steatosis, donor age >60 years, and high-risk behavior such as incarceration, prostitution, and illicit intravenous substance injection is more prevalent in comparison with the US population.8, 9 The scientific limits of our donor identification process create this population. In this environment, disease transmission following solid organ transplantation remains a risk.10, 11 Unfortunately, widespread publicity of these uncommon events supports the misunderstanding of our donor population and results in a tiered nomination of allografts as high-risk or extended criteria donor (ECD) allografts, despite clear evidence that the risks imposed by these allografts is minute versus the morbidity of continued waiting.12 Whether indicated by a donor's behavior or a physical property of the allograft, the ECD classification in liver transplantation is artificial. Unlike renal transplantation, in which the ECD classification is based on outcome differences, widespread acceptance of an ECD classification scheme for liver transplantation has not been possible despite multiple consensus conferences because of a lack of identifiable predictors of posttransplant hepatic function. Instead, the genesis of the ECD classification within liver transplantation results from an allocation system that creates a wide disparity of organ availability. A liver allograft with a history of donor virus exposure or advanced age may be allocated according to match-run priority in certain regions, whereas in other regions, these organs are available for export and allocated at the discretion of the transplant program. Although the numerous publications documenting decreased wait times, increased access to liver transplantation, and transplant center growth with satisfactory results with the use of ECD liver allografts has enhanced the recognition of their utility, the variable ECD definitions developed by authors demonstrate the artificiality of the allograft predictors.13, 14 The increased risk imparted by these allografts, whether it is due to behavioral, serological, or functional ECD indicators, is frequently trumped by the morbidity of continued waiting. The recognition of our donor pool as an environment in which risk is distributed as a continuum requires an accurate predictor of risk. The concept of a DRI, as proposed by Feng et al.,15 is the most comprehensive attempt to establish such a measure. DRI components include age, cause of death, race, DCD, split liver, height allocation, and cold time. Elegant in its conception but impractical in execution, the DRI is inherently limited by its derivation from the Scientific Registry of Transplant Recipients (SRTR). As its name implies, the SRTR is principally a database designed for recipient outcomes with a limited capacity for identifying predictors of donor allograft quality. Donor data fields within the SRTR are crude and, until recently, omitted elements such as biopsy data, estimation of steatosis, and length of intensive care stay, which have been repeatedly implicated by multiple single-center reports to affect donor organ quality. Definitive follow-up regarding donor serologies and disease transmission is also at the discretion of the transplant center, and this has resulted in many missed data collection opportunities. The application of the DRI to the current donor pool demonstrates widespread regional variation in allograft utilization that does not correlate with allograft failure or listing for retransplantation.4 Single centers with a particular interest in increasing patient access to transplantation have reported impressive data that are substantially better than those predicted by the DRI; indeed, there is likely a windfall for centers that import a large percentage of allografts.16 Center-specific reports testify to the success of using these allografts, but the data cannot be extrapolated to larger multicenter inferences on donor and recipient selection. The development of a playbook that we all can follow will require prospective data collection from multiple transplant centers and organ procurement agencies to create a donor information database that would form the basis of a powerful DRI. The design could be similar to the National Institutes of Health-sponsored Adult to Adult Living Donor Liver Transplantation Cohort Study17 with the intent of increasing utilization of expanding categories of the donor pool, such as elderly donors, partial allografts, DCD donors, and steatotic allografts. The systematic analysis of DCD recovery practices with respect to clinical markers, prognostic indicators, and outcomes would be a powerful tool for facilitating allograft utilization. A prospectively derived and validated DRI would reflect the continuum of risk and provide a foundation for informed donor-recipient matching and consent. Widespread dissemination of best practices for using an expanded array of allografts would improve utility and stimulate donor pool expansion. The observation that 54% of consented DCD liver allografts in 2008 either were not recovered or were discarded markedly contrasts with the 684% increase in DCD renal allograft utilization between 1997 and 2005,18, 19 and this highlights the urgent need to increase our understanding of how to best use this important component of the donor pool. The development of a powerful DRI, as described previously, would provide a foundation, but I believe that a review of the current United Network for Organ Sharing (UNOS) allocation policy with respect to DCD allografts could immediately stimulate utilization. A recent retrospective analysis of SRTR data from January 2002 through April 2007 involving 855 DCD donors and 21,089 DBD donors was the first to demonstrate declining utilization of DCD liver allografts.20 Although confirming previous studies demonstrating inferior outcomes from DCD allografts, the study identified an additional impediment to increased DCD liver allograft utilization through an investigation of DCD failure patterns, listing for retransplantation, and outcomes of DCD allograft recipients listed for retransplantation. The temporally different failure pattern between DCD and DBD allografts, identified between approximately 20 and 180 days post-transplantation, generated a unique population of DCD recipients, principally with the diagnosis of ischemic cholangiopathy, who are disadvantaged under current allocation policy by limited access to retransplantation. This required transplant centers to assume greater risk through increased utilization of high-DRI and national-share allografts to achieve retransplantation for the initial DCD allograft recipient.20 The development of an allocation policy recognizing delayed allograft failure patterns of DCD allografts is necessary for further expansion of this important segment of the donor population. UNOS has not taken the lead on this subject in the form of definitive policy but has left resolution of this dilemma to regional review boards on a case-by-case basis. This approach has been ineffective as evidenced by the failure of exceptions to increase access to retransplantation.20 We must create a parachute for the potential DCD recipient and transplant center. The provision of a reliable mechanism for access to retransplantation of DCD liver allograft recipients would immediately reignite DCD utilization, stimulate clinical research, and provide a numeric increase in the transplantation of functional liver allografts. Data dissemination and transparency result in the replication of productive practices and the elimination of underperforming practices. A casualty of this “moving to the middle” process can be invention and the introduction of risk. To continue expansion of the donor pool, the transplantation community must maintain a mechanism to introduce risk into clinical liver transplantation. The development and dissemination of SRTR transplant center program-specific reports (PSRs) has been a notable achievement in promoting improved quality assurance; however, these data have been extrapolated beyond their natural limits by insurers and regulators to affect the dynamics of a transplant center. The failure to achieve expected outcomes has enormous implications for the transplant center and its future operations. Therefore, it is not surprising that when intelligent individuals are given access to a formula and are informed what the intended results should be, behaviors are modified to achieve the expected result. The casualty of moving to the middle is risk, whether it manifests as consideration of a high-risk recipient, utilization of a high-risk allograft, or the introduction of a potential scientific advancement.21 Furthermore, the distribution of risk is not universal but is inversely proportional to the center's annual volume. An open letter from the Executive Committee and Council of the American Society of Transplant Surgeons to the president of UNOS and the chair of the SRTR Technical Advisory Committee highlights these concerns and requests redefinition of the PSR and revision of the statistical methods used to calculate this value.22 Although these efforts are to be applauded, there are methods to maintain an element of risk within clinical liver transplantation regardless of the PSR.21 One technique could be recognition of a research exception status for liver transplant procedures performed under institutional review board-approved protocols in which allograft function is a true scientific variable. The ex vivo machine perfusion protocol currently being used and the future introduction of ischemia/reperfusion modifiers during liver transplantation serve as examples of promising technology that may be impeded by potential PSR concerns.23 Transparency should be maintained by the same data reporting practices for these procedures; however, the collected data should not be included in the calculation of the PSR but should be available as a separate category. Limits to the utilization of research exception status, perhaps manifesting as shortened medical review periods or a percentage of the transplant center's annual volume, should be explored to prevent the perpetuation of unnecessary risk. Another mechanism to maintain risk within clinical liver transplantation would be the development of a high-risk transplant designator employed by the center either before or within a certain time frame following a liver transplant procedure. This would encourage transplant physicians to extend liver transplantation to individuals in unique circumstances with decreased apprehension about affecting the transplant center's dynamics.21 Again, these procedures would be subject to the full reporting practices required by UNOS, but the data would be separate and would be excluded from the calculation of the PSR. A regionally defined percentage cap of this designator based on transplant center volume would distribute risk evenly across centers by consensus, and the transparency of this process may actually prove to be a powerful tool in quality assurance. Lastly, I believe that the current operation of DonorNet is contributing to lost opportunities for allograft utilization. This is apparent from the noticeable increase in the frequency of usable organ offers made with an existing cold ischemic time and no established transport arranged. How often has a center entered a provisional yes behind scores of provisional yes responses only to be called about a procured organ that has cold time, with the recovery surgeon gone and no transport mode established? What is the center to do? The fundamental problem originates from DonorNet being a national allocation scheme, whereas organ recovery remains a local event. The behavior of allocation patterns following organ declination by the primary center must be studied in detail to avoid a domino effect that gathers inertia as the cold time increases. In conclusion, the transplant community recognizes each solid organ as a precious gift and the imperative to use this resource to the fullest extent possible. Although this decade has enjoyed an unparalleled increase in donation awareness, this has not translated into the fullest gains possible from our donor pool. These data are disconcerting and require the immediate attention of the transplant community. Although this problem is multifactorial in origin, the transplantation community must recognize its shortcomings and initiate widespread efforts to maximize utilization of the existing donor pool. The development of a comprehensive research instrument to generate a powerful DRI would enhance rank-list allocation and quality assurance. Furthermore, structured efforts to maintain an atmosphere that supports measured risk and scientific exploration are essential to achieving maximal donor utility and benefit for our patients. This article is dedicated to the memory of John Rice, Michael Tatum, and Fred Hoishman.
No takes yet. Share an insight, caveat, or question.
John F. Renz (2010) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: