Transplantation Supply and Demand
Transplantation data from the United Network for Organ Sharing (UNOS) counts 8,940 total donors for the most recent year reported (1996), of which 5,416 were cadaver donors and 3,524 were living donors. Over the last 9 years (1988 to 1996), the total number of donors (5,910 to 8,940, up 51%), cadaver donors (4,084 to 5,416, up 33%) and living donors (1,827 to 3,524, up 93%) has been increasing slowly (UNOS-Organ Procurement and Transplantation Network (OPTN)/Scientific Registry data as of July 5, 1997). This increase is not due to NHBDs. Their numbers at present are very small and do not influence these figures (see UNOS data later in this chapter).
Increased supply has come about in three principal ways: (1) There are more living donors, and among these spouses and friends who are emotionally as opposed to genetically related donors increased from 3.9% to 12.5% of all living donors from 1988 to 1996. (2) Criteria for accepting cadaver donors have begun to be liberalized and expanded to include, importantly, increasing age (up to 80 years) in some programs, but also donors with diabetes, hypertension, some infections, high-risk social history but negative HIV test, some hemodynamic instability, some chemical imbalances, increased organ preservation time by perfusion or storage using cold solutions, or increased time after death in the body. These expanded criteria are said to have the potential to add 25 to 39% to the cadaver donor supply (Briceno et al., 1997; Jacobbi et al., 1997). In 1996, donors over age 50 comprised 26.5% of all U.S. cadaver donors, double the 1990 percentage (Braslow, 1997). (3) Lastly, there is more efficient use of each donor; that is, more solid organs are procured (and transplanted) per donor. The donor–transplant ratio for living donors, from whom only those organs that are paired or removable in part can be taken, is always about 1. The ratio for cadaver
donors, from whom all solid organs can be taken, rose, in terms of not only organs procured per donor (21%, from 2.98 to 3.60) but as organs actually transplanted per donor (18%, from 2.73 to 3.21 from 1988 to 1996)* (OPTN/Scientific Registry data as of April 15, 1997). Unfortunately, no improvement in the organ-per-donor figures has occurred since 1993, and the organ-per-donor yield for NHBDs is, and likely will remain, lower as that category is expanded (2.5 organs-per-donor, and 1.78 transplants-per-donor, aggregate from 1993 to 1996) (M. Ellison, personal communication, June 27, 1997).
These comparative numbers of transplants or recipients per donor should be part of an assessment of supply and demand. Also the success of transplantation—the survival of the transplanted organ and the recipient patient—depends heavily on the condition of the donor organ (Cosio et al., 1997; Light et al., 1996; Jacobbi et al., 1997). The expansion of criteria for acceptance of organs from donors of greater age and in less satisfactory medical conditions and the increased use of NHBDs exact a price in increased procurement costs of 23% more per donor and thus even more per organ (L. Jacobbi, personal communication, July 17, 1997); increased cost of transplantation; and lower graft (i.e., the transplanted organ) and recipient survival (Alexander et al., 1994; Briceno et al., 1997; Casavilla et al., 1996; Jacobbi et al., 1997; Light et al., 1996; Whiting et al., 1997). This must be weighed against the morbidity, death, and economic and other costs in patients on the waiting list (Hauptman and O'Connor, 1997; Perkins, 1987; Light et al., 1996).
Regardless of donor type, overall patient and graft survival rates have continued to improve over time. One-year graft survivals now range from 74% to 92% depending on the organ, although in subsequent years survivals gradually decrease particularly in expanded donor and NHBD transplants. The best results are with kidneys, which comprised 51% of all cadaver transplants and 98% of all living donor transplants in 1996 (OPTN/Scientific Registry data as of July 5, 1997). Coincident with improvements in the technology available and these better results, the pool of potential recipients has been expanding steadily (Anaise and Rapaport, 1993). These trends are likely to continue.
Changes in supply should be compared with changes in demand. The transplantation waiting list contained 50,047 people waiting on the last day of 1996, 14% more than the previous year and 212% more than at the end of 1988
(16,026). During that 9-year interval, 23,214 people died while on the waiting list—4,083 people in 1996 alone. Most median waits (the number of days elapsing until 50% of those waiting receive a transplant) have been increasing; the most recent are as follows: for kidneys (1994), 842 days; for livers (1995), 254 days; and for hearts (1995), 213 days (UNOS, 1996). These data have recently been superseded by not yet published waits stratified by recipient status: for example, median waits are 609 days, 1,046 days, and "too long to have a median at present" in the least, moderately, and most severely sensitized kidney recipients, respectively. They range from 5 days in the sickest liver recipients to 311 days for those liver candidates in the best condition and, similarly, from 40 to 390 days for the sickest to the healthiest heart recipients. These unpublished data emphasize that waiting times depend largely on candidate medical conditions. They also reflect, for heart and liver, the positive effects of preferences accorded to the most severely ill recipients (M. Ellison, personal communication, July 5, 1997).
In 1996, 33,866 patients were added to the waiting list, and 20,319 organ transplants were performed and removed that number of patients from the waiting list. Death removed 4,083 patients from the list, as noted earlier. An additional 1,019 patients were removed from the list due to advanced disease, and 1,941 patients left the list because of improved status, refusal of transplantation, and other unspecified reasons (OPTN/Scientific Registry data as of July 29, 1997). At a steady state, therefore, it would appear that conforming demand—as measured by the addition of new patients to the list, deaths, and removals from the list due to advanced disease—with supply—as measured by number of transplants performed—would require an increase in organs of around 18,000 (about 90%) and, depending on the kind of donor, because of varying numbers of organs recovered per donor, about the same percentage increase in donors. Without the need to overcome this large backlog, such an equilibrium—even if supply could in some way be increased by the requisite amount—is unlikely for several reasons.
Many observers believe that there are potential recipients who could benefit from transplantation, but because the supply situation is so daunting, through a kind of informal rationing they are not added to the waiting list and therefore are not included on the demand side (Caplan, 1986). Furthermore, 1991 demand estimates based on disease incidence suggested a greater potential need, in the neighborhood of twice the then-current new patient additions to the waiting list (Evans et al., 1992). If estimates are accurate that 30%–40% of patients on hemodialysis are suitable transplant candidates, there should be 60,048 to 80,064 of the 200,160 End Stage Renal Disease Program 1995 roster on the waiting list, but the actual number awaiting kidney transplant on December 31, 1995 was only 31,195 (Anaise and Rapaport, 1993; Health Care Financing Administration, 1997). A recent estimate of suitable heart recipients of 14,000 per year greatly exceeds the 1996 heart transplant (2,368) and waiting list
(3,648) totals (Federle, 1995; OPTN/Scientific Registry data as of July 29, 1997), and the most recent estimate by Roger Evans of unmet need in 1996 was 121,200 (personal communication, August 14, 1997). As noted elsewhere, changes in many parameters—medical technology, disease patterns, expanded indications for transplantation, transplant success rates, funding for health care, the general supply of, and access to transplant services, and the organization of the general health care system and the organ procurement system—will also influence any equilibrium. Better results make transplantation more sought after, more transplant centers make it more easily accessible, and all of these factors influence physicians' perceptions of the indications for referral for this treatment.
Because the available data document the fact that demand is increasing significantly faster than supply and has been for some time (donor increase, 1988–1996, 51%; waiting list increase, 1988–1996, 212%), an equilibrium on the waiting list achieved through increased waiting times and higher death rates, accompanied by increased morbidity and economic and other costs, seems the inevitable, ultimate result of current trends. These adverse effects will occur not only among those on the list, but are likely too among those who receive transplants because they will have waited much longer, will have advanced disability and disease, and will be less medically suitable transplant candidates. Three-year graft survivals for liver and heart patients who are not hospitalized versus those needing life support are 24% and 9% higher, respectively, and transplant procedure charges are respectively 84% and 68% higher for those needing life support (R. Evans, personal communication, August 14, 1997).
Demand could be limited by an explicit decision to ration services by narrowing the indications for transplantation, a decision that would be extremely difficult and controversial in its details. The implicit rationing, noted above, could continue in an informally and tacitly expanded manner as the supply situation becomes even more daunting and life on the waiting list becomes more uncertain and unproductive. Managed care could have a similar effect, although reliable data on effects to date are not available. Nevertheless, the success of managed care is based in part on a smaller and more efficient supply of services and stricter indications for medical interventions, which are conditions that may slow demand. Managed care is said also to be diminishing supply by less often carrying the care of seriously injured or ill patients to the point of potential donation using life support and other critical care interventions (L. Jacobbi, personal communication, July 17, 1997). It seems likely also that increased awareness of the cost of intensive care and increased familiarity with ethical issues surrounding end-of-life care may be having this effect in health care settings in general. If there are fewer declarations of brain death under managed (or conventional) care, this could tip the scales toward NHBDs and demand vigilance that these potential donations were actually obtained (Reiner, 1997). A similar incipient emphasis on NHBDs is suggested also by anecdotal reports that there
may be increasing unwillingness of families to wait for brain death before discontinuing life support (Arnold, 1997).
A more promising alternative on the demand side lies in anticipated long-term advances coming from clinical experience and transplantation biology research. If these advances generate improvements in long-term organ survival, as they have in the past, they should shorten the waiting list and waiting times for various organs by eliminating or postponing the need for retransplantation. Patients awaiting the second transplant or more presently occupy about a fifth of the slots (mostly kidney) on the waiting list (UNOS, 1996). Reductions in demand would result from significant reductions in this category of patient, but only modest reductions are expected as graft survivals slowly improve.
There is little optimism that procurement of organs from either living or brain-dead cadaver donors can be sharply accelerated, although marginal increases should be vigorously pursued. Living donor expansion will provide only one organ per donor and will not address the shortfall in organs that are not paired or transplantable in part. Also, each living donor incurs a risk of death or medical complication. Unpublicized mortality has reached "at least two dozen unpublished but certain donor deaths" (Sheil, 1995). Particularly in partial liver transplantation, the donor risk could be considered a real problem (Sterneck et al., 1995). However, advances in technology, such as laparoscopic nephrectomy, will reduce morbidity, may have a positive effect on mortality and, as reported anecdotally, may increase donation of kidneys from living donors. Increasing the pool of living donors through the use of genetically unrelated donors has raised a question for some that commercial considerations in organ procurement may be introduced if donors without spousal or strong friendship bonds are accepted (Hauptman and O'Connor, 1997). Nevertheless, an increase in living donors by primarily affecting the supply of kidneys affects the severest shortage and the largest category of transplants and helps patients with the longest waiting times.
Evans et al. (1992) have estimated the national, annual, potential number of brain-dead (heart-beating) donors at 6,900 to 10,700; the number of actual donors therefore is currently half this theoretical maximum and close to what one would expect if the upper-bound estimate of potential donors is correct, if all potential donors are approached (which is not the case), and if a traditional consent rate of 50% is achieved. Evans calculated an efficiency in U.S. organ procurement efforts in 1988–1989 from 37% to 59% (percentage of potential donors realized). A more recent study using 1990 data estimated 13,700 potential heart-beating donors, a nonidentification or failure-to-ask rate of 27%, and a consent rate of 45% of those asked (Gortmaker et al., 1996). These figures provide
a somewhat lower assessment of efficiency. In either case, the supply of heart-beating donors does not appear sufficient to satisfy demand.
Current expanded donor criteria or acceptance of higher estimates, including NHBDs, by others (discussed below) (Bart et al., 1981; Nathan et al., 1991) may allow a potential that is greater than the accepted numbers for heart-beating donors. At the same time, seat belt, helmet, drunk driving, and gun safety laws and other public health measures as well as fear of HIV are said to be limiting the growth of or reducing the potential donor pool by lessening serious head trauma and disqualifying otherwise acceptable brain-dead donors (Arnold and Youngner, 1993; Evans et al., 1992). Increasing numbers of donors from the cerebrovascular or stroke category have maintained donor rates, although unfortunately at an older average age, in the face of these changes (UNOS, 1996).
A variously reported 13% to 28% of the population has filled out a donor card (Gallup, 1993; Martin and Meslin, 1994). Despite a combination of laws in the majority of states and rulings from both the Health Care Financing Administration and the Joint Commission on Accreditation of Healthcare Organizations that require essentially all hospitals to ask for donation, a quarter to a third of families of patients with appropriate medical status are never approached about donation and many potential donors are now being missed (Nathan et al., 1991; Siminoff et al., 1995). In addition, 20% of those with the diagnosis of brain death suffer cardiopulmonary arrest and death within 6 hours, and 50% within 24 hours, of admission to the intensive care unit. Thus, a very tight time schedule often frustrates efforts to complete necessary laboratory examinations and carry out other steps required to recover organs. Patients whose relatives were not asked about donation tend to be those who died sooner after admission and who did not die in the intensive care unit (Chapman et al., 1995). Lastly, as many as 35% of patients with very poor prognoses are not started on or are prematurely withdrawn from life support (Hibbard et al., 1992), and a recent survey disclosed that physicians did not recognize or declare brain death in 13% of potential donors, which eliminates the possibility of arranging donations in both situations (Wing and Chang, 1994).
In the past, the organ procurement effort in this country has been criticized for variable and sometimes very low productivity. In the context of a 1989 national rate of about 16 to 17 donors per million population (dmp),* some procurement organizations were achieving rates of about 2 dmp and others about 30 dmp. Other countries have improved procurement through better organization and work effort. Implementation in Spain of a unified, centralized program with an emphasis on improving coordination and performance was associated with a rapid increase from 19 to 28.5 dmp and an actual decrease in the waiting list in that country (Miranda, 1997). When assessing performance, it should be kept in
mind that local conditions other than procurement organization effectiveness—for example, public interest and understanding; local accident rates and trauma patterns; demographic, social, or cultural and religious factors; and possibly, the presence of local transplant services in the community—influence donor rates. Also, on an organization-specific basis, the number of donors is small and a particular year's donor rate may represent a statistical outlier rather than the usual performance.
Nevertheless, a UNOS model of supply and demand in liver procurement and transplantation using 1992 to 1994 data when national procurement numbers had improved to 19.2 dmp (range: 5.0 to 33.3 dmp) (Gortmaker et al., 1996), suggested that if organ procurement organizations' donation rates (for livers) could be raised to the level of the higher recovery OPOs (a potentially achievable level of 30 dmp each year and 80% average liver recovery per donor), this would represent about a 55% increase from the current national average, would cause a proportional decrease in waiting list deaths, and would dramatically reduce regional variations in waiting times (M. Ellison, personal communication, June 27, 1997). A rough approximation of equilibrium appeared within reach, provided there were no other changes on the supply or demand side, if effectiveness could be doubled. Caution should be used in drawing too many conclusions from these estimates, as noted earlier. Demand is unlikely to be stable as supply increases; different organs will have different equilibria; tissue-matching problems will affect the efficiency of kidney utilization; and categories of donor vary in their effectiveness in supplying organs. Nevertheless, the UNOS model does give a rough sense of the magnitude of the current problem and suggests a target that is theoretically not out of reach. Procurement in 1996 stood at 21 dmp nationally with a range of 7.1 to 37.4 dmp.
No rigorous cost-effectiveness analysis has been done to compare the costs and benefits of improving donor procurement. Each organ would require a separate analysis. Different cadaver donors have different efficiencies, ranging on average from 1.78 to 3.2 transplants each, and each transplant if successful takes a patient off dialysis or expensive treatment for organ failure. Of course, each transplant has its cost, and aside from kidney patients who can be maintained on dialysis, each patient who fails to get a transplant ultimately dies and eliminates all cost of any kind of care. Adding donors would generate additional cost (Evans, personal communication, 1997), although others disagree (UNOS, 1997), but a case could well be made for the cost-effectiveness of investing more in a stronger effort to improve organ donation and procurement rates. This assumes a reasonable balance between expenditures made and expenditures avoided, and the assignment of priority to relieving the human suffering, morbidity, and mortality associated with the present state of affairs.
Major improvements in the supply of organs could be made if the public consented to donation at rates higher than those currently experienced. Surveys show that the public consistently gives a high approval rating (85%) to the act of
donation for organ transplantation (Gallup, 1993). Public willingness to donate personally or to consent, as a surrogate decision maker, to a family member's donation, runs from 45% to 50% and 53% to 68%, respectively (Evans, 1992). Gallup found even higher (69%) personal willingness to consider organ donation positively and 92% or 93% willingness to either honor a family member's wish to be a donor, if known, or give permission for donation as a surrogate decision maker, if unknown (Gallup, 1993).
However, the validity of these results has been questioned for a number of reasons, including the important one that a positive answer is the most socially acceptable answer. In Texas, during a time when applicants were required to make a personal decision in writing to obtain a driver's license, consent fell to 20% (Siminoff et al., 1995), and when permission to become a donor for a family member in a brain death situation is actually requested, consent varies from about 40% to 65% both in this country and abroad: 43.7% (Chapman et al., 1995), 46.5% (Siminoff et al., 1995), 50% (Hibbard et al., 1992), 57% (Morris et al., 1989), 62% (Nathan et al., 1991), and 65% (Kootstra, 1997); the U.S. average is about 50% (Braslow, 1997). Public concerns that a seriously ill or injured person known to have consented to organ donation may receive less aggressive care from physicians who are interested in organs for a potentially more promising or younger patient on the transplant waiting list are prominent among reasons for levels of consent that are so much lower than survey approval rates (Caplan, 1993). Some are also worried that organs may be removed before death has definitely occurred (Corlett, 1985). Of relevance to this concern, 21% of the public believes that a brain-dead person can recover from his or her injuries (Gallup, 1993).
At the actual time of request, families are emotionally traumatized. Decisions on transplantation may simply by more than they can handle in a crisis (Perkins, 1987), and the stress and confusion of fast-moving medical events may deny them adequate time to consider the merits of donation and resolve questions or misconceptions that are potential barriers to organ donation (Coolican, 1997; Gallup, 1993). Even after participating in the donation process, about half of donor families, particularly those who do not consent, report an incomplete understanding of brain death (Franz et al., 1997), and 10% to 15% report an inadequate general understanding and insufficient personal attention from transplant personnel (Pearson and Zurynski, 1995). Although evidence to direct and improve the request process suggests that discussions about brain death and discussions about organ donation should be decoupled, that private discussions are best, and that organ procurement organizations' (OPO) participation with hospital staff is advantageous, these methods are not uniformly understood or practiced (Beasley et al., 1997).
Understanding among the general public is undoubtedly less than that reported for donor families, which suggests a need for more and better educational as well as recruitment efforts. A significant minority of the public holds one or
more misconceptions—for example, that donation is deforming or is not possible over age 55 (Gallup, 1993). On the other hand, some have expressed skepticism about the effect of periodic donor campaigns that lack the resources to maintain the intensity necessary to have more than a minimal effect on public understanding and opinion. They have noted that the resources devoted to public education at the organ procurement level averaged only $158,000 per year per organization in 1994 (L. Jacobbi, personal communication, July 17, 1997). This argues for devoting greater resources in general to public education. More specifically, those groups most likely to be responsive could be targeted with carefully focused information. Those involved in deciding about donation could be educated at the time of the decision through making a greater investment in trained, coordinated, and monitored procurement personnel; better-informed medical professionals; and well-planned and directed donor recruitment procedures (Partnership for Organ Donation, unpublished work, June 1997; Perkins, 1987; Siminoff et al., 1995). Reports indicate plenty of room for improvement here (Franz et al., 1997).
Continuing efforts to improve donation within the current structure of the transplantation effort are important. In view of the factors discussed, it is reasonable to assume that real increases in procurement of organs from currently accepted donor categories could be achieved through: better preparation of the public for the eventuality of being confronted with a loved one's death and the possibility of becoming an organ donor; appropriate discussions with potential donors; expanded criteria for cadaver donor acceptance; improved public and professional education; higher request and consent percentages; and expanded governmental and legislative initiatives. Work is needed in all the areas noted above. A number of other helpful suggestions have also been made (Riether and Mahler, 1995), and an array of more recent efforts holds promise. These include legislation establishing a donor registry linked to driver's license renewal in Illinois that does not force a yes or no on donation as was previously the case in Texas, but rather indicates interest in potentially being a donor. This registry has achieved a 38% positive response and is associated with a 52% increase in donors at the Regional Organ Bank of Illinois. An associated intense public campaign has played a key role in this result also (Anderson, personal communication, 1997). In 1994, Pennsylvania enacted legislation that requires routine notification to OPOs of all hospital deaths and places OPO staff in control of requesting consent for organ donation. This legislation also includes penalties for noncompliance. The program has been associated with a 32% increase in transplants from the Delaware Valley (Pennsylvania) Transplant Program. Most recently, New York has enacted similar legislation, but without the noncompliance penalties. Arizona provides an example of a state that repealed presumed donor consent on the driver's license and enacted legislation that statutorily confirms the primacy of advance directives, requires specific training for personnel seeking consent for donation, and requires only distribution of a donor
card and a red ''donor" sticker at application for a driver's license (L. Futral, personal communication, September 17, 1997). The nationally active Coalition on Donation has generated more than $30 million in donated media time and space in the few years since it was established and, with bipartisan congressional support, generated donor cards that were inserted in 70 million income tax refunds in 1997. A new organization, the Redford Institute for Transplant Awareness, using its access to the media and the entertainment industry, is also beginning campaigns.
Current levels of effort including recent positive developments may have a significant effect and, along with further efforts, should be vigorously followed up. Alone, however, they seem unlikely to provide multiple marginal improvements of the magnitude necessary to resolve the demand crisis. The number of variables and the lack of quantitative information that have been cited preclude any certainty about supply and demand. Real demand seems likely to be many times the supply possible with current scenarios. Therefore, transplant programs are being challenged, particularly since the Pittsburgh protocol in 1992 (Arnold, 1995; DeVita and Snyder, 1993), to explore whether a new category of donor, the non-heart-beating donor, might add significantly to supply, although this kind of donor is more a refinement of one of the earliest categories that had fallen into relative disuse than a new concept.