Liver Transplants: How Do We Choose Who Should Live When Not All Can?

With a scarce, non-renewable resource such as livers for transplantation, shouldn’t the individuals who receive organs be the persons who need them most? If recipients could be selected based on need, allocation finally could be divorced from onerous criteria such as social value. Since need in this context can be equated with death (if life-saving treatment is withheld), the manner in which recipients are chosen becomes paramount. On an existential level, the question may be reduced simply to choosing who should live when not all can.

This hard reality is simply a case of demand far outstripping supply. There has been and continues to be a scarcity of donated livers. In fact, nearly 17,000 patients were waiting for one in 2002 (up from 1,676 a decade earlier), while only 5,000 livers were donated that year. As sick as those in need have been, their average wait has increased from 65 to 795 days over ten years! Without a liver, these people die at a rate estimated to be 10% (approximately 1,300 individuals) per year. Addressing justice in life-and-death allocation decisions calls for a frank evaluation of the ethical criteria applied. Considerations include durable concepts, both positive and negative, such as “first come--first served,” medical benefit, social value, progress of science, and favored groups. Recent data has once again suggested that US transplant surgeons are not providing a “level playing field” for liver recipients.1 It has been demonstrated, furthermore, that surgeons are prioritizing their own patients, even when these individuals are less in need than others. Why is this happening?

The Department of Health and Human Services has characterized the development of just criteria for donation as a “final rule mandate.”2 Strong emphasis should be placed on consistent fairness when sharing the “Gift of Life.” In fact, the Department issued recent regulations to ensure “that allocation of scarce organs would be based on common medical criteria, not accidents of geography.” This concern represents two contentious issues. First, in the past, the assessment of illness severity unjustly has included subjective criteria. Second, disparity in the fairness of allocation across geographic regions continues to be an issue.

The problem of subjectivity in the assessment of illness severity seems to have been solved. Medical need can be determined by an evidence-based score, an objective marker of severity. The score, the Model for End-stage Liver Disease (or MELD), is calculated from parameters indicative of liver function (serum creatinine, INR, and bilirubin). The goal of MELD is to provide an objective, medical-benefit standard in order to allocate livers justly. Studies have documented the score’s utility; it identifies those persons who most need a liver. The strategic endpoint of MELD scoring is to identify for transplantation patients who will die within three months rather than those who have been on the waiting list the longest. “First come” should not always be “first served” in the context of liver transplantation, because people on the list the longest are not always the sickest.

Although not perfect, MELD is probably the most objective measure available to prioritize recipients based on need. Following a study, the United Network of Organ Sharing adopted MELD in February 2002.3 Has it worked? The simple answer is that medical benefit criteria per se are no longer at the root of the distribution problem. The current dilemma, so-called “accidents of geography,” which are the direct result of disparately sized Organ Procurement Organizations (OPOs), has not been managed adequately.

In the United States, variation in the size of geographic organ allocation “areas” is a reality. As a result, one might say that the existing OPOs are “bimodal.” From a number of persons served standpoint, they tend to be either very large or very small. Donated organs stay close to the region in which they are procured even if “sicker” patients in other geographic areas are in greater need. Small OPOs have a greater “per capita” supply of organs and fewer people on their list. As a result, less critical patients may receive a transplant more expeditiously. Conversely, large OPOs—those in big cities with large medical centers—often have many critically ill people waiting. Without a timely donation, more of their critical patients die.

The Institute of Medicine has made two recommendations that should affect the contemporary allocation question as it works toward the goal of fairness. Regarding criteria for selection, they caution that an objective score of illness severity, such as MELD, be applied to everyone. Concerning geographic disparity, they strongly suggest that OPOs should serve a population base of 9 million people. Geography should not be the issue; rather, there seems to be an optimal population size that would serve as a better criterion. Data has demonstrated that a census of 9 million would speed transplantation for sicker individuals, without adversely affecting the “less sick” when they later desperately need a transplant.

Liver transplant recipients from February 28, 2002, through March 30, 2003 (4,798 individuals) were stratified by MELD scores.4 Comparing smaller to larger OPOs revealed disturbing results. The proportion of patients receiving liver transplants with a “higher” MELD score (i.e., “sicker”) was lower for the smaller (n=43) than for the larger (n=400) OPOs (19% versus 49%). The bottom line is the numbers demonstrated that recipients in smaller OPOs received organs although they were not in as great a need as individuals in larger OPOs. Despite the availability of the MELD score, sicker individuals were not transplanted because organs were not shared adequately.

The solution—standardizing the OPO populations served at 9 million—should be the next step. Are there ethical downsides to a gerrymander of OPO size? Smaller transplant centers may have to close. This is not a prohibitive price to pay when lives are at stake. Another concern, that patients in smaller OPOs may have to travel farther to get care, may have some merit. However, travel issues are easier to address than the present geographically based disparity.

In All I Really Need to Know I Learned in Kindergarten, Robert Fulghum indicated that sharing, a fundamental lesson to be learned early in life, is an unequivocal societal good. It appears this lesson is one that OPOs and transplant surgeons have yet to learn. If the MELD score is to work, sharing across size-consistent OPOs must be implemented. The Institute of Medicine was right. Let’s not waste MELD, a score that finally may end the wrong of social valuation in organ transplant allocation decisions. In the scheme of things, adversely affecting the viability of small transplant centers pales in comparison to continuing the present alternative.


Institute of Medicine Committee on Organ Procurement and Transplantation Policy. Organ Procurement and Transplantation: Assessing Current Policies and the Potential Impact of the DHHS Final Rule. Washington, D.C., National Academy Press; 1999: 1-29.

JAMA 2004; 291: 1871-1874.