Race and Sex Disparities in Liver Transplantation: Title and subTitle BreakProgress Toward Achieving Equal Access?

Unique among medical specialties, the organ transplantation community has the obligation to explicitly allocate a very limited lifesaving resource. Liver transplantation offers the sole hope for long-term survival for patients with end-stage liver disease. Overall survival rates for transplantation now routinely exceed 90% at 1 year, even among patients with the most advanced liver failure, the majority of whom would die within months without a transplant.^{1 - 2} As stewards of a precious resource, the transplant community has a goal of achieving an equitable, transparent, and efficient system of organ allocation. Meeting these goals is crucial for maintaining confidence in the transplant system and encouraging organ donation.

Before 2002, deceased donor liver allografts were allocated by using a system that was loosely based on severity of illness and time spent waiting. Review of the national outcomes demonstrated that this system consistently failed to achieve the goals of equity, transparency, and efficiency.^{3 - 4} Liver allocation was perceived to be too arbitrary, too easily manipulated, and overly reliant on waiting time. Furthermore, major regional differences were found in the rates of patient death while on the waiting list. To address these issues, the Institute of Medicine (IOM) recommended to the US Department of Health and Human Services that a novel system of organ allocation be defined based solely on medical criteria.³ In response, the transplant community developed and implemented the current system of liver allocation that prioritizes patients on the basis of severity of illness, as assessed by the Model of End-Stage Liver Disease (MELD) score. The MELD score is calculated by using 3 laboratory test results (creatinine, bilirubin, and international normalized ratio), which are objective measures of the degree of organ dysfunction and risk of death. Under the MELD allocation system, the patient with the greatest chance of dying without a transplant is allocated the next available organ regardless of the time the patient has spent on the waiting list.

The clinical benefits of the MELD allocation system have been well documented. Implementation of the MELD system contributed to a reduction in the waiting list death rates from 150 to 125 deaths per 1000 patient-years at risk.¹ However, these waiting list death rates had been declining for the 5 years before MELD at an even greater rate. Furthermore, this reduction in waiting list death rates was accomplished without a significant deterioration in posttransplant survival. The MELD system has also led to a more objective method for estimating the benefit of transplant and subsequent revision of the allocation system. For relatively healthy patients (MELD score of <15), liver transplantation was found to have no survival benefit and was in fact associated with worse survival than remaining on the transplant list.⁵ As a result, the transplant community has required that donated livers be used within a wider geographical region before being used in the local community if no recipients with a MELD score of more than 15 are waiting locally.

In this issue of JAMA, Moylan and colleagues⁶ report that the use of MELD has had another desirable benefit. Retrospective analysis of the United Network for Organ Sharing (UNOS) database confirmed prior findings that in the pre-MELD era, black patients were more likely to die on the waiting list than white patients.⁷ The article by Moylan et al shows that following MELD, there were no significant racial or ethnic differences in time waiting or waiting list deaths. This finding represents an important improvement in the equity of the organ allocation system. However, Moylan et al also identified a persistent disparity between women and men in access to liver transplantation.

The reduction in waitlist mortality for black patients is likely multifactorial and may in fact not guarantee an equal opportunity to obtain a liver transplant. The transplant process requires a complex pretransplant evaluation. Before MELD, early referral for evaluation often resulted in more rapid listing and a correspondingly greater opportunity to accrue waiting time before clinical decompensation. Because patients without private health insurance are significantly less likely to be referred for specialist care, they were often evaluated and listed too late. Under MELD, late listing does not preclude transplantation as it often did in the system based on waiting time. However, as demonstrated by Moylan et al, black patients are still listed with higher MELD scores and for a shorter period, suggesting that the access barriers remain even under the new system. In addition, black patients continue to have a higher rate of liver disease than the national average, but represent a smaller proportion of transplant recipients. Thus, it is likely that many black patients are never listed due to advanced disease at presentation as a result of late referral. This problem is outside of the scope of the allocation system, however, and needs a broader solution.

The sex disparity is likely a reflection of several factors, including a limitation in the MELD calculation, body and organ size considerations, and potentially differences in the etiology of the underlying liver disease. Because women have less body mass, the inclusion of creatinine rather than weight-adjusted glomerular filtration rate in the MELD score is likely to underestimate their degree of renal dysfunction. Thus, MELD scores will be lower in women than in men with the same degree of renal failure, decreasing women's access to liver allografts. Women are also smaller, limiting the pool of available organs. Women may wait longer for a size-appropriate organ because livers from pediatric donors are preferentially allocated to children awaiting transplantation. In addition, a small organ can be used in a larger individual, but the converse is not always possible. Moreover, women are more likely than men to have autoimmune liver diseases, including primary biliary cirrhosis and primary sclerosing cholangitis, both of which are less likely than hepatitis C to lead to kidney dysfunction and higher MELD scores. From the data presented by Moylan et al, it appears logical to adjust the MELD score based on sex to ensure that the risk of death is comparable for men and women who have the same MELD score. Consideration should also be given to expanding access to pediatric donors for women who are less than a certain weight or size.

Although MELD has reduced the disparity based on some demographic characteristics, it has not met the IOM directive to reduce geographic disparity. Under existing practice, the nation is divided into 11 regions by UNOS, each of which is composed of 1 or more donation service areas (DSAs). These regions vary substantially in terms of population size and number of deceased donors. In the analysis by Moylan et al, as well as many other studies, survival with end-stage liver disease is determined in large part by where the patient lives, not by the patient's race or sex. As in politics, right now, all allocation is local. Patients who live in (or have the money to travel to) regions with a higher ratio of donors to transplant candidates can receive organ transplants far earlier and at a lower MELD score.⁸ The disparity is even greater at the DSA level. In small DSAs, the percentage of recipients who receive transplants at low MELD scores (11-18) was 47% compared with only 23% in large DSAs (P < .001).⁹ This difference directly correlates into a higher death rate on the waiting list. By adjusting for these regional differences, Moylan et al⁶ may have diminished the ongoing disparity in access for black patients who are largely concentrated in larger urban areas with longer waiting times. It is likely that women, especially those who are often less able to travel due to other responsibilities and those without health insurance, are likewise systematically less able to access this lifesaving resource.

When compared with the system before 2002, the current MELD system is clearly a step toward achieving the goal of an equitable, efficient, and transparent organ allocation system. The MELD system appears to have reduced, but likely not eliminated, differential access based on race and ethnicity. However, modest differences based on sex persist, but could be addressed by revisions in the organ allocation policy to ensure that MELD scores are comparable across sexes. The real challenge facing the liver transplant community is to tackle the long-standing disparity of geography. If those who can afford to travel can receive organ transplants at lower MELD scores more rapidly, the system will remain fundamentally inequitable.¹⁰ The current construction of DSAs and UNOS regions is largely an accident of history. As recognized by the IOM, achieving the goals of organ allocation requires the creation of equivalent size regions to match supply and demand. The transplant community should be acknowledged for the progress achieved thus far in reducing racial and sex disparities, and should be encouraged to take the next difficult step in further reducing geographic disparities.