| |
Black Patients Have Unequal Access to Listing for Liver Transplantation in the United States
|
| |
| |
Download the PDF here
29 March 2021 Russell Rosenblatt MD, MS; Nabeel Wahid MD; Karim J. Halazun MD; Alyson Kaplan MD; Arun Jesudian MD; Catherine Lucero MD; Jihui Lee PhD; Lorna Dove MD, MPH; Alyson Fox MD, MAS; Elizabeth Verna MD, MS; Benjamin Samstein MD; Brett E
Fortune MD, MS; Robert S. Brown, Jr. MD, MPH
In conclusion, there is significant variability between the ratios of listing for LT, receiving LT, and dying from ESLD in the United States. Importantly, the states that had the most successful outcomes (high transplant to death ratios) had very high rates of listing (LDR). This analysis exposed a significant disparity for Black patients, who had the lowest LDR and poor access to listing for LT, which was previously underappreciated. Identification of this disparity is crucial as it highlights where disadvantage lies - for Black patients, this is prior to transplant listing. Therefore, further efforts including education, outreach, and disease recognition are required in order to improve Black patients' LDR and equalize outcomes for all race/ethnicities.
Abstract
The Model for End‐Stage Liver Disease score may have eliminated racial disparities on the waitlist for liver transplantation (LT), but disparities prior to waitlist placement have not been adequately quantified. We aimed to analyze differences in patients who are listed for LT, undergo transplantation, and die from end‐stage liver disease (ESLD), stratified by state and race/ethnicity. We analyzed two databases retrospectively - the Center for Disease Control Wide‐ranging OnLine Data for Epidemiologic Research (CDC WONDER) and the United Network for Organ Sharing (UNOS) databases from 2014‐2018. We included patients aged 25‐64 years who had a primary cause of death of ESLD and listed for transplant in the CDC WONDER and UNOS databases, respectively. Our primary outcome was the ratio of listing for LT to death from ESLD - listing to death ratio (LDR). Our secondary outcome was the transplant to listing and transplant to death ratios. Chi‐squared and multivariable linear regression evaluated for differences between race/ethnicity. 135,367 patients died of ESLD, 54,734 patients were listed for transplant, and 26,571 underwent transplant. Patients were mostly male and White. The national LDR was 0.40, significantly lowest in Black patients (0.30), p<0.001. The national transplant to listing ratio was 0.48, highest in Black patients (0.53), p<0.01. The national transplant to death ratio was 0.20, lowest in Black patients (0.16), p<0.001. States that had an above‐mean LDR had a lower transplant to listing ratio, but higher transplant to death ratio. Multivariable analysis confirmed Black race is significantly associated with a lower LDR and transplant to death ratio.
Conclusion: Black patients face a disparity in access to LT due to low listing rates for transplant relative to deaths from ESLD.
Introduction
Liver disease is increasingly common in the United States and is associated with significant costs, morbidity, and mortality(1,2). Once complications develop from end-stage liver disease (ESLD), liver transplantation (LT) is the only curative therapy(3). Although, the annual number of LTs are increasing, peaking in 2019 at 8,896(4), access to LT remains limited, unequal (5-7), and does not meet the current demand given dramatic rises in death due to cirrhosis and hepatocellular carcinoma (HCC)(8).
Unfortunately, vulnerable populations face significant disparities in listing and transplant(9). Black patients have long had poor access to LT compared to White patients(10-13), but the implementation of the objective laboratory-based Model for End-Stage Liver Disease (MELD) as the basis for organ allocation neutralized racial disparities for patients already on the waitlist(14).
While most transplant-related disparities research has focused on outcomes involving patients already on the waitlist, it is likely that substantial obstacles are faced prior to waitlist placement. Studies have demonstrated that Black patients were less likely to be referred for LT evaluation(15,16) and even when referred were less likely to pursue care at an LT center(17). While racial disparities on the waitlist for LT have improved(14), those prior to waitlist placement have likely a profound but unmeasured impact. Therefore, this study aimed to quantify racial, ethnic, and state-specific differences in patients who are listed for liver transplant, undergo transplantation, and die from ESLD.
Discussion
Racial and ethnic disparities have plagued the liver transplantation process since its inception, particularly for Black patients(10,11). The adoption of the MELD score appeared to ameliorate these disparities due to its objective nature of using laboratory values, but this benefit was only assessed for patients already on the waitlist(14). Therefore, we hypothesized that because Black patients are referred less often for LT(15-17), among other potential disadvantages, there would be observed disparities in the ratio of patients listed for LT compared to those who died from ESLD (LDR).
Our analysis demonstrates that Black patients continue to face a disparity in access to LT as a result of low listing rates relative to deaths from ESLD. Moylan et al noted that waitlist outcomes in races and ethnicities had resolved because of the MELD score(14), but this study only measured the outcomes in patients who were already listed for transplant, or in other words, the transplant to listing ratio. This is an important metric as most patients with ESLD who do not get transplanted will ultimately die on the waitlist. We did find that Black patients had a transplant to listing ratio above other races suggesting MELD does address outcomes once listed. However, our study expands upon this important work by increasing the denominator to include not only waitlisted patients but all patients with ESLD. By accounting for all deaths from ESLD, the disadvantage faced by Black patients becomes clear as they are placed on the waitlist at a lower rate compared to other race/ethnicities in terms of need as measured by ESLD deaths. Even when controlling for CHS, this significant disparity persists. While having a numerically higher transplant to listing ratio, multivariable analysis demonstrates that Black patients do not have a significant advantage in transplantation rate. The transplant to death ratio is the key and final measurement - it assesses the rate of patients saved from ESLD with the only curative therapy, transplantation. The inability of Black patients to overcome a low listing rate (LDR) is also experienced generally by several states (such as Florida, Tennessee, Alabama, South Carolina, Indiana) who have low listing rates, compensatory high transplant rates of patients on the waitlist, but still and most importantly, low transplant to death ratios.
There are many steps in the process of being placed on the waitlist for LT, and therefore, many potential roadblocks to listing vulnerable patients. ESLD first needs to be identified and patients need to be referred to a transplant center. Black patients are referred at lower rates(15,16) and were much less likely to understand the LT process(17). Beyond the referral itself, appropriate timing is critical, especially in patients with HCC, which presents as more advanced in Black patients(23). Insurance is another key factor in access to LT; Black patients are more likely to be uninsured(24) or have public insurance(25), which is associated with poor access to listing and LT(26). One disadvantage that has been recently overcome is the previously poor outcomes for Black patients with hepatitis C(7) that now is treated equally successfully with direct-acting antivirals in all races/ethnicities. All of these factors, with the exception of hepatitis C, limit transplant candidacy prior to waitlist placement and may be missed in an analysis solely focused on waitlist outcomes.
However, it is unclear why Black patients have higher rates of transplantation while on the waitlist. A possible explanation includes a higher prevalence of hepatitis C and HCC, which was more common in patients who were transplanted because of MELD exception points. However, prior studies show that Black patients with HCC were nearly 50% less likely to be transplanted than White patients(27,28). Additionally, our multivariable analysis demonstrates that the transplant to listing ratio is not significantly different for any race/ethnicity. In contrast, Hispanic patients, who were more frequently listed than White patients, had the lowest likelihood of being transplanted while on the waitlist, again, while acknowledging that this was not borne out in multivariable analysis. Still, tracking outcomes in Hispanic patients will be crucial as they suffer from similar social determinants of health as Black patients(29), have the highest age-adjusted rates of HCC(30), and a high burden from nonalcoholic fatty liver disease(31). The excellent outcomes in Asian patients are likely explained by the high rates of and improved survival with HCC(32).
The efficiency of transplanting patients on the waitlist, the transplant to listing ratio, is reliant on center-specific factors within states such as organ availability, acceptance of organs, and median MELD at transplant, which varies widely(33). For this reason, we included DRI in our multivariable regression and found that it was not associated with an increased transplant to listing ratio. Certainly, in states that have more aggressive transplant centers, increasing transplant efficiency will be extremely challenging, while states that are less aggressive have a potential path forward to increase their transplantation rates. This is further bolstered by the finding that increased DRI was significantly associated with a higher transplant to death ratio. Additionally, because a high transplant to listing ratio was not shown to be associated with a higher transplant to death ratio, we hypothesize that the impact of the new acuity circles policy on transplant to death ratio, the most important metric, will not be significant. However, future research on its this topic is needed to prove this hypothesis correct.
Our analysis is limited in that it cannot differentiate why patients were not listed for LT, why patients were not transplanted, and what organ offers were turned down. States that listed smaller proportions of their populations who died of ESLD tended to transplant their waitlist more efficiently, but this did not overcome the initial lower LDR as measured by the transplant to death ratio. The most desirable outcome is a high transplant to death ratio - which was present in states like Maryland (very high LDR but low transplant to listing ratio) as well as Georgia, Nebraska, and Wisconsin (all with both a high LDR and high transplant to listing ratio). Clearly, the LDR needs to be high to have a high transplant to death ratio.
Furthermore, the finding of an inverse relationship of rates of listing for transplant to transplant efficiency of patients already on the waitlist carry major implications for how disparity research can be performed in the field of LT. There is reliance on the UNOS database(7,14,26), which is excellent at characterizing disparities once patients are placed on the waitlist, but this type of analysis will miss disparities occurring prior to listing. Black patients are only one of likely several groups that suffer from having a low LDR. Other potentially vulnerable groups include patients with Medicaid insurance, who are sicker at time of listing(26), those from rural areas, who live farther from LT centers, and women who already face unequal access to LT while on the waitlist(5,34-36). Patients who live in rural areas and far from LT centers likely have poor access to subspecialty care, reducing the likelihood of initial referral for LT. These vulnerable groups warrant more attention in terms of frequency and timeliness of referral for LT.
The major limitations of this study include a lack of granularity with patient data and geographic data. Due to database limitations, we were unable to obtain reasons for why patients were not listed for LT and were unable to confirm which patients identified in the UNOS database and died were present in the CDC WONDER database. There were likely some patients on the waitlist who may have died for other reasons (such as an accident, myocardial infarction, or non-liver malignancy) and were coded as dying of ESLD. In addition, administrative error in coding for causes of death could be present. But these differences are likely to be non-directional and equal between states and races/ethnicities. Our geographic limitations were notable for the heterogeneous nature of the states themselves. Patients may also live in areas where it is geographically closer to an LT center in a different state, which can alter any of the three ratios. Additionally, because a state has an LT center does not mean it is geographically convenient for patient to receive care, and distance from an LT center plays an important role in outcomes(34). Measuring state statistics alone may under or overestimate the impact of distance to an LT center. Finally, it is worthwhile noting that our analysis excluded patients at both low and high ages (< 25 years old and > 64 years old) and therefore the numbers seen in this study may not align with previously published numbers that include the entire transplant or death due to ESLD population.
In conclusion, there is significant variability between the ratios of listing for LT, receiving LT, and dying from ESLD in the United States. Importantly, the states that had the most successful outcomes (high transplant to death ratios) had very high rates of listing (LDR). This analysis exposed a significant disparity for Black patients, who had the lowest LDR and poor access to listing for LT, which was previously underappreciated. Identification of this disparity is crucial as it highlights where disadvantage lies - for Black patients, this is prior to transplant listing. Therefore, further efforts including education, outreach, and disease recognition are required in order to improve Black patients' LDR and equalize outcomes for all race/ethnicities.
| |
| |
| |
|
|
|
