Balancing Efficiency and Fairness in Liver Transplant Access: Tradeoff Curves for the Assessment of Organ Distribution Policies. Transplantation. 2020;104:981-987. [PMID: 31644494 DOI: 10.1097/tp.0000000000003017]

Common definitions and variables are needed for the United States to join the conversation on acute-on-chronic liver failure

Acute-on-chronic liver failure (ACLF) is a variably defined syndrome characterized by acute decompensation of cirrhosis with organ failures. At least 13 different definitions and diagnostic criteria for ACLF have been proposed, and there is increasing recognition that patients with ACLF may face disadvantages in the current United States liver allocation system. There is a need, therefore, for more standardized data collection and consensus to improve study design and outcome assessment in ACLF. In this article, we discuss the current landscape of transplantation for patients with ACLF, strategies to optimize organ utility, and data opportunities based on emerging technologies to facilitate improved data collection.

Heterogeneous donor circles for fair liver transplant allocation

The United States (U.S.) Department of Health and Human Services is interested in increasing geographical equity in access to liver transplant. The geographical disparity in the U.S. is fundamentally an outcome of variation in the organ supply to patient demand (s/d) ratios across the country (which cannot be treated as a single unit due to its size). To design a fairer system, we develop a nonlinear integer programming model that allocates the organ supply in order to maximize the minimum s/d ratios across all transplant centers. We design circular donation regions that are able to address the issues raised in legal challenges to earlier organ distribution frameworks. This allows us to reformulate our model as a set-partitioning problem. Our policy can be viewed as a heterogeneous donor circle policy, where the integer program optimizes the radius of the circle around each donation location. Compared to the current policy, which has fixed radius circles around donation locations, the heterogeneous donor circle policy greatly improves both the worst s/d ratio and the range between the maximum and minimum s/d ratios. We found that with the fixed radius policy of 500 nautical miles (NM), the s/d ratio ranges from 0.37 to 0.84 at transplant centers, while with the heterogeneous circle policy capped at a maximum radius of 500 NM, the s/d ratio ranges from 0.55 to 0.60, closely matching the national s/d ratio average of 0.5983. Our model matches the supply and demand in a more equitable fashion than existing policies and has a significant potential to improve the liver transplantation landscape.

A transformer-based deep learning approach for fairly predicting post-liver transplant risk factors

Liver transplantation is a life-saving procedure for patients with end-stage liver disease. There are two main challenges in liver transplant: finding the best matching patient for a donor and ensuring transplant equity among different subpopulations. The current MELD scoring system evaluates a patient's mortality risk if not receiving an organ within 90 days. However, the donor-patient matching should also consider post-transplant risk factors, such as cardiovascular disease, chronic rejection, etc., which are all common complications after transplant. Accurate prediction of these risk scores remains a significant challenge. In this study, we used predictive models to solve the above challenges. Specifically, we proposed a deep learning model to predict multiple risk factors after a liver transplant. By formulating it as a multi-task learning problem, the proposed deep neural network was trained to simultaneously predict the five post-transplant risks and achieve equal good performance by exploiting task-balancing techniques. We also proposed a novel fairness-achieving algorithm to ensure prediction fairness across different subpopulations. We used electronic health records of 160,360 liver transplant patients, including demographic information, clinical variables, and laboratory values, collected from the liver transplant records of the United States from 1987 to 2018. The model's performance was evaluated using various performance metrics such as AUROC and AUPRC. Our experiment results highlighted the success of our multi-task model in achieving task balance while maintaining accuracy. The model significantly reduced the task discrepancy by 39 %. Further application of the fairness-achieving algorithm substantially reduced fairness disparity among all sensitive attributes (gender, age group, and race/ethnicity) in each risk factor. It underlined the potency of integrating fairness considerations into the task-balancing framework, ensuring robust and fair predictions across multiple tasks and diverse demographic groups.

Removing geographic boundaries from liver allocation: A method for designing continuous distribution scores

BACKGROUND

The Organ Procurement and Transplantation Network (OPTN) is eliminating geographic boundaries in liver allocation, in favor of continuous distribution. Continuous distribution allocates organs via a composite allocation score (CAS): a weighted sum of attributes like medical urgency, candidate biology, and placement efficiency. The opportunity this change represents, to include new variables and features for prioritizing candidates, will require lengthy and contentious discussions to establish community consensus. Continuous distribution could instead be implemented rapidly by computationally translating the allocation priorities for pediatric, status 1, and O/B blood type liver candidates that are presently implemented via geographic boundaries into points and weights in a CAS.

METHODS

Using simulation with optimization, we designed a CAS that is minimally disruptive to existing prioritizations, and that eliminates geographic boundaries and minimizes waitlist deaths without harming vulnerable populations.

RESULTS

Compared with Acuity Circles (AC) in a 3-year simulation, our optimized CAS decreased deaths from 7771.2 to 7678.8 while decreasing average (272.66 NM vs. 264.30 NM) and median (201.14 NM vs. 186.49 NM) travel distances. Our CAS increased travel only for high MELD and status 1 candidates (423.24 NM vs. 298.74 NM), and reduced travel for other candidates (198.98 NM vs. 250.09 NM); overall travel burden decreased.

CONCLUSION

Our CAS reduced waitlist deaths by sending livers for high-MELD and status 1 candidates farther, while keeping livers for lower MELD candidates nearby. This advanced computational method can be applied again after wider discussions of adding new priorities conclude; our method designs score weightings to achieve any specified feasible allocation outcomes.

Logistical burden of offers and allocation inefficiency in circle-based liver allocation

Recent changes to liver allocation replaced donor service areas with circles as the geographic unit of allocation. Circle-based allocation might increase the number of transplantation centers and candidates required to place a liver, thereby increasing the logistical burden of making and responding to offers on organ procurement organizations and transplantation centers. Circle-based allocation might also increase distribution time and cold ischemia time (CIT), particularly in densely populated areas of the country, thereby decreasing allocation efficiency. Using Scientific Registry of Transplant Recipient data from 2019 to 2021, we evaluated the number of transplantation centers and candidates required to place livers in the precircles and postcircles eras, nationally and by donor region. Compared with the precircles era, livers were offered to more candidates (5 vs. 9; p  < 0.001) and centers (3 vs. 5; p  < 0.001) before being accepted; more centers were involved in the match run by offer number 50 (9 vs. 14; p  < 0.001); CIT increased by 0.2 h (5.9 h vs. 6.1 h; p  < 0.001); and distribution time increased by 2.0 h (30.6 h vs. 32.6 h; p  < 0.001). Increased burden varied geographically by donor region; livers recovered in Region 9 were offered to many more candidates (4 vs. 12; p  < 0.001) and centers (3 vs. 8; p  < 0.001) before being accepted, resulting in the largest increase in CIT (5.4 h vs. 6.0 h; p  < 0.001). Circle-based allocation is associated with increased logistical burdens that are geographically heterogeneous. Continuous distribution systems will have to be carefully designed to avoid exacerbating this problem.

Assessment of donor quality and risk of graft failure after liver transplantation: The ID2 EAL score

Accurate assessment of donor quality at the time of organ offer for liver transplantation candidates may be inadequately captured by the donor risk index (DRI). We sought to develop and validate a novel objective and simple model to assess donor risk using donor level variables available at the time of organ offer. We utilized national data from candidates undergoing primary LT (2013-2019) and assessed the prediction of graft failure 1 year after LT. The final components were donor Insulin-dependent diabetes mellitus, Donor type (DCD or DBD), cause of Death = CVA, serum creatinine, Age, height, and weight (length). The ID² EAL score had better discrimination than DRI using bootstrap corrected concordant index over time, especially in the current era. We explored donor-recipient matching. Relative risk of graft failure ranged from 1.15 to 3.5 based on relevant donor-recipient matching by the ID² EAL score. As an example, for certain recipients, a young DCD donor offer was preferable to an older DBD with relevant comorbidities. The ID² EAL score may serve as an important tool for patient discussion about donor risk and decisions regarding offer acceptance. In addition, the score may be preferable to succinctly capture donor risk in future organ allocation that considers continuous distribution (www.iddealscore.com).

The Effect of Acuity Circles on Deceased Donor Transplant and Offer Rates Across Model for End-Stage Liver Disease Scores and Exception Statuses

Acuity circles (AC), the new liver allocation system, was implemented on February 4, 2020. Difference-in-differences analyses estimated the effect of AC on adjusted deceased donor transplant and offer rates across Pediatric End-Stage Liver Disease (PELD) and Model for End-Stage Liver Disease (MELD) categories and types of exception statuses. The offer rates were the number of first offers, top 5 offers, and top 10 offers on the match run per person-year. Each analysis adjusted for candidate characteristics and only used active candidate time on the waiting list. The before-AC period was February 4, 2019, to February 3, 2020, and the after-AC period was February 4, 2020, to February 3, 2021. Candidates with PELD/MELD scores 29 to 32 and PELD/MELD scores 33 to 36 had higher transplant rates than candidates with PELD/MELD scores 15 to 28 after AC compared with before AC (transplant rate ratios: PELD/MELD scores 29-32, _2.34 3.32_4.71 ; PELD/MELD scores 33-36, _1.70 2.51_3.71 ). Candidates with PELD/MELD scores 29 or higher had higher offer rates than candidates with PELD/MELD scores 15 to 28, and candidates with PELD/MELD scores 29 to 32 had the largest difference (offer rate ratios [ORR]: first offers, _2.77 3.95_5.63 ; top 5 offers, _3.90 4.39_4.95 ; top 10 offers, _4.85 5.30_5.80 ). Candidates with exceptions had lower offer rates than candidates without exceptions for offers in the top 5 (ORR: hepatocellular carcinoma [HCC], _0.68 0.77_0.88 ; non-HCC, _0.73 0.81_0.89 ) and top 10 (ORR: HCC, _0.59 0.65_0.71 ; non-HCC, _0.69 0.75_0.81 ). Recipients with PELD/MELD scores 15 to 28 and an HCC exception received a larger proportion of donation after circulatory death (DCD) donors after AC than before AC, although the differences in the liver donor risk index were comparatively small. Thus, candidates with PELD/MELD scores 29 to 34 and no exceptions had better access to transplant after AC, and donor quality did not notably change beyond the proportion of DCD donors.

Artificial Intelligence in Liver Transplantation

BACKGROUND

Advancements based on artificial intelligence have emerged in all areas of medicine. Many decisions in organ transplantation can now potentially be addressed in a more precise manner with the aid of artificial intelligence.

METHOD/RESULTS

All elements of liver transplantation consist of a set of input variables and a set of output variables. Artificial intelligence identifies relationships between the input variables; that is, how they select the data groups to train patterns and how they can predict the potential outcomes of the output variables. The most widely used classifiers to address the different aspects of liver transplantation are artificial neural networks, decision tree classifiers, random forest, and naïve Bayes classification models. Artificial intelligence applications are being evaluated in liver transplantation, especially in organ allocation, donor-recipient matching, survival prediction analysis, and transplant oncology.

CONCLUSION

In the years to come, deep learning-based models will be used by liver transplant experts to support their decisions, especially in areas where securing equitability in the transplant process needs to be optimized.

Evolution of liver transplant organ allocation policy: Current limitations and future directions

Since the adoption of the model for end-stage liver disease (MELD) score for organ allocation in 2002, numerous changes to the system of liver allocation and distribution have been made with the goal of decreasing waitlist mortality and minimizing geographic variability in median MELD score at time of transplant without worsening post-transplant outcomes. These changes include the creation and adoption of the MELD-Na score for allocation, Regional Share 15, Regional Share for Status 1, Regional Share 35/National Share 15, and, most recently, the Acuity Circles Distribution Model. However, geographic differences in median MELD at time of transplant remain as well as limits to the MELD score for allocation, as etiology of liver disease and need for transplant changes. Acute-on-chronic liver failure (ACLF) is a subset of liver failure where prevalence is rising and has been shown to have an increased mortality rate and need for transplantation that is under-demonstrated by the MELD score. This underscores the limitations of the MELD score and raises the question of whether MELD is the most accurate, objective allocation system. Alternatives to the MELD score have been proposed and studied, however MELD score remains as the current system used for allocation. This review highlights policy changes since the adoption of the MELD score, addresses limitations of the MELD score, reviews proposed alternatives to MELD, and examines the specific implications of these changes and alternatives for ACLF.

Liver transplantation in the United States during the COVID-19 pandemic: National and center-level responses

COVID-19 has profoundly affected the American health care system; its effect on the liver transplant (LT) waitlist based on COVID-19 incidence has not been characterized. Using SRTR data, we compared observed LT waitlist registrations, waitlist mortality, deceased donor LTs (DDLT), and living donor LTs (LDLT) 3/15/2020-8/31/2020 to expected values based on historical trends 1/2016-1/2020, stratified by statewide COVID-19 incidence. Overall, from 3/15 to 4/30, new listings were 11% fewer than expected (IRR = 0.84 0.890.93 ), LDLTs were 49% fewer (IRR = 0.37 0.510.72 ), and DDLTs were 9% fewer (IRR = 0.85 0.910.97 ). In May, new listings were 21% fewer (IRR = 0.74 0.790.84 ), LDLTs were 42% fewer (IRR = 0.39 0.580.85 ) and DDLTs were 13% more (IRR = 1.07 1.151.23 ). Centers in states with the highest incidence 3/15-4/30 had 59% more waitlist deaths (IRR = 1.09 1.592.32 ) and 34% fewer DDLTs (IRR = 0.50 0.660.86 ). By August, waitlist outcomes were occurring at expected rates, except for DDLT (13% more across all incidences). While the early COVID-affected states endured major transplant practice changes, later in the pandemic the newly COVID-affected areas were not impacted to the same extent. These results speak to the adaptability of the transplant community in addressing the pandemic and applying new knowledge to patient care.