Making dollars and sense out of liver transplantation

Healthcare costs in the United States are rising at an alarming rate. The cost and quality of healthcare delivery are coming under increased scrutiny by both public and private payer sectors with the clear intent of increasing the ability of the consumer to use value-driven decisions in purchasing healthcare services. Achieving this objective will require higher levels of transparency for both cost and quality than are currently available. The cost of transplant services is diminutive in comparison with the overall cost of healthcare, amounting to 4% of gross healthcare expenditures. However, there are particular peculiarities associated with both the cost accounting and reimbursement that differentiate transplantation from other health disciplines. These include the high per-unit cost, organ-acquisition cost centers (the last bastion of cost-based reimbursement), and finally “carve-out” contractual agreements resulting in case rates from private payers for various phases of transplantation care.1, 2 LOS, length of stay; MELD, Model for End-Stage Liver Disease; OPTN, Organ Procurement and Transplantation Network. Because the Model for End-Stage Liver Disease (MELD) score reflects the severity of illness of patients undergoing liver transplantation, it is intuitive that patients with high MELD scores require a higher utilization of resources and therefore that the MELD score is a significant cost driver. Previous reports have established an association between the MELD score and length of stay (LOS), which is also an established cost driver, and have shown a strong correlation between the MELD score and costs. Axelrod et al.3 showed that the cost of transplant admission correlated directly with the MELD score, and they also demonstrated that the increased cost was attributable almost directly to longer LOS for the patients with the higher MELD scores. Washburn et al.4 demonstrated a correlation between the MELD score and hospital charges, including charges 3 months pre-transplant, charges for the transplant admission/episode, and charges 3 months post-discharge. Both reports comprised single-center studies that correlated granular clinical outcomes with institutional financial outcomes. In this issue of Liver Transplantation, Buchanan and colleagues5 analyze the MELD score as a predictor of total liver transplantation cost. In an attempt to validate previous observations in the broader context of multiple institutions, the authors have used a novel database link between billing claims from a large private payer and the Organ Procurement and Transplantation Network (OPTN) registry. In order to overcome the flaws inherent to large registries such as the OPTN registry, the authors have analyzed patient-specific claims for 1 year preceding and 1 year following transplantation for patients identified in both data sets. This approach, which adds granularity and context to OPTN data, confirms the previously demonstrated correlation between the MELD score, the pre-admission cost, and the transplant episode cost, which is driven mainly by LOS. The study fails to show a correlation between the MELD score and posttransplant and re-admission costs. Despite the obvious novelty and merits of linking relevant data sets, the study does not fully leverage the use of claims data to address some important issues. For instance, although the current study confirms the previously demonstrated correlation between the MELD score and both LOS and cost, there are varying underlying reasons that LOS might exceed established clinical care pathways. Moreover, there are indirect cost drivers associated with the higher MELD score that may not be explained by LOS. Because claims data are specific to resource utilization, it would be helpful to dissect the anatomy of these costs. Also, the finding that postdischarge costs are unrelated to the MELD score may be related to the fact that patients with higher MELD scores have a higher chance of being discharged to a chronic rehabilitation facility, and these costs would not be captured as an institutional charge but would be included in the patient's claims data under a different provider. Nonetheless, the study reiterates 3 important concepts. First, in the setting of fixed reimbursement, understanding and optimizing cost is essential in achieving sustainable financial outcomes. Cost patterns for liver transplantation do not seem to follow traditional statistical patterns, showing instead normal distributions associated with significant skews to the right.6 These patterns do not conform to standard contractual formats with fixed reimbursement complemented by outlier stop loss protection from financial risk. Clearly, the magnitude of cost depends on the severity of illness (MELD score); therefore, clinical care pathways should include an assessment of the severity of illness in order to justify deviations. Second, reimbursement schemes and pricing strategies should incorporate an evaluation of the severity of illness because reimbursement-to-cost ratios translate directly into financial outcomes. As mentioned by the authors, the Center for Medicare and Medicaid Services has recently issued a revised classification of diagnosis-related group reimbursement that will drastically change the assignment of resource utilization based on patient comorbidities and the presence or absence of postoperative complications. Among transplant codes (72 CFR 47130), this will affect heart and liver transplantation. However, the list of comorbidities does not include variables that reflect the severity of liver disease, particularly the MELD score. The authors allude to the possibility of a tiered approach to reimbursement based on disease severity but provide little insight into how the claims data could be used to this end. Also, both the clinical and financial outcomes following liver transplantation depend on the quality of the graft. Although no standardized measure of graft quality exists, there appears to be a correlation between the donor risk index and clinical outcomes.7 We have previously observed that, regardless of the MELD score, as the donor risk index increases, so do LOS and cost.8 Therefore, it would seem important to recognize the impact of both recipient and donor variables on cost. In our own analyses, outlier thresholds do little to protect transplant center profitability, although these can soften the financial impact of high-risk cases (unpublished data). It is also evident that despite risk adjustment, expected clinical outcomes, including utilization of resources, may not take into account all the relevant variables that can exacerbate the discrepancy between expected and actual outcomes, not only clinically but also financially. Third, although the authors have analyzed the costs associated with transplantation, the cost of not undergoing transplantation is not addressed. The study provides a glimpse into nontransplant costs by including pretransplant costs. However, these cost data are included only for patients who ultimately undergo transplantation. A recent report9 has compared the cost of a living donor to that of not undergoing living donor liver transplantation, which includes the cost of managing patients on the waiting list as well as patients who are allocated a deceased donor organ while on the waiting list. Most transplant centers must deal with patients who undergo liver transplantation as well as patients who are supposed to undergo liver transplantation but do not either as a result of the organ shortage or because they develop a contraindication to liver transplantation. The costs involved in the care of these patients are an integral part of caring for this patient population. In most transplant centers, the same personnel, resources, and cost centers are involved for both scenarios. Therefore, a complete analysis of liver transplant costs should include the cost of care for patients who do not ultimately undergo transplantation. These data would help address the question of the cost effectiveness of transplantation, the answer to which has remained elusive. The study tangentially addresses reimbursement, although it should be implicit that any focus on cost must take reimbursement into consideration. Thus, the question remains: does the MELD score affect transplant center profitability? Others have shown that biliary complications following liver transplantation are costly and that the burden of cost (and risk) falls primarily on the payer.10 Current trends point toward increasing transparency by providers with respect to clinical outcomes and cost. Also, federal regulatory forces dictate the use of marginal organs for transplantation that are known to be associated with higher cost and worse clinical outcomes. As the payer community continues to seek value-based purchasing of transplant services, it will be increasingly important for transplant centers to understand both the numerator and denominator of the value equation. Value equals quality divided by unit cost. Value-based purchasing implies buying more quality at a lower cost. In reality, however, reimbursement has more to do with pricing than with cost, and most transplant centers have little appreciation of true costs; this makes cost-based pricing difficult. As a result, reimbursement is based increasingly on market-based strategies versus cost-based strategies. This commoditization of transplantation services threatens to limit transplantation to good-risk recipients and to the selective use of optimal grafts. In a value-driven market, one of the unintended consequences might be that transplant centers are financially rewarded for selecting low-risk recipients and high-quality donors and financially penalized for transplanting high-risk recipients and using marginal donors, whereas current federal regulations mandate the latter. The overriding transplant regulations on organ allocation (42 CFR Parts 405, 482, 488, and 498) mandate that organs be allocated to the sickest patients first, and recent regulatory efforts have also mandated the use of marginal donors (42 CFR Parts 413, 441, 486, and 498). Given these clinical and economic dynamics, reports such as the current study are needed to increase awareness about cost drivers in transplantation. However, this type of analysis should be viewed in the context of reimbursement and the financial implications of regulatory mandates, including national recipient/organ selection and allocation policy. As this field of socioeconomic study evolves, it is clear that cost data will need to be reconciled with reimbursement schemes and pricing strategies so that transplantation can remain both financially and economically sustainable while adhering to the relevant regulatory mandates.