ACOs and Bending the Cost Curve for Health Care Spending for People with Kidney Failure

The US government spends more than $45 billion annually on the care of Medicare beneficiaries with kidney failure requiring dialysis or transplantation (1). Over the past decade, the Centers for Medicare and Medicaid Services has implemented a series of value-based payment initiatives to reduce the high costs of kidney failure. In 2011, the ESRD Prospective Payment System (PPS) was expanded to incorporate injectable medications into bundled dialysis payments in an effort to reduce the high costs of these medications. In 2015, Centers for Medicare and Medicaid Services implemented the Comprehensive ESRD Care (CEC) model to encourage dialysis clinics, nephrologists, and other care providers to deliver high-value, coordinated care, by offering shared savings from judicious use of resources. In 2019, the Advancing American Kidney Health (AAKH) initiative called for five new nephrology-specific alternative payment models that include incentives to delay the onset of dialysis and encourage home dialysis and kidney transplantation. In 2021, Medicare beneficiaries with kidney failure requiring dialysis or transplantation will have the option to enroll in Medicare Advantage plans.

A strength of this multifaceted approach to improving the value of care delivered to patients with kidney failure is that it creates incentives to control costs across a wide range of care settings. For instance, expansion of the ESRD PPS focuses on reducing outpatient dialysis costs, which account for approximately 25% of Medicare’s fee-for-service ESRD expenditures. The CEC model, and new payment models resulting from the AAKH, focus more on limiting inpatient and physician costs, which account for 32% and 14% of Medicare’s ESRD expenditures, respectively (1). When combined, these payment models have a potential to expand value-based payment incentives to the majority of providers caring for patients with kidney failure.

Unlike the ESRD PPS, which applies to all Medicare fee-for-service beneficiaries, many of the newer value-based payment initiatives are more limited in scope. Although early evaluations suggest the CEC model led to fewer hospitalizations in its first 2 years, only 12% of dialysis facilities participated in the model (2). Strict eligibility criteria specifying the minimum numbers of treated patients with CKD and kidney failure, and the requirement for physicians and dialysis facilities to voluntarily join, together create barriers to participation in many of the newer payment models. Other methods of introducing value-based payment into dialysis care are, therefore, necessary. In this issue of CJASN, Bakre et al. (3) examine whether Accountable Care Organizations (ACOs) can lead to cost savings in patients with kidney failure who receive long-term dialysis.

ACOs are voluntary, primary care–focused alternative payment models that can involve physicians, hospitals, and other health care providers. The ACO model offers financial incentives for participating entities to invest in care coordination and quality improvement. Total medical expenditures for aligned beneficiaries are compared with expected expenditures in a similar “benchmark” population. Participating providers are eligible to share cost savings when patients incur lower costs than expected. In two-sided ACOs, providers have the potential to benefit from more shared savings, but must assume financial risk if expenditures exceed expectations. In 2018, an estimated 10% of the US population was part of an ACO, including roughly 12 million individuals with Medicare (4). As the increasing role of ACOs in US health care results in a sizable proportion of dialysis beneficiaries being aligned to ACOs, they might represent another avenue for the broad application of value-based payment in long-term dialysis care.

The authors compared spending among Medicare fee-for-service beneficiaries who were aligned to ACOs in 2012 through 2015 with spending among beneficiaries who were not aligned to ACOs. During this period, there were 26,694 beneficiary-years with ACO alignment and 167,817 beneficiary-years without ACO alignment. Cost data for all Medicare fee-for-service beneficiaries in the 3 years before ACOs (2009–2011) were also used to examine baseline characteristics. A multivariable regression model accounted for observed differences in patient characteristics between the comparison groups. On October 1, 2015, the ESRD-specific CEC model started. Although ACO models have continued since 2015, the analysis focused on the period before implementation of the CEC model to avoid bias due to the transition of some patients from ACO to CEC models.

There are two main findings from this study. First, the number of ACO beneficiaries who received dialysis increased over time, and growth in the proportion of patients receiving long-term dialysis who were in an ACO mirrored growth in the proportion of all fee-for-service Medicare beneficiaries who were in an ACO. Even after some patients with kidney failure began transitioning from ACO to CEC models, 23% of long-term dialysis beneficiaries were aligned to ACOs in 2016, similar to 27% of general Medicare beneficiaries. This suggests payment adjustments for kidney failure in ACO models are adequate, and ACO providers do not avoid enrolling beneficiaries who receive long-term dialysis.

Second, in the fully adjusted regression model, spending on ACO-aligned beneficiaries was slightly less than spending for nonaligned beneficiaries ($143 less per beneficiary-quarter; 95% confidence interval, $5 to $282). Although the difference in costs from ACOs was statistically significant, cost savings attributed to ACOs were relatively small, representing only 1% of the reported average quarterly spending. Selected ACO characteristics that might be associated with savings or that are associated with savings in the general population—such as whether the ACO was physician led and whether nephrologists participated in the ACO—were not associated with savings, although the study may not have sufficient statistical power to detect these differences.

According to the authors, lower costs among ACO-aligned long-term dialysis beneficiaries indicate ACOs may successfully achieve cost savings in this population. They note how estimated savings in ACO-aligned dialysis recipients are higher in absolute terms than average savings observed across all ACO beneficiaries. They also suggest the observed magnitude of savings may be smaller than in more recent years. This is because the earlier ACOs examined in the study were primarily one-sided risk models, and savings might be higher in more recent years when increased adoption of two-sided risk models created stronger incentives to reduce costs.

Although ACOs might serve as an additional mechanism to apply value-based payment to long-term dialysis care, it remains unclear if ACOs currently achieve this objective. In additional analyses, Bakre et al. found that costs were only lower for dialysis recipients enrolled in ACOs when compared with costs incurred by non-ACO dialysis recipients who regularly saw primary care physicians (PCPs). Yet, having a PCP was independently associated with additional spending of $3000 per quarter, which far outweighs the $143 savings from ACOs. Expenditures for patients enrolled in ACOs were not lower when compared with patients on long-term dialysis who do not see PCPs regularly. Because many ACOs are structured around primary care, aligning patients to ACOs generally involves receiving care from a PCP. This study does not provide evidence of cost savings through alignment in ACOs of patients receiving dialysis who do not regularly see a PCP. Meanwhile, an analysis of dialysis expenditures in more recent ACOs that were both one- and two-sided risk models did not find substantial cost savings, suggesting the assumption of increased risk in two-sided risk models has not made ACOs more effective at lowering dialysis costs (2).

One reason why savings among ACO beneficiaries receiving long-term dialysis are small—or nonexistent—may be because many ACOs do not focus specifically on patients with kidney failure. ACOs are generally focused on improving primary care delivery. The relatively small proportion of patients receiving long-term dialysis who are enrolled in any given ACO may limit interest in investing in the infrastructure and devoting the resources necessary to effectively coordinate care in this population. In fact, a given dialysis provider may have patients enrolled in multiple ACOs, further complicating efforts to develop systems of care coordination that involve dialysis providers (5). The finding of limited cost savings in long-term dialysis beneficiaries is similar to findings in patients with cardiac problems, where participation in ACOs did not affect the rates of low- or high-value coronary revascularization (6).

In summary, the study by Bakre et al. highlights the potential for ACOs to become an influential mechanism for incorporating value-based payment into long-term dialysis care. At the same time, it suggests ACOs do not currently achieve this goal in a meaningful way. ACOs have been associated with modest savings in the general population (7). To achieve similar success in long-term dialysis, health care providers participating in ACOs will need to identify ways to incorporate dialysis care in their broader chronic-care management programs. Some health systems participating in ACOs have begun to consider ways to coordinate care in patients with kidney failure (8). If these programs are effective, future ACOs will need to adopt similar practices. Meanwhile, stopping the progression of CKD to kidney failure and promoting kidney transplantation, as envisioned by AAKH, may be the ultimate solution to addressing the high cost of dialysis care. Because ACOs care for many patients with advanced CKD, they are uniquely well positioned to reduce dialysis costs by optimizing advanced CKD care. Future studies will need to determine whether ACOs achieve dialysis cost savings through these alternative mechanisms.

Disclosures

K. Erickson reports providing consulting services for Acumen LLC; receiving honoraria from Dialysis Clinic, Inc., Satellite Healthcare, and University of Missouri; serving on the editorial boards of American Journal of Kidney Diseases, Seminars in Dialysis, and CJASN and as a member of the American Society of Nephrology Public Policy Board; and receiving research funding from Health Care Service Corporation’s Affordability Cures grant. The remaining author has nothing to disclose.

Funding

None.