A cornerstone of the Patient Protection and Affordable Care Act (ACA), the accountable care organization (ACO) model was expected to flourish under its implementation. Through the ACA’s Medicare Shared Savings Program (MSSP), ACOs were envisioned as vehicles for reducing the cost of care for Medicare, while improving the quality of care provided to Medicare patients. However, despite early interest and enthusiasm on the part of healthcare organizations and healthcare suppliers, ACO participation has declined in recent years.

Casual observers of healthcare policy might have wondered if this model of care was disappearing for good. It wasn’t. In fact, CMS is taking a bold look at how to reenergize the concept.

Writing in the New England Journal of Medicine earlier this year, Douglas Jacobs, M.D., M.P.H., chief transformation officer of the Centers for Medicare & Medicaid Services (CMS) and Purva Rawal, Ph.D., chief strategy officer at the CMS Innovation Center (Innovation Center) characterized ACOs as “essential to achieving the Centers for Medicare & Medicaid Services’ (CMS) goal of having all beneficiaries in the traditional Medicare program cared for by providers who are accountable for costs and quality of care by 2030.”

This renewed enthusiasm for ACOs as part of CMS’s larger strategy of value-based care —a decade after the implementation of the ACA—has taken some by surprise. But, in fact, while ACOs haven’t received much media attention in recent years, CMS never ceased running MSSP or testing new ACO model types through their Innovation Center.

What is an ACO

At its most basic, an ACO is a voluntary, but formal, alliance of hospitals, physicians, and/or other providers which assumes responsibility for the coordination and delivery of high-quality care for a defined population of a minimum of 5,000 Medicare Fee-for-Service (FFS) beneficiaries. The number of beneficiaries assigned to an ACO varies. In part, it is dependent on how an ACO is configured. The American Journal of Managed Care estimated that in 2018, ACOs led by hospital systems had an average enrollment of 44,000 beneficiaries, whereas physician-led systems had an average enrollment of 20,000.

An ACO must have a governing body with operating authority and a fiduciary duty. It may be self-managed or may contract with a management services organization for administrative oversight.

In contrast to the capitation payment model that characterizes the Medicare Advantage program, ACOs continue to be paid on a FFS basis. However, unlike the traditional FFS model, which has been characterized as incentivizing volume in the use of services, ACO reimbursement is structured to reward cost efficiencies. To incentivize controlled spending, ACOs are eligible to share in savings—and may be accountable for sharing losses—based on per capita spending relative to a specified benchmark.

But it isn’t just about cutting costs. To qualify for shared cost savings, an ACO must meet its program’s quality performance standards. These include improving the quality of care provided to Medicare beneficiaries across a set of quality measures.

While beneficiaries in Medicare Advantage programs are restricted to using providers within a defined network, beneficiaries in ACOs retain the freedom to see providers of their choice outside of their ACO network. Further, beneficiaries in ACOs are not self-selected, but are aligned with or “attributed to” individual ACOs based on Medicare claims data. Beneficiaries are advised of their enrollment status through a written notice of attribution—a notification that may be ignored by as many as it confuses. To help ACO beneficiaries understand the implications of their attribution—including the data sharing arrangements of ACOs—the Innovation Center offers a Beneficiary Engagement Toolkit.

CMS also requires that ACOs include either a beneficiary representative or consumer advocate on their boards.

ACO risk models

The most critical distinction in ACO models to date may have been the degree of risk assumed by individual ACOs.

In one-sided risk models, an ACO shares in a percentage of healthcare cost savings but are not penalized for exceeding cost-of-care benchmarks. In two-sided risk models ACOs are eligible to receive a larger percentage of any cost savings but, importantly, are also required to refund to Medicare a portion of payments when spending exceeds benchmarks. In short, they incur both “downside” and “upside” risk.

In a Final Rule published in the federal register at the end of 2018, CMS took a definitive step towards ultimately moving all ACOs into two-sided risk models. ACOs were given two options for participating in MSSP: through a “basic” track under which they start with one-sided risk and move to two-sided risk over time through a defined “glide path” or through an “enhanced” track under which they are eligible for a greater percentage of cost-sharing with the immediate assumption of two-sided risk.

MSSP’s BASIC track glide path allows participants to take on differing degrees of risk depending upon their participation at levels ranging from A through E. Under levels A and B,  providers share in a percentage of healthcare cost savings but are not penalized for exceeding cost-of-care benchmarks. Levels C through E include increasing levels of risk and rewards. ENHANCED E  has the highest level of risk and rewards for meeting benchmarks and the quality performance standard.

Performance to date

With approximately $1 billion allocated annually for testing alternative payment models, CMS and the Innovation Center have explored several models of ACOs in the past decade.

In its 2020 report to Congress, the Innovation Center listed 38 separate service delivery model initiatives operated under its authority in the two-year period from October 1, 2018, through September 30, 2020.

The greatest participation has been in MSSP, which is now 10 years old. With 483 current participant ACOs, serving over 11 million Medicare beneficiaries, the one-sided risk model has achieved high scores for quality and beneficiary satisfaction, but collectively failed to demonstrate significant cost savings.

ACO operating under the Innovation Center’s demonstration models have achieved mixed results. The Pioneer ACO Model resulted in fewer emergency department visits and inpatient admissions and the ACO Investment Model achieved $381.5 million in net Medicare savings in a two-year period. By contrast, the Advance Payment ACO Model and the Next Generation ACO Model failed to demonstrate cost efficiencies or improvements in quality.

Data analysis and evidence-based care

The ACO model emphasizes the use of technology for data management and analysis, reflecting the push for utilizing certified electronic health record technology (CEHRT) in both the ACA and the American Recovery and Reinvestment Act of 2009, (ARRA). For example, under MSSP, depending upon the level—A through E—at which it participates, an ACO must certify that either 50 percent or 75 percent of the eligible clinicians within the ACO utilize CEHRT to communicate clinical care to their patients or other health care providers.

HIT and CEHRT simplify the tracking of patient visits, conditions, and care which is key to achieving financial benchmarks and quality targets. To enable analysis of service use and facilitate coordination of care for such chronic conditions as heart disease, diabetes, and chronic obstructive pulmonary disease, CMS will share an attributed beneficiary’s claims data for visits outside the organization with an ACO at its request. Beneficiaries are advised that an ACO may request access to this information and are given the opportunity to decline.

However, technology can have high initial costs and present unique administrative burdens and there is evidence that these concerns shaped early participation in ACOs.

Shifting participation

Bolstered by their administrative staff, existing investments in HIT, and previous experience with risk-based payment models, hospitals and large health systems dominated the early years of ACOs. But recently physician-led networks are entering the space at a greater pace.

Some believe that the goal of minimizing hospital admissions among ACO beneficiaries better aligns with the interests of clinicians, whose bottom-lines are not directly impacted by reduced admissions. The American Academy of Family Practitioners goes so far as to argue that “to be successful an ACO must engage physicians in leadership roles to promote culture change.”

Concerns over the ACO models

The National Association of ACOs (NAACOS), the country’s largest association of Medicare ACOs presaged the recent decline in ACO participation in a 2016 white paper, ACOs at a Crossroads: Costs, Risk and MACRA. At the time the organization expressed concern that two-sided ACO risk models “set the bar much too high in terms of financial risk” and called upon CMS to account for the costs incurred in establishing ACOs.

In its June 2022 report to Congress, the Medicare Payment Advisory Commission (MedPAC) reiterated its 2021 call for a reduction in the number of ACO models as a means of “harmoniz(ing) Medicare’s portfolio of alternative payment models.” It has also cautioned against mandating participation in ACOs and called for the development of better and less confusing financial incentives.

Finding the “sweet spot” for incentives and risk remains the greatest challenge in successful implementation of ACOs. And, failing that, the Innovation Center may be leaning towards mandating participation, despite MedPAC’s express wishes to the contrary. In its Strategy Refresh white paper, the Innovation Center observed that voluntary participation has resulted in selection bias amongst ACOs, with organizations entering and exiting programs based upon projected losses and not committing to systems transformation. In response, the Innovation Center has stated an intention to examine “whether mandatory models can increase quality and access for beneficiaries.”

On the horizon

The latest entrant in the expanding list of ACO models is the ACO Realizing Equity, Access, and Community Health (REACH) Model—a redesign of the controversial Global and Professional Direct Contracting Model adopted under the Trump administration. The REACH model, which will be instituted in 2023, is significant for its incorporation of CMS’s expanding institutional focus on health equity. Participating plans will be required to identify and address health inequities within populations served.

An equity focus is also a key component of the Value in Health Care Act, introduced by Reps. Peter Welch (D-Vt.), Suzan DelBene (D-Wash.), Darin LaHood (R-Ill.), and Brad Wenstrup (R-Ohio). If passed, the legislation would increase shared savings under MSSP and assist ACOs with startup and operational costs. The bill, which has been endorsed by NAACOS, the American Hospital Association and the American Medical Association, also calls for the study of health inequities and racial disparities under its payment model.

A decade after their implementation, ACOs continue to hold their original promise of cost efficiencies and quality improvement. But the conversation about ACOs is now steeped in data that was lacking at the outset. Applied Policy will be watching closely as CMS leverages the ACO model and “lessons learned” towards its goal of high-quality, patient-centered care within a sustainable model for Medicare.