Rising healthcare costs and declining reimbursements have placed immense pressure on providers and health systems to manage the health and costs of the populations they serve more effectively. These pressures coincided with the advent of the Medicare Shared Savings Program (MSSP) in the Affordable Care Act, which couples population health approaches with financial incentives to support the transition from volume to value.
The MSSP was met with a fair degree of enthusiasm and expectations (as well as some cynicism) when it launched in 2012. Some anticipated that the MSSP, along with the Pioneer ACO program, could save $1.5 billion nationwide over three years.
Manatt has supported more than a dozen MSSP applicants, ranging from independent practice associations (IPAs) to large health systems. Our experience has given us a unique perspective on the characteristics and development approaches of the ACOs participating in the program.
An Open and Transparent Program Design
The diversity of participating organizations reflects the MSSP’s open and transparent design, which is somewhat unique among federal programs. The MSSP established an annual open application period during which any healthcare organization that serves Medicare fee-for-service (FFS) beneficiaries and meets a basic set of requirements may apply. Those requirements are uniform regardless of the type or size of the applicant.
To be eligible, ACOs must serve at least 5,000 beneficiaries, report on 33 clinical quality measures, and meet cost benchmarks established by the Centers for Medicare and Medicaid Services (CMS). In exchange for meeting these and a host of other clinical, financial and organizational requirements, ACOs have the opportunity to realize shared savings if they outperform cost targets established by CMS and, in some cases, realize losses if they don’t.
Diverse Participating Organizations and Populations
The MSSP has generated significantly more interest than was first anticipated. More than 360 Medicare ACOs already have been established since the program’s launch. Among the MSSP’s participants are IPAs, academic medical centers (AMCs), regional hospitals, multistate delivery systems, and federations of all of the above. Each organization’s patient population reflects differing demographics, disease burdens, severity of illness and utilization patterns.
For example, one AMC was managing a population with more comorbidities and complex conditions than a community hospital located in a retirement community. As a result, the approaches that both institutions take to manage care should not necessarily be the same. The AMC needed more specialist and specialized care, while the community hospital focused more on primary care.
The ACOs that Manatt supported also showed significant differences in their clinical and information technology (IT) infrastructures, as well as their experience managing risk. One participating entity had no existing care coordination or care management infrastructure, requiring it to consider significant investments to support the ACO’s launch. Another already had a population health management system in place and National Committee for Quality Assurance (NCQA) certified patient-centered medical homes within a number of its primary care clinics, leaving the ACO better prepared for managing patient risk. Despite the differences between these two organizations, both applied for and were accepted into the MSSP. While the MSSP established a level playing field, it is decidedly uneven with respect to each organization’s capabilities and patient population profiles.
ACO Implementation Approaches Vary – A Lot
Given the wide variations among entities, it is impractical to create a single strategy for successful MSSP participation. Each organization must develop its own plan that takes into account both the specific characteristics of its patient population and its existing capabilities. Since an organization’s clinical and IT capabilities have implications for how effectively the ACO can manage care and achieve savings, it is critical that organizations engage in forecasting and assessments, as they develop their ACO implementation strategies:
- Patient Population Characteristics. The CMS beneficiary attribution methodology assigns members to an ACO retrospectively by identifying where patients in the ACO’s geographic region receive the plurality of their primary care services. ACOs can use forecasting to anticipate the size and characteristics of the attributed beneficiary population, while also identifying hot spots and areas for potential savings. While beneficiary attribution is primarily driven by primary care services, Manatt’s forecasting experience suggests that it also may be driven by specialty care in certain settings. In one case, an AMC had twice as many attributed beneficiaries from specialists than a community hospital and 60 percent more than an IPA-based ACO. For some ACOs, higher specialty care utilization may require additional investments in care coordination for high-risk populations (e.g., specialist-based medical home models). Forecasting also may be used to predict with greater accuracy whether the ACO will meet the CMS minimum beneficiary requirement, which informs whether applicants should conduct further provider recruitment. Forecasting the expected attributed beneficiaries also provides insight into the ACO’s IT infrastructure needs to support more complex populations. For example, it can reveal whether the ACO needs more sophisticated clinical decision support and population health management tools.
- Existing capabilities. Manatt’s experience in supporting MSSP applicants with gap assessments also uncovered significant differences in existing capabilities among different types of organizations. For instance, while some were more advanced than others, many federated provider groups, such as IPAs, did not have robust IT or care coordination infrastructures deployed across groups. Although IPA-based ACOs were typically developed through a physician-led governance model, they outsourced IT functions and relied on a “pod”-based care management model to coordinate care across practices. Since the IPAs lacked the kind of capital reserves available to hospital systems, they looked to national and regional payer partners to help support the ACOs’ implementation. These relationships do not preclude the ACOs from entering into contracts with other payers. In fact, there may be benefits to investors for the ACO to have multiple partners, if they can share in downstream savings. Larger institutions, such as AMCs and multistate delivery systems had more robust clinical and IT infrastructures and could centralize services to support their patient populations. However, these infrastructures often were fragmented across inpatient and ambulatory care settings. Manatt’s assessments revealed differing strategies and protocols across inpatient, ambulatory and post-acute care settings that need to be resolved. The MSSP pressed these ACOs to address fragmentation issues, reducing the likelihood of gaps in care as patients transition across settings.
According to the National Association of ACOs, the 114 MSSP ACOs that joined the program in 2012 have invested more than $400 million to build and operate their ACOs. However, as noted previously, the overall level of investment for each ACO varies based on its existing capabilities. Manatt’s business planning experience with MSSP applicants shows that ACO investments typically range from $1.3 million per year to $6 million per year, depending on an organization’s existing infrastructure and scale. These variations in costs underscore the need for accurate assessment and forecasting of needed resources.
Preliminary Program Results
Given that the existing capabilities, patient populations and ACO implementation approaches of MSSP participants varied so dramatically, it should come as no surprise that their performance also showed significant differences. In January 2014, CMS released preliminary performance results for the 114 MSSP ACOs that began operation in 2012.
Within this initial participant cohort, 27 of the 110 Track 1 participants, which are not liable for downside risk, generated shared savings. Of the four Track 2 participants, which are liable for downside risk if spending exceeds expectations, two generated shared savings and two generated shared losses. While only a small portion of the initial program participants generated shared savings during the first year, more than half of the 114 participating ACOs had lower than projected expenditures, generating $128 million in savings for the Medicare trust fund.
Following the release of the initial shared savings results, CMS released data in February on 5 of the 33 quality measures used to evaluate the clinical care that ACOs provide. The five released measures—which evaluate ACOs’ results with respect to patients with diabetes and coronary heart disease—provide additional insight into the performance of the initial participant cohort.
Similar to the shared savings results, the data indicate that performance varied across ACOs. For example, on the measure that assessed the percentage of diabetic ACO patients with documented daily aspirin use, one ACO achieved 81 percent while another achieved only 7 percent. ACO quality performance also varied dramatically across measures assessing blood pressure, tobacco use and hemoglobin control.
The 2015 MSSP application cycle will begin in June and conclude in the fall, when CMS selects new program participants. While the 2015 cycle will likely not be affected, it is anticipated that CMS will release additional guidance for the MSSP program sometime this fall. Significant regulatory changes could potentially have an impact on how future ACOs are measured, as well as how they receive shared savings. Manatt will continue to monitor changes at the federal level that are of interest to current and prospective MSSP participants.
As current ACOs further develop care coordination and population health management capabilities, it is likely that some organizations will use the MSSP as a stepping stone toward assuming more risk for their beneficiary populations (i.e., Medicare Advantage and commercial HMOs). By waiving certain fraud and abuse requirements, the MSSP provides a pathway for healthcare organizations to be deemed clinically integrated by the Federal Trade Commission, adding a degree of protection from federal antitrust and antikickback rules.
While ACOs are still subject to state law, some are leveraging the MSSP waiver to engage in contracts with commercial payers. This approach allows ACOs to spread their clinical and IT infrastructure investments over commercial and public payer patient populations. As more organizations leverage this approach, it is likely that the spread of Medicare and commercial ACO arrangements will continue.
The preliminary results from the MSSP’s first year provide some insight into the varied performance of the initial MSSP ACOs. CMS intends to release final results later this year, which may shed further light on early experiences with the program. Given that ACOs in the initial cohort varied significantly in their existing capabilities and patient populations, it is noteworthy that some were able to outperform their cost projections and achieve shared savings. Time will tell whether other MSSP participants will be able to improve care while driving unnecessary costs out of the system. Their continued investment and participation in the program, however, suggest a longer-term commitment to adopting a care model that moves them away from traditional FFS medicine. That in itself may be the most significant achievement of the program to date.