Across the country, policy experts are heralding accountable care organizations (ACOs) as the way to rethink the delivery of higher quality and more efficient care. Yet Medicaid, which cares for many of the nation’s sickest and highest-cost patients, has been largely absent from the ACO conversation. Now that the June 6 deadline for comments to CMS’ proposed Medicare ACO rule is behind us, it is time for the Center for Medicare and Medicaid Innovation (CMMI) to consider initiating a demonstration project to test safety-net ACOs, with a set of eight core considerations as its starting point.
By providing a stable structure for medical, behavioral health, and social service providers to build collaborative interventions, safety-net ACOs offer significant potential for improving care coordination and curbing spending for Medicaid’s highest-risk, highest-cost patients. This is particularly timely in preparation for 2014, when Medicaid opens its doors to an additional 16-20 million low-income adults, including many childless adults with multiple chronic illnesses, physical disabilities, and mental health conditions as well as an array of complicated social needs. In every state, a small proportion of this high-risk population uses a disproportionate amount of health care resources, often in repeat visits to emergency departments (ED) and preventable hospital stays.
In poor communities, the primary insurers are public — Medicaid and Medicare — making an all-payer safety-net ACO demonstration project easier to implement. The Camden Coalition of Healthcare Providers (CCHP), featured in Atul Gawande’s January 2011 New Yorker article “The Hot Spotters,” offers an innovative ACO prototype. In Camden, NJ, one of the nation’s poorest cities, one percent of patients account for 30 percent of medical costs. The leading causes of emergency department (ED) visits – head colds, ear and viral infections, sore throats, and asthma — are all primary care issues. Over five years, some $650 million was spent caring for this city of 79,000 people; in just one Camden apartment tower, 600 residents accrued over $12 million in hospital and ED expenses.
The CCHP effort focuses on this subset of patients who are in and out of emergency rooms and hospitals far too frequently. Together with faith-based community organizers, part of the PICO National Network, CCHP is uniting local health care providers, hospitals, social service agencies, and patients to build a population-based model that improves care and controls costs for these ED ‘super-utilizers.’
While CCHP has had preliminary successes and offers potential long-term savings, such community-based endeavors are difficult to initiate and sustain without start-up financing, ground-level technical assistance, and buy-in from state and local policymakers, health plans, patients, and community members. CMS could jump start investments in safety-net ACOs by proposing a national demonstration to bolster programs similar to the Camden pilot.
A safety-net ACO pilot could be loosely based on the Medicare Shared Savings Program, yet structural and population differences between the two programs warrant rethinking many of the underlying provisions of the Medicare ACO rule. For example, whereas a Medicare ACO would assign individual beneficiaries to hospital/health system providers, a safety-net ACO would necessitate a population-based, geographic focus encompassing both Medicaid and Medicare recipients. Safety-net patients, particularly the costliest and most complex, are often highly mobile and move from hospital to hospital, making a multi-hospital, community-wide approach more sensible.
Perhaps most importantly, while Medicare ACOs focus on reducing excessive use of specialty care and expensive procedures, Medicaid ACOs would need to address inadequate access to effective primary care and poor coordination of care, both of which lead to overuse of costly, preventable ED visits and hospital stays.
Eight core considerations can help shape a future Safety Net ACO Program:
- Provide upfront demonstration funding support. Under the proposed Medicare Shared Savings Program, no upfront funding is available (with the exception of the Pioneer ACO Model) for provider organizations to invest in the infrastructure and additional staff necessary to get an ACO off the ground. As opposed to the hospital/health system base of most Medicare ACOs, ACOs focused on safety-net populations are more likely to emerge from community-based providers without comparable start-up capital. Before getting started, ACOs must meet specific patient-centeredness requirements, including building a multidisciplinary team of primary care physicians, care managers, community health workers, health educators, and other staff. For many Medicaid providers operating with low reimbursement and limited reserves, it will be financially difficult, and in many cases prohibitive, to hire additional staff without a revenue stream to support them. As such, Medicaid providers that would otherwise be ideally suited to become ACOs may not opt into this new model.
- Help Medicaid providers meet information technology (IT) requirements. The Medicare proposed rule sets a high bar for the electronic prowess of ACO sites. Eligible ACO providers must be able to “electronically exchange summary of care information when patients transition to another provider or setting of care, both within and outside the ACO.” At least 50 percent of an ACO’s primary care providers must be “meaningful electronic health record users” by year two of the program. Yet many Medicaid providers – including one- and two-physician practices and small rural health centers – still use paper-based systems and have scant resources to invest in IT upgrades. Although funding from the HITECH Medicaid Incentive program and technical support from the Regional Extension Centers have helped, additional time and funding would be necessary for Medicaid ACOs to meet enhanced IT demands.
- Include federally qualified health centers (FQHCs) and rural health centers (RHCs). FQHCs and RHCs form the backbone of care delivery for a large proportion of Medicaid beneficiaries, yet the Medicare-proposed ACO rule excludes these critical providers from serving as ACOs because limitations in Medicare data collection prevent provider-level attribution for providers working in FQHCs. While only a small percentage of Medicare beneficiaries receive care in these community centers, they are often the costliest dual eligible patients. A safety-net ACO demonstration project could use local data collection to address the limitations of Medicare’s current FQHC/RHC data set.
- Support ACOs in multiple delivery system models. Across states, there are broad variations in how Medicaid services are delivered and financed, with many states using some type of risk-based managed care system. Today, roughly 70% of Medicaid beneficiaries are in managed systems of care, and states will want to enroll most of the expansion population into managed care in 2014. A safety-net ACO program would need to work seamlessly with managed care organizations, which are typically the primary delivery system in the urban areas most conducive to population-based approaches for serving low-income patients.
- Support technical assistance and collaborative learning. Financial incentives alone will not enable safety-net health care providers to establish ACO models of patient-centered, lower-cost care. Practices serving high-need Medicaid populations will likely need intensive technical assistance to support population-based management, including EHR implementation, care process redesign, and cultural shifts to sustain continuous quality improvement. Peer-to-peer exchange through a “learning collaborative” network as well as onsite support will be critical to help time- and resource-strapped Medicaid practices adopt ACO tenets of care.
- Engage patients and community organizations. Community members and patients who are desperate for a more coordinated health care system can be key catalysts for reform. As such, community participation is vital to the successful start-up of safety-net focused ACOs. Although the Medicare proposed rule requires one beneficiary on the ACO governing board, Medicaid ACOs would need to engage both beneficiaries and community organizations to represent and meaningfully engage the broader community.
- Extend expectations for return on investment. Because Medicaid providers typically do not have the levels of IT infrastructure, staff, and economies of scale necessary to quickly transition to an ACO organizational structure, a three-year program akin to the proposed Medicare ACO pilot may not be sufficient. A five-year demonstration program to establish budget neutrality, with upfront grants and/or alternative payment methods that allow shared-savings across Medicare and Medicaid, may be more appropriate. Additionally, it would be disastrous in low-income communities if a safety-net ACO caused financial distress and failure of a crucial safety-net hospital. To encourage hospital participation, Disproportionate Share Hospital and Graduate Medical Education funding could be used as transitional sources to cushion excessive reductions in safety-net hospital reimbursements. These funding sources would have to be retooled in a transparent and accountable way.
- Use states as neutral intermediaries. Safety-net ACOs can take advantage of their state’s position as a neutral convener to mitigate antitrust concerns. States can also ask for safe harbors from federal laws like Stark, anti-kickback, and civil monetary penalty.
ACOs have gotten ample attention in Medicare; now, equal focus and support should be given to the development of safety-net ACOs. Addressing the aforementioned core issues would provide a solid foundation for the support of safety-net ACO start-ups. We encourage CMS to take initial steps to create opportunities for ACOs to flourish in Camden and other low-income communities across the country with peak concentrations of the nation’s highest-need, highest-cost patients.