Leveraging Medicaid in a Multi-Payer Medical Home Program: Spotlight on Rhode Island's Chronic Care Sustainability Initiative
Key Themes
The common contract is central to the multi-payer model. The common developmental contract was not easily developed, but its use is central to the CSI. The consistency across all payers of a core set of required metrics, and fixed supplemental payments to practices tied to performance expectations, align the incentives and goals that practices face, and encourage practices to pool the supplemental revenues from all their payers to invest in improvements that benefit all their patients. Interviewees representing practices favored having common metrics over different metrics for four or five different payers, and they pointed out that common metrics also reduce overhead costs for practices.
The common requirements were developed through a consensus process involving participating stakeholders (largely plans and providers). Plans had to be willing to modify their existing contract requirements, at least to some extent. For example, a plan might have to give up “home-grown” requirements of its own in favor of the common contract’s emphasis on NCQA accreditation. Similarly, national firms with common requirements across their health plans had to be willing to make exceptions for a plan operating in Rhode Island if the firm’s and the CSI’s requirements differed. Because of anti-trust laws that limit health plans’ ability to discuss pricing with each other, the deliberations that produced consensus on certain requirements of the common contract would not have been feasible except for the health insurance commissioner’s central convening authority. Specifically, the common developmental contract requires health plans to make a uniform contribution to practices to finance infrastructure-building. Such payments are legally considered to be pricing-related, but the discussion could take place as long as it was convened by OHIC.
Medicaid participation in a multi-payer system presents opportunities and challenges.
Leverage and experience. Because Medicaid is such a large purchaser in Rhode Island – it accounts for a substantial share of health care costs in the state – the ability to leverage the program is integral to the CSI’s impact. Although the majority of adults in Rhode Island are commercially insured, Medicaid is the dominant source of coverage in some CSI practices and covers a substantial share of patients in others. As Medicaid plan officials see it, an important gain from the CSI is that, with all payers at the table, Medicaid and commercial plans can learn from each another. Small plans get what one interviewee described as “a brain trust of medical directors committed to improving quality and re-engineering change in office-based care.” CSI principals say that the initiative has benefited significantly from the experience of plans and safety-net providers serving Medicaid patients, as they have been leaders in quality improvement and have a greater understanding of what it takes to coordinate care, often among many providers, for populations with complex needs for health services and other supports. The CSI provides a framework and environment for consultation and sharing of knowledge and ideas, and helps to avoid the problem of “reengineering a system that works for only one-third of [a] health plan’s members.”
Patient mix differences and fair payment. CSI plans and practices with a sizeable share of Medicaid patients face some unique challenges. Performance metrics are not currently risk-adjusted, possibly because of delays in data acquisition and a lack of resources and expertise to develop a methodology. The RIQI is now taking on the methodological work for risk-adjustment through a subcontract with Research Triangle Institute. Compared with the privately insured, Medicaid patients are more likely to have mental health co-morbidities as well as socioeconomic hardships that contribute to poorer health outcomes. In addition, a substantial share of the patients served by community health centers participating in the CSI are uninsured, making management of their care more challenging and performance benchmarks harder to achieve. (Health centers generally include the uninsured in their performance metrics because they operate on a philosophy that all patients are treated the same and that performance metrics need to be practice-wide, not payer-specific.) Recognizing the potentially higher risk profile of Medicaid and uninsured patients, and seeking to be fair to practices, the CSI established a “half-way to the goal line” approach that rewards practices with incentive payments if they get half-way to the performance benchmark relative to where they started. Stakeholders speaking from a Medicaid perspective praised this measure as a constructive compromise that rewards practices and providers in the safety-net space for their progress toward CSI goals. They expressed concern that plans not view the practices to which this policy applies as “slackers,” but highlighted that the consensus-oriented governance of the CSI provides a structure for discussing and resolving issues that have different implications for different stakeholders.
Combining a collaborative model with a mandate is useful.
Consensus helps legitimate authority. Getting multiple payers to work together on a unified strategy requires a lot of buy-in. Having the CSI pilot begin on a voluntary basis almost certainly helped foster willingness to participate. But even the voluntary initiative might well have been harder to achieve without Rhode Island’s long history of collaboration and what stakeholders describe as the talented leadership of the health insurance commissioner. The health plan representatives who were interviewed do not like the requirement to participate, but acknowledged that the mandate was probably valuable to the effort to extend the chronic care initiative to all adults statewide. By pairing a mandate for commercial insurer participation with a consensus-oriented approach to governance, the state is able to exert leverage that is needed to direct health plans to do some things differently and to spend money in ways that, except for the primary care spend requirement, they would not. It is likely that one reason private insurers have accepted the mandate is that, while their participation is required, they are also part of the consensus process that determines the specific requirements of the initiative.
The locus of authority determines its reach. Because the mandate to participate in the CSI operates through the OHIC, it does not reach all plans or purchasers in the state. The OHIC’s authority covers commercial insurers, but does not extend to self-insured employers or to public purchasers. Although the legislative mandate does not apply to the Medicaid plan, this plan has participated voluntarily in the CSI from the beginning, and, since 2012, the state has required all health plans with Medicaid contracts to pay the supplemental care management fees to practices on behalf of their Medicaid as well as their commercial enrollees, if the practice serves at least 200 of their Medicaid members. The CSI mandate does not apply to the Medicare program, but Medicare participates under the aegis of the Multi-payer Advanced Primary Care Practice (MAPCP) demonstration.
Collaboration requires time and effort and leadership.
Stakeholder commitment. The development and operation of a governance process that is based on consensus decision-making require heavy investments of time by many parties, and a strong commitment to work together to hammer out common goals and objectives and implement the initiative. Ongoing operation of the CSI succeeds because its key stakeholders — plans, practices, and purchasers — continually commit significant time, focus, and effort to the enterprise. To illustrate, the administrator of one participating family medicine practice with eight providers reported that she routinely attends five different CSI-related meetings monthly, as well as ad hoc meetings. Collectively, two or three other individuals in the practice attend another five to six meetings each month on data, reporting, care management, and other topics.
Engaged leadership. Leadership from all the stakeholder groups has been critical to the CSI’s success. Even in a state with a collaborative tradition, there exist competing interests, gaps in trust, and other sources of tension between parties that must be addressed. Individual stakeholders with different perspectives must, through dialogue and other interaction, gain confidence in the other participants and the fairness of the project to perceive an alignment between their own goals and interests and the initiative’s aims. The CSI management team highlighted plan and provider engagement in the initiative as a key to building increased trust between these two groups. Conversations facilitated by the learning collaboratives helped to give plans confidence in providers’ commitment to practice transformation and the use of metrics to drive change, and to give providers confidence that plans are paying them fairly and giving them the data they need to improve care management.
Broad-scale initiatives and practice transformation both require infrastructure.
Overhead costs. Convening meetings, collecting and reporting data centrally for aggregation across payers, and providing assistance to help practices change, are all overhead costs that require funding. Perhaps because the CSI started out as a small pilot, such costs were underestimated initially. Also, because administrative and staff support for the initiative was limited early on, development of a formal organizational structure (e.g., by-laws) was delayed. Funding constraints also led the CSI to rely on different financing approaches for learning collaboratives at different times, rather than on an established and stable means of support.
The timing of the Beacon grant to Rhode Island was fortunate. The overlap between the needs of the CSI and the purposes of the grant – to link development of EHRs to clinical practice improvement and enhanced outcomes – gave the state access early on to substantial support to build infrastructure. However, such grants are not routine and they are typically time-limited. With the Beacon grant’s termination, the CSI health plans are providing additional funding for infrastructure (over and above their care management payments to practices) for at least one year, but there is concern about their ability and willingness to maintain this support over time, particularly if the amount required of them grows. Some expressed the view that, because more work on the data infrastructure was not accomplished earlier, progress was delayed and opportunities were lost, adding to health plans’ and practices’ current costs. For example, if the infrastructure for the exchange of clinical data had been more developed, CSI would not have to invest as much now in claims-based “work-arounds” to analyze costs across all payers.
Costs of practice-level infrastructure-building. Like other efforts to develop patient-centered models of primary care, the CSI generates new needs for spending – to manage the multi-payer effort, collect data, create metrics and reporting processes, and support practice transformation. Primary care practices need to set up systems to track data relevant to the metrics they must report and integrate these metrics into their performance improvement programs. The supplemental payments to participating practices provide critical additional funding to support other required practice changes as well, such as expanded use of nurse care managers who can spend more clinical time with patients. One practice manager reported that the supplemental payments have allowed her to fund a nurse care manager, hire a full-time quality assistant who can do “deeper dives” into the data (e.g., to understand outliers), and keep the practice open longer hours. Some practices had already made some of the needed investments before the supplemental payments became available, but even for these practices that are farther along, the care management fees help finance intensified efforts and additional staff time to participate in learning collaboratives with other practices.
At least some of the CSI practices are finding the supplemental payments insufficient to cover all the costs involved in participating. They explained that they continue to participate for the good of their patients, whom they see getting better care because of the practice changes spurred by the initiative. Consensus is lacking on what level of support participating practices require. Sponsors of similar initiatives elsewhere can also anticipate debate about the appropriate levels of payment to support the practice-level infrastructure needed to support patient-centered care.
Maintaining payer support for CSI may require evidence of savings.
Those we interviewed noted that the health care environment is very fluid in Rhode Island, as it is elsewhere in the nation. The practices that participated in the CSI from the outset were early adopters of change who were philosophically committed to its success; the practices joining now may not be as committed to doing what is necessary internally to develop clinical data and use them to improve performance, although they should be able to benefit from the experience of the established practices. Health plans say that their willingness to provide practices with additional support ultimately depends on a demonstrated return on investment. In today’s payment reform environment, they say, they need evidence that investing in primary care is not only improving patient care and patient experience, but also generating savings somewhere in the system. Some payers wondered aloud whether practices should have more “skin in the game” – that is, whether their payment should be more strongly tied to their performance on clinical process and outcome metrics. It could be that the ability to maintain support for these kinds of initiatives will require settling the return-on-investment question, challenging though that may be.
Practice transformation is the beginning, not the end, of system change.
There was agreement that, ultimately, performance should be measured not just at the level of individual practices, but also at the system level, using metrics like aggregate ED, hospitalization, and readmission rates. Clearly, hospitals are integral to true systemic change. At this writing, hospitals were not yet at the CSI table, and stakeholders reported that they were still relatively traditional, with a primary focus on filling beds; they added that ACO development and hospital acquisition of practices remained limited in the state. As a consequence, the CSI has been focused mainly on what goes on in primary care practices. More recently, however, the state says it has been engaging hospitals, and has added performance measures on hospitalization and ED use that are tied to payment. One interviewee placed the CSI in the broader context of system transformation this way: “I think we’re working towards a system that is more integrated and will address the whole triple aim. Has that been proved yet? I don’t know. I think we’re still walking down the road. We certainly can’t continue in this [current] system. I think this (medical home model) has more optimism around it than other types of approaches. But…we’re not there yet. This is new territory.”