Several approaches beyond the procompetitive strategies discussed above could be used to constrain prices. Hands-off strategies include requiring providers to publicly justify high prices or having the state discourage, but not prohibit, high prices. Providers could be publicly identified as expensive and possibly shamed into restraining their prices. It also may be possible to disadvantage high-price providers in any dealings they have with the state.
Alternative approaches involve more direct forms of regulation. One approach is to set prices. Maryland used this approach for decades in conjunction with other strategies; more recently, the state has moved toward more of a global budget model (see box). Price setting can be difficult, because incentivizing the optimal allocation of services requires setting thousands of prices relative to one another. By setting prices as a function of an external benchmark, such as Medicare prices, the process can be simplified; this approach also can help deal with the introduction of new services.
Yet, even in this model, regulators must worry about differences across providers. For example, should the prices at academic medical centers be the same as at community hospitals? Inevitably, if rates are set to be the same across all providers, many prices will rise in many cases, diminishing any savings. If they vary by provider, analytical challenges and equity issues will arise. In addition, because rate setting supplants market dynamics, it limits the ability of competition to drive efficiencies or better quality. For example, providers generally cannot increase their prices if they improve quality.
Maryland Evolves from Price Setting to Global Budgeting to a Total Cost of Care Model
In 1971, Maryland established the Health Services Cost Review Commission (HSCRC), an independent agency that set hospital rates for payers in 1974 with the following goals:
- Constraining hospital cost growth
- Ensuring hospital solvency
- Promoting equity in hospital financing.
Later, in 1977, Maryland received a waiver from the Centers for Medicare and Medicaid Services that allows the state to pay HSCRC-approved rates, which were set prospectively, to both Medicare and Medicaid. This waiver gave the state significant regulatory authority over health care spending.
Since then, Maryland’s all-payer system has continued to evolve and HSCRC’s scope has widened. In 2014, the state added global hospital budgets in response to concerns about excessive utilization of care. Budgets have generally been set based on a hospital’s historical revenue trended forward, with factors such as inflation, quality, and market area changes being taken into consideration. Adjustments could be requested, but penalties are applied to subsequence budgets if the revenue varies more than the allowed percentage over a year.
When global budgets were launched, Maryland also started monitoring, but not regulating, nonhospital cost growth. Then, in 2019, the state began piloting the Maryland Total Cost of Care Model to hold the state at risk for total Medicare costs. It includes three parts:
- The Hospital Payment Program to test population-based payments for Maryland hospitals
- The Care Redesign Program that allows hospitals to make incentive payments to nonhospital health care providers if it has attained savings under its fixed global budget
- The Maryland Primary Care Program that is structured to incentivize primary care providers to offer advanced services and care management to their patients.
Through legislative mandate, all acute-care hospitals in Maryland are required to submit confidential patient-level administrative data on discharges and visits to HSCRC. This includes demographic, financial, and clinical information. These data are then managed by the Chesapeake Regional Information System.
The HSCRC itself is composed of members that have been appointed by the governor. They represent independent experts, payers, providers, and consumers, including those under the commission’s purview.
A more modest approach is to cap prices. Defining price caps as a function of Medicare rates is likely the simplest approach. Caps that reflect a multiple of relatively low rates nationally, or in the state or local market, are more responsive to market conditions or changes in delivery costs and less dependent on changes in Medicare policy. This allows for growth of fees to match more closely that of input costs, such as nurses’ wages or medical supplies, as opposed to scheduled Medicare fee increases, which are set below input cost growth. This concern is particularly salient if fee caps are extended to physician services because current Medicare policy has virtually no fee increases scheduled for physicians in nominal terms for many years.
From a state perspective, the advantages of using Medicare rates likely outweigh the disadvantages. However, from a federal perspective, Medicare policy would be more complicated if all rates were based on Medicare fees. That is because the stakes associated with changes in Medicare prices would be higher, placing more pressure on Congress any time prices change.
Typically, the price caps that have been discussed by states have been tied to a relatively low percentage of Medicare rates.13 However, higher, or more generous, caps have been proposed that would minimize hospital closures, potential job losses, and other market disruptions and could be lowered gradually as outcomes are monitored.14 Another option is a limited price cap. In 2016, Montana capped hospital prices for state employees. The State of Montana Benefit Plan limited both inpatient and outpatient hospital prices to, on average, 234 percent of Medicare.15 The policy has reduced disparities between the least and most expensive hospitals in prices charged to state employees by 28 percentage points.16 Because the caps were limited to state employees, its impact on providers is less than an analogous cap for all commercial coverage.
A related strategy is to cap price growth. This can supplement price caps with less disruption. Focusing on price growth may be easier and face less political opposition but can lock in existing inefficiencies or price differentials across providers (unless price convergence is built in). In 2010, as part of a broader effort to improve affordability, Rhode Island implemented inflation caps that were equal to the Medicare price index plus 1 percentage point for both inpatient and outpatient services.17 These caps, along with a switch from per diem to diagnosis-related group (DRG) payments were largely responsible for the state’s reduction in fee-for-service spending. In addition, price growth caps gave insurers greater leverage when negotiating with providers, suggesting that price growth caps may be more effective in areas where providers have disproportionate market power.18 Still, it is also worth noting that caps on price or price growth do not directly influence utilization, potentially limiting the impact of these policies on total spending.
Commission Support of Price Reduction Policy
Addressing high prices is a narrower charge than addressing overall health care spending because prices reflect a single transaction that policymakers could directly regulate. In practice, however, price regulation raises a number of practical problems that commissions may help address. For one, the commissions could develop and implement soft strategies to control prices, such as holding hearings to shed light on high prices or requiring providers to publicly justify their prices.
State commissions also can play an important role if regulatory strategies are used. For example, if states decide to set, as opposed to cap, prices, commissions (like the one in Maryland) could help determine the appropriate price. While this may not be needed if prices are based on a multiple of Medicare or Medicaid rates, in many cases some variation across facilities may be important because of variation in markets or quality. Commissions can perform or support the related analysis and even be granted authority to set the prices.
Commissions also may play a similar role if caps are placed on prices or price growth, particularly given the logistical challenges associated with any price regulation (setting or capping). Specifically, health care services are paid for in a variety of ways. For example, hospitals may be reimbursed on a per diem or DRG basis. States focused on price regulation must be able handle the various payment structures used (or require uniformity in pricing systems, which may lock in existing, suboptimal models). States also must recognize the ways providers and insurers can circumvent price regulations. For example, quality-based payment models could be altered in ways that are designed not to promote quality but rather to circumvent price regulations, for example, by lowering the performance needed to meet quality targets.
Commissions can help monitor the effects of price regulations and adapt regulations accordingly to adjust for payments occurring outside of traditional current procedural terminology (CPT) or DRG bases. Since this sort of analysis requires greater expertise and resources, commissions may be valuable in helping states navigate potential policy options and find the tools that work best for them. Additionally, they may help reduce prices, to some degree, by facilitating lower administrative costs. This could be done by standardizing billing and quality metrics or by streamlining processes for reporting data to states. Indiana, for example, operates a health information exchange that enables data to be more easily shared among hospitals, payers, and public health officials.19 Still, there is little evidence such initiatives will significantly reduce spending or spending growth.
Reducing Utilization of Low-Value Care
Discussions surrounding health care spending often focus on reducing the use of low-value care. These are services that offer patients with certain clinical presentations no benefit, or benefit less than cost, leading to unnecessary spending and potentially even patient harm. This is important given that overtreatment is estimated to cost between $158 billion and $226 billion each year.20 Eliminating even a fraction of that spending has the potential to yield substantial savings without significantly impacting quality or outcomes.
Identifying low-value care is difficult. The utility of a clinical service ultimately depends on the context in which it is provided. For example, imaging is often unnecessary for headaches and is, therefore, of low value in many cases. Yet for a patient who has risk factors or symptoms that may signal a more serious underlying etiology, imaging is of much greater value. This distinction is often difficult to make with administrative data, which can lack clinical detail and nuance.
Despite these measurement challenges, initiatives such as Choosing Wisely, implemented by the American Board of Internal Medicine Foundation, have created lists of typically low-value care and private vendors sell software to measure how often these services are provided. These lists have been used to analyze Medicare claims and develop measures of low-value services across a variety of categories, such as imaging and surgical procedures.21 In one study, the most sensitive versions of these measures found that 42 percent of patients received low-value care, totaling 2.7 percent of overall annual spending. More specific versions of these measures, however, found that 25 percent of patients received low-value care, totaling 0.6 percent of overall spending.22 That difference suggests that the impact of any policy targeting low-value care will depend significantly on the measures being used.
Examples of Low-Value Care
- Unnecessary diagnostic testing prior to low-risk surgery
- Vitamin D screening
- Prostate-specific antigen screening in men age 75 and older
- Imaging during the first six weeks of low back pain
- Branded drugs when identical generics are available
Efforts to measure and raise awareness about low-value care, often driven by state data agencies, have been growing. For example, states such as Colorado, Maine, Virginia, and Washington have been able to use all-payer claims databases to track spending on potentially wasteful services. In Colorado, the Center for Improving Value in Health Care found that 13 services accounted for 81 percent of all low-value spending, with eight of those services having the potential to harm patients.23 The Virginia Center for Health Innovation (VCHI) identified more than 2 million low-value services that contributed to almost $750 million in spending, such as baseline laboratory studies in patients without systemic disease undergoing low-risk surgery.24
Efforts to eliminate low-value care focus on education or incentives. For example, with regard to education, Colorado plans to generate and share provider-specific data to encourage improvement at the local level.25 In addition, VCHI created the Smarter Virginia Health Care Coalition and is working with the state’s major health systems to reduce the use of low-value services. This involves training and educating physicians and developing strategies, in conjunction with large health systems, to reduce low-value care.
With regard to incentives, payment models, such as accountable care organizations (ACOs), reward providers for providing high-value care and reducing unnecessary spending. While not targeted exclusively at low-value care, these efforts have been shown to reduce the utilization of low-value services. Value-based insurance design (VBID) plans, which typically incentivize patients to choose services whose clinical benefits exceed the costs or services that are high-value, sometimes also actively discourage the use of low-value care.
Commission Support for Reduction in Low-Value Care
Commissions may be useful in supporting efforts to reduce low-value care. They can help interpret and publicize state measurement activities. They can develop priorities regarding which low-value care services to focus on. In some cases, they may develop more aggressive strategies, or even penalties, designed to discourage use of low-value care.
Address Spending Holistically
Holistic strategies are designed to influence spending overall, as opposed to particular components such as prices or utilization. The broadest holistic strategies target total cost of care or premiums, but more limited versions may target episodes of care or spending from specific provider types, such as inpatient hospitals. These approaches typically assign individuals to accountable entities, such as insurers or provider systems, and hold those entities accountable, to varying degrees, for spending. Examples of these strategies would include premium regulation and encouraging alternative payment models or global hospital budgets (particularly those that extend responsibility for care delivered outside of the hospital).
Soft holistic approaches could involve monitoring and reporting spending or encouraging policies that promote alternative payment models. For example, states could use alternative payment models for Medicaid or state employees, which could induce participation in such models in the commercial sector. In other cases, states could use soft sanctions if spending targets are exceeded. For example, in Massachusetts, the state attributes patients to providers and can require providers to file improvement plans if spending exceeds the target. In these approaches, provider systems are accountable for total spending, which makes them responsible for care delivered by organizations outside of their system. While provider systems may be able to influence that care, or which providers their patients visit, it may be problematic to severely penalize providers for actions that they may have only limited control over.
Other states have adopted stronger approaches. For example, Arkansas previously used a multipayer, episode-based payment model. This had been facilitated by a collaboration between the state’s Medicaid program and the state’s largest insurer, which had a dominant market share. However, the state’s program was limited by the number of conditions suitable for episode payment — those with clear episode definitions and sufficient volume to adequately adjust for risk differences. Their experience suggests that episode-based payment may be more useful for a limited and targeted set of conditions. Maryland has been successful with a broader hospital budgeting model. Though the state’s model was initially limited to price regulation, it evolved to a hospital global budgeting model and has recently expanded to include some responsibility for total costs of care.
An alternative approach would be to hold insurers accountable for the total costs of care. Insurers are a natural target because insurers aggregate all spending (claims costs, other financial transfers, and insurer administration) to the population level and, therefore, have the broadest scope. To the extent that the core concern is high premiums, insurers are the most proximate organization to the policy target.
There are complications to using a holistic approach, however. For example, insurers must negotiate prices with providers, and, if providers have significant market power, insurers may not be able to successfully control spending. Moreover, some tools that insurers may use to control spending, such as narrow networks or utilization management, may not be acceptable to customers (or policymakers), making it difficult for insurers to meet any premium targets without pressure on providers. If the premium targets are too strict relative to available tools, some insurers may leave a market, further complicating cost-containment activities.
Finally, there are some regulatory barriers to focusing on insurers. In 2019, 61 percent of Americans with employer-sponsored health coverage were enrolled in self-insured plans.26 As a result of the federal Employee Retirement Income Security Act (ERISA) and its preemption clause, states have long considered it more difficult to regulate these plans.27 However, this may be changing, given the recent U.S. Supreme Court decision in Rutledge v. PCMA.28 In Rutledge, which involved state regulation of pharmacy benefit managers, the Court stated that “ERISA does not preempt state rate regulations that merely increase costs or alter incentives for ERISA plans without forcing plans to adopt any particular scheme of substantive coverage.”
Many holistic strategies also require attribution of patients to providers. For example, in Maryland, the state attributes patients to hospitals using a multistep process based on first attributing the patients to physicians and then attributing the physicians to hospitals. In cases where patients cannot be attributed using that method, other approaches are used, such as using geography.
Similarly, holistic approaches also must address risk adjustment. Risk adjustment is included in many existing population-based programs, such as in health insurance marketplaces, the Medicare Advantage program, and the Medicare ACO program. Although adjustment models seem to perform acceptably, there are concerns about both the fit of the models and the susceptibility of the models to changes in provider coding behavior. For example, Massachusetts and public programs such as Medicare Advantage have seen very large increases in risk scores.
Strategies to combat coding increases are therefore very important. The Centers for Medicare and Medicaid Services adopts several such strategies for ACOs, including normalizing risk scores each year and capping risk-score growth. Moreover, holistic models, such as population-based payment models, are best suited for large provider groups. Holding small practices accountable for total spending is more problematic, even when risk adjustment is working well, since risk adjustment accounts for expected utilization and there may be wide variation in actual utilization.
The choice among holistic strategies depends on many factors. If the delivery system is very fragmented, soft strategies that focus on episodes may be preferred to population-based models, such as an ACO, because small providers may not be well suited to manage risk. Moreover, relative to population-based approaches, episode-based models are less reliant on attribution (for many episodes) and risk adjustment. Still, episode-based strategies often only cover a portion of spending, often missing patients with multiple, complex chronic conditions. Population-based and global budget models are more comprehensive and are well suited to address both the medical and social needs of high-risk and high-cost patients. But these broader models are best suited for larger delivery systems. And though broader payment models do not need to address overlap in episodes or medical conditions, the issues related to risk adjustment and attribution are more salient. Maryland’s model has evolved over time from a price-setting model to a hospital-budgeting one and now has begun to transition closer to a population-based payment model.
It is worth noting that holistic strategies do not have to be pursued alone and may complement other approaches. For example, states taking a holistic approach may adopt any of the other strategies with respect to competition, prices, or even utilization to support the ultimate goal of lower statewide spending. Rhode Island not only used inflation caps but had hospitals transition from per diem to DRG-based payments to slow the growth of health care spending.29 The state also combined these efforts with requirements for commercial insurers to increase their spending on primary care and care coordination services (see box).
Rhode Island’s Price Controls and Affordability Standards
In 2004, Rhode Island created the Office of the Health Insurance Commissioner (OHIC) to review commercial health insurance rates.30 Given broad statutory language and regulatory authority, the commission’s role became more extensive over time, including mandating increases in primary care spending, adding quality incentives to payment systems, and creating a multipayer medical home. By 2010, OHIC began to adopt affordability standards and price controls, including inflation caps and DRG payments. OHIC also required commercial insurers to increase spending on primary care and care coordination services. In 2018, the commissioner established the Working Group on Healthcare Innovation to develop recommendations for a global state spending cap. The group’s priorities also include standardization and incentivizing quality.
Rhode Island has an all-payer claims database, but given limited resources, these data are managed by researchers at Brown University. OHIC is led by the state health insurance commissioner, who is appointed by the governor.
Commission Support for Holistic Strategies
Given these challenges, state commissions can be very useful for holistic strategies. In soft-touch approaches, they can analyze data, recommend specific policy actions, and support any sanctions, if applicable. In more regulatory approaches, the commission can support development of important policy parameters.
The experiences of different states are instructive. Massachusetts operates a soft-touch holistic system where the Health Policy Commission (HPC) acts as an independent agency, with the goal of making care more affordable to state residents. Though the HPC does not have direct regulatory authority, the HPC has a number of influential functions, including conducting and sharing analyses about drivers of health care spending, making evidence-based recommendations for public policy, and supporting regulatory actions by other agencies.31 Although the commission may not be able to penalize payers or providers, it can “encourage, cajole, and if needed, shame them into doing their part to control costs.”32
Without the direct authority to regulate, these commissions also must be creative when building credibility. A simple, but important, way to do this is by being transparent. Commissions can share their analysis with the public through regular reports or offer perspectives and recommendations by speaking with news and media outlets, therefore influencing debate. If more people know about the commission and are engaged in its work, the harder it becomes to ignore recommendations. Commissions also can collaborate. Many, such as the one in Massachusetts, derive authority by working with other regulatory agencies. Thus, even if the commission itself cannot impose penalties, it can pass on information to agencies that can.
Conversely, states may prefer to create commissions with greater holistic authority, giving them the ability to force changes in the health care sector. For example, Maryland has focused on hospital budgets and total cost of care measures. This is feasible in Maryland because its commission has a long and well-established history. It is also worth noting that stakeholders have significant interest in working with the commission, given the state’s Medicare waiver. Without it, Medicare and Medicaid payment rates would drop significantly, since the waiver created all-payer rates that increased fees from public payers. Although the difference could be offset as commercial rates rise, that would take time, creating something akin to golden handcuffs that incentivize hospitals to make the current system succeed.
Part II. Implementing Policy Goals
What Data and Analytic Capability Do States Need?
To understand trends in spending and implement many of the strategies discussed, states need data. States with the proper regulatory authority can compel organizations, particularly payers, to report data. This is useful not only because it requires less resources from the state, but also because payers are likely better equipped to make accurate estimates about spending for their members.
Aggregate data supplied by payers may be enough, but more detailed, claims-level data allow for more thorough and accurate analyses. To this end, many states now have (or are creating) all-payer claims databases, which are particularly useful for understanding prices or pursuing in-depth examinations of specific services or patient populations.33 However, these databases are resource-intensive, and states may lack the funds and expertise to analyze the data. All-payer claims databases not only require an intake platform and servers, but experienced statisticians and programmers.
In Massachusetts, the state manages its own data with the Center for Health Information Analysis (CHIA). CHIA manages the state’s all-payer claims database and also collects financial data from payers and providers. Notably, CHIA functions as an independent entity that collaborates with state agencies, such as the Health Policy Commission. Though this separation has the potential to create inefficiencies, it also builds trust in CHIA’s work. The data are aggregated and analyzed without regard for what other agencies hope to do with the results. Oregon has a similar Health Policy and Analytics Division. Rhode Island, on the other hand, does not leverage its own data. The state instead works with researchers at Brown University who receive funding through grants. Therefore, the design of a state’s policies may be dependent on its analytic capabilities and available funding sources.
Who Should Be Part of a Commission?
The composition of the commission is a key consideration and dependent on the role that the commission will play. While insights from payers and providers are generally crucial for substantive and political reasons, some states may find it better to minimize the number of commission members employed directly by providers, insurers, or groups funded by those sectors to avoid conflicts of interest. Inclusion of people retired from those segments, or who work for organizations that engage those sectors, may be useful, which is what Massachusetts has done. But other states differ. The commission in Oregon includes stakeholders such as large hospital systems and strong consumer advocates. Their inclusion arguably increases buy-in since they are able to participate in the decisions being made. However, it is important that stakeholders have a strong incentive, like in Maryland, for the commission to succeed. For example, a statutory price reduction, if goals are not met, could foster constructive engagement.
Regardless, expertise in hospital care, nursing care, behavioral health, physicians, and long-term care may all be important. Academics with health policy experience, particularly related to health care markets and financing, are valuable, as well as representatives of employers and organized labor and consumer or patient advocates. In addition, it is important to have some state government representatives, even if ex officio. The Health Policy Commission in Massachusetts has 11 members spanning relevant areas of expertise, while the Heath Services Cost Review Commission in Maryland also includes providers and insurers, among others.
Conclusion
Given barriers to significant federal action, substantial policy changes to control spending growth are likely to be limited to state policy. States can pursue many different strategies, ranging from efforts to promote competition, reduce prices, or decrease utilization of low-value care to broader strategies that address overall spending.
State health policy commissions can play a central role in supporting any of those efforts. However, given differences in political economy and resources, there is likely no one approach to designing commissions for success. Some states may choose to use commissions to provide analytic, convening, and advisory input while other states may give commissions more regulatory authority. States must reflect on their respective priorities, decide on the basic approach they want to take to address health care spending, and then design the commission that works best to support those approaches.
