New Rules for Medicaid Managed Care – Do They Undermine Payment Reform?

Replacing fee-for-service (FFS) financing with value-based payment methods is a central goal of payment reform efforts. The perverse incentives associated with FFS are well known: they reward greater volume, more intensity, and generally contribute to the fragmentation of care.¹ Medicare’s Accountable Care Organizations (ACOs) and bundled payment initiatives represent the most prominent efforts to move away from FFS, with the hope of slowing the growth of healthcare spending and improving patient outcomes. State Medicaid programs are following suit, increasing the use of risk-based managed care, Medicaid ACOs, and Delivery System Reform Incentive Payment (DSRIP) programs. However, new Medicaid managed care rate setting rules threaten to entrench volume-based financing by requiring that payment rates be closely tied to historical utilization data. While broad CMS policies move away from FFS and towards value-based payment, official actuarial requirements, as they apply to Medicaid managed care, require adherence to FFS methodologies, potentially undermining the promise of payment and delivery system reform for many states.

The rules in question refer to new language around “actuarial soundness.” Since 2002, federal regulations have required states to pay Medicaid Managed Care Organizations (MCOs) rates that are actuarially sound, meaning that the funding provided to MCOs is adequate to cover expenditures for healthcare services and related administrative expenses. Policies requiring actuarially sound rates act, in part, as a check against the potential for states to underfinance MCOs and compromise enrollee access to care.

Early regulations on actuarial soundness largely focused on the process that states needed to use in rate setting. They did not set specific standards on methodologies, data, or accuracy,² although CMS indicated that managed care rates should be independent of state budget constraints.^3,4 Further guidance for interpreting these rules was issued in 2005 by the American Academy of Actuaries, which declared that rates should be based on “assumptions that are individually reasonable and appropriate” but refrained from setting explicit standards, noting that “other approaches may also be reasonable and may currently be in common use.”⁴ These guidelines allowed for considerable flexibility in rate setting. Furthermore, actuaries often proposed a range of rates. In selecting from a potentially wide range, states had the option to choose rates that were influenced by their budgets, even if this practice was at odds with federal intentions. Compliance with these flexible rules was overseen by regional Medicare offices that typically gave rates cursory reviews.

CMS recently approved regulations⁵ that aim to eliminate previous ambiguities around actuarial soundness and make the rate setting process more transparent. The new rules explicitly require rates to be developed based on utilization as measured through historical managed care claims or encounter data and to target a minimum medical loss ratio (MLR) of 85% as a component of actuarially sound rate setting. The regulations make it plain that MLRs are meant to gauge the appropriateness of rates: a high MLR could indicate excessively low capitation rates, while a low MLR could indicate rates that are overly generous. In an indication of their increased interest, CMS has moved jurisdiction for Medicaid rate review from regional offices to the CMS Office of the Actuary and has reserved the authority to withhold federal financial participation if rates are deemed to be noncompliant with actuarial soundness standards. This threat considerably raises the stakes for close adherence to the rules of actuarial soundness.

The proposed new regulations would give the federal government greater oversight and control of Medicaid spending as it invests billions of additional dollars in Medicaid expansion. One interpretation of the rules is that they algorithmically and mechanically set rates, eliminating the ability of actuaries to put forth a range of plausible rates from which states can choose. As a consequence, states would have limited flexibility in setting their rates, and the process would be more explicitly insulated from state budgeting targets.

While the goals of these proposed rules may be well-intentioned, a reliance on detailed claims and encounter data as the basis for rate setting is ultimately in conflict with the goals of payment reform. Even when subsumed within capitation payments or global budgets, reliance on historical claims preserves FFS incentives and its attendant problems that new payment models aim to replace. For example, an MCO that invested in a successful care coordination program to reduce hospital readmissions would actually be penalized, as reductions in utilization would lead to reductions in per-member payment rates.

In theory, some of the deleterious FFS incentives could be reduced under a competitive bidding regime, where MCOs could offset their lost revenue in capitation rates with more members. However, there are currently few states that use competitive bidding, and their numbers appear to be shrinking.^6,7

Strict adherence to actuarially sound rates also provides relatively little opportunity for states or health plans to pay for quality if those programs do not generate an equivalent claim or encounter, a particularly vexing issue facing creative population health efforts and also for calculating medical loss ratios. These rules would further embed the reliance on claims-based payments as a way of demonstrating actuarial soundness.

A claims-based method of rate setting also creates challenges to establishing explicit “sustainable” growth rates, a goal embodied in reforms in Massachusetts, Maryland, and Oregon. Ideally, a lower growth rate decoupled from the fee-for-service, volume-based payment system has the potential to lead to greater financial stability for states by removing some pressure from state budgets. Furthermore, models that focus on explicit spending growth rates, decoupled from claims utilization, create a focus on delivery systems that improve value and reduce waste. In contrast, a claims-based method of rate setting reduces incentives for investing in cost-effective health related services (such as wellness programs or supportive housing) as a way of achieving spending targets.

New rules for actuarial soundness are appropriately focused on ensuring that state and federal payments are sufficiently funding Medicaid MCOs while also safeguarding their joint investment, but they should also support state attempts to experiment with new delivery systems. It is possible that a limited federal capacity to monitor adherence to these rules will diminish their practical implications. However, the specter of violating these regulations has the potential to temper the ambitions of states eager to move away from FFS payments. For example, CMS issued a guidance in 2014 to the State of Oregon that led Oregon to abandon its previous rate setting methodology in favor of one that aligned with actuarial requirements.⁸ In particular, whereas the state had envisioned a Medicaid ACO model that would allow for “flexible” spending not tied to medical claims and a restricted spending growth rate of 3.4%,^9,10 CMS instructed the state to use claims data to set rates and to move away from an aggregate spending target. As a result, the state retroactively reset rates, which resulted in rate increases for some ACOs and reductions for others that were then liable for previous (although newly determined) “overpayments.” Oregon is currently undergoing a waiver renewal to request latitude in aligning the new rate methodology with “flexible” spending. These actions suggest new risks for states contemplating innovative payment methods.

There are a number of ways that CMS can provide adequate funding to ensure access without tying states to FFS or claims based encounters. For example, states that have current actuarially sound rates might be allowed to grow those rates by a targeted amount, if states could demonstrate that they were maintaining appropriate levels of access and quality of care. This would allow for reasonable inflationary increases, which would enable states to pay plans on the basis of quality and outcomes, irrespective of the volume of services.

A second option might be to give states flexibility to test rate adequacy through a budget-driven model, as opposed to the standard utilization based approach. Rather than using historical claims data and unit cost data to project estimates of per-member expenditures, which then determine how much states must budget for their program, states could set their Medicaid budgets and then assess rate adequacy within that budget. In this process, modeling would begin with the budget, converting overall budget targets into per-member expenditures and deriving unit costs (e.g., payment rates for medical services) from those estimates. Those unit costs could then be compared to other benchmarks – e.g., some percentage of the Medicare payment rate – to assess adequacy. Whereas actuarial soundness rules currently specify a “bottom-up” approach, building up rates from individual claims experiences, which then defines the required budgeting amounts, this proposed model could be seen as a “top-down” method, starting with an assessment of how much states have budgeted and how this would be allocated to the population of patients. While this model would represent a significant departure from the ways CMS has historically required that states set rates, a top-down approach is aligned with the growing emphasis on population health management, accountable care models, and global budgets.

A third option is to integrate rules on actuarial soundness with new rules on network adequacy. Both network adequacy and actuarial soundness are tied to concerns about the growth in Medicaid managed care and the implications for appropriate access to a variety of types of care. CMS has indicated that states may have flexibility in how network adequacy is assessed, but should include measures of time and distance standards for primary care, specialists, and hospital services, expected utilization, and the number of new providers accepting Medicaid patients. These measures could be benchmarked to private or Medicare markets to ensure comparable access standards. Broadening the ways in which actuarial soundness could be validated may be in simpatico with the goals of assessing network adequacy and may further reduce the financial and infrastructure burdens imposed on states for managing each set of standards separately.

In order to promote experimentation with new payment models and delivery system reforms, prompt federal action is needed to support payment innovation occurring at the state level. This will require improved coordination between states and CMS as well as consideration of other mechanisms besides claims or encounter data that can be used to quantify quality or value. In the absence of greater flexibility, states will be unable to move away from volume-based purchasing, limiting their ability to create optimal, efficient programs for their vulnerable Medicaid populations.