Strategies to Reduce Medicaid Spending: Findings from a Literature Review
Strategies to Reduce Medicaid Acute Care Spending
Premiums, Cost sharing, and Enrollee Wellness Incentives
Research on premiums and cost sharing in Medicaid does not show reductions in Medicaid spending apart from savings associated with lower enrollment. Premiums in Medicaid increase the likelihood of disenrollment and lead to fewer new enrollments and shorter periods of coverage, and less enrollment can generate savings. In addition, there is evidence that enrollees with chronic conditions who experience interruptions in their enrollment have higher Medicaid spending after re-enrolling in the program, due largely to exacerbations of the condition that lead to more use of emergency department (ED) and inpatient hospital care. Cost sharing reduces enrollees’ use of both low-value, high-cost services and high-value, low-cost services. The net impact of cost sharing on Medicaid spending depends on the mix and volume of services used and spending per user. For example, even if cost sharing results in reduced hospital admissions, the associated savings could be reduced if the average cost per admission increases. There are few studies examining the impact of cost sharing on total Medicaid spending, but those studies show that cost-sharing has no impact on total Medicaid spending or results in higher spending. The administrative costs of collecting, tracking, and enforcing premiums and cost sharing also are part of the total spending equation.
The jury is still out on the impact of wellness incentives on Medicaid spending. Research indicates that the use of consumer incentives holds some potential to motivate healthy behaviors and use of preventive care in Medicaid, but evidence that these types of incentives can reduce Medicaid spending is not available.
Complex Care Management
Research on the impact of complex care management on Medicaid spending is limited and mixed. While complex care management (CCM) programs targeted to Medicaid “super-utilizers” are promising, rigorous evaluations have produced mixed findings regarding reduced utilization and Medicaid costs. Some evaluations have shown reduced ED visits, hospital admissions, inpatient expenditures, and total Medicaid expenditures, while others have shown no impact or reductions in use and expenditures that did not reach statistical significance Evidence that CCM is more successful in reducing costs for some subgroups than for others indicates that effective targeting of CCM resources is important, especially because CCM can be costly and large spending reductions are necessary to yield net savings. Data on net savings attributable to CCM is lacking. Initiatives to reduce Medicaid spending by funding housing-related services for Medicaid enrollees in supportive housing are relatively new, and evidence is limited. However, several small programs have reported some improvements in utilization or spending.
Patient-Centered Medical Homes
Research on the impact of patient-centered medical homes (PCMH) on spending is mixed. Few evaluations of PCMH look at the Medicaid experience specifically, but the broader body of research is informative. Studies show that PCMH improve quality of care, but evidence for reduced ED and hospital use and lower total costs of care is mixed. Some studies show improvements, while others show no or minimal change or show that change is limited to individuals with significant health care needs. Evidence from Medicare shows that PCMH can increase program spending if savings from reduced ED and hospital use are not sufficient to offset PCMH payments to practices and if better coordination of care leads to increased use of services. Current findings are based on spending just a few of years after PCMH implementation, so whether or not additional savings accrue over a longer time period is not known. Even if PCMH themselves do not reduce Medicaid spending, the model may be integral to broader reforms to lower costs.
The impact of Medicaid health homes on Medicaid spending is not yet established. Health homes are similar to PCMH, but they specifically target individuals with chronic conditions and emphasize physical and behavioral health integration as well as improved linkages to community and social supports and long-term services and supports. A couple of state-funded evaluations of Medicaid health home programs for enrollees with multiple chronic conditions or a serious and persistent mental health condition have shown significant decreases in use of and spending for specific services. However, the federal five-year evaluation of Medicaid health homes has not concluded, so the question of their net impact on Medicaid spending is still unknown.
Alternative Payment Models
The use of accountable care organizations (ACO), episode-based payments, and global budgets is still new in Medicaid, and research regarding the effectiveness of these models in containing Medicaid costs is very limited. Early evidence from Medicare ACOs, which may inform expectations for Medicaid, has shown modest savings in total costs of care and improved quality for Medicare ACO participants relative to a control group, but thus far, the savings have not been sufficient to offset Medicare bonus payments to high-performing ACOs. At this time, evidence from rigorous evaluations of episode-based payments and global budgets is also too limited to support expectations that these models can reduce total Medicaid spending.
Strategies to Reduce Medicaid Long-Term Care Spending
Tightening Financial Eligibility Rules for Long-Term Care Services
Empirical evidence indicates that any program savings from state strategies to tighten financial eligibility rules for individuals seeking Medicaid long-term care services would be small and offset to some extent by additional administrative expenses. These strategies include curtailing asset transfers, increasing estate recovery efforts, and including retirement accounts as countable assets. Further curtailing asset transfers is unlikely to produce substantial savings as most empirical research about individuals seeking Medicaid long-term care eligibility finds that relatively little asset transfer takes place or the amount transferred is small. Increasing estate recovery efforts also is unlikely to produce program savings as states’ current efforts to recover Medicaid long-term care expenditures from deceased beneficiaries’ estates have not yielded large amounts of funds, and more aggressive estate recovery programs would require additional staffing, are likely to be time-consuming and expensive, and may be politically unpopular. Finally, including retirement savings as countable assets when determining long-term care eligibility is unlikely to yield significant savings because few Medicaid beneficiaries are likely to have high account balances. Survey data suggest that seniors with the most severe disabilities have few assets. Similarly, the financial wealth of adults with disabilities, who are most likely to use long-term care services, is much lower compared to those without disabilities.
Promoting Private Long-Term Care Insurance
Microsimulation studies of the potential impact of private long-term insurance on Medicaid long-term care expenditures find little to no effect even far into the future. Private long-term insurance is expensive and without large costly subsidies is unlikely to be affordable for many middle-class families, who might spend down to Medicaid. Any Medicaid savings realized by greater reliance on private long-term care insurance would be offset by subsidy costs. Studies of the Long-Term Care Partnership approach, in which Medicaid pays costs beyond the services covered in an approved private long-term care insurance policy, differ as to whether this method would modestly reduce or increase Medicaid LTSS spending. In addition, the number of companies offering private long-term care insurance and the number of policies sold have sharply declined, casting doubt on the future role of the product as a source of long-term care financing for the general population.
Expanding Home and Community-Based Services (HCBS)
Cost savings from HCBS programs show mixed results, although more recent studies have more favorable findings. Earlier studies found that costs increased when HCBS were expanded because large increases in home care use more than offset relatively small reductions in nursing home use. More recent studies show more favorable results. Several studies have shown small or nonstatistically significant reductions in nursing home use resulting from expanded use of HCBS, but other studies have found a stronger impact. A 2013 review of 38 studies of publicly funded HCBS programs showed wide variation among programs in different states, but promising results related to cost savings. The Money Follows the Person program appears to achieve savings, but does not transfer many people from institutions to the community. Evidence of overall Medicaid savings on acute and post-acute expenditures as a result of HCBS expansion remains limited. Importantly, states have steadily increased use of HCBS over time under the current financing system. Thus, any savings from increasing use of these services will occur without any change in financing systems.
Increasing Use of Managed Long-term Services and Supports
Studies find mixed evidence on overall cost savings from the use of managed long-term care delivery systems. There is limited evidence that the use of managed long-term services and supports can lead to lower utilization of some services like preventable emergency room visits, hospital stays, and institutional services. Evaluations of managed long-term care programs from the mid-2000s found inconclusive and inconsistent evidence of savings. Results from the current Centers for Medicare & Medicaid Services Financial Alignment Initiative demonstrations are not yet available.
There is little empirical evidence to support the potential to achieve substantial Medicaid cost savings through strategies aimed at improving programmatic and administrative efficiencies in delivering acute and long-term services and supports. Evidence about cost savings from the eight potential strategies examined in our literature review ranges from limited to mixed to too early to tell. The strategy with the largest evidence base to date, requiring premiums, has been shown to lower program enrollment, thereby generating savings; however, reducing enrollment is inconsistent with Medicaid’s goal of promoting health coverage and access. Less is known about some newer strategies, such as alternative payment models, and based on early trends about their use in Medicare, it is unclear that these approaches will dramatically slow Medicaid spending or produce net savings. Whether additional savings are achievable from more recent initiatives to pursue strategies in combination, such as PCMH within an ACO under a global budget, is not yet known. To the extent that savings are achievable from strategies such as managed long-term services and supports or expanding HCBS, states already are implementing these policies, making further savings by changing the financing structure unlikely. Moreover, policymakers may choose to implement certain strategies in pursuit of other goals, such as improved health outcomes or increased enrollee satisfaction.
In sum, none of the strategies we reviewed demonstrates the potential to achieve savings at a level to replace the loss of $834 billion in federal Medicaid funding from 2017-2026, estimated by the CBO under the American Health Care Act as passed by the House. If faced with funding reductions of that magnitude, states will be left primarily with the options of reducing provider payments or eliminating optional services or coverage pathways to control Medicaid costs, which are likely to have negative effects on beneficiaries and providers.