Note: This analysis was updated on May 5, 2023 clarify an error about verification for exemptions.

On April 26, 2023, the House of Representatives passed a Republican debt ceiling bill (HR 2811, the Limit, Save, Grow Act of 2023) that includes a requirement for states to implement work requirements for certain Medicaid enrollees. Data show that 91% of non-elderly Medicaid enrollees who are not on Supplemental Security Income or Medicare are working or face barriers to work. We estimate that if the proposal were fully implemented in 2024 and the rate of Medicaid eligibility loss was as the Congressional Budget Office (CBO) estimated, then 1.7 million enrollees would not meet work or reporting requirements and potentially face disenrollment in that year. States could continue to provide Medicaid to those enrollees but would not receive federal matching funds for doing so. It is unclear if any states would choose to do that, though CBO estimated over half of enrollees would continue to be covered at the states’ expense. If states chose to keep all 1.7 million people enrolled, $10.3 billion of Medicaid spending would shift from the federal to state governments in 2024. A small number of states with the largest share of enrollees under the Affordable Care Act (ACA) Medicaid expansion would account for almost half of the increased state spending or coverage losses.

What is the work requirement policy in the debt ceiling bill?

Under the Medicaid work requirement plan, certain adult enrollees ages 19-55 would need to work or participate in other qualifying activities (like community service or job training) for at least 80 hours per month. There would be exemptions for those who are physically or mentally unfit for employment (as determined by a physician or other medical professional), pregnant, the parent or caretaker of a dependent child or incapacitated person, complying with a work requirement under a different federal program, participating in drug or alcohol treatment or rehabilitation program, or enrolled in school at least half time. If enrollees fail to meet the work or reporting requirements for three or more months, the federal government would cease paying the federal share of Medicaid for their expenses. States could disenroll them or continue their coverage but pay 100% of the costs. Eligibility for federal funds could resume at the start of the following calendar year.

How would the policy affect Medicaid enrollment or spending?

The CBO cost estimate included national estimates of coverage loss and changes in federal spending but did not include state-specific estimates or details about how individual states would respond or why. CBO estimates that once work requirements were implemented, each year an average of 15 million enrollees would be subject to the new requirements and about 1.5 million of them would lose eligibility for federal funding, resulting in federal savings of $109 billion over the period. CBO anticipates that about 60% (900,000) of the people who lost eligibility would live in states that maintained coverage with state-only funds and the remaining 600,000 would become uninsured because they lived in states that did not maintain coverage. In a follow-up estimate, CBO estimated that state costs would increase by $65 billion over the 2023-2033 period in states that opt to continue coverage without federal funds (or an average of $6.5 billion annually). In summary, CBO stated that “under those requirements, federal costs would decrease, the number of people without health insurance would increase, the employment status of and hours worked by Medicaid recipients would be unchanged, and state costs would increase.”

Assuming the policy is fully implemented in 2024 and CBO’s estimated rate of eligibility loss applies in all states, we estimate that 1.7 million people could be determined ineligible in 2024 and provide state-by-state estimates of the number of people who could potentially lose coverage. To estimate state-by-state coverage loss, we use state-by-state estimates of 2024 projected enrollment that reflect coverage losses attributable to unwinding the continuous enrollment period, which applied during the COVID pandemic but ended on April 1, 2023. We adjust those estimates to account for new enrollment during the months from April 2023 to May 2024, assuming new enrollment is similar to that of the past six months. We also adjust those estimates to account for re-enrollment of people who are disenrolled during the unwinding (e.g., “churn”) and South Dakota’s new expansion – an estimated 40,000 adults. North Carolina also recently adopted a new expansion program but implementation of the expansion is contingent on the 2023-2024 budget, so our estimates do not include North Carolina.

Consistent with the Department of Health and Human Services, we assume the work requirements would only apply in expansion states and only apply to enrollees in the expansion group, since other categories of enrollees would be exempt (e.g., children, older people, parents who qualified under pre-ACA eligibility rules, and people who cannot work due to a disability). Wisconsin is a non-expansion state and not included in the analysis; however, they do have a waiver to cover adults up to 100% of the federal poverty level (FPL) who could be subject to the work and reporting requirements.

We estimate 16.7 million enrollees in the expansion group would be between ages 19 and 55 in May 2024 using the age distribution of expansion adults in administrative data (“T-MSIS”). This estimate includes some parents of dependent children, who would be exempt from the work requirement but still potentially subject to reporting requirements. If 10% fail to meet the work or reporting requirements, as CBO assumed, 1.7 million enrollees could lose eligibility for federal matching funds in 2024.

If all states elected to maintain coverage for the 1.7 million people (rather than the 40% losing coverage as assumed by CBO), we estimate that the policy would shift $10.3 billion from federal to state spending in 2024 (state-by-state results in Appendix Table 1). It is unknown if any states would choose to maintain coverage at full state cost, making up for the loss of federal funds, and CBO did not provide any details about which states would do so under their estimates. We estimate for each state what the coverage loss would be based on CBO’s national rate, and also what it would cost each state to prevent the loss of coverage. The distribution of costs across the states is closely related to the size of states’ expansion populations and five states (CA, IL, NY, PA, WA) account for nearly half of the total costs (Figure 1). We estimated the per-person costs for expansion group enrollees in 2024 by growing 2021 per-person costs from the Medicaid CMS-64 administrative data and adjusting them upward slightly based on trends in spending through the current year to date from the U.S. Treasury outlays. We assume per-person costs in 2024 will grow at the same rate as costs in 2023. The new costs to states equal 90 percent of total spending for affected enrollees—reflecting the 90% of costs that the federal government pays for expansion enrollees who remain eligible.

What to watch?

We used CBO assumptions to estimate the share of people who could lose eligibility for federal matching funds, but that estimate is highly uncertain. For example, if adults with dependent children who are eligible through the expansion pathway are automatically exempted from work and reporting requirements, the share that could lose eligibility could be lower than 10 percent. However, when Arkansas implemented work requirements, 25 percent of those subject to the requirements lost coverage, suggesting that the percent who lose eligibility could be well above the 10% assumed by CBO and used in our state-by-state estimates. The most likely outcome is that the actual rate will vary considerably across the states. Our estimates are also specific to H.R. 2811.

Another source of uncertainty is that the policy could be applied to non-expansion eligibility groups, which would make many more people subject to reporting requirements, such as parents and people who cannot work due to a disability. Although we anticipate that most of those people would qualify for an exemption, some could lose Medicaid eligibility on account of being unable to comply with reporting requirements. We estimate that there will be 15.3 million adults ages 19-55 in non-expansion eligibility groups in 2024, compared with 16.7 million expansion adults (state-by-state results in Appendix Table 1, tab 2).

While the debt ceiling bill passed the House, it is not expected to pass in the Senate. Leaders in the Senate and the Administration support passage of a clean debt ceiling bill; however, proposals to reduce federal spending could continue to be debated as part of the debt ceiling or as part of the regular budget process. It is unclear whether this policy—or similar related policies—could end up being debated as negotiations over the debt ceiling and federal spending continue.

According to recently released data from the Centers for Medicare & Medicaid Services (CMS), Medicare Advantage now provides Medicare coverage for just over half of eligible beneficiaries. In January 2023, 30.19 million of the 59.82 million people with both Medicare Part A and Part B were enrolled in a private plan (Figure 1).

Medicare Advantage, the private plan alternative to traditional Medicare, covers Medicare Part A and B benefits (and typically Part D benefits), often for no additional premium (other than the Part B premium). Insurance companies contract with the Medicare program and receive payments for providing these services.

Enrollment in Medicare Advantage has increased dramatically in recent years. In 2007, less than one in five (19%) eligible Medicare beneficiaries were enrolled in a private plan. The growth in enrollment is due to a number of factors, including the attraction of extra benefits offered by most plans, such as vision, hearing, and dental services, and the potential for lower out-of-pocket spending, particularly compared to traditional Medicare without supplemental coverage. Medicare Advantage plans also offer the simplicity of one-stop shopping, in that enrollees do not need a separate Part D prescription drug plan or supplemental coverage.

At the same time, Medicare Advantage plans typically use tools to manage utilization and costs that may limit access to care, such as prior authorization requirements and referrals for specialists and mental health providers. For example, in 2021, Medicare Advantage enrollees submitted 35 million prior authorization requests. In addition, Medicare Advantage plans generally require enrollees to receive care from in-network providers or pay more out-of-pocket for out-of-network care.

As the role of Medicare Advantage grows, so will interest in understanding how well the program serves the increasingly diverse group of enrollees who receive their Medicare coverage from private insurers, including a disproportionate share of Black, Hispanic and Asian and Pacific Islander beneficiaries. A recent review of evidence on how Medicare Advantage compares to traditional Medicare found few differences between the programs. Additionally, gaps in data make it difficult to evaluate plan performance, including assessments of the program’s impact on value and equity.

Mental health and substance abuse disorder services are needed by people of all ages, including older adults. In February 2023, 20% of adults 65 and older reported symptoms of anxiety and/or depression. Medicare covers a range of mental health and substance use disorder services, and Medicare Advantage plans are required to cover the same set of services as traditional Medicare. With increasing Medicare enrollment, the need for mental health and substance use disorder care will likely continue to grow. Members of Congress and the Biden Administration have introduced proposals to address concerns about barriers to mental health and substance use disorder care in Medicare. However, despite the rise in enrollment in Medicare Advantage plans, which now cover about half of people on Medicare and typically offer extra benefits beyond traditional Medicare, not much is known about the scope of mental health and substance use disorder benefits covered by these plans.

This brief describes the scope of mental health and substance use disorder benefits in Medicare Advantage plans, including extra benefits, cost sharing, and prior authorization and referral requirements, based on an analysis of the Centers for Medicare and Medicaid Services (CMS) Medicare Advantage plan benefit and enrollment files for 2022. We limited the analysis to individual Medicare Advantage plans that are generally available to all beneficiaries for enrollment (which excludes plans sponsored by employers/unions and special needs plans (SNPs) that are offered to a defined group of beneficiaries).

Key Findings

• A small share of plans provided extra benefits specifically for mental health and substance use disorders in 2022: 12% of enrollees were in plans that provided access to additional inpatient hospital psychiatric services, while 6% were in plans that offered tailored extra benefits, such as reduced cost sharing, to those with mood disorders or opioid use disorders.
• For in-network outpatient mental health and substance use disorder services, Medicare Advantage plans usually require cost sharing, which were typically copays rather than coinsurance, and varied across plans in 2022.
• About 60% of Medicare Advantage enrollees were in plans that did not cover out-of-network outpatient mental health and substance use disorder services in 2022; for the 40% who were enrolled in plans with some out-of-network coverage, the most common cost-sharing approach for out-of-network coverage of these services was coinsurance rather than a copay, typically 50%, though cost sharing also varied across plans.
• Prior authorization requirements are common: 98% of Medicare Advantage enrollees were in plans that required prior authorization for some mental health and substance use disorder services in 2022.
• About one in four (26%) Medicare Advantage enrollees were in plans that required referrals for some mental health and substance use disorder services in 2022.
• There are data gaps that limit the ability of beneficiaries and policymakers to understand the full extent of how mental health and substance use disorder benefits vary across Medicare Advantage plans and are working for enrollees. For example, there is not much information on whether Medicare Advantage enrollees are experiencing difficulty finding (and getting treated by) mental health providers in their plan’s network, the extent to which enrollees use in-network and out-of-network providers for these services, or their out-of-pocket liability for these services. Further, there is no data on prior authorization requests, denials, and appeals by type of mental health and substance use disorder service, by enrollee characteristics, or at the plan-level, such as for special needs plans versus plans available for general enrollment.

Coverage and Cost Sharing for Mental Health and Substance Use Disorder Services

People on Medicare have the option of receiving their Part A and Part B Medicare benefits through traditional Medicare or a Medicare Advantage plan. For Medicare Advantage plans, the federal government contracts with private insurers to provide Medicare benefits to enrollees, and plans are required to meet federal standards. For example, Medicare Advantage plans are required to cover all Part A and B benefits covered under traditional Medicare, and Medicare Advantage plans are required to provide an out-of-pocket limit.

Medicare covers a range of mental health and substance abuse disorder services, both inpatient and outpatient services, and Part D plans cover outpatient prescription drugs used to treat these conditions. Medicare Part A covers inpatient care for beneficiaries who need treatment for a mental health condition in either a general hospital or a psychiatric hospital, though Medicare beneficiaries are only covered for up to 190 days in a lifetime for stays in a psychiatric hospital. Medicare Part B covers several outpatient services, including one depression screening per year, a one-time “welcome to Medicare” visit, which includes a review of risk factors for depression, and an annual “wellness” visit, where beneficiaries can discuss their mental health. Part B covers individual and group psychotherapy with doctors or with certain other licensed professionals, psychiatric evaluation, medication management, partial hospitalization, and opioid use disorder treatment services, among others.

Medicare also covers some telehealth services under Part B, including for mental health and substance use disorder services as well as non-mental health related services, on both a permanent basis and on a temporary basis as part of the COVID-19 public health emergency. Medicare Advantage plans have additional flexibilities to provide telehealth benefits, and since 2020, Medicare Advantage plans have been permitted to include costs associated with telehealth benefits (beyond what traditional Medicare covers) in their bids for basic benefits. This benefit includes telehealth visits provided to enrollees in their own homes and services provided outside of rural areas (benefits not covered by traditional Medicare prior to the COVID-19 pandemic) and can be provided for mental health and substance use disorder services. In 2022, 98% of Medicare Advantage enrollees in individual plans had a telehealth benefit.

Medicare Advantage plans have flexibility when designing plan benefits and use rebate dollars to provide non-Medicare covered benefits, such as dental, vision, and hearing, as well as reduce cost sharing, and expand Medicare-covered benefits, such as by offering additional days of coverage in a psychiatric hospital. While Medicare Advantage plans may provide modified or reduced cost sharing compared to traditional Medicare, Medicare Advantage cost sharing for Part A and B services cannot exceed cost sharing for those services in traditional Medicare on an actuarially equivalent basis. According to the Medicare Payment Advisory Commission (MedPAC), in 2023, about 39 percent of rebate dollars or $912 annually were used to lower cost sharing for Medicare services, although the allocation of lower cost sharing by benefit type, including for mental health, is not shown.

Most people with traditional Medicare have some kind of supplemental coverage (e.g. Medicaid, employer-sponsored coverage or Medigap) which helps with Medicare cost sharing. Out-of-pocket costs may differ between traditional Medicare and Medicare Advantage plans, and vary from one Medicare Advantage plan to another, though are subject to limitations.

Medicare Advantage plans can and do apply cost management tools to mental health and other services, such as prior authorization requirements, referrals, and limited networks that can restrict beneficiary choice of in-network physicians and facilities.

Plan Benefit Design

A small share of plans provided extra benefits specifically for mental health and substance use disorders in 2022: 12% of enrollees were in plans that provided access to additional inpatient hospital psychiatric services, while 6% were in plans that offered tailored extra benefits, such as reduced cost sharing, to those with mood disorders or opioid use disorders.

EXTRA benefits

In 2022, 12% of enrollees were in plans that provided access to additional inpatient hospital psychiatric services as an extra benefit, including either additional days of coverage per benefit period and/or coverage of non-Medicare-covered stays. As noted above, Medicare beneficiaries are generally only covered for up to 190 days in a lifetime for stays in a psychiatric hospital. However, there is no publicly available data on how frequently extra benefits are used by enrollees.

While not necessarily specific to mental health and substance use disorders, there are other supplemental benefits that can be used toward services to address these conditions. For example, in 2022, 39% of enrollees were in plans that provided transportation benefits, which could be used to get to and from medical appointments for mental health and substance use disorder services. Because many supplemental benefits are applicable to all types of services, not just for mental health and substance use disorders, we do not go into detail on those benefits here.

uniformity flexibiltiy

As of 2019, Medicare Advantage plans may tie the availability of supplemental benefits to certain subgroups of beneficiaries, such as those with mental health disorders or diabetes, making different benefits available to different enrollees, as long as similarly situated enrollees are treated the same. This option, known as uniformity flexibility, allows Medicare Advantage plans to tailor supplemental benefits to enrollees with mental health or substance use disorders as well as reduce or eliminate cost sharing for certain services, and offer lower deductibles if they meet certain criteria.

In 2022, more than 1.1 million enrollees, or about 6% were in a Medicare Advantage plan that offered either reduced cost sharing or supplemental benefits to those with mood disorders (depending on the plan criteria) or opioid use disorders.

Chronic condition special needs plans (C-SNPs)

There are also other types of Medicare Advantage plans, called Special Needs Plans (SNPs), that are able to target plan benefits to enrollees with certain conditions. C-SNPs, which are for people with severe chronic or disabling conditions, can restrict enrollment to specific types of beneficiaries with significant or relatively specialized care needs, including beneficiaries with mental health conditions. In 2022, four firms offered C-SNPs for people with mental health conditions, covering about 1,800 enrollees. These plans are expected to provide benefits that meet the needs of enrollees, which can include but are not limited to, parity between medical and mental health benefits and services, no or lower cost sharing for mental health and substance use disorder services, and longer benefit coverage periods for inpatient services or other types of services.

Cost Sharing for Outpatient Services

For in-network outpatient mental health and substance use disorder services, Medicare Advantage plans usually require cost sharing, which were typically copays rather than coinsurance, and varied across plans in 2022. About 60% of Medicare Advantage enrollees were in plans that did not cover out-of-network outpatient mental health and substance use disorder services in 2022; for the 40% who were enrolled in plans with some out-of-network coverage, the most common cost-sharing approach for out-of-network coverage of these services was coinsurance rather than a copay, typically 50%, though cost sharing also varied across plans.

Under traditional Medicare, there is a $233 deductible (in 2022) and 20 percent coinsurance that applies to most services covered under Medicare Part B after the deductible is met. However, some mental health and substance use disorder services have different cost-sharing amounts, such as opioid treatment program services which do not require cost sharing, though the Part B deductible still applies for supplies and medications a beneficiary gets through an opioid treatment program provider.

Medicare Advantage plans can charge either copays or coinsurance for these services or no cost sharing at all. In 2022, when plans imposed cost sharing, nearly all required copays for in-network services and more frequently coinsurance for out-of-network services. Medicare Advantage plans’ cost sharing for outpatient mental health and substance use disorder benefits varied across plans, whether the service was in-network and out-of-network, and by service category (Table 1).

About six in ten (60%) Medicare Advantage enrollees did not have any coverage of mental health and substance use disorder services received from out-of-network providers in 2022, meaning enrollees in these plans would have to pay completely out-of-pocket to receive these services.

Therapy Sessions with a Psychiatrist or With Another Type of a Mental Health Provider

In-Network: These services consist of either individual or group therapy sessions with a psychiatrist or with other mental health providers, such as clinical psychologists and clinical social workers and other professional authorized by states to furnish mental health services. About 9 in 10 Medicare Advantage enrollees (90%) were in plans that charge cost sharing for therapy sessions, with virtually all charging copays (88%) rather than coinsurance (2%). The most common copay amount across plans was $40, though copays ranged from $2 to $40. For the small share of enrollees in plans with coinsurance, the most common coinsurance amount was 20% and ranged from 15% to 20%. In total, about 10% of Medicare Advantage enrollees were in plans that did not require cost sharing for therapy sessions. (For more detailed data on each therapy session type, see Table 1).

Out-of-Network: About 4 in 10 Medicare Advantage enrollees (41%) were in plans that provided coverage for out-of-network individual or group therapy sessions. More enrollees were in plans that charged coinsurance (22%) rather than copays (18%), with the most common coinsurance being 50%, and the most common copay being $40. About 6 in 10 (59%) Medicare Advantage enrollees were plans that did not provide coverage for out-of-network therapy sessions. Individual and group sessions are displayed together for out-of-network services due to how data are input in the Medicare Advantage benefit files.

Partial Hospitalization

In Network: These services include a more structured program of individualized and multidisciplinary outpatient psychiatric treatments that is more intensive than in a doctor or therapists’ office, as an alternative to an inpatient stay. Nearly 9 in 10 Medicare Advantage enrollees (86%) were in plans that charged cost sharing for each day of partial hospitalization, with virtually all charging daily copays (84%) rather than coinsurance (2%). The most common copay amount was $55, though copays ranged from $5 to $55 across plans. The most common coinsurance amount was 20% and ranged from 14% to 20%. In total, about 14% of Medicare Advantage enrollees were in plans that did not require daily cost sharing for partial hospitalization.

Out-of-Network: About 4 in 10 Medicare Advantage enrollees (41%) were in plans that provided out-of-network coverage for partial hospitalization. More enrollees were in plans that charged coinsurance (27%) rather than copays (13%). Coinsurance of 50% per day was the most common across plans, while the most common copay amount was $75 per day. About 6 in 10 (59%) Medicare Advantage enrollees were in plans that did not provide coverage for out-of-network partial hospitalization services.

Opioid Treatment Program Services

In Network: Medicare Advantage enrollees with opioid use disorder can receive services through an opioid treatment program, which may consist of medication-assisted treatment (MAT), substance use counseling, individual and group therapy, toxicology testing, intake activities, and periodic assessments. About 6 in 10 Medicare Advantage enrollees (63%) were in plans that charged cost sharing for various opioid treatment program services, with mostly all charging copays (57%) rather than coinsurance (4%). The most common copay amount was $40, though copays ranged from $5 to $250 across plans. The most common coinsurance amount was 20%, and ranged from 10% to 50%. About 2% of enrollees may have either had a copay or coinsurance. In total, 37% of Medicare Advantage enrollees were in plans that did not require cost sharing for opioid treatment program services.

Out-of-Network: About 4 in 10 Medicare Advantage enrollees (40%) were in plans that provided coverage for opioid treatment program services. More enrollees were in plans that charged coinsurance (22%) rather than copays (6%). The most common coinsurance amount was 50%, while the most common copay amount was $65 across plans. About 6 in 10 (60%) Medicare Advantage enrollees were in plans that did not provide coverage for out-of-network opioid treatment program services.

For information on individual and group therapy sessions for substance use disorder as well as more detailed information on other services, see Table 1.

Prior Authorization and Referrals

In 2022, 98% of Medicare Advantage enrollees were in plans that required prior authorization for some mental health and substance use disorder services. About one in four (26%) Medicare Advantage enrollees were in plans that required referrals for some of these services.

Prior Authorization

Medicare Advantage plans can impose prior authorization requirements, which entail enrollees getting approval from their plan prior to receiving a service, and if approval is not granted, then the plan generally does not cover the cost of the service. According to a KFF analysis, over 35 million prior authorization requests were submitted to Medicare Advantage plans in 2021.

In 2022, virtually all Medicare Advantage enrollees (98%) were in plans that required prior authorization for some mental health and substance use disorder services. More than 9 in 10 Medicare Advantage enrollees were plans that required prior authorization for inpatient stays in a psychiatric hospital (93%) and partial hospitalization (91%). Slightly more than 8 in 10 Medicare Advantage enrollees were in plans that required prior authorization for opioid treatment program services (85%), therapy sessions with other mental health providers besides psychiatrists (sometimes referred to as mental health specialty services; 84%), therapy sessions with a psychiatrist (84%), and outpatient substance abuse disorder services (83%) (Figure 1).

Health insurers use prior authorization to both contain spending and prevent enrollees from receiving unnecessary or low-value services, though there are some concerns these requirements may create barriers and delays to receiving necessary care. However, data are not available to analyze prior authorization and appeal rates by type of service, including mental health and substance use disorder services, so it is unclear to what extent prior authorization creates barriers to obtaining care.

To address some concerns related to the use of prior authorization in Medicare Advantage, CMS finalized a rule in April 2023 that would clarify the coverage criteria Medicare Advantage plans can use when making prior authorization determinations, including that they follow traditional Medicare coverage guidelines when making medical necessity determinations, which would apply to both mental health and non-mental health related services. Further, CMS clarified that mental health services needed in emergency medical situations cannot be subject to prior authorization.

A proposed rule issued by CMS in December 2022 would institute an electronic prior authorization process in Medicare Advantage and increase the speed at which Medicare Advantage plans must respond to prior authorization requests, which would also apply both to mental health and non-mental health related services. This proposal is similar to legislation that passed the House during the 117^th Congress.

referRals

Medicare Advantage plans can also impose referral requirements, in which a primary doctor must provide a written letter in order for a patient to see a specialist for services, and if not provided, then the plan generally does not cover the cost. In contrast, traditional Medicare generally does not require prior referrals for services.

About one in four Medicare Advantage (26%) enrollees were in plans that required referrals for some mental health and substance use disorder services in 2022, including inpatient hospitalization in a psychiatric hospital (22%), partial hospitalization (21%), opioid treatment program services (20%), therapy with a psychiatrist (19%), outpatient substance abuse disorder services (19%), and therapy sessions with other mental health providers besides psychiatrists (18%) (Figure 1).

While some data on the use of referrals in Medicare Advantage is available, little is known if and how they impact enrollees’ access to care or create barriers to care.

Data Gaps and Limitations

There are data gaps that limit the ability of researchers and policymakers to understand the full extent of how mental health and substance use disorder benefits in Medicare Advantage plans are working for enrollees.

Access to Mental Health and Substance Use Disorder Providers: There is not much information on whether Medicare Advantage enrollees are experiencing barriers accessing mental health providers in their plan’s network and the extent to which enrollees use in-network and out-of-network providers for these services. Prior KFF research has shown that access to psychiatrists was more restricted than for other physician specialties in Medicare Advantage plans, but the analysis did not examine network inclusion of other types of mental health providers, such as clinical psychologists and clinical social workers. Psychiatrists are also less likely than other specialists to take new Medicare (or private insurance) patients, and they’re more likely than other specialists to “opt out” of Medicare altogether, meaning they do not participate in the Medicare program; this may impact availability of these providers in Medicare Advantage networks and also affects access to those providers for people in traditional Medicare.

Research from the Government Accountability Office (GAO) has also shown that consumers with various types of insurance (not specific to Medicare Advantage plans) face challenges accessing mental health care in-network, which may be due in part to providers not accepting new patients or long wait times to see providers. CMS recently finalized a rule that seeks to address some of these barriers, including applying Medicare Advantage network adequacy requirements to clinical psychologists and clinical social workers (in addition to psychiatrists), and notifying enrollees when their mental health providers are dropped mid-year from networks. However, little is still known about the extent Medicare Advantage enrollees specifically are experiencing barriers accessing providers in-network and how often enrollees seek out-of-network providers for these services.

Out-of-Pocket Liability for Enrollees’ Use of Mental Health and Substance Use Disorder Services: CMS does not publish data on Medicare Advantage enrollees’ out-of-pocket liability for services, including mental health or substance use disorder services. Having information about enrollee liability would facilitate comparisons of out-of-pocket spending both across plans and compared to traditional Medicare. It could additionally help compare whether there are differences in cost-burdens across subgroups of beneficiaries, including by race/ethnicity, sex, age, or diagnosed health conditions.

Prior Authorization Requests by Type of Mental Health and Substance Use Disorder Service: Data is not reported to CMS on prior authorization requests, denials, and appeals by type of service, enrollee characteristics, or by plan type, including for mental health and substance use disorder services. Therefore, it is not possible to assess whether prior authorization requests for certain types of mental health and substance use disorder services are denied more often by some plans than others, or whether prior authorization requests tend to be denied more for some types of services than others. Additionally, plans are not required to report prior authorization data by demographic characteristics of Medicare Advantage enrollees, which make it impossible to assess whether prior authorization requirements have a disproportionate impact on certain subpopulations of enrollees, which could affect access to care, out-of-pocket costs, and health outcomes. Further, the aggregate-level data reported by Medicare Advantage insurers to CMS and made available in a public use file are only available at the contract, rather than plan level. Contracts can include multiple types of Medicare Advantage plans – for example, sometimes combining those available for individual purchase with SNPs. By aggregating data in this way, it is not possible to assess variations in prior authorization practices across plans within a contract, including across plans that serve different populations.

Discussion

Medicare Advantage plans cover all mental health and substance use disorder services covered under Medicare Parts A and B. Coverage of these services may differ between traditional Medicare and Medicare Advantage plans because Medicare Advantage plans have flexibility to offer additional benefits, impose different cost-sharing amounts (if actuarially equivalent), employ utilization management tools such as prior authorization and referrals, and rely on defined networks of providers, all of which may vary from one plan to the next. Differences between traditional Medicare and Medicare Advantage, and among Medicare Advantage plans, can pose challenges for beneficiaries to decipher.

While there is great interest in how well Medicare Advantage plans are serving enrollees’ mental health and substance use disorder needs, lack of data limits the extent to which policymakers, researchers and beneficiaries can fully answer that question. There is little data on the breadth of provider networks and what that means for access to services or how frequently enrollees seek care out-of-network. There is no data available on out-of-pocket liability for these services in Medicare Advantage plans, nor how out-of-pocket spending compares to non-mental health services both in Medicare Advantage and compared to traditional Medicare. There is also no data on how often plans impose prior authorization for various mental health and substance use disorder services and how often that results in delays or denials of care by type of service, by enrollee characteristic, or by plan type.

As part of the Consolidated Appropriations Act, 2023, the GAO is directed to conduct a study comparing the mental health and substance use disorder benefits offered by Medicare Advantage plans to traditional Medicare and to other benefits offered by Medicare Advantage plans, which may provide more insight on some of these issues. The results of this study are expected in 2025.

Given the number of people aging onto Medicare, increasing enrollment in Medicare Advantage, and mental health needs among the Medicare population, ensuring access to affordable mental health and substance use disorder services will be an ongoing issue.

In response to the unprecedented nature of COVID-19, the federal government declared numerous types of emergencies, Congress enacted several pieces of legislation, and various executive actions were taken and waivers issued, which, collectively, established time-limited flexibilities and provisions designed to protect individuals and the health system during the pandemic. The effective end dates of many, though not all, of these flexibilities and provisions are tied to the public health emergency (PHE) declaration made pursuant to Section 319 of the Public Health Service Act, first declared in January of 2020.  Others are linked to the public health emergency declaration made under Section 564 of the Federal Food, Drug and Cosmetic (FD&C) Act; the declaration made under the Public Readiness and Emergency Preparedness (PREP) Act; and emergency and major disaster declarations made under the Stafford Act. In some cases, subsequent legislation has either delinked provisions from these declarations or otherwise changed their duration.

The Biden Administration recently announced that it will end the PHE on May 11, 2023 and FEMA has announced that the emergency incident period under the Stafford Act will also end on that date. Other related emergency declarations or provisions have already ended or are ending soon. The following table (Table 1) provides a timeline identifying key health-related flexibilities and provisions specified by these various measures, the specific measure that determines their end date, and their end date (an end date for the Section 564 declaration has not yet been announced).

In addition to the end of the flexibilities detailed in Table 1 below, there are also expectations that the federal supply of COVID-19 vaccines could be depleted or need to be replaced by an updated booster dose sometime this year and, similarly, the federal supply of COVID-19 treatments will also be depleted. At a result, COVID-19 vaccines and treatments are transitioning to the commercial market. Importantly, this change is not tied to the end of the public health emergency.

Jen Kates, Senior Vice President and Director of KFF’s Global Health & HIV Policy program presented “The End of the Public Health Emergency Declaration for COVID-19,” at the 2023 Preparedness Summit on April 24. The presentation describes key health-related flexibilities that will end because of the ending of the federal Public Health Emergency (PHE) declaration, as well as other emergency declarations and measures.

Zachary Levinson, Project Director of KFF’s Program on Medicare Policy, testified before the U.S. House of Representatives’ Committee on Ways and Means on April 26, 2023 as part of a hearing on Tax-Exempt Hospitals and the Community Benefit Standard. His testimony describes the value of tax exemption for hospitals, federal oversight of community benefits, concerns about the adequacy of government requirements, proposed policy solutions, and general tradeoffs of policies that seek to strengthen requirements for tax-exempt status.

A new KFF analysis finds that between 8 and 24 million people across the U.S. could be disenrolled from Medicaid during the unwinding of the program’s continuous enrollment provision.

The estimates draw on data collected through KFF’s recent survey of state Medicaid and CHIP officials, conducted with the Georgetown University Center for Children and Families. The survey focused on states’ eligibility and enrollment policies, and their approaches to the unwinding of pandemic-era protections that prevented states from disenrolling people from Medicaid during the public health crisis.  It included questions about how many people states estimate will lose Medicaid coverage in the months following the expiration of those protections in March.

Among states that responded to the survey, the midpoint estimate was that 18 percent would be disenrolled, suggesting that total Medicaid enrollment could fall by about 17 million people nationally during the unwinding, including 5 million children and 12 million adults. The new analysis offers three illustrative scenarios for how state-level Medicaid enrollment could decline between March 2023 and May 2024, ranging from 8 percent to 28 percent of total enrollees, which is plus or minus 10 percentage points from the midpoint and similar to the varying projections reported by states. While neither the lowest nor highest rate of disenrollment is an expected national outcome, the range may be helpful in examining what the variation could be across states.

The new findings are consistent with HHS estimates that as many as 15 million people will be disenrolled during the unwinding process, including 6.8 million who will likely still be eligible. (KFF’s analysis does not estimate how much of the Medicaid coverage loss would be among people who are still eligible versus those who are no longer eligible.)

The unwinding of the continuous enrollment provision will play out differently across the states based on policy choices states have made and variation in their administrative infrastructures. Some states have adopted multiple policies that are more likely to promote continued coverage among those who remain eligible. Other states have adopted fewer of these policies, which will likely lead to a larger number of people losing Medicaid coverage, including some who remain eligible.

According to another new KFF analysis, eight states meet at least eight of the nine key metrics that will support continued Medicaid coverage for those who remain eligible. Seven states meet only three or four of the metrics. The nine metrics come from the same survey of state Medicaid and CHIP officials and include things such as taking 12-14 months to complete all renewals, following up with enrollees who haven’t responded to a renewal request before terminating coverage, and completing 50 percent or more of renewals using ex parte, or automated, processes.

Among the 17 million children and adults who could be disenrolled from Medicaid, it remains uncertain how many will transition to other health coverage or become uninsured. A recent KFF analysis found that nearly two-thirds of people experienced a period of uninsurance after being disenrolled from Medicaid or the Children’s Health Insurance Program (CHIP).

The two new analyses, “How Many People Might Lose Medicaid When States Unwind Continuous Enrollment?” and “State Policy Choices Are Likely to Affect the Extent of Medicaid Enrollment Declines During the Unwinding Period,” as well as other data and analyses about the Medicaid unwinding, are available at kff.org.

Starting April 1, 2023 some states resumed disenrolling people from Medicaid after a three-year period during which states provided continuous enrollment in exchange for enhanced federal funding. Between February 2020 and March 2023, Medicaid enrollment grew by an estimated 20 million people, contributing to declines in the uninsured rate, which dropped to the lowest level on record in early 2022. As states unwind the continuous enrollment provision, we estimate, based on a recent KFF survey, that 17 million people could lose Medicaid coverage – including some who are no longer eligible and others who are still eligible but face administrative barriers to renewal. Decreases in Medicaid enrollment could reverse the recent gains in insurance coverage overall.

This analysis estimates the number of people who could lose Medicaid during the unwinding period under three possible rates of Medicaid coverage loss, and shows for each illustrative rate, state-by-state coverage reductions among Medicaid children and adults. In practice, rates of Medicaid coverage loss will vary across the states, depending on states’ approaches to the unwinding and the extent to which they engage in outreach and assistance activities to minimize disenrollment among people who are still eligible.

The analysis uses a combination of enrollment data from the Centers for Medicare and Medicaid Services (CMS) Performance Indicator Project (PI data), Medicaid claims data (T-MSIS data), and some state-specific sources (see Methods for a detailed explanation of the analysis). We exclude people enrolled in the Children’s Health Insurance program (CHIP) because it is expected that some of the people who disenroll from Medicaid will enroll in CHIP and CHIP enrollment may increase during the unwinding. (In some states, CHIP enrollment decreased during the continuous enrollment period, potentially because some people who were eligible for CHIP remained instead on Medicaid.)

Our estimates are based on the best available public data on states’ Medicaid enrollment, but predicting the future here with incomplete data is highly uncertain and the scenarios described in this data note are intended to be illustrative only. The scenarios presented here represent a range of outcomes that states predicted in the recent KFF survey, but the range reflects predictions reported across all states that responded to the question. The analysis excludes estimates of people who may newly enroll in Medicaid during the unwinding period and among those who disenroll, how many will re-enroll within a short time (e.g., “churn”). Our analysis projects the rate of disenrollment, but net changes in Medicaid enrollment—which reflect new enrollment and churn—will be smaller than the number of people who lose coverage.

How Many People Might Lose Medicaid?

If Medicaid enrollment decreased by 18% in all states between March 2023 and May 2024, as suggested by a recent KFF survey, 17 million would lose Medicaid coverage (Figure 1). In KFF’s recent survey of states about their Medicaid eligibility policies, just over one-third of states were able to report their projected coverage losses associated with the unwinding. The midpoint of state responses was that 18% of Medicaid enrollees could be disenrolled over the course of the unwinding, responses were quite varied, ranging from 7% to 33%. We use 18% for this analysis to illustrate how much enrollment could decrease. But, recognizing the uncertainty and the expected variation across states, we present disenrollment numbers if the rate is 10 percentage points above or below the median, a range of 8% to 28% which is similar to the range coming from state responses.

An 18% decline in enrollment would translate to 17 million people losing Medicaid, of whom 5 million are children and 12 million are adults. These projected coverage losses are consistent with estimates from the Department of Health and Human Services (HHS) suggesting that as many as 15 million people will be disenrolled, including 5 million children and 10 million adults. HHS estimates that among the 15 million people who lose coverage, nearly 7 million will still be eligible. (KFF’s analysis models disenrollment but does not attempt to estimate the number of people who are still eligible.)

Although the midpoint of states’ estimated disenrollment rate was 18%, rates will range considerably across the states. If rates range from 8% to 28%, 8 million to 24 million people could lose Medicaid, including 2 million to 7 million children and 5 million to 17 million adults (see Appendix Table 1 for state-specific estimates). While neither the lowest nor highest rate of disenrollment is an expected national outcome, the range may be helpful in examining what the variation could be across states.

Expected disenrollments during the unwinding could reverse more than half of the Medicaid enrollment gains experienced during the continuous enrollment period (Figure 2). Comparing the number of people who could lose Medicaid with estimated enrollment growth during the continuous enrollment period, shows that the number of people losing Medicaid coverage would be nearly three-quarters of the enrollment increase under the midpoint scenario. Under this midpoint scenario, just over one-quarter of the enrollment increase during the pandemic would persist, meaning Medicaid coverage would be higher than pre-pandemic levels. The higher-range estimate of Medicaid coverage loss—which is not expected to occur in most states—suggests that in some states, Medicaid disenrollments during the 14-month unwinding period could exceed the enrollment gains from the 3-year continuous enrollment period. In some states, the number of people losing Medicaid could exceed the coverage gains because all current Medicaid enrollees will have to go through a renewal process.

What Might Happen to the Uninsured Rate During the Unwinding?

As states resume renewals for all Medicaid enrollees, there is substantial uncertainty regarding how many people will lose Medicaid and of those, how many will transition to other coverage or become uninsured. A large share of people are covered by Medicaid, and one analysis shows that over half of children are covered by Medicaid and CHIP. Significant declines in Medicaid enrollment are likely to increase the number of uninsured, and other estimates suggest that there will be increases in the uninsured rate, in addition to increased enrollment in private health insurance and CHIP. The extent to which the uninsured rate may rise is difficult to predict based on current data sources. For example, because enrollment in Medicaid has been automatically sustained during the pandemic, it is possible that some people who are identified as enrollees with administrative data may have since transitioned to a job with employer-provided health coverage. And, federal surveys that are used to measure the uninsured rate – which rely on respondents self-reporting their health insurance status – show smaller increases in Medicaid enrollment than are observed in administrative data.

Even if people who disenroll from Medicaid are eligible for CHIP or for subsidies to purchase nearly-free private insurance through the ACA, they may not know they are eligible or transition to other coverage without a gap in coverage. Our recent analysis of coverage outcomes after disenrolling from Medicaid or CHIP found that nearly two-thirds of people experienced a period of uninsurance. Actual outcomes will vary across states depending on an array of state policy decisions. State efforts to provide outreach and enrollment assistance can help ensure that those who remain eligible for Medicaid—an unknown share of current enrollees—are able to retain coverage, and that those who are no longer eligible transition to other coverage.

Appendix Table

The Nation Overall Also Experienced Higher Rates of Premature Deaths than Peer Countries

During the COVID-19 pandemic, people of color on average died at younger ages than White adults, resulting in substantial racial disparities in premature death and years of life lost, a new analysis finds.

The analysis examines the Centers for Disease Control and Prevention’s excess deaths statistics during the pandemic, which captures not only deaths caused by COVID-19 but also higher-than-expected deaths from other causes such as drug overdoses, suicide, health disease and liver disease. 

During the pandemic, all racial and ethnic groups experienced an increase in their premature mortality rate (defined as per capita deaths among people under age 75), though this rate rose more sharply among people of color than among White people. 

Expressed another way, White people who died prematurely (under age 75) during the pandemic on average lost 12.5 years of life, significantly less than the average years lost among American Indian and Alaska Native (22), Hispanic (19.9), Native Hawaiian and Other Pacific Islander (18.8), Black (18.3), and Asian (14) people.  

Collectively, the analysis estimates that people in the United States lost 14.8 million years of life due to excess deaths from March 2020 when the pandemic began through December 2022, with a disproportionately large share of those lost years affecting people of color.A separate analysis compares the rates of premature deaths in the United States with other large and wealthy peer countries through 2021 (the most recent year for which international data are available). The U.S. had the highest rate of premature death among the group of 12 nations during the two-year period. The U.S. on average had more than two times the average years of life lost per 100,000 people as the United Kingdom, the country with the next highest rate. The two analyses, “Racial Disparities in Premature Deaths During the COVID-19 Pandemic” and “Premature Mortality During COVID-19 in the U.S. and Peer Countries,” are both available through the Peterson-KFF Health System Tracker, an information hub dedicated to monitoring and assessing the performance of the U.S. health system.