Budgetary Effects of Policies to Modify or Eliminate Medicaid’s Institutions for Mental Diseases Exclusion

Medicaid is a joint federal-state health insurance program for people with low income. States administer the program pursuant to certain federal rules and regulations, and the federal government makes matching payments to the states to cover a share of the costs. Under a policy known as the institutions for mental diseases (IMD) exclusion, the federal government does not make matching payments to states for expenditures for services provided to Medicaid enrollees ages 21 to 64 who are in certain types of inpatient facilities. Federal reimbursement is available, however, under several exceptions to the IMD exclusion. States make extensive use of those exceptions.

In this report, the Congressional Budget Office estimates the budgetary effects of two options, each with three variants, for expanding federal Medicaid payments for those excluded services.

• Under current law, states may amend their Medicaid plan and receive federal matching funds through September 30, 2023, for care for Medicaid enrollees ages 21 to 64 with at least one substance use disorder (SUD) in eligible IMDs if several criteria are met.1 Permanently extending that option (referred to as the “state plan option” throughout this report) would increase federal Medicaid expenditures by $155 million to $560 million, on net, over the 2024–2033 period; the range reflects three alternative specifications of the option that CBO examined.
• Eliminating the IMD exclusion would increase federal Medicaid expenditures by larger amounts. Eliminating the exclusion for stays for SUDs would increase those expenditures by $7.7 billion, on net, over the 2024–2033 period; eliminating the exclusion for stays for mental health disorders would increase those expenditures by $33.5 billion, on net; and eliminating it for both types of stays would increase those expenditures by $38.4 billion, on net.

Under all of the options that CBO examined, outlays would increase because of greater federal spending for inpatient and long-term care services. Those costs would be partially offset by slightly less spending for emergency department visits. The estimates are uncertain because state-level policy decisions and the prevalence of the disorders are difficult to project, among other reasons.

Each policy option could affect people’s access to care and their ability to afford it, providers’ capacity, and the quality of care. A detailed analysis of those effects is outside the scope of this report.

Behavioral Health Care for Medicaid Enrollees and the IMD Exclusion

In calendar year 2019, about 35 percent of adult Medicaid enrollees had a behavioral health condition, which can include mental illnesses (such as anxiety disorder, depression, bipolar disorder, and schizophrenia) or an SUD (such as alcohol use disorder and opioid use disorder). Treatment for those conditions can be provided in outpatient, residential, and inpatient settings. The severity of a person’s behavioral health needs, along with other factors, can affect which care setting is most appropriate and how long it is needed. Outpatient care may be used by people who are willing to participate in recommended services, whose living arrangements are stable, who have access to transportation, and who have a supportive social network. Residential treatment can be appropriate for people with serious behavioral health conditions who have not improved in outpatient settings or whose work or living arrangements are not stable and who have limited or no support from their social network. Inpatient care can be appropriate for people experiencing severe behavioral health symptoms who require continuous care.2 That care is more structured and medically oriented than residential care.3

Medicaid’s IMD Exclusion

Medicaid covers many behavioral health services, including physicians’ services, inpatient and outpatient hospital services, and prescription drugs. One limitation to that coverage is the IMD exclusion, which has been in place in the Medicaid statute since 1965.4 The IMD exclusion prohibits the federal government from making matching payments to state Medicaid programs for services provided to enrollees residing in IMDs. According to the law, IMDs are defined as hospitals, nursing facilities, and other institutions that have more than 16 beds and are primarily engaged in diagnosing, treating, or caring for people with mental diseases, including SUDs.

The IMD exclusion applies to adults ages 21 to 64, and it applies not only to services furnished by an IMD but also to services delivered outside the facility to current patients of IMDs.5 States have the option to cover services in IMDs for people younger than 21 or older than 64 and to have the federal government cover part of the cost.6 States can also receive federal matching payments for services provided outside of IMDs for pregnant and postpartum people who are eligible for Medicaid on the basis of being pregnant and who are receiving SUD treatment in IMDs. Analysis of the under-21, over-64, and pregnant and postpartum populations is beyond the scope of this report.

Exceptions to the IMD Exclusion

Four exceptions to the IMD exclusion make federal financing available for IMD stays and are used by many states. (States may also pay, without a federal match, for services Medicaid enrollees ages 21 to 64 receive while in IMDs.) Three of the exceptions are permanently available: the use of disproportionate share hospital (DSH) payments, section 1115 demonstration waivers under the Social Security Act, and Medicaid managed care “in-lieu-of” authority.

DSH Payments. Federal law requires state Medicaid programs to make DSH payments to qualifying hospitals that serve a large number of Medicaid enrollees and uninsured individuals. States can make such payments for uncompensated care at IMDs.7 As of 2022, 32 states and the District of Columbia used that authority; the majority of states used less than 20 percent of their total DSH payments for that purpose.8

Section 1115 Waivers. Section 1115 of the Social Security Act gives the Secretary of Health and Human Services authority to approve experimental, pilot, or demonstration projects that the Secretary finds to be likely to assist in promoting the objectives of the Medicaid program. Section 1115 waivers are available for mental health and for SUDs; they allow states to receive federal reimbursement for treatment for enrollees who are patients in IMDs. For waivers to be approved by the Centers for Medicare & Medicaid Services (CMS), states must commit to achieving certain milestones, such as use of evidence-based placement criteria and improved care coordination and transition between levels of care. Furthermore, federal policy guidance indicates that states adopting section 1115 waivers for SUDs should aim for a 30-day statewide average length of stay for residential treatment. For mental health, the section 1115 waivers have an expected statewide average length of stay of 30 days, with federal reimbursement limited to stays of no more than 60 days.9 As of December 2022, 34 states and the District of Columbia, representing 73 percent of the Medicaid enrollees ages 21 to 64, had a section 1115 waiver. Of those states, one state had only a mental health waiver, 25 states had only an SUD waiver, and 8 states and the District of Columbia had both types of waivers.

Managed Care in-Lieu-of Services. States with Medicaid managed care plans can pay for treatment in IMDs as an in-lieu-of service, which is a service that is not included under the state plan but that is a clinically appropriate, cost-effective substitute for a similar, covered service. Under that authority, federal matching funds are available for the monthly payments to managed care plans for enrollees ages 21 to 64 who have an IMD stay if certain criteria are met. For example, the services in the IMD must be medically appropriate, the enrollee must voluntarily select them, the plan must offer the services on an optional basis, and they must be limited to no more than 15 days per month.10 The Substance Use-Disorder Prevention that Promotes Opioid Recovery and Treatment for Patients and Communities Act (SUPPORT Act, Public Law 115-271) recently codified that provision (and any successor regulation) into statute. As of 2020, 32 states and the District of Columbia used that authority.

State Plan Option. A fourth authority, a state plan option, is available to states on a temporary basis: Effective from October 1, 2019, to September 30, 2023, section 5052 of the SUPPORT Act allows states to amend their Medicaid plan and receive federal matching funds for care for Medicaid enrollees ages 21 to 64 with at least one SUD in eligible IMDs if several criteria are met. For example, states must meet maintenance-of-effort requirements and cover early intervention, outpatient, intensive outpatient, partial hospitalization, residential, and inpatient services.11 IMDs must follow evidence-based practices and provide at least two forms of medication for opioid use disorder to be eligible. Federal reimbursement is available for up to 30 days per 12-month period per eligible enrollee. By the end of 2022, three states had adopted that state plan option.

Combinations of Exceptions. States have adopted varying combinations of optional exceptions to the IMD exclusion. According to data available as of February 2023, two states had none of the exceptions to the IMD exclusion (see Figure 1). Nine states had one exception. Of those states, one used DSH payments, three used any section 1115 waiver, and five used in-lieu-of authority. Twenty-four states and the District of Columbia had two exceptions, and 15 states had three exceptions—all a combination of DSH payments, section 1115 waivers, and in-lieu-of authority. The three states with the state plan option available under the SUPPORT Act also used another exception policy: One state also had a section 1115 waiver, a second also used DSH payments, and a third state also used in-lieu-of authority.12 No states had all four exceptions.

Behavioral Health Care for Medicaid Enrollees

Despite Medicaid’s coverage of behavioral health services, including exceptions to the IMD exclusion, many Medicaid enrollees with such conditions do not receive treatment. Among Medicaid enrollees age 12 or older with an SUD in calendar year 2019, fewer than one in five reported receiving any treatment in the past year, and approximately 85 percent reported needing treatment for alcohol or drug use in the past year but not receiving care at a mental health center, inpatient treatment at a hospital, or in- or outpatient treatment at a rehabilitation facility.13 Of Medicaid enrollees age 18 or older with serious mental illness in calendar year 2019, about 70 percent reported receiving any treatment in the past year, and about 45 percent reported perceiving an unmet need for treatment. Low treatment rates for behavioral health conditions are not limited to Medicaid. In calendar year 2019, for example, about 90 percent of adult Medicare enrollees with an SUD reported having a treatment gap, and about 30 percent of adult Medicare enrollees with serious mental illness reported perceiving an unmet need for treatment.14

Policy Options

This report considers two options, each with three variants, for modifying the federal payment policy for IMDs. Changes to that policy could expand the options available for care of SUDs and mental health disorders and thus increase the ability of people to receive care in the most appropriate setting for their symptoms and circumstances. Better access to a continuum of care could reduce unnecessary use of services, such as emergency department admissions. For example, CMS identifies reduced use of the emergency department as a goal of section 1115 waivers.15

Option 1: Permanently Extend the State Plan Option Available Under the SUPPORT Act

The first option would permanently extend section 5052 of the SUPPORT Act, which allows federal reimbursement for services for enrollees with SUD in eligible IMDs under a state plan option. States choosing to adopt the state plan option—available for diagnosis, treatment, or care for SUDs—would have to develop such a plan and have it approved by CMS. In addition, states would be subject to maintenance-of-effort requirements, and qualifying IMDs would have to follow evidence-based practices and offer at least two forms of medication for opioid use disorder.

CBO analyzed the budgetary effects of three alternatives for permanently extending the state plan option available under the SUPPORT Act:

• Permanently extend the state plan option, with no change to the allowed duration of stays (up to 30 days per 12-month period per eligible individual) or size of facilities (more than 16 beds) eligible for federal reimbursement;
• Permanently extend the state plan option and expand federal reimbursement to stays of up to 60 days per 12-month period but keep the size of the facilities unchanged (at more than 16 beds); or
• Permanently extend the state plan option, with no change to the allowed duration of stays (up to 30 days per 12-month period), but change the size of the facilities eligible for federal reimbursement to include only those with 17 to 39 beds.16

Option 2: Eliminate the IMD Exclusion

CBO also analyzed the effects of a second option, which would eliminate the exclusion for IMD stays in three possible ways:

• Eliminate the IMD exclusion for SUD stays,
• Eliminate the IMD exclusion for mental health stays, or
• Eliminate the IMD exclusion for both types of stays.

In CBO’s estimates, the policy options would be implemented beginning in October 2023 (the start of fiscal year 2024).

Effects on the Federal Budget

Modifying or eliminating the IMD exclusion would increase Medicaid outlays. CBO estimates that increases in federal outlays between 2024 and 2033 would be between $155 million and $38.4 billion depending on the alternative (see Table 1). (The estimates are relative to CBO’s February 2023 baseline budget projections.) In 2033, the estimated increase in outlays would represent between less than 0.01 percent and 0.5 percent of projected federal Medicaid spending in that year. The basis for the estimates is described in the next section.

Permanently Extend the State Plan Option Available Under the SUPPORT Act

CBO estimates that permanently extending the state plan option available under section 5052 of the SUPPORT Act would increase federal outlays by between $155 million and $560 million over the 2024–2033 period. Those estimates reflect CBO’s assessment of how many states would take up the permanent state plan option and the effects on Medicaid spending of allowing for federal reimbursements of IMD stays in those states. Each alternative applies only to IMD stays for the diagnosis, treatment, or care of people with SUDs.

Permanently Extend the Current State Plan Option. CBO estimates that making the state plan option permanently available to states as originally enacted—for stays of up to 30 days per 12-month period per eligible individual and for facilities with more than 16 beds—would increase outlays by $535 million, on net, over the 2024–2033 period. Spending on inpatient and long-term care would increase by $540 million and would be partially offset by a reduction of $5 million in spending for care in emergency departments.

Permanently Extend the State Plan Option for Stays of up to 60 Days per 12-Month Period. CBO estimates that making the state plan option permanent but increasing the number of days that could be eligible for federal reimbursement would increase outlays by $560 million over the 2024–2033 period, $25 million more than the estimate for the first variant. Spending on inpatient and long-term care would increase by $570 million, which would be partially offset by a reduction of $10 million in spending for care in emergency departments.

Permanently Extend the State Plan Option for Facilities With 17 to 39 Beds. CBO estimates that making the state plan option permanent but restricting the number of beds for a facility to qualify for federal reimbursement would increase outlays by $155 million over the 2024–2033 period, $380 million less than the estimate for the first variant. Spending on inpatient and long-term care would increase by $155 million, which would be partially offset by a reduction (of less than $5 million) in spending for care in emergency departments during that same period.

Eliminate the IMD Exclusion

Eliminating the IMD exclusion would have a much larger budgetary impact. CBO estimates that such a change would increase federal outlays by between $7.7 billion and $38.4 billion over the 2024–2033 period, depending on the types of stays encompassed by the policy. Those estimates reflect CBO’s assessment of the effects on Medicaid spending of allowing for federal reimbursement of IMD stays, and they account for varied effects among states with different policies under current law and adjustments to account for how people with different behavioral health diagnoses use services.

The effects on the federal budget of repealing the IMD exclusion only for SUD stays are expected to grow over the 2024–2033 period; by contrast, the effects of repealing the exclusion only for mental health stays are projected to decline over the period. In each case, growth in enrollment and cost of services over time boosts the size of the estimated effect. However, the effects are attenuated in the later years by CBO’s expectation that the share of Medicaid enrollees living in states with waivers will grow under current law. The effects of policies that eliminate the IMD exclusion would be smaller in states that are expected to adopt waivers, because most of the associated increase in federal reimbursement would occur under current law.

CBO expects the increase in the share of enrollees living in states with a mental health waiver to be larger than the increase in the share of enrollees living in states with an SUD waiver: Those shares are projected to rise from 13 percent in 2022 to 85 percent in 2033 and from 73 percent to 85 percent, respectively. Altogether, the mitigating effect of the increased adoption of section 1115 waivers under current law for mental health stays is larger than the effect of increased enrollment and costs over time, so the projected effects of repealing the exclusion only for mental health stays decline over the coming decade. In the case of repealing the IMD exclusion for SUD stays, the effects of enrollment and costs are not fully mitigated, and the effect of the policy grows over time.

Eliminate the IMD Exclusion for SUD Stays. Eliminating the IMD exclusion only for SUD stays would increase federal Medicaid outlays by $7.7 billion, on net, over the 2024–2033 period, CBO estimates. Spending for inpatient and long-term care would increase by $7.9 billion, and spending for care in emergency departments would fall by $235 million. That net increase in outlays reflects changes in the use of SUD services that would result from the policy, taking into account state policies and the use of SUD services that CBO projects under current law.

Eliminate the IMD Exclusion for Mental Health Stays. Eliminating the IMD exclusion only for mental health stays would increase federal Medicaid outlays by $33.5 billion, on net, over the 2024–2033 period, resulting from an increase of $33.6 billion in spending for inpatient and long-term care and a reduction of $100 million in spending for care in emergency departments. The net increase in outlays reflects changes in the use of mental health services that would result from the policy, taking into account state policies and the use of mental health services that CBO projects under current law.

Eliminate the IMD Exclusion for SUD Stays and Mental Health Stays. CBO estimates that eliminating the IMD exclusion for SUD stays and mental health stays would increase federal Medicaid outlays by $38.4 billion, on net, over the 2024–2033 period. The net spending increase would be the result of an increase of $38.7 billion in spending for inpatient and long-term care and a decrease of $330 million in spending for care in emergency departments. The $38.4 billion in additional spending under this alternative is lower than the sum of the $7.7 billion and $33.5 billion in costs estimated for the two alternatives discussed above because some Medicaid enrollees receive treatment for both SUD and mental health disorders.

Analytic Approach

The estimated effects of the policies account for two important factors. The first is the share of Medicaid enrollees in states with section 1115 waivers under current law, because policies that modify or repeal the IMD exclusion would have a larger effect in states without those waivers. Between 2017 and 2022, the share of Medicaid enrollees in states with only a section 1115 SUD waiver rose from about 30 percent to about 60 percent, the share of Medicaid enrollees in states with only a section 1115 mental health waiver rose from zero to about 1 percent, and the share of Medicaid enrollees in states with section 1115 waivers for both SUD and mental health rose from zero to about 13 percent. According to CBO’s estimates, if current laws remained in place in 2033, 5 percent of enrollees ages 21 to 64 would live in states with SUD waivers only, 5 percent would live in states with mental health waivers only, and almost 80 percent would live in states with SUD and mental health waivers. The rising share of enrollees ages 21 to 64 living in states with waivers means that the effects of future policies that further relaxed or eliminated the IMD exclusion would be smaller because some of those effects are expected to happen under current law.17

The second key factor is CBO’s estimates of how Medicaid outlays would change in response to the policy. Those estimates are based on the agency’s analyses, using data from the Transformed Medicaid Statistical Information System (T-MSIS), of the effects of state adoption of section 1115 SUD waivers. Specifically, CBO compared changes in certain Medicaid claims between 2016 and 2021—those for inpatient, emergency department, and long-term care services with an SUD diagnosis—in states that adopted waivers in 2019 and states that did not have waivers.18 The methodology, which captured the effects of waivers among states with and without in-lieu-of policies, can provide an estimate of the effects of relaxing or eliminating the IMD exclusion because such waivers allow participating states to receive federal payments for the diagnosis, treatment, or care of SUD for Medicaid enrollees ages 21 to 64 in IMDs.19

Effects of Adoption of Section 1115 SUD Waivers

CBO examined changes in Medicaid claims with an SUD diagnosis for inpatient care, which includes care provided in IMDs and other facilities such as acute care hospitals and psychiatric hospitals that do not qualify as IMDs. The agency also examined SUD claims for long-term care facilities and emergency department services. The analyses thus capture the direct effects of the adoption of waivers on IMDs as well as spillover effects on other types of services and facilities.20

Those analyses indicated that section 1115 waivers resulted in increased federal spending for Medicaid enrollees after 2019. Specifically, they showed that the number of days of SUD care covered by Medicaid in inpatient and long-term care settings among enrollees with SUD claims increased by more than 50 percent in states that adopted section 1115 waivers in 2019 relative to the trend in states without waivers. In CBO’s assessment, the increase in covered days of care (and the associated increase in federal spending) stemmed from shifts in the payer for inpatient and long-term care (a larger federal share because of the availability of federal matching funds), increased acceptance of Medicaid payment among providers, and increased use of such care. The analyses also showed a reduction in visits to emergency departments. Reduced spending for those services is expected to slightly offset the increases for inpatient and long-term care.21

Effects of Policy Options

Because section 1115 SUD waivers differ from the options examined in this report, CBO adjusted the estimates from the analysis of section 1115 SUD waivers before using them to gauge the effects of the policy options discussed in this report.

Permanently Extend the State Plan Option. Although they have several characteristics in common (such as the use of evidence-based practices and a focus on transitions to community-based services), section 1115 waivers and the state plan option available under the SUPPORT Act differ in certain respects. The state plan option allows for federal reimbursement for fewer days of care than waivers do and requires IMDs to provide at least two forms of medication for opioid use disorder. In addition, waivers require an implementation plan, monitoring of performance measures, and evaluation of their effects—steps that are not required when amending state plans.

Because of the similarities and differences between section 1115 waivers and the state plan option, CBO expects that, if the state plan option became permanently available, some states might adopt that option rather than section 1115 waivers. In addition, a few other states that are not expected to adopt section 1115 waivers under current law would adopt the state plan option to receive federal payments for IMD services. CBO projects that the share of Medicaid enrollees in states with a state plan option—states that would not have adopted a section 1115 waiver under current law—would reach 5 percent in 2033 (an increase from 1 percent in 2022).

To estimate the effects of extending the state plan option on SUD claims for services for Medicaid enrollees, CBO made downward adjustments to the estimated changes in spending in states that adopted section 1115 SUD waivers. The overall size of the downward adjustment ranged from about one-third to three-quarters, depending on the variant. The largest reduction applied to the variant that would restrict the number of beds for eligible facilities, because that is the most restrictive variant.

CBO’s assessment was informed by several factors. For example, for all three versions of the policy, CBO made a downward adjustment to the estimated effect of waivers to account for the state plan option’s requirement of at least two forms of medication for opioid use disorder. According to CBO’s analysis of the 2019 National Survey of Substance Abuse Treatment Services conducted by the Substance Abuse and Mental Health Services Administration, approximately 60 percent of relevant facilities provided at least one form of such medication, and roughly half provided two.

Further, because the allowable length of stay is, on average, longer for section 1115 waivers than under the SUPPORT Act policies, CBO reduced the estimated changes in spending in states that adopted section 1115 SUD waivers on the basis of an analysis of stays in residential facilities among all Medicaid enrollees using T-MSIS claims. CBO found that, in 2021, 98 percent of inpatient stays and 87 percent of long-term care stays for SUD were less than 60 days; 5 percent of inpatient SUD stays and 8 percent of long-term care SUD stays were between 30 days and 60 days.

CBO applied a third downward adjustment to the third alternative, which would extend the state plan option permanently for facilities with 17 to 39 beds, on the basis of its analysis of data collected by the Substance Abuse and Mental Health Services Administration. According to those data, approximately 27 percent of days of care for patients with SUD were in facilities with 17 to 39 beds in 2016. By comparison, approximately 14 percent of days of care for patients with SUD were in facilities with 16 beds or fewer. In addition, CBO expects that some facilities would reduce their number of beds to qualify for federal reimbursement, contributing to the increase in Medicaid spending.

Eliminate the IMD Exclusion for SUD Stays. The effects of eliminating the IMD exclusion for SUD stays would vary depending on whether states have section 1115 SUD waivers, because those waivers allow for federal reimbursement for IMDs under certain conditions under current law.

States Without Waivers. In states without section 1115 SUD waivers under current law, CBO expects that eliminating the IMD exclusion would result in percentage changes in spending that are larger than the estimated changes in spending in states that adopted section 1115 SUD waivers because such repeal would not have the waivers’ restrictions, such as limits on length of stays in IMDs.22 In CBO’s assessment, that increase in spending would equal 115 percent of the estimated changes in spending associated with adoption of section 1115 SUD waivers.

States With Waivers. In states with section 1115 SUD waivers under current law, CBO expects that eliminating the IMD exclusion would result in some increase in spending because the waivers’ restrictions, such as limits on length of stays in IMDs, would no longer be in effect. In CBO’s assessment, that increase in spending would equal 15 percent of the estimated changes in spending associated with adoption of section 1115 SUD waivers.

Eliminate the IMD Exclusion for Mental Health Stays. Because CMS guidance on and, thus, state adoption of section 1115 waivers for mental health stays came after the guidance on and adoption of section 1115 SUD waivers, CBO could not directly quantify their effects. Instead, the agency based its estimates on the percentage changes in spending associated with the adoption of section 1115 SUD waivers, with certain adjustments.

States Without Waivers. In CBO’s estimate, in states without section 1115 mental health waivers under current law, eliminating the IMD exclusion would result in percentage changes in spending for mental health stays that are smaller than the effects of eliminating the exclusion for SUD stays described above. CBO expects that, similar to the adjustment for eliminating the IMD exclusion for SUD stays, the increase in spending associated with repeal would equal 115 percent of the estimated changes in spending associated with adoption of section 1115 SUD waivers.23 However, because more people with serious mental illness report receiving treatment than do people with SUD under current law, CBO multiplied that estimated effect by 75 percent. That adjustment reflects a shift in the payer of care—from state governments to the federal government, like the one estimated for states adopting section 1115 SUD waivers—and also accounts for smaller increases in use of care. Altogether, in CBO’s estimate, eliminating the exclusion for mental health stays would result in an increase in spending that equals 86 percent of the estimated changes in spending associated with adoption of section 1115 SUD waivers.

States With Waivers. In CBO’s estimate, in states with section 1115 mental health waivers under current law, eliminating the IMD exclusion would result in a small increase in spending because the waivers’ restrictions, such as limits on length of stay in IMDs, would no longer be in effect. In CBO’s assessment, that increase in spending would equal 11 percent of the estimated changes in spending associated with adoption of section 1115 SUD waivers after accounting for a greater share of people with serious mental illness than with SUD who report receiving treatment under current law.24

In addition to the difference in the estimated percentage changes in spending for policies that would eliminate the exclusion on the basis of whether the policies applied to SUDs or mental health disorders, the estimated effects of the policies would differ because Medicaid spending on SUD is less than Medicaid spending on mental health disorders. According to CBO’s analysis, in 2022, Medicaid spending on inpatient, emergency room, and long-term care was about $5 billion for SUD claims and about $14 billion for mental health disorders claims.

Uncertainty About the Estimates

An important source of uncertainty stems from the difficulty in predicting whether and which of the exceptions to the IMD exclusion states may adopt over the 2023–2033 period under current law. If more or fewer states adopted policies that allow federal payment to be made for services delivered in IMDs than CBO projects, the budgetary effects of modifying the IMD exclusion would be smaller or larger than estimated in this report. For options that extend the state plan option available under the SUPPORT Act, there is additional uncertainty about the adoption of the state plan option. The specifics of legislation or regulations that modify the IMD exclusion could also affect the results.

Further sources of uncertainty about the estimated budgetary effects include whether existing and new IMDs can accommodate all Medicaid enrollees needing that type of care.25 For that reason, how health care utilization among enrollees with behavioral needs would change overall and by type of service is uncertain. Increased use of IMDs could change receipt of care in different ways than CBO anticipates, resulting in more or less health care spending. Additional uncertainty arises because section 1115 SUD waivers not only affect IMD payment policy but also require states to achieve certain milestones, such as use of evidence-based practices and improved coordination of care. The overlap of the IMD policy and other requirements of a single waiver prevented CBO from distinguishing between the effects of easing the IMD exclusion on facilities and other requirements of the waiver.

The prevalence of behavioral health conditions is another source of substantial uncertainty in CBO’s projections of behavioral care spending under current law.26 That uncertainty affects the agency’s estimates of policy options related to the IMD exclusion. The number of people with behavioral health conditions is currently elevated—a result, at least in part, of the opioid crisis and the coronavirus pandemic—and it is unclear whether it will remain elevated over the next decade. In addition, there is considerable uncertainty about the effects of actions that the federal and state governments have taken in recent years, or could take in future years, to address behavioral care needs.27 For example, the recently introduced 988 Suicide & Crisis Lifeline, which provides support to people in crisis, could change how some people access care.28 Finally, future changes in how people who interact with the criminal justice system receive health care services could affect health care spending for those people.29

Other Effects

In addition to their budgetary effects, options in this report could affect access to care and its affordability, how care is delivered, providers’ capacity, and quality of health care. Those effects are difficult to measure and are likely to vary depending on whether a given policy modifies the IMD exclusion by adopting the SUPPORT Act provision permanently or eliminates the exclusion.

Policies that modify or eliminate the IMD exclusion would probably result in increased access to inpatient and residential care for Medicaid enrollees because of increased acceptance of Medicaid by treatment facilities. One recent study found that acceptance of Medicaid at residential treatment facilities increased by 6 percentage points in the year after adoption of section 1115 SUD waivers and by 11 percentage points two years after adoption compared with facilities in states that did not adopt waivers.30 The same study found that acceptance of Medicaid at intensive outpatient treatment facilities increased by 5 percentage points to 7 percentage points after adoption of section 1115 SUD waivers compared with facilities in nonadopting states. Wider acceptance of Medicaid could improve affordability of services by reducing the amount that patients would need to pay out of pocket. However, it is also possible that the treatment facilities may not be able to accommodate the new demand generated by the policies because of constraints on the number of beds or providers.31

Modifications to the IMD exclusion, and any changes in access to care that result from those modifications, might affect the setting (or type of facility) in which Medicaid enrollees with SUDs or mental health conditions receive care, which could have spillover effects on providers’ capacity. Services provided in IMDs are part of a continuum of behavioral health care delivered in outpatient, residential, and inpatient facilities. With the IMD exclusion in place, some Medicaid enrollees may be receiving care in settings (such as general acute-care hospitals) that are available with federal reimbursement rather than in settings that are most appropriate for their conditions. Increased access to care in IMDs might shift where care is delivered, which could affect the capacity of certain providers. Such effects are in line with the decreased use of emergency department services for SUD that CBO observed in states that adopted section 1115 SUD waivers. Increased availability of federal funds might also affect the capacity of inpatient and residential service providers by enabling facilities to expand, although shortages in the behavioral health workforce could limit such effects.

The options discussed in this report could change the use of outpatient services. In CBO’s assessment, however, the evidence in its analysis and the academic literature is insufficient to quantify such changes and their effects on spending. Increased use of inpatient services could increase use of outpatient services if follow-up outpatient care is provided after an inpatient hospital stay. But use of outpatient services could decrease if utilization shifts from outpatient to inpatient or residential settings. Use of outpatient services could also decrease if care for certain individuals is delivered more efficiently in an inpatient or residential setting. In addition, relaxing the IMD exclusion could increase federal spending for outpatient services because the IMD exclusion prohibits (with certain exceptions) federal reimbursement for services delivered to people staying in IMDs, even if those services are delivered in an outpatient setting. CBO will continue to monitor the literature on this topic.

The quality of care could increase or decrease under policies that would change the IMD exclusion, depending on the IMD policies states currently have in place as well as the details of the policy options. For states with few or no exceptions to the IMD exclusion or that have adopted the temporary SUPPORT Act provision, a permanent adoption of that provision could increase the quality of SUD care. People in those states would gain or maintain access to IMDs, potentially expanding the continuum of care recommended for some people with SUD. Furthermore, quality of care might improve because of SUPPORT Act requirements that IMDs follow evidence-based practices. A report by the Medicaid and CHIP Payment and Access Commission found that several Medicaid officials reported improvements in the quality of care in states with section 1115 SUD waivers that required providers to meet evidence-based treatment guidelines similar to those in the SUPPORT Act.32 Another study found that section 1115 waivers for treatment of SUD were associated with increased use of medications for opioid use disorder at outpatient treatment facilities.33

The quality of care could decline for people in states with section 1115 waivers under current law if those states relaxed requirements for evidence-based care in response to the elimination of the IMD exclusion. In addition, quality of care could decline under policies that limit the length of stay in IMDs to a level that did not align with evidence-based clinical practice, or it could remain the same if current-law restrictions did not align with those practices.34 Evidence on how the quality of mental health care might change under those policies is not available.

---