SB0597: SUMMARY OF BILL ON THIRD READING (Date Completed: 3-14-22)

CONTENT

Senate Bill 597 (S-3) would amend the Social Welfare Act to do the following:

-- Require the Department of Health and Human Services (DHHS), by January 1, 2023, to develop and begin implementation of a phased-in plan to integrate the administration and provision of Medicaid physical health care services and behavioral health specialty services for behavioral health populations through the creation of specialty integrated plans (SIPs).

-- Require the plan to provide for full integration and administration of physical health care services and behavioral health specialty services and supports through SIPs by 2030.

-- Require the integration plan to meet certain criteria, such as requiring a SIP to contract with each community mental health services program (CMHSP) within its service area to provide behavioral health specialty services, and requiring a CMHSP to contract with each SIP in its service area to provide, directly or indirectly, behavioral health specialty services.

-- Require the integration plan to provide for the phased-in transition and enrollment of all eligible Medicaid beneficiaries from a specialty prepaid health plan (i.e., a public prepaid inpatient health plans (PIHP)) into a SIP within the timeline prescribed in the bill.

-- Require the DHHS, in consultation with one representative from each of the interested parties (listed in the bill), to develop key metrics used to determine whether an implementation phase was successful.

-- Require the DHHS, in consultation with the Behavioral Health Accountability Council, to monitor each implementation phase and to complete a formal evaluation of each phase within 20 months after its effective date.

-- Specify that the DHHS, except in a case of malfeasance or misfeasance, would have to require the PIHP system and the CMHSPs to maintain all current provider contractual arrangements throughout the transition phase.

-- Specify that a provision requiring Medicaid-covered specialty services and supports to be delivered by PIHPs chosen by the DHHS would apply until SIPs were available to provide the specialty services for all eligible Medicaid beneficiaries in accordance with the integration plan.

-- Require the DHHS, within two years after the bill's effective date, to consolidate the 10 specialty prepaid health plans into a single statewide entity that would have to manage Medicaid-covered specialty services and supports.

-- Require the DHHS to establish the administrative board structure requirements for the statewide entity and prescribe the administrative board's general composition.

-- Require the DHHS to seek a waiver from the Federal government to allow, and if the waiver were granted, allow a SIP to manage and arrange the delivery of comprehensive physical and behavioral health care services for Medicaid beneficiaries.

-- Modify the Act's provisions pertaining to performance bonus incentive pools to exclude or include SIPs, as applicable.

Senate Bill 598 (S-3) would amend the Mental Health Code to do the following:

-- Specify that a SIP would not be responsible for duties set forth in the bill until after completion of a successful transition, as determined at each integration phase by the Behavioral Health Accountability Council.

-- Specify that a SIP would be a separate entity that supported the CMHSPs and either was a managed care organization or a specified system of health care delivery and financing.

-- Require procedures and policies for SIPs to be set by December 1, 2022.

-- Require SIPs, upon implementation, to ensure that services were delivered in a manner to that demonstrated that they were based on recipient choice and involvement, and designed to divert individuals with serious mental illness, serious emotional disturbance, or developmental disability (DD) from possible incarceration when appropriate.

-- Require a SIP to ensure services were available for individuals with substance use disorder (SUD), and allow a SIP to contract with a Department-designated community health entity to provide SUD services in its service area, instead of requiring the entity to coordinate the provision of those services.

-- Make multiple changes to the Code to include SIPs as well as local public behavioral health entities as eligible providers of publicly funded behavioral health services.

-- Require a SIP (or a CMHSP, as currently required) to participate in the development of school-to-community transition services for individuals with serious mental illness, serious emotional disturbance, or DD.

-- Require the Director, beginning by January 1, 2026, or after implementation of a SIP, to designate a SIP to assume the responsibilities of overseeing the provision of SUD services for a county or region.

-- Require a multidisciplinary council to be established to select a Director of the Office of Recipient Rights on the date a SIP was implemented.

-- Establish the Office of the Behavioral Health Ombudsman as an autonomous entity within the DHHS and prescribe the duties of the Ombudsman.

-- Create the Behavioral Health Accountability Council within the Office of the Behavioral Health Ombudsman and prescribe its membership and duties.

Senate Bill 598 (S-3) is tie-barred to Senate Bill 597. Senate Bill 598 (S-3) also would repeal Section 269 of the Mental Health Code. (Section 269 allows the Department-designated community mental health entity and its CMHSP provider network to contract for and spend funds for the prevention of SUD and for the counseling and treatment of individuals with SUD.)

MCL 400.105d & 400.109f (S.B. 597) Legislative Analyst: Jeff Mann

330.1100d et al. (S.B. 598)

FISCAL IMPACT

Introduction

The bills gradually would shift responsibility for the provision of Medicaid mental health and SUD services (known collectively as "behavioral health") to SIPs. This shift would be phased in with at least 24 months between each phase. Services for children with severe emotional disturbance and those in foster care would be integrated in the first phase, followed by integration of services for seriously mentally ill adults in the second phase, followed by the integration of services for those with SUD needs in the third phase, then integration of services for those with intellectual or developmental disability in the final phase. Progression to each phase would depend on meeting metrics established by the DHHS as determined by evaluations completed by the Behavioral Health Accountability Council and the DHHS. If the DHHS determined that a phase was not successful, the next phase could be delayed and the current phase either could be extended or terminated.

The Senate Fiscal Agency (SFA) makes the following comments on the legislation and its fiscal impact if it were enacted:

-- There would be minor costs associated with implementation of the bills, including costs for the establishment of a behavioral health ombudsman.

-- There is no compelling evidence that implementation of the bills would lead to a significant change in administrative costs; data presented on this front comparing PIHP administrative costs to Medicaid Health Plan administrative costs are "apples to oranges" comparisons and not particularly revealing.

-- The identification of savings that occur to support expanded behavioral health services, at minimum, would be based on State savings from the two-way risk corridor, with SIP risk corridor savings accruing to the SIPs; this effectively would prevent the State from realizing savings from the integration.

-- The ability to achieve significant savings would be limited for some key behavioral health populations and, because of possible underserving of populations, services could expand, and costs could increase.

-- The rate-setting process could lead to an increase in rates to reflect expanded services that would be sufficient to allow for savings to occur, but the cost would be about $10.0 million GF/GP for each 1.0% increase in overall SIP rates.

-- While there are no explicit financial reserve or capital requirements for the SIPs in the legislation, the SIPs would be more likely to avoid financial problems than PIHPs are currently.

-- After full implementation, the bills (assuming baseline increases in other nonbehavioral health payments to local units of government) would put the State in danger of being below the "Headlee Amendment" minimum for payments to local units of government.

-- There would be a risk that the Insurance Provider Assessment (IPA), which provides a net benefit to the State's bottom line of about $450.0 million per year, would no longer meet a mandatory Federal statistical test and could be invalidated in the short or long term; revisions to the IPA to ensure it continued to meet the Federal statistical test could result in changes in net revenue benefit for the State and to the tax rate applied to private insurers.

General Aspects

The proposed legislation includes a provision allowing "reinvestment of realized savings into the integrated behavioral health system to further promote and expand access to clinically integrated services and locations", which would appear to allow for reinvestment of identified savings into program expansion. The legislation not only would shift administration and management of the provision of Medicaid behavioral health services to the SIPs but would make the SIPs administer and manage the provision of non-Medicaid behavioral health services that community mental health (CMH) boards and SUD coordinating agencies currently provide exclusively. Senate Bill 598 (S-3) Section 203 appears to terminate, upon the final transition to the integrated system, the role of the CMH boards and SUD coordinating agencies in the provision and management of non-Medicaid services as well as Medicaid services. The bills also would create an autonomous Office of the Behavioral Health Ombudsman in the DHHS.

The bills would have an unclear fiscal impact because, under the provision directing that the State's savings from the two-way risk corridor be reinvested in programming, there would appear to be no potential for State savings on services. This provision requiring transfer of identified savings (rather than identified savings reducing State costs) would mean that the bills' fiscal impact either would be neutral or result in increased costs to the State. The bills would lead to increased administrative costs for the DHHS because of the process of implementing integration and creating the Office, with a minimum estimated annual cost for the proposed Office of $100,000 General Fund/General Purpose (GF/GP). There also could be changes in administrative costs for managed care services tied to the likely shift of responsibility to managed care entities that have more ties to the physical health side, but, there are no relevant data indicating a meaningful difference in administrative costs. There also are considerations tied to solvency of current and potential future providers of Medicaid behavioral health services.

General considerations on savings and ways to achieve savings also must be considered. Many other states have partially integrated behavioral health services, with some services fully integrated with physical health and other services still carved out (in particular services to the intellectual/developmentally disabled population). The opportunities for savings may be more limited or more challenging to achieve for some populations. There also is the possibility of increased costs if populations currently covered by PIHPs are underserved. Past experience with the expansion of Medicaid managed care to populations with significant pre-existing conditions indicates that these populations may well be underserved and costs for behavioral health services could increase, with the trade-off being the likelihood of better outcomes and possibly reduced physical health costs. This has been evident in the shift of Children's Special Health Care Services to a mandatory managed care model and the initial shift of Medicaid behavioral health services to a managed care model. In the former case, the growth in cost per case paralleled medical inflation (albeit during a period when Medicaid fee screens were not being increased so cost pressures should have been less than general medical inflation). In the latter case there was significant cost growth from the outset in fiscal year (FY) 1998-99 and FY 1999-2000 before growth rates abated.^{^[1]}

There are two key potentially problematic fiscal considerations indirectly related to the legislation, one tied to payments to local units of government and the other related to the continued feasibility of a relatively new insurance provider tax. These will be discussed in detail later in this analysis.

Headlee Amendment Considerations. One of the most notable considerations relates to the requirement in the Michigan Constitution (the so-called "Headlee Amendment") that the percentage of State Spending from State Resources paid to local units of government be at least as great as the percentage in FY 1978-79. Shifting spending away from the local PIHPs to the new entities would reduce State payments to local units of government by well over $1.2 billion by the time the population was shifted fully from the PIHPs to the SIPs. This issue is further complicated by the July 2021 ruling by the Michigan Supreme Court in Taxpayers for Michigan Constitutional Government v. State of Michigan.^{^[2]} In that case, the Court held that payments to public school academies (PSAs, also commonly known as charter schools) should not count as payments to local units of government but also instructed the Court of Appeals to consider whether payments to the subset of PSAs chartered by local school districts or community colleges should be counted as payments to local units of government.

The Supreme Court ruling combined with the eventual shift of spending from PIHPs and non-Medicaid CMHSP and SUD funding to the SIPs could put the State below the Headlee limit by the time the integration legislation was fully implemented. There is still significant uncertainty because full implementation would not occur for several years. Over that time, there could be major changes above or below inflationary baseline to revenue sharing payments, the School Aid foundation allowance, community colleges funding, or local transportation funding that make any definitive statement on the Headlee impact impossible. The SFA can state, however, that the State's "surplus" in the Headlee calculation (about $2.9 billion in the enacted FY 2021-22 budget before any adjustments for the Supreme Court ruling), barring significantly above-inflation increases in other payments to local units, would nearly or completely disappear (see below for further discussion of this consideration).

Potential IPA Impact. There also is the possibility that the approach integration outlined in the bills could lead to difficulties in maintaining Federal approval of the IPA. Because of the prospective elimination of the PIHPs, the State could no longer be able to design a tax framework that meets the Federal statistical test needed for Federal approval of provider taxes that are not broad-based. If the IPA does not meet the Federal statistical test in any given year, there would be a negative net change to the State's balance sheet of over $450.0 million GF/GP in that year. It also is possible (and arguably likely) that the IPA could be adjusted in ways that would allow the IPA to meet the Federal statistical test, but these adjustments could reduce the State's net benefit from the tax or change the IPA tax rate for private insurance. The magnitude of these possible issues with the IPA is far more significant than any other fiscal issue and could overshadow any of the other cost or savings estimates (see below for further discussion of this consideration).

"Headlee Amendment" Implications

In 1978, Michigan voters adopted Proposal E, commonly known as the "Headlee Amendment" after its primary advocate. Proposal E amended the Michigan Constitution. One of the proposal's provisions requires that the proportion of State spending from State resources paid to local units of government be at least as great as that percentage was in FY 1978-79. That percentage eventually was calculated to be 48.97%. For purposes of ensuring compliance with this 48.97% minimum, each appropriations bill contains a Section 201 that includes estimates of State spending from State resources and the amount of State spending from State resources that is provided in that bill to local units of government. The State has consistently exceeded this 48.97% minimum since the Headlee Amendment was adopted.

School Aid, revenue sharing, payments to community colleges, and local transportation funding are among the largest components of payments to local units. Payments made in the DHHS budget to PIHPs and CMHSPs in the Autism Services, Healthy Michigan Plan—Behavioral Health, Medicaid Mental Health Services, and Medicaid Substance Use Disorder Services line items as well as other line items in the Behavioral Health portion of the DHHS budget are another significant source of these payments. At present, the State amount paid to the PIHPs is less than the long-term baseline because of the pandemic-related enhancement in the Medicaid match rate, which reduced the GF/GP dollars needed to support these line items. The SFA estimates that by FY 2022-23 State spending from State resources for line items that provide funding to PIHPs will be at least $1.29 billion. The SFA estimates that by FY 2022-23 State spending from State resources for line items that provide funding to CMHSPs will be $135.5 million. If the bills were enacted, the $1.29 billion in funding to PIHPs would be shifted to new nonlocal entities over the subsequent few years, which would reduce State spending from State resources paid to local units of government by up to that amount. Furthermore, because Section 203 of Senate Bill 598 (S-3) would eventually require the SIPs to take over the administrative and management functions for services provided by CMHSPs, State spending to local units of government would be reduced by another $135.5 million to reflect the transfer of all administrative and management responsibility and funding for non-Medicaid behavioral health services from local CMHSPs to the SIPs.

In the initial FY 2020-21 budget, State Spending from State Resources was $35.4 billion, with estimated payments to local units of government being $19.9 billion, leaving the State $2.6

billion over the Headlee amendment minimum of 48.97% (the Headlee minimum in this instance is calculated by taking 48.97% of $35.4 billion, which is $17.3 billion; $19.9 billion exceeds $17.3 billion by $2.6 billion).^{^[3]} Governor Whitmer's proposed FY 2021-22 budget featured a similar "Headlee surplus" of $2.6 billion. The FY 2021-22 budgets signed by Governor Whitmer in September 2021, thanks to a significant increase in School Aid funding, reflect a Headlee surplus of $2.91 billion. When adjusted for the eventual expiration of the COVID-19-related Federal enhancement of the Medicaid match rate and the official change in the FY 2022-23 Medicaid match rate , the baseline Headlee surplus is about $2.82 billion (see Table 1, "Adjusted for baseline FMAP" column). This $2.82 billion Headlee surplus estimate should not be viewed as a long-term precise number, but rather a best guess estimate of the medium-term baseline Headlee surplus before potential implementation of the bills.

As noted above, this issue is further complicated by the July 2021 ruling by the Michigan Supreme Court in Taxpayers for Michigan Constitutional Government v. State of Michigan.^[4] The Michigan Supreme Court's ruling could reduce State payments to local units of government by $1.30 billion to $1.49 billion, depending on whether payments to public school academies chartered by local school districts or community colleges would continue to count as payments to local units of government.

Removing this funding from the Headlee surplus reduces the baseline Headlee surplus to $1.33 billion to $1.52 billion. After full implementation, the $1.29 billion baseline State funding to PIHPs would be greatly reduced (and likely would be eliminated), which would reduce the Headlee surplus by a comparable amount and could well put the State close to the Headlee limit. The shift of non-Medicaid behavioral health services funding from the CMHs and coordinating agencies to the SIPs (which appears to be mandated upon the final phase of integration under the bills) would further reduce payments to locals by $135.5 million, which could put the State below the Headlee limit. The SFA notes that these changes, as shown in Table 1, could put the State below the limit because the calculation is a rough estimate and the bills would not be fully implemented for several years. As noted above, the Headlee baseline could be adjusted up or down by hundreds of millions if there were significant above- or below-inflation changes in the school foundation allowance, funding for community colleges, revenue sharing payments, or local transportation funding.

The SFA also notes that the same concerns about the Headlee surplus would apply to the package of behavioral health bills being considered by the House of Representatives (House Bills 4925 through 4929). Because of the more rapid implementation timeline for those bills, there would be a potential Headlee concern much sooner than would be the case with the Senate bills, although the House bills would not transfer responsibility for services currently provided by CMHSPs so the risk of being below the Headlee limit would be less.

Table 1 shows total payments to locals in the State budget, State spending from State resources, the Headlee limit, and the Headlee surplus or deficit in multiple columns. These columns show various adjustments that are made to derive baseline Headlee estimates by the time the bills would be fully implemented. The Headlee limit equals 48.97% of the State spending from State resources and the surplus/deficit is the payments to locals less the Headlee limit.

Savings Opportunities & Cost Concerns

One of the arguments made in support of greater behavioral health integration is the potential for efficiencies and other cost savings due to better coordination of care. This potential can vary due to the different types of populations served and their specific coverage situation. Generally speaking the most common behavioral health services are usually provided to five different categories of Medicaid recipients: mentally ill adults (MI-A), mentally ill children (MI-C), individuals with intellectual and developmental disabilities (IDD) or DD, dually diagnosed MI/IDD, and those facing an SUD. Several years ago, Medicaid began covering autism services as well, though for reporting purposes most of the payments for these services are in the IDD category.

Table 1: Impact of Senate Bills 597 (S-3) and 598 (S-3) on Headlee Amendment Calculations (in millions)
	FY 2021-22 Enacted	Expiration of Enhanced FMAP^*	Adjust for FY 2022-23 FMAP^**	Adjusted for Baseline FMAP	Remove Full Amt for PSAs^***	Adjusted for PSA Ruling
Payments to Locals...............	21,484.4	53.9	22.4	21,560.7	(1,488.0)	20,072.7
Total State Spending (TSS)...	37,930.5	240.0	100.0	38,270.5	0.0	38,270.5
Headlee Limit (48.97% of TSS).	18,574.5	---	---	18,741.0	---	18,741.0
Headlee Surplus/(Deficit).....	2,909.9	---	---	2,819.7	---	1,331.7
	Adjusted for PSA Ruling	Remove Adjusted PIHP Funding	Remove Adjusted CMHSP Funding	Adjusted for PSA PIHP & CMHSP	PSA Fund: K12/CC Chartered ^****	Adjusted for Minim. PSA/PIHP/CMHSP
Payments to Locals...............	20,072.7	(1,293.8)	(135.5)	18,643.4	193.4	18,836.8
Total State Spending (TSS)...	38,270.5	0.0	0.0	38,270.5	0.0	38,270.5
Headlee Limit (48.97% of TSS).	18,741.0	---	---	18,741.0	---	18,741.0
Headlee Surplus/(Deficit).....	1,331.7	---	---	(97.6)	---	95.8
* - The Federal government is providing a 6.2% enhancement to the Medicaid match rate in all calendar quarters of the COVID-19 public health emergency. This would presumably expire prior to the full implementation of the bills in 2026 or later.
** - The base Medicaid match rate for FY 2022-23 is now estimated to decrease by 0.77%, which would result in an increase in GF/GP costs.
*** - The full amount of public school academy (PSA) funding counted for Headlee purposes is estimated to be $1,488.0 million.
**** - The amount of K-12/community college chartered PSA funding is estimated to be $193.4 million; courts could rule that some or all of this subset of funding would count for Headlee purposes.

Entities Eligible to Become SIPs

Based on information from other states, it is clear that integrated plans generally are either Medicaid physical health HMOs that expand to include behavioral health services or specialized behavioral health entities that focus on behavioral health services. The legislation includes criteria for the SIP procurement process, including network adequacy, staffing, financial plans and cost sharing, five years of behavioral health experience, five years of physical health experience, and five years of managed care experience.

The term "five years of behavioral health experience" does not specify the type of behavioral health experience. The Medicaid health plans in Michigan deal with what is commonly known as "mild to moderate" behavioral health issues but do not provide the broad spectrum of behavioral health services that specialized entities or PIHPs provide. If the term is intended to be interpreted to treat the "mild to moderate" coverage as the requisite behavioral health experience, then the current Medicaid health plans certainly would be eligible to bid for those contracts.

The Challenge of Identifying Savings

The bills outline a process for identifying savings that is clearer than the process outlined in the bills as introduced. The introduced bills' savings language amounts to a tautology ("reinvestment of realized savings into the integrated behavioral health system to further promote and expand access")—if there were savings, those savings would accrue to the SIP because of lower expenditures and would be available for expansion of services. The substitute bills specify that, at minimum, the State share of savings (if any) from the two-way risk corridor required under the bill would be savings for the purpose of transfer of savings.

A two-way risk corridor, as noted above, usually makes the managed care entity responsible for the first few percent of any costs or savings above or below the capitated rates paid. After that point, the State covers all excess costs and accrue any additional savings. For instance, a hypothetical managed care risk corridor could be designed so that the managed care entity would cover all costs and accrue all savings between 95.0% and 105.0% of the rates, with the State covering any costs above 105.0% and accruing any savings below 95.0%. This would mean, if costs were 107.0% of what was set up under the rates, the managed care entity would have to absorb the first 5.0% of excess costs, with the State covering the remaining 2.0%. Similarly, if costs were 92.0% of what was set up under the rates, the managed care entity would accrue a surplus of the first 5.0% of savings, with the State in theory accruing the remaining 3.0%. Under the substitutes, in the latter situation, the 3.0% savings that the State normally would accrue instead would be invested in the integrated behavioral health system.

The substitutes specify that the savings to be transferred are "at a minimum" the risk corridor savings, which would allow for potential transfer of additional unspecified savings to support services expansions and enhancements. There is a challenge in identifying any savings beyond savings to the State resulting from the risk corridor. The legislation does not outline a specific process for identification of any other savings. Estimating any other savings could prove to be challenging as the populations served are not static and average health care costs for most Medicaid populations tend to increase from year to year, so savings estimates may face a signal-to-noise ratio. The ideal approach would be to have randomly chosen experimental groups and a control groups but the legislation, with a full transfer of responsibility, understandably does not take that approach.

The Section 904 Report

For over two decades the former Department of Community Health and the current DHHS have reported extensive data on services provided by CMHSPs and PIHPs to these populations. The most recent report, covering FY 2019-20, was required by Section 904 of Public Act 166 of 2020.^{^[5]} An examination of the data in the Section 904 report provides some insights into savings opportunities, primary services provided to given covered populations, and potential cost concerns.

Administrative Costs

The data in the Section 904 report indicate that FY 2019-20 PIHP administrative costs for behavioral health services were about 5.62% of total costs, but this does not reflect a medical loss ratio calculation and thus comparisons to potential SIP or current MHP administrative costs are not particularly valid or meaningful.

Expansion of Services

As previously noted, shifts to managed care for populations with significant pre-existing conditions appear to have led to increases in costs. This should not necessarily be viewed as a negative; these cost increases generally imply that the population is underserved and a broader array of services certainly could be expected to lead to more positive outcomes.

Solvency

Several PIHPs (one in particular) have faced serious financial challenges in recent years. These year-to-year issues with finances and potential for future insolvency create challenges for PIHPs, which has frequently led to discussions of additional funding for those PIHPs as well as disputes over State and local financial responsibility to address these situations. The bills would not establish any specific capital requirements for the SIPs; there would be only a requirement for "financial plans". This could leave SIPs financially vulnerable and could risk creating political pressure for additional financial support from the State. The Michigan Association of Health Plans (MAHP), which has publicly supported the bills, has noted that the definition of eligible providers includes "managed care organizations" or a system of health care delivery as defined under the Insurance Code. The MAHP notes that the latter entities do have some basic requirements for net worth so those latter groups would face a basic capital requirement, though it appears that any entity that bids as a "managed care organization" would not be subject to such a basic capital requirement.

Given that Medicaid health plans and private behavioral health managed care entities generally have greater access to capital than do PIHPs and given the Insurance Code requirements for many of these entities, it appears highly likely that the risk of insolvency for the SIPs under the bills would be less than what is currently the case with PIHPs.

Expenditures for the MI-A Population

The Section 904 report includes extensive data on spending in FY 2019-20 by PIHPs on the mentally ill adult population. Services were provided to over 115,000 individuals, with roughly half living in a family situation and others in more congregate settings ranging from adult foster care to homeless shelters to incarceration. Expenditures on this population were $922.7 million, with almost 45% ($408.5 million) of the funding being spent on inpatient and outpatient services and over one quarter ($238.8 million) going to living supports (generally housing), with case management ($152.1 million), day services ($60.7 million), and assertive community treatment ($54.0 million) making up the bulk of the rest of the spending.

The opportunity for savings on these services under any new financial model of course would not be limited to the behavioral health side of services. The challenge and opportunity for any new managed care entity is not simply a matter of finding ways to reduce costs on the behavioral health services that are provided, but rather implementing a prevention-based approach that would provide better coordination of care, especially on the medication side, that could avert both physical and behavioral health issues.

A 2011 study in World Psychiatry notes that the prevalence of "modifiable [health] risk factors", in particular, obesity, smoking, diabetes, and hypertension, is greater in adults with schizophrenia and bipolar disorder.^{^[6]} These risk factors are clearly associated with higher physical health care expenses, in particular costs related to cardiovascular disease. The study notes that there is some correlation between the use of some antipsychotic medications and obesity. The study states that "an increase in well-established [diabetes] risk factors in these patients partially accounts for much of the increased risk. However, additional factors (disease, treatment) are important as well."

These studies indicate that there is an opportunity to help abate ongoing physical health concerns for individuals with severe mental illness through effective treatment of these risk factors—but it also appears that mental health treatment, particularly medication, can cause or exacerbate some of these risk factors so the opportunities to abate these risk factors and perhaps achieve cost savings on the physical health side would be partially limited.

One of the complications of a coordinated approach is a primary payer issue for a portion of the Medicaid population served by PIHPs. While the Federal Medicare program often is viewed as being strictly a program for the elderly, disabled individuals with a substantial work history become Medicare eligible 24 months after meeting Social Security disability standards and certain disabled children, even in adulthood, can become Medicare eligible based on their parents' work history. The Section 904 report indicates that about 15% of the MI-A population is on Medicare (and thus dually eligible for Medicare and Medicaid; these individuals are known colloquially as "dual eligible"). Medicare is the primary medical payer for most of these individuals, thus the potential for Medicaid savings on the physical health side for the dual eligibles is more limited than one may expect initially: Medicaid serves as gap coverage for behavioral health services for dual eligibles while the cost of their physical health services is really dictated by Medicare policies, policies that cannot be changed at the State level (with the exception of those dually eligible clients who are enrolled in the Michigan dual eligible waiver, known as MiHealthLink).

Expenditures for the MI-C Population

The Section 904 report also includes data on spending for services to the MI-C population (also known as "Children with Serious Emotional Disturbance" or "SED"). Total spending in FY 2019-20 was $261.1 million on 35,000 children, with about one-third of the funding ($85.3 million) going to outpatient services, one-third ($83.8 million) to family support, and the rest split almost evenly between inpatient/crisis services ($36.8 million) and case management ($44.6 million).

One of the concerns over the years regarding services to the MI-C population has been the anecdotal belief that the population is underserved and that children are not receiving sufficiently intense or thorough services, so the potential for savings may be more limited than with the MIA population. The lawsuit KB v. Lyon (Eastern District of Michigan, case 2:18-cv-11795-BAF-SDD), which was settled recently by the State of Michigan, would require significant expansion of community-based behavioral health services for children because of concerns that these services had been consistently underfunded. The enacted FY 2021-22 budget includes $91.0 million Gross and $30.0 million GF/GP to cover the administration's estimated cost of expanded services under the settlement. Given this and other concerns, meaningful behavioral health savings due to integration of behavioral health services to the MI-C population may not be feasible.

Expenditures for the Population Facing IDD

Expenditures on the IDD population comprise almost half of total PIHP spending, $1.627 billion for about 34,000 individuals. The fact that IDD spending is by far the largest portion of PIHP spending is often not well understood. About half of the cost for the IDD population ($858.0 million) is for living supports. Decades ago, many IDD individuals lived in State facilities and other large congregate settings. The State took action over the years to move IDD individuals who were in these congregate settings into group homes and other community-based settings. While, due to economies of scale, this resulted in an increase in average housing cost per client, most would argue that it was a preferable environment for the individuals being served and that those individuals had a higher quality of life than they did in more congregate settings.^{^[7]}

Most of the PIHP expenditures on this population are for the aforementioned living supports, day supports ($308.4 million), and case management ($160.3 million). There is another category of expenditures for the IDD that has grown quickly in recent years, services to individuals with autism, which will be discussed separately below.

Because so much of the cost for the IDD population goes toward housing and day services, and because of research (see next paragraph) indicating that physical health costs for the IDD population (unlike the MI-A and MI-C populations) are similar to those of the general non-IDD population, the opportunities for physical health cost savings for the IDD may be relatively limited. This challenge is even more apparent because over one-third of the IDD population is comprised of dual eligibles and thus Medicare is the primary payer for physical health costs for much of the IDD population.

A recent study examined health care costs for the IDD population in the United States as compared to costs for the non-IDD population.^{^[8]} The study found that "per user costs across for services combined were higher in the general [non-IDD] population [than in the IDD population], driven primarily by greater spending for office visits and hospitalizations.^{^[9]} The study did note higher medication costs for the IDD population, particularly those under the age of 40, but this is not unexpected given the dually diagnosed mentally ill/IDD population, which is a significant subset of the IDD population. The study also offered a caveat that "lower or equivalent costs could very well be an indicator of unaddressed needs or biases in determining who in the [IDD] population can access services".^{^[10]} However, that would indicate the possibility of increased costs under a different model of IDD services rather than the possibility of savings. The possibility of significant Medicaid physical health savings for the IDD population appears to be rather limited.

The other challenge is finding savings opportunities on the behavioral health side for the IDD population. These opportunities exist but may be difficult to achieve; one could produce significant savings by moving individuals with IDD into more congregate settings, shifting back to the old practice of large-scale facilities for many IDD individuals. One could reduce reimbursements to group homes. Or one could reduce reimbursement or access to day activities. Living supports and daytime services represent well over 80% of the nonautism spending on the IDD population, so there are not many other options for savings.

Dually Diagnosed MI/IDD Individuals

Some individuals serviced by the PIHPs, about 16,000, are dually diagnosed with mental illness and intellectual/developmental disabilities. The bills would integrate services for these individuals later in the process, in the final stage with the IDD population rather than with the MI population. The same caveats on the IDD population apply to the potential for savings on the dually diagnosed MI/IDD population.

Expenditures on SUD Services

Medicaid SUD services for the "SUD population" were once provided by substance abuse coordinating agencies, but almost a decade ago these services were effectively transferred to the PIHPs, in large part because of the perceived benefit of integrating mental health and SUD services. The Section 904 report indicates that total spending, split between traditional Medicaid and the Healthy Michigan Plan, was about $144.7 million. About 32% of that was spent on residential treatment, 26% on outpatient services, 20% on methadone treatment, and just under 10% on detoxification. Overall, out of the $144.7 million, over $11.0 million was spent on integrated treatment (in this case, integrated treatment means combined SUD and mental health treatment). Because of the prevalence of limited integrated care that already exists in the SUD population, the opportunities for savings from further integration may be relatively limited.

Expenditures on Autism Services

Autism services expenditures have grown at double digit rates annually since Medicaid autism services were first covered nearly a decade ago (a recent SFA issue paper discussed this growth).^{^[11]}

It is likely that expenditures will exceed $300.0 million Gross in FY 2020-21 after being $70 million Gross as recently as FY 2015-16. Section 959 of Public Act 67 of 2019 required a workgroup report on autism spending and included the following statement: "There continues to be a lack in standardization of the diagnostic process and limited access of providing further assessments with complex youth due to the limitations on services from the PIHPs/CMHSPs." There would appear to be fertile ground for a more comprehensive approach that could produce significant savings relative to the baseline expenditure trend and better outcomes in this program.

Non-Medicaid Community Mental Health & SUD Services

Senate Bill 598 (S-3) would amend the Mental Health Code to specify that the SIPs "take over the administrative and management functions set forth in [Chapter 2 (County Community Mental Health Programs)] and the community mental health services program is responsible only for providing services." Chapter 2 outlines both Medicaid and non-Medicaid behavioral health services, so the substitute's provision appear to outline the eventual transfer of CMH non-Medicaid and SUD coordinating agency funding from the local public entities such as CMHSPs to the SIPs. This provision would not have a fiscal impact as the non-Medicaid services are not an entitlement and the funding level is set through the appropriations process. Any shift would reduce payments to locals by $135.5 million, which would directly affect the Headlee payments to local calculation.

It should also be noted that the substitute does not speak to the local share of funding for CMHSPs. The Mental Health Code requires counties to pay up to 10.0% of non-Medicaid mental health costs (MCL 330.1302 sets a 10.0% payment). A maintenance of effort provision added in the 1995 Mental Health Code rewrite froze the cost for CMHSPs that become mental health authorities at the amount paid in the year before the CMHSP became an authority (MCL 330.1308). These provisions mean that counties contribute toward the provision of non-Medicaid mental health services in their area. This funding is not reflected in the DHHS budget but is rather a separate payment to CMHSPs appropriated by individual counties. It appears that, if a SIP took over administration of the non-Medicaid services, counties would be required to shift these local share payments from the CMHSPs to the SIPs.

Impact on Current Integration Efforts

The State has established several integration programs that do not clearly mesh with the proposal outlined in the bill. This analysis has already noted the Governor's proposal to fund services related to the KB v. Lyon settlement. The State has a Health Homes demonstration project, which works to coordinate physical and behavioral health care for Medicaid beneficiaries who have high rates of hospital utilization, funded at $33.0 million Gross and $2.6 million GF/GP. The State is beginning to implement a Federal demonstration project called the Certified Community Behavioral Health Clinic (CCBHC). The CCBHC would be a two-year demonstration project at 14 sites with comprehensive services including behavioral health and mobile crisis services and supportive services. Governor Whitmer's FY 2021-22 budget would fund CCBHC at $26.5 million Gross and $5.0 million GF/GP and the House and Senate DHHS budget bills also included funding for this initiative.

The goal of both aforementioned demonstration projects is to improve outcomes with the potential of reducing costs. At this point, with the programs not really having started, it is impossible to estimate their fiscal impact let alone their fiscal impact relative to the integration model outlined in Senate Bills 597 (S-3) and 598 (S-3). If the integration process outlined in the bills were more fiscally "effective" than these integration efforts, there would be a net cost reduction, but if a new approach that supersedes current integration efforts were less fiscally effective, net costs could increase.

Transition from Phase to Phase & Legislative Oversight

The legislation would require the DHHS, in consultation with the stakeholders, to create evidence-based metrics to evaluate efficacy of each phase including cost and efficiency. The legislation would direct the Behavioral Health Accountability Council, 18 months into a given 24-month phase, to complete a formal evaluation of the phase. The DHHS then would have complete a formal evaluation 20 months into a given 24-month phase. Within 60 days of that evaluation, the DHHS would have to submit a report to the Legislature determining whether the phase had been successful, unsuccessful, or undetermined. If the evaluation indicated the phase was unsuccessful or had an undetermined outcome, the DHHS would have to recommend continuation of the phase, extension of the duration of the phase for further evaluation, or propose to reform, modify, or terminate the phase. If a phase were unsuccessful but its duration was extended and one of the cost metrics in the evaluation indicated a negative cost (fiscal) impact, that negative fiscal impact could continue indefinitely until the metrics were met.

The legislation also directs that, "five years after implementation of the program, the legislature may review the program's success and efficacy to determine if the program shall continue". This provision is permissive. If the legislation were enacted, the Legislature could attempt to amend or repeal the statute at any time even without the specific five-year review provision.

Summary of Potential Savings & Costs

Some areas of PIHP expenditures are more likely to lead to actual savings under an integration model than others. There would appear to be more opportunities for efficiencies and savings in services to MI adults and autism services than in services to the IDD population. However, the use of the two-way risk corridor to define potential savings that could be transferred would make the potential savings directly affected by the initial rate-setting process. If rates were set high enough to cover potential higher costs, the likelihood of savings occurring would be greater. It should be noted that higher rates would carry a cost to the State—each 1.0% increase in overall SIP rates would cost about $10.0 million GF/GP (so a 10.0% increase in rates would cost about $100.0 million GF/GP).

Because any identified savings would be transferred to support expansion of integrated behavioral health services, the bills would not lead to savings to the State on behavioral health costs. There is the potential for a net increase in costs to the State if the populations involved were determined, in the rate-setting process, to be effectively underserved, with the corresponding likelihood of a better level of services being provided to clients.

The legislation also likely would reduce potential solvency concerns for the managed care entities because many of the SIP bidders would be subject to the Insurance Code capital reserve requirements and thus likely would have better capital reserves than the PIHPs. At present, there are PIHPs that have clear solvency issues; this risk and potential liability for the State could be alleviated. However, this likely better solvency situation would not be as certain to occur as it would if the legislation included specific capital reserve requirements for all bidders.

Potential Impact on the IPA

The Health Insurance Claims Assessment (HICA) was repealed at the start of 2019 and was effectively replaced by the IPA. The IPA is a multi-tiered tax on health insurers, with significantly higher tax rates on Medicaid health plans than on private insurers and PIHPs. Because of actuarial soundness requirements, the Medicaid health plans' and PIHPs' tax costs are reimbursed so the higher tax rate has no net impact on the Medicaid health plans' and PIHPs' bottom line. Because the IPA met a certain Federal statistical test it was approved and it resulted in a meaningful tax reduction for private health insurers when compared to the HICA, with the State also seeing a significant windfall due to the structure of the tax. The SFA's analysis of the HICA repeal and IPA creation legislation has more specific details on the HICA and IPA tax structure, the rates charged to private insurers, Medicaid health plans, and PIHPs, and the Federal statistical test.^{^[12]}

The inclusion of the PIHPs in the tax base was necessary for the IPA to meet the Federal statistical test. It is uncertain, if integration occurred, whether the IPA could be redesigned to meet the Federal statistical test if the PIHPs had a more limited (or no) role at all in Medicaid behavioral health. The SIP approach at least would allow for a structure similar to the IPA structure in place now and that structure could still meet the Federal statistical test (in contrast it would be difficult for a full integration model with behavioral health funding going to the present Medicaid health plans to meet the Federal statistical test). The phased-in approach would further complicate matters; even if there were a way to maintain adherence to the Federal statistical test over the long term, the IPA statute likely would have to be amended at the start of each phase. The structure of the substitutes could have an impact on this process. Because the phases last 24 months it would be possible to ensure each phase would be implemented at the beginning of a fiscal year. Phases that do not start at the beginning of a fiscal year would feature more complex revised IPA calculations and could face more difficulties in gaining Federal approval, so aligning the beginning of each phase with the start of a fiscal year would make the IPA approval process less difficult. The IPA could be at risk of Federal disapproval at the start of any phase as well as in the longer term. It is possible, for instance, that the IPA could be structured in a way that met the Federal statistical test after full integration but could not be so structured during the transition period, in which case, the State would lose the IPA revenue during the transition period. It is possible that the IPA could be structured to meet the Federal statistical test during the transition period but not after full integration, in which case the State would lose the IPA revenue in every year subsequent to the full integration.

The structure of the IPA, with a varying per-member month tax being applied to nonexempt insurers, makes it probable (albeit not at all certain) that a revised IPA meeting the Federal statistical test could be implemented at any new phase of the implementation of the integration outlined in the bills (and because "probable" does not mean certainly, it is possible that the Federal test would be met during most phases but not in all phases, with the corresponding large negative change to the State's balance sheet in any phase where the test could not be met). A secondary challenge involves the potential adjustments to per-member month tax rates to ensure the Federal statistical test is met. The adjustments could lead to significant changes in the net revenue generated by the IPA and could lead to changes in the IPA tax rate for non-Medicaid (private) insurers. So, even if the Federal statistical test could continue to be met, there could be a significant fiscal impact for the State (with a secondary impact in which changes in the IPA rate for private insurers would have an indirect fiscal impact on State and Local government employee benefit costs).

If the IPA did not meet the Federal statistical test for a given year and were not in effect that year, there would be a net reduction in the State's bottom line of at least $450.0 million during that year. This net revenue loss would have a major impact on the State's balance sheet, possibly forcing budget restraint, budget reductions, or alternative health care tax structures, such as the return of a more broad-based replacement health insurance tax.

A Note on Approaches in Other States

While the approach in other states does not directly affect the fiscal analysis of these bills, it is relevant to note that a narrow majority of states have moved toward integration of behavioral and physical health for the mentally ill population, but about three-quarters of those states have chosen to carve out (not integrate) services to the IDD population and do not use an integrated model for the IDD population.

Conclusion

The fiscal impact of these bills cannot be precisely estimated, in large part because the potential changes in expenditures are tied to allocation and policy decisions made by the new integrated entities, which cannot be predicted. Because savings that otherwise would accrue to the State under the two-way risk corridor instead would be transferred to expand or enhance services, the bills would not lead to net reduction in costs for the State.

The potential for efficiencies exists, in particular for services provided to the mentally ill adult and autistic populations. On the other hand, to the extent that these populations are underserved, there is the potential for cost increases (and potentially corresponding improvements in outcomes). Because the SIPs likely would have better capital reserves than the PIHPs, there is clear potential of avoiding the behavioral health managed care entity solvency issues that have faced several PIHPs of late. The administrative cost impact is more difficult to quantify. As noted, the comparisons of administrative costs between the MHPs and PIHPs are not particular useful.

The bills would have a major impact on State payments to local units of government that, when fully implemented, could well put the State below the Headlee amendment limit. There would be the opportunity, over the period during which the bills would be implemented, to adjust State payments to locals in other areas to avoid this problem. However, that would require policy changes independent of the bills' provisions, and so these adjustments cannot be assumed.

There also is the potential, if a revised approach to the IPA did not meet the Federal statistical test, of a short-term or long-term negative change to the State's bottom line in the range of $450.0 million GF/GP in each year that the statistical test was not met. Revisions to the IPA made to meet the Federal statistical test likely would alter the State's net revenue from the IPA.

Date Completed: 3-14-22 Fiscal Analyst: Ellyn Ackerman