0414: FLOOR SUMMARY (Date Completed: 3-13-12)

INSURANCE: AUTISM COVERAGE S.B. 414 (S-3), 415 (S-3), & 981 (S-1): FLOOR SUMMARY
[Please see the PDF version of this analysis, if available, to view this image.]

Senate Bill 414 (Substitute S-3 as reported)
Senate Bill 415 (Substitute S-3 as reported)
Senate Bill 981 (Substitute S-1 as reported)
Sponsor: Senator Mike Green (S.B. 414) Senator Tupac A. Hunter (S.B. 415) Senator Randy Richardville (S.B. 981)
Committee: Health Policy

CONTENT
Senate Bills 414 (S-3) and 415 (S-3) would amend the Nonprofit Health Care Corporation Reform Act and the Insurance Code, respectively, to require a policy, certificate, or contract to provide coverage for the diagnosis and treatment of autism spectrum disorders (ASDs).

Senate Bill 414 (S-3) would apply to a Blue Cross Blue Shield of Michigan (BCBSM) group or nongroup certificate. Senate Bill 415 (S-3) would apply to an expense-incurred hospital, medical, or surgical group or individual policy or certificate delivered, issued for delivery, or renewed in this State, as well as a health maintenance organization (HMO) group or individual contract.

An insurer, HMO, or BCBSM could not do any of the following:

-- Limit the number of visits a member, insured, or enrollee could use for treatment of ASDs covered under the bills. -- Deny or limit coverage on the basis that it was educational or habilitative in nature.
-- Subject autism coverage to dollar limits, copays, deductibles, or coinsurance provisions that did not apply to physical illness generally, except as provided for applied behavior analysis.

Coverage for applied behavior analysis could be subject to a maximum annual benefit of $50,000 per covered insured, enrollee, or member, and could be limited to an individual up to age 18.

Senate Bill 415 (S-3) also would prohibit an insurer and an HMO from terminating coverage or refusing to deliver, execute, issue, amend, adjust, or renew coverage solely because an individual was diagnosed with, or had received treatment for, an ASD.

If a member, insured, or enrollee were receiving treatment for ASD, the bills would allow BCBSM, an insurer, or an HMO to request a review of that treatment consistent with current protocols and to require a treatment plan.

Beginning January 1, 2014, a qualified health plan offered through an American health benefit exchange established in Michigan pursuant to the "Federal act" would not have to provide the required ASD coverage to the extent that it exceeded the essential health benefit requirements of the Federal act. ("Federal act" would mean the Patient Protection and Affordable Care Act as amended by the Health Care and Education Reconciliation Act, and any regulations promulgated under those Acts.)

Senate Bill 981 (S-1) would create the "Autism Coverage Incentive Act" to require the Department of Licensing and Regulatory Affairs (LARA) to create and implement an autism coverage incentive program through which carriers and third-party administrators could seek reimbursement for paid claims for the diagnosis and treatment of ASDs. The bill also would do the following:

-- Create the "Autism Coverage Fund" to provide the reimbursement.
-- Require LARA to review and consider applications for reimbursement in the order they were received, and approve or deny an application within 30 days after receiving it.
-- Require LARA to submit annual reports to the State Budget Director and the Legislature on the funding awarded under the program and the program's administrative costs.
-- Limit the amount of the annual appropriation to the Fund that could be used for administrative expenses to 1%.
-- Specify a State policy to encourage autism coverage and a legislative intent to further this policy by providing annual appropriations.
-- Prohibit LARA from making a commitment or exercising its authority under the proposed Act until the Legislature had appropriated sufficient funds to cover it.

If there were insufficient money in the Fund to reimburse a carrier or third-party administrator for approved paid claims, reimbursement could not be made. Applications that were approved but not reimbursed could be paid, however, if Fund revenue became available.

All of the bills are tie-barred to each other.

Proposed MCL 550.1416e (S.B. 414) Legislative Analyst: Julie Cassidy
Proposed MCL 500.3406s (S.B. 415)

FISCAL IMPACT
The estimated fiscal impact of this legislation includes the cost to State and local governments as insurers and potential savings to State and local governments due to positive results from intervention, particularly reductions in special education and mental health expenditures. This estimation process is further complicated because many if not most State and local government employees work for entities that self-insure and thus are not subject to State regulation. (Instead, health benefits provided for employees of self-insured entities are regulated at the Federal level under the Employee Retirement Income Security Act, or ERISA.) This legislation would not affect ERISA plans.

Among the services covered would be behavioral health treatment, pharmaceutical services, and what is known as "applied behavior analysis" (ABA). When provided to individuals diagnosed with autism spectrum disorders, ABA services are intensive services intended to improve an ASD individual's ability to function.

Estimating the cost of the services to State and local government requires taking a universal view, in which the potential cost to all insurers (including ERISA plans) is measured against total expenditures by all insurers (including ERISA plans) to derive a percentage cost increase. That percentage increase can then be used on estimates of State and local insurance costs to derive an estimated fiscal impact. Including ERISA plans in this estimate would be problematic only if the likelihood of ASDs among that segment of the child population were significantly different from the likelihood among non-self-insured individuals. There appears to be no reason to expect a significant difference in the ERISA population. The estimation process also includes the cost for services to Medicaid and MIChild recipients and, given that Supplemental Security Income (SSI) disability results in categorical eligibility for Medicaid, the total cost might well be somewhat overstated.

The Department of Community Health (DCH), in its presentation to the House Subcommittee on the Department of Community Health on February 28, 2012, estimated that about 4,500 Michigan children between the ages of 2 and 5 have ASDs. The bills, however, would provide coverage up to age 18; thus, the universe of those potentially eligible for services would be much greater than 4,500. The DCH estimate for the average cost of expanded treatment for Medicaid- and MIChild-eligible ASD children is $17,000 per year per child.

Using this number (which may be low due to Medicaid provider rates being lower than average), and assuming a universe of eligible children closer to 15,000 than 4,500 due to the 18-year age limit, leads to an estimated cost to all insurers of $255.0 million per year. It is important to note that this $255.0 million figure is not the actual increase in health insurance costs that could occur due to this legislation. This number is clearly overstated as a significant number of the children involved would be publicly insured by Medicaid or MIChild or would belong to self-insured plans and would not be subject to this legislation, and the actual number is likely less than half that amount. The $255.0 million figure is a useful number, however, to estimate an upper bound on the percentage cost increase.

Looking at the process used by the Senate Fiscal Agency (SFA) to derive the Health Insurance Claims Assessment (HICA) legislation revenue estimate, health expenditures in Michigan are close to $80.0 billion. Once out-of-pocket costs and Medicare costs are subtracted (as Medicare rarely covers children), the expenditure base is closer to $50.0 billion. The $255.0 million figure represents about 0.5% of total health care expenditures. Therefore, a reasonable conclusion is that implementation of the bills would increase health insurance costs by a maximum of 0.5% and likely less than that.

Research in other states resulted in similar conclusions, with an estimated cost increase of $50 per policy per year reflected in analyses in South Carolina and Wisconsin, with a significantly lower cost estimate of $5 to $20 per policy per year in Indiana. These estimates correspond to rate increases of 0.1% to 0.5%.

In 2011, the SFA estimated that paid health claims for State and local government in Michigan are about $3.6 billion in State and local revenue. If the 0.5% figure is correct and none of the State and local entities are self-insured, then the legislation would increase State and local expenditures by $18.0 million per year, with about half of those costs accruing to school districts. Given the general belief that at least half of the public employees in the State work for entities that self-insure, the most useful estimate is that the legislation would increase State and local expenditures by $9.0 million per year.

The net cost savings from improved outcomes is far more difficult to estimate. Studies have varied on the benefits of ABA and similar interventions. Replication of results has been difficult to achieve and most studies have not been randomized controlled trials. Randomized controlled studies are the most useful studies in evaluating programs and interventions. There do not appear to be any randomized longitudinal trials that would help provide a clear answer to the long-term impact of ABA.

Some of the randomized controlled trials show improvement in intelligence quotient (IQ) and adaptive behavior for a significant portion of the population served (source: British Medical Journal, Evidence-Based Mental Health, November 2011). There was less evidence in the randomized controlled studies to support improvement in symptom severity scores and other measures. A nonrandomized Pennsylvania study indicated improvement in symptoms and language and socialization skills (Penn State College of Medicine). Improvements in various areas should lead to lesser need for non-health-related services, in particular special education, in the longer term. Providing a precise estimate of such savings is impossible; the most useful studies are randomized controlled studies and the results speak to impact on measures, not specific expenditures that likely would be reduced due to changes in those measures. As noted, long-term randomized controlled longitudinal studies would be necessary to examine the demand and cost of services and the results of the intervention.

Given the cost of special education and mental health services, minor improvements for a small portion of the ABS population served would result in savings that would almost certainly more than offset the increased insurance cost for State and local governments. Therefore, one would expect the bills to lead to an indeterminate but positive fiscal impact for State and local governments. A more definitive fiscal impact statement would have to be based on a well-designed long-term peer-reviewed randomized controlled longitudinal study.

Senate Bill 981 (S-1) would create some new costs for the Department of Licensing and Regulatory Affairs related to the development and administration of the autism coverage incentive program. Administrative costs could include processing applications for the program, administering payments, and likely other operational expenses. The bill also would establish the Autism Coverage Fund, which would be funded by a legislative appropriation made at a later date. The bill specifies that not more than 1% of the annual appropriation made to the Fund could be spent on administering the autism coverage incentive program. It is not known at this time how much money would be appropriated to the Fund, how much (up to 1%) of that would be used for administration, or if 1% of the appropriation would be sufficient to administer the program. Any administrative costs in excess of 1% of the Fund's annual appropriation would be borne by existing LARA resources.

The cost of Senate Bill 981 (S-1) would equal the amount of public money deposited in the Fund. The bill does not specify the amount as it would be subject to annual appropriations, so the cost of the Fund would be determined by the Legislature through the budgetary process.

Date Completed: 3-13-12 Fiscal Analyst: Steve Angelotti
Josh Sefton

Analysis was prepared by nonpartisan Senate staff for use by the Senate in its deliberations and does not constitute an official statement of legislative intent. sb414,415,981/1112