How States Use Medicaid to Fund Community-Based Services to Children With Autism Spectrum Disorders

The 2005 Medicaid Analytic eXtract (MAX) “other therapies” file, the most recent year available to us, was the data source. It comprises paid Medicaid fee-for-service claims and data on managed care encounters for outpatient services in 48 states. Poor data quality prevented inclusion of Maine, Colorado, and the District of Columbia. Claims for children age three to 17 associated with primary ICD-9 codes for autistic disorder (299.00) or Asperger’s disorder or pervasive developmental disorder not otherwise specified (299.8) were identified (12). The study received institutional review board approval at the University of Pennsylvania.

To limit the analysis to commonly used services in each state, a service category minimum of 100 claims or 1% of all autism-related claims per state—whichever represented fewer claims—was set. The threshold for assessment services was lowered to 50 because these services usually are conducted annually. Service descriptions associated with specific procedure codes were identified through software developed by the Agency for Healthcare Research and Quality and supplemented with searches on state Medicaid Web sites and communications with state agencies providing services for children with intellectual and developmental disabilities (13). Many states have developed their own terminology for comparable services. Therefore, we created broad categories of similar services for cross-state comparisons. Development of broad service categories began with a review of literature on evidence-based and effective services for children with autism spectrum disorders and children with other special health needs (7,14). We then created final categories by revising initial categories in light of empirical findings from the claims-based analysis and conversations with clinicians and experts. Funding of services in 16 categories was examined: individual therapy, physical and occupational therapy, in-home supports, speech therapy, diagnostic assessment, behavior modification, family therapy, case management, targeted case management, respite, day treatment, social skills training, habilitation services, treatment planning, family education and training, and assistive communication devices. A list of the specific procedure codes by state is available under “Tools for Researchers” on the Web site of the Center for Mental Health Policy and Services Research, University of Pennsylvania (www.med.upenn.edu/cmhpsr/resources.html).

In 2005, the number of services for children with primary autism spectrum disorders that were paid for by state Medicaid agencies ranged from a low of two in Connecticut to a high of 11 in Maryland and Missouri (Table 1). States paid for an average of seven services. The five most commonly reimbursed services for autism spectrum disorders were individual therapy (45 states); occupational and physical therapy (42 states); in-home supports (42 states), including medical and nonmedical supports such as nursing, one-to-one assistance, and supported-living services to help children remain in communities; speech therapy (37 states); and diagnostic assessment (31 states). Between 20 and 30 states reimbursed for the following services: behavior modification, including ABA (29 states); family therapy (27 states); and case management (26 states). Between ten and 20 states provided the following services: targeted case management (18 states), which provides staff who interact with the child and his or her family to coordinate and find services; respite (15 states); and day treatment (11 states). Ten states or fewer provided the following services: social skills training (nine states); habilitation services (eight states) to improve self-help, socialization, and adaptive skills; treatment planning (seven states); family education and training (seven states); and assistive communication devices (four states).

This study highlights the types of services that state Medicaid agencies reimburse for children with a primary diagnosis of autism. No state Medicaid agency paid for services in all 16 categories described here. Only six states funded all four commonly used services to address core deficits of autism: physical and occupational therapy, speech therapy, behavior modification, and social skills training.

Some study limitations should be noted. First, some services may have been missed; our methods did not examine services billed with broad EPSDT service codes because no specific service was identified. Second, states may use other funding sources to deliver services to children with autism. For example, the finding that less than half of states paid for diagnostic assessment may be explained, in part, by the fact that children receive a diagnosis of autism during the process of applying for Supplemental Security Income before Medicaid enrollment (1). This variation may reflect a state’s intention to fund other services or may result from a Medicaid policy decision not to fund or provide these services to children with autism. In other words, the variation does not mean that a state lacks knowledge about what is covered by Medicaid and about the services that would therefore receive a federal match.

Our study revealed considerable differences in state use of Medicaid to reimburse services for children with autism, indicating that some states may have opportunities to receive the federal Medicaid match for funding autism services. Few states funded social skills training, habilitation services, family education and training, and assistive communication devices. States have been slow to add services needed by children with autism spectrum disorders to address the growing increase in prevalence of these disorders among children.

This study was funded by grant MH077000-01 from the National Institute of Mental Health (NIMH) (“Interstate Variation in Health Care and Utilization among Children with Autism Spectrum Disorders”). The authors thank Michele DeFelice Haverly, M.S., Deb Dunn, J.D., Maureen Davey, Ph.D., Steve Eiken, B.A., M.P.A., Chris Koyanagi, Debra Langer, M.P.A., M.Sc., and Gail Stein, M.S.W., M.Ed., for helpful suggestions and comments on earlier drafts. All views expressed are those of the authors and do not necessarily reflect the views of NIMH.

The authors report no competing interests.