Changes in Service Use After a State Autism Coverage Mandate: Evidence from Kansas

Autism spectrum disorder (ASD) is characterized by persistent deficits in social communication and interactions. ASD treatments are often costly and historically have been restricted in private insurance plans. In response, 39 states now have laws mandating coverage of ASD treatment. Research on the effect of state ASD mandates is limited. While ASD mandates are hypothesized to increase utilization, effects on out-of-pocket spending and on types of treatments used are unknown. Coverage mandates may alter use of non-mandated services (for example, either increase or decreasing use of prescription medications).

We evaluate changes in spending and utilization following the implementation of a January 2011 law in Kansas mandating comprehensive coverage of ASD treatments in the State Employee Health Plan (SEHP). The mandate required the SEHP insurer to pay for ASD care up to $36,000 per year. We used unique data from the Kansas All-Payer Claims Database (2009-2013). The sample was comprised of children 0-19 with an ASD diagnosis. We estimated differences-in-differences regression models comparing changes in spending and service use after the mandate among ASD-diagnosed children in the SEHP to ASD-diagnosed children with other private health insurance. Models adjusted for comorbid physical and mental health diagnoses, region of residence, sex, and age.

Spending on ASD treatments increased for both groups over the study period, but the increase was markedly higher for children in the SEHP. Differences-in-differences models reveal increased per child annual spending on ASD care of $1,183 in the SEHP after the mandate (from a pre-mandate baseline of $866), and increased out-of-pocket spending of $166 (from a baseline of $180). A large portion (as much as two-thirds) of this increased spending was offset by reduced by decreased non-ASD outpatient care and on reduced use of psychotropic medications. At the mean, the annual number of outpatient visits for ASD treatment increased by 13.4 visits relative to the privately insured group. The largest changes in service use were detected among the highest-utilizing children (i.e., those in the 95^th percentile of the service utilization distribution). Therapy and behavioral services in particular increased substantially after the mandate. In a sub-analysis, we found children who joined the SEHP after the mandate were costlier in the post-mandate period than those joining in the pre-mandate period.

Our study is one of the first to demonstrate potential effects of state ASD coverage mandates. The large and significant increase in ASD treatments suggest that the mandate facilitated greater service provision, especially for those children who may have been previously constrained by service caps. A large portion of increased ASD care may be offset by reductions in other service categories, including prescription drugs, since some children may reduce pharmacotherapy when outpatient treatments are available. Children joining the program after the mandate were costlier on average, which could reflect selection effects. To fully evaluate policy implications, it is important to identify whether the mandate altered quality of care or affected child health status. Planned analyses for ASHE will consider changes in emergency department visits and treatment continuity.