Unmet Need for Autism Treatment: Variation by State Prior to Autism Insurance Reform

Background:  

Early treatment for autism relies heavily on the use of speech and behavioral therapies.  National data has found that children with autism spectrum disorder (ASD) are significantly more likely to have unmet health care needs and to live in families reporting greater financial burden and medical expenditures compared with other children (Kogan et al., 2008; Liptak et al., 2006).  It is unknown to what extent the unmet health care needs are specifically associated with unmet need for autism therapies and how unmet needs vary by state.  This question is particularly important to answer empirically given efforts to enact health insurance reform around autism at the state level. 

Objectives:  

The purpose of this analysis is to use existing national data to measure the association between a parental report of ASD and unmet need for speech, occupational, and physical therapy.  In addition, the association is examined by insurance type and state.  The data will present a nationally representative picture of U.S. children in the period prior to the enactment of any state autism insurance reform laws.

Methods:  

Weighted logistic regression is used to measure associations in the National Survey for Children with Specially Health Care Needs (NS-CSHCN), 2005-2006.  The subpopulation examined in this analysis includes all children age 3 and older whose parent indicated the need for speech, occupational, or physical therapy in the survey screener.   All analyses accounted for the complex sampling design of the survey.  

Results:  

Within the subpopulation, 1,457 children had parent-reported ASD compared with 4,817 children without ASD.  85% of children with parent reported need for speech, occupational, or physical therapy received services (95% confidence interval (CI), 84%, 89%).  After controlling for an array of potential child and family-level confounders, children with ASD had 0.76 times the odds of receiving needed speech, occupational, or physical therapy compared with children without ASD (p<0.077, [CI, 0.56, 1.03]).  Controlling for the type of health insurance had no substantive effect on the adjusted odds ratio, and there were no significant interactions between insurance type and ASD.  There was some variation by state in the prevalence of receiving needed therapies.  Estimates ranged from only 66% receiving needed therapy in Arizona (CI [55%, 77%]) to 94% in North Dakota (CI [89%, 995]).  While the NS-CSHCN is powered to provide state level estimates overall, specific subpopulations may not have adequate power to detect differences.  In regressions of ASD status on receipt of needed therapy by state, only four states had statistically significant odds ratios (p<.1).  In each of the four states, children with ASD had lower odds of receiving needed therapy compared with children without ASD. 

Conclusions:  

National data indicate that children with ASD are less likely to receive needed therapy services than children with parent reported need without ASD.  While the majority of special needs families needing therapy report that they receive therapy, there is some significant variation by state.  The NS-CSHCN could be used to evaluate the impact of insurance reform in the future.