Toward better outcome measurement for insomnia in children with autism spectrum disorder

Abstract

This report presents a new parent-rated outcome measure of insomnia for children with autism spectrum disorder. Parents of 1185 children with autism spectrum disorder (aged 3–12; 80.3% male) completed the first draft of the measure online. Factor and item response theory analyses reduced the set of 40 items to the final 21-item Pediatric Insomnia Autism Rating Scale. In this online sample, Pediatric Insomnia Autism Rating Scale mean was 21.8 ± 15.0 (range 0–63; coefficient alpha = 0.94). Item response theory results indicated excellent reliability across a range of scores. Next, we conducted a telehealth assessment of 134 children with autism spectrum disorder (aged 3–12; 73% male). In this clinical sample, Pediatric Insomnia Autism Rating Scale mean was 28.9 + 14.62 (range 0–61; coefficient alpha = 0.93). Pearson correlations of Pediatric Insomnia Autism Rating Scale with the PROMIS Sleep Disturbance and the modified Children’s Sleep Habits Questionnaire was 0.74 for each. This was significantly higher than correlations of parent-rated disruptive behavior, hyperactivity, repetitive behavior, and anxiety measures (range 0.29–0.59)—supporting convergent and divergent validity. Parents of 63 children completed the Pediatric Insomnia Autism Rating Scale three times over 4 weeks. Intraclass correlation was excellent (0.88) over three time points. Pediatric Insomnia Autism Rating Scale is a valid dimensional measure of insomnia in autistic children that produces reliable scores. Lay Abstract Insomnia, trouble falling asleep or staying asleep, is common in autistic children. In a previous report, we described the results of focus groups with parents of autistic children toward the development of the Pediatric Autism Insomnia Rating Scale. In this article, we report on the steps taken to complete the Pediatric Autism Insomnia Rating Scale. With help from the Simons Foundation registry, we collected information from parents on 1185 children with autism spectrum disorder to test the new measure. These results were evaluated using standard statistical methods such as factor analysis. To confirm the validity of the new measure, we enrolled a separate sample of 134 autistic children for a detailed assessment by video conference. This step showed that the Pediatric Autism Insomnia Rating Scale is clearly measuring symptoms of insomnia in children with autism spectrum disorder and not related problems such as hyperactivity, repetitive behavior, or anxiety. We also showed that the total score on the Pediatric Autism Insomnia Rating Scale is stable when repeated over a brief period of time. This is important because a measure that is not stable over a brief period of time would not be suitable as an outcome measure. In summary, the Pediatric Autism Insomnia Rating Scale is a brief and valid measure of insomnia in children with autism spectrum disorder that provides reliable scores.