Prevalence and Consequences of Elopement in Autism Spectrum Disorders

Background: Death, injury, and major family burden due to elopement behavior in children with autism spectrum disorders (ASD) have been reported, yet there has been virtually no research on this topic. In response to the Interagency Autism Coordinating Committee’s urgent call for data to address this gap, the Interactive Autism Network (the largest web-based autism registry), in partnership with diverse advocacy organizations, deployed an online Elopement and Wandering Questionnaire to families of children with ASD.

Objectives: The aim of this study was to provide a preliminary estimate of elopement prevalence among children with ASD. Predictors and consequences of elopement were also evaluated. 

Methods: Elopement was assessed in children age four through seventeen. In addition to elopement frequency and age of occurrence, parents were asked if their child had ever gone missing long enough to cause concern, to describe consequences of the behavior, and to report what motivated their child’s elopement. The final study sample included 1,367 children with ASD. The association between child sociodemographic and clinical characteristics and elopement risk was estimated using a Cox proportional hazards model.

Results: Forty-eight percent (n=661) of survey respondents reported that their child had attempted to elope at least once at age four or older; 26% (n=358) were missing long enough to cause concern. Of these, 65% reported a close call with traffic injury and 26% a close call with drowning. Police were contacted in 36% of cases. Elopement rate decreased with age but began to increase during adolescence. Compared with affected children, unaffected siblings had significantly lower rates of elopement across all ages. After adjusting for other characteristics in the model, children who screened positive on the Social Communication Questionnaire had more than a two times increased risk of elopement compared with children with negative screens (RR 2.05, 95% CI: 1.17, 3.61) and, on average, the risk of elopement increased by one percentage point for every one point increase in Social Responsiveness Scale t-score (RR 1.01, 95% CI: 1.00, 1.02). Motivation for affected children’s elopement was typically goal-directed. Children whose parents believed they eloped because they enjoyed running/exploring were significantly more likely to have a diagnosis of autistic disorder or “other ASD” (p<.001), while parents of children with Asperger’s disorder were more likely to report that elopement was driven by the need to escape an anxious situation (p<.001).

Conclusions: Nearly half of children with ASD engage in elopement behavior, a much higher rate than that of their typical siblings, and more than a quarter go missing long enough to cause concern. Likelihood of elopement is positively correlated with autism severity, and elopement behavior is often goal-directed. This critical information is already being used by families, advocates, and policy makers, and has played a role in the adoption of an ICD-9 code for ASD-related “wandering.” The speed with which data were collected and employed demonstrates the potential of an online registry of engaged families working in partnership with committed advocacy groups.