Joint and Independent Contributions of Autism Spectrum Disorder and Intellectual Disability on Emergency Department Utilization during Adolescence

Background: Frequent utilization of emergency department (ED) services places considerable burden on medical resources and is costlier than treatment provided through primary and urgent care settings. Adolescents with autism spectrum disorder (ASD) and/or intellectual disabilities (ID) are at particularly elevated risk for high ED utilization due to co-occurring psychiatric and medical comorbidities and poor continuity of care. This study will build upon previous literature by determining the joint and independent contributions of ASD and ID on ED utilization during adolescence.

Objectives: The objective of this study was to compare ED utilization and monetary charges during adolescence between four mutually-exclusive cohorts: ASD without co-occurring ID (ASD-only), ASD with co-occurring ID (ASD+ID), ID without ASD (ID-only), and population controls (PC).

Methods: A cross-sectional, non-experimental design was used to analyze administrative records from ED visits that occurred in the state of South Carolina between the years 2004-2015. This study included 2,555 individuals aged 12-17 years. A majority of participants were Black Non-Hispanic (59%) or White Non-Hispanic (37%). Approximately 2% of participants had racial/ethnic backgrounds classified as other Non-Hispanic, and 1.8% were Hispanic. Similar to other studies of healthcare utilization, we stratified participants by high and low ED utilization, as high utilizers accounted for a disproportionate amount of ED visits. Participants were classified as high utilizers if they had at least 10 ED visits during adolescence and low utilizers if they had < 10 ED visits. Table 1 provides descriptive information as a function of cohort and ED utilization. Generalized linear models were used to compare cohorts on the number of ED visits and total charges, stratified by high and low ED utilization.

Results: Among high utilizers, adolescents with ID-only had the most frequent ED visits, with 1.2 times more visits than the PC cohort (95% CI = 1.0-1.3). However, average total ED charges were significantly higher for the ASD-only cohort (mean=$30k USD, 95% CI=$18-50k) than the ID-only cohort (mean=$19k USD, 95% CI=$13-29k). Descriptive examination of primary diagnoses revealed that the ASD-only cohort tended to be seen for psychiatric conditions, while those with ID-only tended to use the ED for low-cost, low-acuity conditions (e.g., upper respiratory infections). High utilizers with ASD+ID did not significantly differ from other cohorts on the number of ED visits or ED charges. Among low ED utilizers, no differences between cohorts were observed for ED utilization or charges.

Conclusions: ED utilization is an important indicator of health outcomes and access to quality primary care. In this study, high utilizers with ID-only had the most frequent ED visits, but those with ASD-only incurred the highest ED charges. This may be because individuals with ASD-only tended to be seen for psychiatric concerns, which require longer length of stay and are costlier to treat. Further research is warranted to better characterize and meet the healthcare needs of individuals with ASD and/or ID during adolescence.