Prolonged Analysis of ASD Assessment Methodologies Utilizing the ADDM Dataset

Background: Appropriately identifying ASD can be challenging for practitioners. Differences in training, experience, and cultural/linguistic responsiveness can contribute to delayed or misidentification of ASD (AAP, 2016). Disparities in ASD identification by gender, socioeconomic status and race/ethnicity exist (Mandell et al., 2007). These disparities can significantly impact the educational, social and vocational outcomes for children with ASD. Limited research has been conducted investigating trends in ASD assessment practices over time, and how these practices relate to ASD identification trends by population.

Objectives: This analysis will investigate ASD assessment practices (e.g. types of ASD specific assessments, IQ tests, and adaptive measures) over time (2000 to present). The study will describe what types of assessments providers (medical, educational; as well as by specialty, psychologist, developmental pediatrician, etc.) utilize. The focus of this project will be to investigate changes in ASD assessment practices and whether they vary by provider, source type (education vs. health), and ADDM site.

Methods: Data acquisition has been proposed and approved by the CDC ADDM PI. The data is currently being processed and will be delivered to the lead author by January 15^th, 2019. Following the acquisition of the dataset, the authors will determine the appropriate analyses given sample sizes. It is predicted that some analyses will need to be conducted using basic frequencies as the usage of certain assessments or numbers in populations may be limited. However, for more robust analyses the authors intend on utilizing multiple regression methodologies. Additional comparisons will be conducted to explore, e.g., 1) score profiles in relation to diagnostic/eligibility statements and ADDM case status, including the level of confidence ADDM clinician reviewers have in confirming case status; 2) patterns over time (2000-2014) in the use of assessment instruments for children at different ages, and the age at which children are diagnosed with ASD or other coexisting conditions; and 3) variation in testing practices in relation to children's sex, race/ethnicity, and socioeconomic status. Analyses will be limited to children in the "8-year-old" cohorts.

Results: This project will explore changes over time in the assortment of instruments used to diagnose ASD. Special attention will be given to the types and specialties of providers evaluating children with ASD, the frequency and variety of assessment tools utilized over time, the settings in which these instruments are used (education vs. health), and differences in practices among ADDM Network sites. Additional analyses will focus on differences in these assessment practices by child's age at the time of assessment, sex, race/ethnicity, presence/age of ASD diagnosis, coexisting diagnoses, and socioeconomic status.

Conclusions: ADDM data provide an ideal source of information to explore long-term changes in formal assessment practices among providers conducting diagnostic evaluation of children with ASD. These instruments are useful for diagnosing conditions such as ASD and intellectual disability (ID), to determine eligibility for services, and to inform strategies for service delivery. Changes in assessment practices may be related to trends in early identification, racial/ethnic health disparities, and socioeconomic factors.