Reliability and Validity of the Autism Spectrum Addendum to the Anxiety Disorders Interview Schedule (ADIS/ASA)

Background: Assessing anxiety in autism spectrum disorder (ASD) is inherently challenging due to overlapping (e.g. social avoidance, repetitive behavior) and ambiguous symptoms across these conditions. Ambiguous features include social fearfulness without awareness of social evaluation, excessive worry about novelty and change, and phobic responses to unusual stimuli, such as men with beards, specific sounds, or toilets. How such symptoms are conceptualized across measures and clinicians may have an impact on research findings, especially given that the majority of anxiety measures are neither designed nor validated in ASD samples. Notably, data regarding the prevalence and characteristics (e.g. cognitive ability, age, ASD severity) associated with anxiety in youth with ASD is inconsistent (Van Steensel et al., 2011), partially due to inconsistent measurement and differentiation of symptoms across studies (Kerns & Kendall, 2012). The Autism Spectrum Addendum (ASA) is a set of clinical guidelines and supplementary items designed to facilitate use of the Anxiety Disorders Interview Schedule – parent version(ADIS; Silverman & Albano, 1996), a semi-structured diagnostic interview, for children with ASD. The ASA provides a structured approach to differentiating traditional anxiety disorders (i.e. Social Phobia, Separation Anxiety, Generalized Anxiety) in ASD and also measures the ambiguous fears and worries that arise in this population.

Objectives: To assess inter-rater reliability as well as convergent and discriminant validity of the ADIS/ASA. 

Methods: The parents of 70 cognitively-able (IQ>70) children (ages 8-13 years, Mage=10.67, SD=1.64) with ASD completed the ADIS/ASA along with a battery of other behavioral measures as part of a screening evaluation for a randomized clinical trial comparing two cognitive-behavioral therapies for anxiety. A second rater independently observed and scored recordings of the original interviews. Inter-rater reliability as well as convergent and discriminant validity of the ADIS/ASA with other measures were assessed.

Results: Inter-rater agreement for specific ASA items (ICC=.82-.96), traditional anxiety (ICC=.85-.98), and ambiguous anxiety severity ratings (ICC=.87-.95) was excellent. Agreement was also good to excellent regarding principal diagnoses (κ =.82), the presence of clinically significant traditional anxiety (κ =.67-.91), and ambiguous anxiety (κ =.77-.90). Pearson correlations indicated convergence of the most severe ADIS/ASA anxiety severity ratings (traditional or ambiguous) with Pediatric Anxiety Rating Scale totals (PARS; r=.52, p<.01). ADIS/ASA traditional anxiety ratings converged with the Child Behavior Checklist (CBCL) anxiety subscale (r=.35, p<.01), but diverged from Autism Diagnostic Observation Schedule, Second Edition (ADOS-2) comparison scores (r=-.02, n.s.) and CBCL Attention (r=.023, n.s.) and Aggression subscales (r=-.16, n.s). By comparison, the severity of ambiguous ADIS/ASA symptoms was significantly correlated with ADOS-2 severity ratings (r=.31, p<.02), but not the CBCL anxiety (r=.03, n.s.), Attention (r=.08, n.s.) or Aggression (r=.11, n.s.) subscales.

Conclusions: Findings indicate that the ADIS/ASA provides consistent measurement of traditional as well as more ambiguous anxiety-like symptoms in children with ASD across different raters. Additionally, the ADIS/ASA shows adequate convergent and discriminant validity with other measures.  Implications for future studies of the treatment, prevalence and phenomenology of anxiety in ASD will be discussed.