(PS2-A6) Creating a Short-form for the Screen for Adult Anxiety Related Disorders

Background: The Screen for Adult Anxiety Related Disorders (SCAARED), an adult version of the Screen for Child Anxiety Related Disorders, was created to allow clinical continuity in the assessment of anxiety disorders across the lifespan. The SCAARED is a measure capable of screening for specific Diagnostic and Statistical Manual of Mental Disorders, 5th Edition anxiety disorders in adults (Angulo et al., 2017). Despite its demonstrated clinical utility as a cost-effective and well-validated screening tool, the length of the SCAARED is a barrier for its use in clinical settings, where time constraints are a leading concern among providers (Combs & Markman, 2014). Therefore, we created a short-form of the SCAARED with the goal of further addressing barriers to accessibility of evidence-based assessment.

Methods: The SCAARED is a self-report questionnaire composed of 44 items each rated on a three-point Likert-type scale. De-identified responses of the SCAARED were collected from an international sample of 3,852 adults in English, (77.5% female; 76.7% White) spanning ages 18 to 90 (M = 32.4, SD = 12.9). Responses were collected between 2018 and 2021 by Helping Give Away Psychological Science (HGAPS) and the Depression and Bipolar Support Alliance (DBSA). Our analytic plan combined exploratory factor analysis (EFA), confirmatory factor analysis (CFA), expected parameter change (EPC), and Item Response Theory (IRT).

Results: After running a scree plot on the 44-item SCAARED, individual PCA scree plots were run for the four literature-derived subscales (Angulo et al., 2017). Both generalized and separation anxiety subscales were not unidimensional. Five hyper-correlating items were identified via EPC. With one item from each pair removed, these two subscales met the test of unidimensionality. An EFA was run for each subscale. Three items for each subscale were selected for the short form using strength of loading and clinical expertise to decide. A CFA was run on the short form in a confirmatory sample, showing excellent fit statistics (CFI=0.99, TLI=0.99, RMSEA=0.04) and reliability (Lambda6 = 0.85, Omegatotal = 0.88). Bland-Altman plots showed strong similarities between the long and short form of the SCAARED. Lastly, IRT analyses were run (threshold parameters ranging from 1.30 to 4.79). The subscales show greater than rxx(θ) = .80 between approximately θ = -2.1 and 1.75.

Conclusion: The new SCAARED short-form has excellent fit while maintaining excellent reliability and content validity. Our findings also suggest that the clinical utility of the full-length SCAARED could be improved by removing redundant items. Short-forms expand accessibility for busy clinicians and reduce burden on patients while retaining the validity of the original questionnaire (Youngstrom et al., 2020). Further research assessing the continued utility of the short-form in both community and clinical settings is needed, including evaluation of its screening utility and sensitivity to treatment effects.