1 - (SYM 26) Maximizing Gains in Computerized Interpretation Bias Modification: A Multi-session, Personalized Approach

Anxiety is the most common mental health disorder in youth (Merikangas et al., 2010). Interpretation bias (IB) – automatic, uncontrolled attribution of threat to ambiguity – is an underlying anxiety mechanism, and computerized cognitive bias modification for IB (IBM) has demonstrated preliminary efficacy in adults (Cristea, Kok, & Cujpers, 2015). IBM may have high potential for accessibility given its digital format. However, studies with youth are limited, focusing almost exclusively on unselected samples, useing standardized stimuli across all participants regardless of symptoms endorsed, and using the same stimuli for IB assessment and IBM, which does not address whether youth simply learn to complete a specific task with specific stimuli or actually experience bias change.

Here, we address each limitation in a just-completed (February 2023) NIMH R61 double-blind randomized clinical trial. Fifty youth diagnosed with Separation, Social, or Generalized Anxiety Disorder were assigned to a four-week 16-session IBM or an interpretation control condition (ICC). Stimuli were personalized to youths’ symptoms, and IB assessment included two performance-based tasks (word-sentence association paradigm, ambiguous faces) and a self-report IB questionnaire (Children’s Automatic Thoughts Scale; CATS; Schniering & Rapee, 2007). Anxiety assessments included clinician-rated Pediatric Anxiety Rating Scale (PARS; RUPP 2002) and Clinical Global Impressions – Severity and Improvement (CGI-S/I; Guy, 1976), and youth-/caregiver-report on the Screen for Child Anxiety Related Emotional Disorders (SCARED, Birmaher et al., 1999). IB and symptom assessments were conducted weekly after every 4 training sessions.

The computerized intervention approach was feasible, with only 6% attrition, and more than 90% of at-home trainings completed. All assessors utilized the computerized personalized stimulus selection procedure successfully. Acceptability ratings for youth and caregivers was high. Significantly more youth in the IBM group (68%) were treatment responders on the CGI-I compared to ICC (24%; χ²(1)13.29, p< .001).

The final talk will include primary outcomes for IB and youth- and caregiver-reported anxiety. A formal mediation analysis with temporal precedence will determine whether and at what dose IB reduction might precede and predict anxiety reduction. Research and clinical implications, including experimental methods for future treatment personalization with CBM-I as a technology-based intervention, will be discussed.