Project Summary Although exposure and response prevention (ERP) is well-established as the first line of treatment for obsessive- compulsive disorder (OCD), 40-50% of patients who receive this intervention do not recover. Novel augmentations are needed to more efficiently target the mechanisms that maintain OCD symptoms in this population. One candidate mechanism is interpretation bias, the misinterpretation of ambiguous stimuli as threatening, contributing to the development and maintenance of OCD. In the absence of interventions that more efficiently and effectively target interpretation bias, it will likely remain difficult to increase rates of treatment response and decrease relapse. The objective of this proposal is to test the feasibility, acceptability, adherence, target engagement, and clinical outcomes of an intervention targeting interpretation bias, with the ultimate goal of contributing to the development of novel, scalable, technology-driven augmentations to ERP. Improvements in interpretation bias have been associated with OC symptom reduction. Therefore, an accessible intervention which directly targets interpretation bias may be an ideal augmentation to improve clinical outcomes during ERP. Cognitive Bias Modification for Interpretation Bias (CBM-I), a computerized intervention, has shown reliable effects for engagement of interpretation. Studies of CBM-I in OCD have been largely conducted in analogue samples, and studies with clinical samples have been limited. This proposal will address the critical need of developing interventions to augment ERP by utilizing CBM-I, with primary aims to: 1) test whether CBM-I induces changes in interpretation bias in OCD and to determine if these changes are associated with clinical outcomes across multi-modal assessments, and 2) leverage advances in machine learning to develop personalized predictions of which individuals with OCD are best-suited for CBM-I. We hypothesize that a multivariate model incorporating pre-treatment clinical, behavioral and demographic characteristics will predict patient-specific probability of responding to CBM-I. These aims map onto the candidate’s training goals, with critical new training and mentorship provided in the areas of: 1) effectiveness trials methodology with an experimental therapeutics approach (Co-Primary Mentor Dr. Courtney Beard; Co-Mentor Dr. Sabine Wilhelm); 2) ecological momentary assessment (Co-Primary Mentors Dr. Christian Webb and Dr. Beard; Collaborator Dr. Justin Baker); 3) machine learning techniques (Dr. Webb and Collaborator Dr. Boyu Ren); and 4) career development (Co-Mentor Dr. Kerry Ressler). This K23 Award will support an innovative program of patient-oriented research and provide the candidate with the skills necessary to become an independent investigator focused on advancing the understanding, prediction, and treatment of non-response and relapse, to optimize outcomes of exposure therapy for refractory patients with OC-related diso...