(PS7-20) OCD Symptom Dimensions as Predictors of Treatment Response During In-Person and Telehealth CBT for Pediatric OCD

Abstract Text:

Background: Pediatric intensive outpatient programs (IOP) for OCD serve severely impaired youth, yet given the cost and intensity of IOPs, identifying whether specific presenting symptoms predict differential treatment response is crucial for anticipating trajectories of improvement and tailoring interventions. Prior research, primarily using data from in-person pediatric programs, has identified roughly four factors of OCD symptom dimensions (e.g., contamination concerns, sexual/aggressive intrusions), but rarely examined these as predictors of treatment response, yielding mixed results. There is also limited evidence on whether treatment format (in-person vs. telehealth) interacts with symptom dimensions in predicting response. Telehealth may involve different mechanisms (i.e., less social pressure) which could impact treatment response. The present study examined OCD symptom dimensions as predictors of clinician-rated OCD symptom change across nonrandomized in-person and telehealth formats of IOP exposure-based CBT for pediatric OCD. We hypothesized that four symptom dimensions would emerge, and that higher scores on each dimension would predict higher baseline levels of clinician-rated OCD severity and lower treatment response, particularly in the telehealth format.

Method: Adolescents with severe OCD (N = 123) participated in an exposure-based CBT IOP for approximately 12 weeks via in-person (n = 52) or telehealth (n = 71) format. They were assessed via Children’s Yale-Brown Obsessive Compulsive Scale clinician assessments at baseline and every four weeks until program completion (508 assessments).

Results: Exploratory factor analysis did not replicate the four-factor from other studies compared to models with more additional factors. In multilevel growth models, clinician-rated OCD severity declined markedly across participants in both treatment formats, following a pattern of linear improvement that accelerated over time. As expected, OCD dimensions related to harm/aggression/superstition and mistakes/perfection/embarrassment were associated with higher clinician rated severity. However, baseline symptom dimensions neither predicted treatment response nor interacted with treatment format.

Conclusion: Findings suggest that the factor structure of pediatric OCD symptoms may not always generalize from past samples. Nonetheless, our result imply that intensive CBT was associated with marked clinical improvements regardless of which OCD symptom dimensions predominated at baseline.