Date of Award

December 2014

Degree Type


Degree Name

Doctor of Philosophy


Educational Psychology

First Advisor

Karen Callen Stoiber

Committee Members

Razia Azen, Markeda Newell, Han Joo Lee, Bradley C. Riemann


Adolescent, CY-BOCS-SR, Exposure Response Prevention, Obsessive Compulsive Disorder, Residential, Treatment Outcome


Exposure and response prevention (ERP) paired with psychopharmacological interventions are considered first line treatments for pediatric obsessive compulsive disorder (OCD). Recent literature has emphasized the importance of investigating effectiveness and treatment outcomes for difficult-to-treat cases of pediatric OCD who do not respond to outpatient treatment. Effectiveness studies have found that adolescent patients treated in residential settings have demonstrated gains comparable to those patients included in published outpatient outcomes studies (Bjorgvinsson et al., 2008; Leonard et al., 2014). Current research efforts are needed to 1) better predict gains in real-world clinical settings and 2) identify risk factors for difficult-to-treat patients who have previously failed effective treatment strategies. The purpose of this study was to investigate predictors of residential treatment outcome in 196 adolescent (13-17 years old) patients receiving intensive combined treatment for a primary diagnosis of OCD. Clinical factors within five relevant domains were investigated to determine whether predictors above and beyond baseline symptom severity are able to predict OCD treatment response in this population. Patients were assessed at admission and discharge with standardized self-report measures of OCD symptom severity, depression, anxiety sensitivity, and anxiety-related symptoms. Admission data were used for the prediction of OCD severity at discharge as measured by the Children's Yale-Brown Obsessive-Compulsive Scale - Self Report (CY-BOCS-SR; Piacentini, Langley, & Roblek, 2007). Results indicated the variables of having received partial hospitalization/day treatment prior to admission to residential and higher severity at admission significantly predicted greater OCD severity at discharge. The relation between treatment history and outcome deserve additional exploration to uncover factors related to previous treatment experience and subsequent treatment outcomes. Differences among subtests on the CY-BOCS-SR in predicting treatment outcome provided evidence for the need for additional investigations of the link between severity of compulsory behaviors at admission and response to treatment. The challenge remains to systematically identify program components that match unique patient needs to increase prescriptive efficacy with this difficult-to-treat population.The challenge remains to systematically identify program components that match unique patient needs to increase prescriptive efficacy with this difficulty-to-treat population.