Empirically-derived response trajectories of intensive residential treatment in obsessive-compulsive disorder: A growth mixture modeling approach.

Background: This study investigated distinct trajectories of treatment response in a naturalistic intensive/residential treatment (IRT) program for adults with severe obsessive-compulsive disorder (OCD). We hypothesized that: (1) distinct trajectories would emerge and (2) demographic variables, psychiatric comorbidity, OCD symptom subtype, level of insight, previous exposure and response prevention (ERP) treatment, and quality of life, would differentially predict assignment to these trajectories.
Methods: Participants included 305 individuals with primary OCD admitted for IRT.
Results: Two trajectories emerged over the course of the first eight weeks of treatment, with the vast majority of participants demonstrating treatment response. The first trajectory (96%, n = 292) showed a negative, linear treatment response (a.k.a. "linear responders") and more severe OCD symptoms at admission. The second trajectory (4%, n = 13) had less severe OCD symptoms at admission and did not exhibit a significant overall change in symptoms over the course of treatment. More specifically, this second trajectory or "u-shaped responders" show a non-significant linear response through week four of treatment, followed by slightly increased symptoms in week five. Assignment to these classes was not differentially predicted by hypothesized predictor variables.
Limitations: Our final model had inconsistent fit indices and small class prevalance of the u-shaped responder group; therefore, model selection was based on both fit indices and substantive meaning.
Conclusions: This study emprically derived two distinct trajectories of OCD symptom severity over the course of IRT. These findings have the potential to refine IRT for patients with severe OCD, and to potentially guide future investigation into the optimal delivery of ERP treatment for OCD generally.
(Copyright © 2018. Published by Elsevier B.V.)