Safety behaviors, compulsive behaviors, and therapist extensiveness as predictors of outcome in behavior for youth OCD

The efficacy of behavior therapy in treating adult OCD is well-documented and early trials in youth populations are encouraging (March et al., 1994; Piacentini et al., 2002; POTS Team, 2004), however, residual symptoms often persist. The study of process variables can provide suggestions for improving the development, delivery, and outcomes of psychological therapies (Kazdin & Nock, 2003). One potentially important variant in exposure procedures identified by the literature has been the use of safety behaviors. However, it has been difficult to draw conclusions regarding the effect of safety behaviors on OCD treatment because the term has been interpreted broadly across studies and results have been mixed. The current study uses observational coding of existing data to assess three predictors of outcome within two early exposure sessions and two late exposure sessions of a manualized exposure and response prevention program. Participants include 43 youth (ages 8 – 17 years) diagnosed with a principal OCD diagnosis who received a 12-week exposure and response prevention program. Predictors assessed via observational coding include safety behaviors (avoidance and escape), compulsive behaviors, and therapist extensiveness (therapist ability to engage the client in difficult exposures). Outcome was measured by symptom report (CY-BOCS) at midtreatment and post-treatment. It was hypothesized that higher average occurrence of safety and compulsive behaviors during exposure tasks would predict higher OCD symptoms while lower ratings of therapist extensiveness would predict higher OCD symptoms at mid-treatment and post-treatment. Interaction effects were also investigated. Multiple regression analyses demonstrated a significant relationship between safety behaviors exhibited during the early exposure phase and OCD symptoms at mid-treatment in the expected direction, B = 0.31, t = 2.09, p<.05, R2=.25. However, safety behavior in the late phase was not related to OCD symptoms at post-treatment nor was safety behavior throughout treatment significantly related to OCD symptoms at posttreatment. Significant findings involving compulsive behaviors, therapist extensiveness, and interaction effects did not emerge. Methodological limitations, such as restricted range of the independent variables and the need for a more frequently administered dependent variable, and recommendations for future studies are discussed.Psy.DIncludes bibliographical referencesby Daniela Colognor