Background: Previous meta-analyses of paediatric obsessive-compulsive disorder (OCD) have shown much higher effect size for standard individual cognitive behaviour therapy (SI-CBT) compared with control conditions than for serotonin reuptake inhibitors (SRIs) compared with placebo. Other factors, such as systematic differences in the provided care or exposure to factors other than the interventions of interest (performance bias) may be stronger confounders in psychotherapy research than in pharmacological research.
Aims: These facts led us to review SI-CBT studies of paediatric OCD with the aim to compare the effect estimates across different comparisons, including active treatments.
Method: We included only randomized controlled trials (RCTs) or cluster RCTs with treatment periods of 12-16 weeks. Outcome was post-test score on the Children's Yale-Brown Obsessive Compulsive Scale (CYBOCS).
Results: Thirteen papers reporting from 13 RCTs with 17 comparison conditions were included. SI-CBT was superior to wait-list and placebo therapy but not active treatments. Effect estimates for SI-CBT in wait-list comparison studies were significantly larger than in placebo-therapy comparison studies. In addition, the SI-CBT effect estimate was not significantly different when compared with SRIs alone or combined SRIs and CBT.
Conclusions: Performance bias may have inflated previous effect estimates for SI-CBT when comparison contingencies included wait-list. However, the calculated SI-CBT effect estimate was lower but significant when compared with placebo therapy. The effects of SI-CBT and active treatments were not significantly different. In conclusion, our data support the current clinical guidelines, although better comparisons between SI-CBT and SRIs are needed.
Keywords: Child and adolescent; Cognitive behaviour therapy; Comparison; Meta-analysis; Obsessive–compulsive disorder; Outcome; Paediatric; Placebo; Psychotherapy; Wait-list.