CBT and pharmacotherapy after nonresponse for anxiety and obsessive-compulsive disorders in youth: a systematic review and meta-analysis

Approximately half of children with anxiety disorder(s) or obsessive-compulsive disorder (OCD) derive insufficient benefit from first-step evidence-based treatments, including cognitive behavioural therapy (CBT) and pharmacotherapy. Few studies have examined second-step treatment options, and clinical guidelines offer limited direction on subsequent care. This systematic review and meta-analysis synthesises current evidence on interventions for children and adolescents with anxiety disorders or OCD, who showed insufficient response to prior evidence-based treatment. We examined how studies operationalised partial response and non-response, which second-step treatments were provided, and the efficacy of second-step treatments. Five databases were systematically searched up to 21 March 2025, and 40 studies with 49 treatment arms (N = 1,373) were included, most of which were uncontrolled (k = 35 of 40 studies) and were identified as a high risk of bias. There was substantial heterogeneity among the included studies, reflecting variety in operationalisations of partial and non-response, and among first-step and second-step treatments. Included second-step interventions were CBT, pharmacotherapy, or combined treatments. Sixteen of the 40 included studies examined second-step interventions in a stepped-care context. Within-group meta-analyses showed large, statistically significant pooled effects of second-step treatments for symptom reduction and general functioning from pre-treatment to post-treatment and to follow-up, both for anxiety disorders and OCD. Remission rates ranged from 6% to 82.4%. Pooled remission rates were 49% at post-treatment and 67% at follow-up for anxiety studies, and 39% at post-treatment and 53% at follow-up for OCD studies. Drop-outs during second-step treatments ranged from 0% to 36.4%, with a pooled drop-out rate of 8% for anxiety studies and 9% for OCD studies. Subgroup analyses on second-step treatment type (CBT, pharmacotherapy, combined) revealed no statistically significant difference between subgroups on any outcome. Overall, the lack of well-designed, controlled stepped-care studies precludes reliable conclusions regarding the efficacy of second-step treatments for youth who benefited insufficiently from initial treatment, despite preliminary indications of positive pre-post changes in symptoms and remission from uncontrolled studies.