**2. Methods**

A systematic review of studies catalogued in PubMed was performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (www.prismastatement.org). The search was done over the full time span up until 17 April 2020. Only papers in English were included. We used the following broad Booleian search strategy: "(deep brain stimulation) AND (obsessive compulsive disorder) AND ((exposure and response prevention) OR (behavioral therapy) OR (cognitive behavioral therapy))". Given the limited yield of a narrower, exploratory search, all papers that addressed any form of postoperative CBT in patients receiving DBS for therapy-resistant OCD were included. This not only comprised clinical trials, cohort studies, case series and case studies, but also systematic and narrative reviews on DBS for OCD and position papers if they also commented on CBT after DBS. Reference lists of the included studies were checked for additional papers. The Evidence Project risk of bias tool was used to assess the quality of the included studies (not being reviews) [11]. Two authors rated the quality (MG and AL) and discrepancies were resolved by consensus. Due to the very limited yield of our search, no minimum quality score was applied for inclusion. The e ffectiveness of CBT added to DBS was quantified by looking at the changes in scores on the Yale–Brown Obsessive Compulsive Scale (Y-BOCS) before and after CBT. Although the initial intention was to perform a meta-analysis, this was not possible because of the limited number of included studies that, in addition, used di fferent indication criteria and di fferent forms and ways of delivery of CBT. Timing and procedural aspects of CBT in the studies are reported in a descriptive way.
