A systematic review of potential publication bias in U.S. National Institutes of Health-funded trials of psychological treatments for obsessive-compulsive disorder in adults

Cognitive-behavioral therapy (CBT) with exposure and response prevention is the most empirically supported psychological treatment for obsessive–compulsive disorder (OCD). Healthcare organizations and practice guidelines recommending CBT for OCD typically rely on meta-analyses that usefully aggregate treatment outcome findings. However, a critical methodological aspect of meta-analyses is the potential impact of publication bias— that is, the tendency for significant findings to be published more frequently than non-significant ones, which can inflate overall effect sizes and lead to inaccurate conclusions about treatment efficacy. Identifying unpublished studies is difficult without a complete record of all studies actually conducted. To address this issue, we searched databases of United States National Institutes of Health (NIH) grants between 1980–2018 that funded comparator randomized controlled trials (RCTs) examining the efficacy of psychological treatments for adults with OCD and determined whether identified grants resulted in publications. Results indicated that 100 % of the ten identified grants resulted in publications, suggesting an absence of publication bias in this area of the literature. Quantitative aggregation of findings from the ten identified grants revealed a large effect size (g = 0.74, 95 % CI = [0.48, 0.99)], τ = 0.44). for comparisons between identified psychological treatments and a range of comparator conditions for reducing OCD symptomology measured by the Yale-Brown Obsessive Compulsive Scale (Y-BOCS). Findings are discussed in the context of meta-science issues and methodological challenges related to publication bias in hopes of improving clinical research methods and intervention development for treatments of OCD.