Discovering Common Elements of Empirically Supported Self-Help Interventions for Depression in Primary Care: a Systematic Review.

Abstract

Although the efficacy of self-help cognitive-behavioral therapy (CBT) for depression has been well established, its feasibility in primary care settings is limited because of time and resource constraints. The goal of this study was to identify common elements of empirically supported (i.e., proven effective in controlled research) self-help CBTs and frameworks for effective use in practice. Randomized controlled trials (RCTs) for self-help CBTs for depression in primary care were systematically identified in Pubmed, PsycINFO, and CENTRAL. The distillation and matching model approach was used to abstract commonly used self-help techniques (practice elements). Study contexts associated with unique combinations of intervention elements were explored, including total human support dose (total face-to-face, telephone, and personalized email contact time recommended by the protocol), effective symptom domain (depression vs. general psychological distress), and severity of depression targeted by the study. Relative contribution to intervention success was estimated for individual elements and human support by conditional probability (CP, proportion of the number of times each element appeared in a successful intervention to the number of times it was used in the interventions identified by the review). Twenty-one interventions (12 successful) in 20 RCTs and 21 practice elements were identified. Cognitive restructuring, behavioral activation, and homework assignment were elements appearing in > 80% of successful interventions. The dose of human support was positively associated with the proportion of interventions that were successful in a significant linear fashion (CPs: interventions with no support, 0.20; 1-119 min of support, 0.60; 120 min of support, 0.83; p = 0.042). In addition, human support increased the probability of success for most of the extracted elements. Only social support activation, homework assignment, and interpersonal skills were highly successful (CPs ≥ 0.60) when minimal support was provided. These findings suggest that human support is an important component in creating an evidence-informed brief self-help program compatible with primary care settings.