(PS10-A6) Self-guided Behavioral Activation Using a Cbt-based Single-session Intervention

Introduction: Depression is a leading cause of disability worldwide. Behavioral activation (BA) is an empirically supported treatment for depression. However, traditional mental health care (e.g., one-on-one psychotherapy with a professional) cannot address the public health burden of untreated mental disorders, including depression. Low-intensity interventions, which may not require the use of a professional, have the potential to address the public health burden of mental disorders by serving as a more accessible tool for mental health care. These low-intensity interventions can be self-guided or use paraprofessionals to guide the intervention. One example of a low-intensity intervention is an online single-session intervention (SSI). The Common Elements Toolbox (COMET) is an SSI delivered online and self-guided (e.g., without the guidance of a professional or paraprofessional). COMET contains CBT and positive psychology elements, including BA. It is unclear, however, to what extent self-guided BA reflects BA as traditionally presented in face-to-face CBT. Machine learning approaches may yield novel insight into how to further personalize BA in self-guided treatments.

Methods: Using data from an 8-week randomized controlled trial of COMET-SSI with online workers, we conducted a content analysis of the activities generated from the BA portion of the intervention. 409 participants were asked to brainstorm three possible activities, then choose one to schedule in the future. Two experts in CBT used qualitative content analysis to code activity types into established BA themes. Discrepancies between raters were resolved through discussion until complete consensus was achieved. Additionally, we compared the self-generated activities to a BA dictionary using Linguistic Inquiry and Word Count (LIWC), an established natural language processing tool, to quantify the similarity of the self-guided BA content to those used in online text-based counseling sessions.

Results: Results yielded 9 different activity types, including sedentary hobbies (41%), physical (29%), active hobbies (18%), social (6%), task-oriented (3%), psychological/emotional (1%), spiritual (0.6%), volunteering (0.2%), and academic (0.1%). The initial codebook contained one code for “hobbies”, which included a large degree of heterogeneity. The most frequently generated activities were sedentary hobbies, including crafting, listening to music, and reading. Therefore, the raters divided the “hobbies” theme into “active” vs. “sedentary” hobbies. LIWC analysis found an average lexicon overlap of 17% between the self-guided BA activities and those in the BA dictionary.

Discussion: Results suggest that there may be considerable differences between self-guided BA compared to traditional BA psychotherapy. Specifically, users of the self-guided SSI may schedule more sedentary activities, and ones that involve less social support compared to BA in traditional psychotherapy settings. Implications of this work include providing more psychoeducation and specificity in what constitutes BA. Future work should examine how different activity types of self-generated BA may be differentially associated with depression treatment outcomes.