Guided Self-help for Common Mental Disorders

Leveraging Computational Social Sciences and Natural Language Processing to Optimize Engagement and Response to Low-intensity CBT for Depression and Anxiety

Common mental disorders (CMDs) like depression and anxiety account for a large proportion of disability worldwide. Access to effective treatments like cognitive-behavioral therapy (CBT) is limited and has not reduced the public health burden of psychopathology. For patients with mild-moderate CMDs, lower-intensity treatments like guided self-help CBT (GSH-CBT) are effective and more scalable (e.g., via the internet). The advent of social media has opened avenues for dissemination of GSH-CBTs and allows for passive sensing of mood, thinking, behavior, and social networks. We propose to leverage a social media platform used by over a fifth of the United States (Twitter) as a recruitment tool to virtually screen over 150 individuals, recruit N=60 to a 5-week course of GSH-CBT, and extract social media data from individuals engaged in GSH-CBT. Sociodemographic and social media data will be used to predict engagement, outcomes, and processes in GSH-CBT.

Inclusion Criteria:

• At least mild distress: K6 score ≥ 6
• Having reasonably regular access to the internet or a telephone

Exclusion Criteria:

- Suicidality: Patient Health Questionnaire-9 (PHQ9) item 9 ("thoughts that you would be better off dead, or of hurting yourself ") ≥ 2 ("more than half the days")