Depression intervention using AI chatbots with social cues: a randomized trial of effectiveness.

Area of Science:

• Digital mental health
• Human-computer interaction
• Clinical psychology

Background:

• Depression is a prevalent issue among college students.
• Chatbots are increasingly used for mental health interventions.
• The impact of social cues in chatbot design for depression treatment requires further investigation.

Purpose of the Study:

• To compare the effectiveness of high-social-cue (HSC) chatbots versus low-social-cue (LSC) chatbots in treating depressive symptoms.
• To evaluate the influence of social cues on treatment outcomes, user satisfaction, and adherence.

Main Methods:

• An 16-week open-label randomized controlled trial involving 84 college students with moderate to severe depressive symptoms (PHQ-9 scores ≥9).
• Participants were assigned to either an HSC group (text, voice, animations) or an LSC group (text-only).
• Outcomes included changes in PHQ-9, GAD-7, PANAS scores, user satisfaction (CSQ-8), therapeutic alliance (WAI-SR), and adherence.

Main Results:

• The HSC group showed significantly greater reductions in depression (PHQ-9) and anxiety (GAD-7) scores compared to the LSC group.
• Higher adherence rates, user satisfaction (CSQ-8), and therapeutic alliance (WAI-SR) were observed in the HSC group.
• No significant differences in baseline characteristics were found between the groups.

Conclusions:

• Chatbots incorporating high-social-cue designs are more effective in alleviating depression and anxiety symptoms in college students than text-only chatbots.
• HSC chatbots enhance user adherence, satisfaction, and the development of a therapeutic alliance.
• Findings support the integration of richer social cues in chatbot design for improved mental health interventions.