Background: Mobile Behavioral Activation (BA) is efficacious for the treatment of perinatal depression, however, the effect of comorbidity on symptom trajectories remains underexplored which is important given that at least 10% of women in the perinatal…

Background: #MentalHealth chatbots are increasingly used to support people with #depressive symptoms, and large language models make these systems more flexible than rule-based chatbots. However, it remains unclear how well large language model–based chatbots deliver structured psychological interventions. Objective: This study examined how well a GPT-4o–based chatbot delivered a behavioral activation intervention for young people with #depression using sessions with artificial users and clinical expert assessment. It also identified limitations and potential refinements. Methods: We implemented a GPT-4o (gpt-4o-2024-08-06; OpenAI)–based chatbot using a structured system prompt to deliver a single-session behavioral activation intervention for people with #depression aged 14 to 29 years. We generated 48 sessions with GPT-4o–based artificial users derived from clinical vignettes varying across 7 characteristics. Ten clinical experts, either licensed psychotherapists or advanced psychotherapy trainees, independently assessed the sessions using the 14-item Quality of Behavioral Activation Scale (Q-BAS), rated from 0 to 6, supplemented by rating therapeutic capabilities, artificial user authenticity and difficulty, and qualitative feedback. Results: The chatbot completed all 7 intervention phases in every session. The mean holistic session quality rating was 3.94 (SD 1.23), and the mean Q-BAS rating was 4.03 (SD 1.18). Thirteen of 14 Q-BAS components exceeded the satisfactory threshold of 3. Ratings were highest for mood assessment (mean 5.42, SD 1.09) and activity planning (mean 4.98, SD 1.41) and lowest for explaining positive reinforcement (mean 2.92, SD 2.30) and supporting activity-mood monitoring (mean 3.02, SD 2.04). Therapeutic capability ratings were highest for message safety (mean 5.90, SD 0.37), message clarity (mean 5.56, SD 0.77), and objective, nonjudgmental communication (mean 5.17, SD 1.04) and lowest for therapeutic r#Apport (mean 4.12, SD 1.45) and natural conversation flow (mean 4.25, SD 1.42). Artificial users were rated below the scale midpoint for authenticity (mean 2.75, SD 1.41) and difficulty (mean 1.23, SD 1.46). Clinical experts described the chatbot as structured, clear, and safe but identified insufficient clinical reasoning as the main limitation, particularly in evaluating the therapeutic suitability and feasibility of activities, barriers, solution strategies, and rewards. Artificial users were often highly compliant, especially when identifying positive activities. Conclusions: In expert-rated sessions with artificial users, the chatbot delivered the behavioral activation intervention as intended and performed strongest on procedural components. It performed less well on positive reinforcement and activity-mood monitoring, indicating refinement needs in clinical reasoning, follow-up questioning, and evaluating whether proposed activities, plans, barriers, solution strategies, and rewards are therapeutically #Appropriate and feasible. The findings identify targets for improvement before testing with human users, while the artificial user design and expert ratings limit conclusions about real therapeutic interactions.

Background: Large Language Models (LLMs) have rapidly advanced across numerous fields, including mental health care. A shortage of trained therapists and mental health care providers has driven informal use of LLMs for therapeutic support. However, their clinical utility remains poorly defined. This study aimed to systematically evaluate the capabilities and limitations of LLMs in single-turn therapeutic interactions compared to psychotherapists-in-training. Objective: To systematically evaluate and compare the therapeutic knowledge and single-turn response capabilities of LLMs versus psychotherapists-in-training in the context of Behavioral Activation therapy for depression, and to assess how both groups' performance changes when provided with structured therapeutic training materials. Methods: Participants (n=6 LLMs, n=8 human) completed a questionnaire on depression and Behavioral Activation (BA) with 20 multiple choice items, and 10 therapy scenarios with 3 open-ended items each that postulated empathic response, use of validation strategies, and Theory of Mind capabilities. Human participants completed the questionnaire prior and post to a 5-hour workshop and five-week period with learning material. LLMs received identical training content as context during the second test. All open-ended questions were rated on 5-point scales by two experts. Results: At baseline, LLMs demonstrated higher knowledge scores than human participants (61.0 vs. 52.0 out of 100 points) and were rated higher in empathy (U=2.0, p=.005, r=.917), validation quality (U=2.5, p=.006, r=.896), anticipation of cognition (U=0.0, p=.002, r=1.000), and anticipation of emotion (U=0.0, p=.002, r=1.000). Following BA training, LLMs maintained their performance advantage across multiple choice and open-ended items. Conclusions: The results suggest that LLMs may generate high-quality therapeutic single-turn responses that integrate clinical knowledge with empathetic communication. The findings hint at LLMs' potential as valuable tools in mental health care, though further clinical trials are needed to evaluate their performance in ongoing therapeutic relationships and clinical outcomes.

Background: Brief psychosocial interventions, such as the Healthy Activity Program (HAP), which are based on behavioral activation and delivered by nonspecialist providers (NSPs), have emerged as cost-effective solutions for the treatment of depression. HAP treatment outcomes are improved by the engagement of patients in activation-focused homework assignments and their adherence to these assignments during therapy. Currently, patients are expected to complete these homework assignments using a paper workbook. Objective: The aim of this study was to describe the user-centered development process of PracticePal, a chatbot designed to enhance patient engagement and homework adherence, and to evaluate its #feasibility and acceptability as a therapy aid in India. Methods: We used a user-centered approach to co-develop PracticePal, incorporating conversational flows and video scripts in Hindi. The chatbot was piloted with 30 participants having depression who were receiving the HAP from 15 nonspecialist counselors in primary care in rural Madhya Pradesh, India. The #feasibility and acceptability of PracticePal were assessed through engagement data, in-depth interviews with a subset of 6 participants, and focus group discussions with 11 counselors. Treatment completion rates and changes in depressive symptoms were explored as secondary outcomes. Results: Average patient engagement spanned 29 days (95% CI 24-34) during the 60-day treatment period. The engagement of patients with PracticePal increased as their treatment progressed, particularly after the third HAP session. Of the 30 patients, 20 (67%) accessed more than half of the multimedia content available on the chatbot. On average, there was a greater frequency of self-initiated engagement (1558 out of a total of 1835 times, 84.9%) across all sessions compared with reminder prompts (277 out of 1835 times, 15.1%). All 30 patients completed treatment and experienced a reduction in the mean Patient Health Questionnaire-9 score from 13 (95% CI 12.6-13.6; signifying moderate severity) to 4 (95% CI 2.9-4.7; signifying none/minimal severity). Patients found the chatbot’s reminders for activities, mood tracking, and video messages helpful and observed that it could help others in their social network. NSPs also reported improved participation of patients in the homework tasks compared with the paper workbook. A few patients faced challenges with low internet bandwidth, and those with limited literacy suggested increasing the amount of video content for easier accessibility. Conclusions: The PracticePal chatbot is a feasible and acceptable therapy aid to complement a psychological treatment, with promising potential to enhance the effectiveness of NSP-delivered psychosocial interventions in low-resource settings. Future steps include conducting a fully powered randomized controlled trial to assess its effectiveness in improving mental health outcomes.