Large Language Model-Powered Conversational Agent Delivering Problem-Solving Therapy (PST) for Family Caregivers: Enhancing Empathy and Therapeutic Alliance Using In-Context Learning

2025-06-13

AI Summary

Family caregivers face elevated mental health risks due to role overload and limited psychosocial resources, necessitating accessible, evidence-based psychological support. Method: This study introduces the first tripartite clinical framework integrating Problem-Solving Therapy (PST), Motivational Interviewing (MI), and Behavioral Chain Analysis (BCA), embedded within a large language model (LLM)-driven conversational agent. We propose a novel clinical prompt engineering approach combining few-shot learning with retrieval-augmented generation (RAG) to deliver personalized, evidence-informed interventions. The system is built upon an expert-annotated dataset and structured dialogue templates. Contribution/Results: In a randomized trial with 28 caregivers, the optimal configuration significantly improved empathic perception (p < 0.01) and therapeutic alliance scores. Ninety-two percent of participants rated the agent as emotionally validating and its strategies as feasible. Results demonstrate the clinical viability and innovative potential of LLMs for lightweight, highly adaptable psychological support tailored to family caregivers.

Technology Category

• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Multiagent Systems: Agent Communication
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Family caregivers often face substantial mental health challenges due to their multifaceted roles and limited resources. This study explored the potential of a large language model (LLM)-powered conversational agent to deliver evidence-based mental health support for caregivers, specifically Problem-Solving Therapy (PST) integrated with Motivational Interviewing (MI) and Behavioral Chain Analysis (BCA). A within-subject experiment was conducted with 28 caregivers interacting with four LLM configurations to evaluate empathy and therapeutic alliance. The best-performing models incorporated Few-Shot and Retrieval-Augmented Generation (RAG) prompting techniques, alongside clinician-curated examples. The models showed improved contextual understanding and personalized support, as reflected by qualitative responses and quantitative ratings on perceived empathy and therapeutic alliances. Participants valued the model's ability to validate emotions, explore unexpressed feelings, and provide actionable strategies. However, balancing thorough assessment with efficient advice delivery remains a challenge. This work highlights the potential of LLMs in delivering empathetic and tailored support for family caregivers.

Problem

• Addressing mental health challenges in family caregivers
• Evaluating LLM-powered therapy for empathy and alliance
• Balancing thorough assessment with efficient advice delivery

Innovation

• LLM-powered conversational agent for mental health
• Combines PST, MI, and BCA techniques
• Uses Few-Shot and RAG prompting methods