PAL: Designing Conversational Agents as Scalable, Cooperative Patient Simulators for Palliative-Care Training

2025-07-02

AI Summary

In end-of-life and palliative care, clinical communication skills training is constrained by the scarcity of standardized patients. This study introduces PAL—a large language model–based multimodal conversational agent system that simulates emotionally nuanced patient dialogues and delivers structured, empathy-informed feedback via both text and speech modalities. Its contributions are threefold: (1) it establishes the first emotion-sensitive paradigm for medical dialogue generation; (2) it implements a collaborative AI-augmented training architecture with a multimodal human–AI closed-loop learning mechanism; and (3) it yields a scalable, high-stakes clinical training system. In a mixed-methods study involving 17 U.S. medical students and clinicians, PAL significantly enhanced depth of reflection and self-assessed communication competence. These findings provide empirical support for AI-enhanced clinical communication education.

Technology Category

• Natural Language Processing: Conversational AI/Dialog Systems
• 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
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Effective communication in serious illness and palliative care is essential but often under-taught due to limited access to training resources like standardized patients. We present PAL (Palliative Assisted Learning-bot), a conversational system that simulates emotionally nuanced patient interactions and delivers structured feedback grounded in an existing empathy-based framework. PAL supports text and voice modalities and is designed to scaffold clinical skill-building through repeated, low-cost practice. Through a mixed-methods study with 17 U.S. medical trainees and clinicians, we explore user engagement with PAL, evaluate usability, and examine design tensions around modalities, emotional realism, and feedback delivery. Participants found PAL helpful for reflection and skill refinement, though some noted limitations in emotional authenticity and the adaptability of feedback. We contribute: (1) empirical evidence that large language models can support palliative communication training; (2) design insights for modality-aware, emotionally sensitive simulation tools; and (3) implications for systems that support emotional labor, cooperative learning, and AI-augmented training in high-stakes care settings.

Problem

• Simulating emotionally nuanced patient interactions for palliative-care training
• Providing structured feedback using an empathy-based framework
• Scaling access to low-cost, repeated clinical skill practice

Innovation

• Conversational agent simulates nuanced patient interactions
• Supports text and voice for clinical practice
• Uses empathy framework for structured feedback