Enhancing Pediatric Communication: The Role of an AI-Driven Chatbot in Facilitating Child-Parent-Provider Interaction

Research Background and Problem

• Problem or Challenge: The authors highlight that effective communication with pediatric cancer patients is a significant yet complex challenge due to children's limited ability to express emotions and symptoms at their developmental stage. Additionally, caregivers often lack sufficient time to provide parents with adequate resource support. In this context, children's communication needs are frequently overlooked, with more focus placed on parents' needs. While existing technologies offer some communication aids for children, they fail to address the dynamic and complex communication requirements between children and parents, and do not effectively promote information sharing and understanding among children, parents, and healthcare providers.
• Importance: Clear and sensitive communication in pediatric cancer care can improve disease management outcomes for children and positively impact their psychological and emotional well-being. Addressing this issue not only enhances the treatment experience for patients but also alleviates parental stress and improves the efficiency of the entire care team.
• Research Motivation and Related Work: Previous studies have demonstrated the potential of technology in pediatric health management, such as intelligent drawing tools or animation-based interactive devices. However, these technologies primarily focus on enhancing children's expressive abilities and fail to dynamically adapt to the complex communication needs among children, parents, and providers. The authors propose that AI-driven chatbots may better address the communication challenges between children, parents, and healthcare providers.

Solution

• Proposed Approach: The authors designed and developed a prototype chatbot system named ARCH (Agents for Reinforcing Child-parent Health communication), which consists of two sub-chatbots:

  • Child Bot: Acts as a "peer" to help children express concerns and emotions.
  • Expert Bot: Generates anonymized scenarios to provide parents with communication advice and guidance.

  The system connects these two chatbots through a linear dialogue process to meet the distinct needs of children and parents.

• Innovations:

  • ARCH introduces a dual-bot architecture that protects children's privacy while meeting parents' needs for critical information.
  • The Child Bot incorporates a "peer emotional support" design, providing a non-judgmental space where children feel more comfortable sharing.
  • The Expert Bot uses information collected from children to generate practical case scenarios, offering parents specific communication guidance and opportunities for reflection and validation of their communication behaviors.

• Implementation Steps and Key Technologies:

  • Phase 1: Design Conversations: Conducted design sessions with 12 pediatric cancer patients and their parents to explore their expectations and needs for the chatbot.
  • Prototype Development and Implementation:
    • The Child Bot's dialogue process is divided into three stages (Explore, Develop, Summarize), while the Expert Bot focuses on single-stage communication guidance.
    • Built a unified system using a ReactJS and FastAPI-based front-end and back-end framework, integrating the GPT-3.5 large language model to support natural language interaction.
  • Phase 2: Expert Interviews: Interviewed 15 experts in the field of child communication to further validate the chatbot's design and identify potential challenges.
  • Design Decisions: Focused on emotional capture, privacy protection, and adapting to diverse cultural and care environments in chatbot design.

Research Outcomes

• Specific Results:

  • The ARCH prototype was designed as a communication aid between children and parents, supporting children's emotional expression, parents' behavioral reflection, and healthcare providers' assessment of communication needs.
  • Three core roles were clearly defined: serving as an emotional outlet for children, providing parents with tools for communication reflection and validation, and offering healthcare providers an assessment aid.

• Advantages:

  • Child-Centric Design: The Child Bot's "peer role" effectively reduces the psychological burden of discomfort for children, enabling them to naturally share feelings and concerns.
  • Parent-Focused Functionality: The Expert Bot provides opportunities for reflection and validation, helping parents better understand their children's needs.
  • For Healthcare Teams: Offers an assessment tool that saves time for healthcare teams while expanding available data sources.

• Experimental or Evaluation Results:

  • Expert feedback indicated that ARCH provides an innovative approach to fostering understanding and communication among children, parents, and healthcare providers.
  • Experts highly appreciated the emotional support function of the Child Bot and the scenario-based guidance provided by the Expert Bot.
  • Some experts pointed out potential cultural adaptation needs and privacy protection risks, offering directional guidance for further improvements.

• Limitations and Future Directions:

  • Limited to cancer care settings: The system's adaptability to other disease contexts (e.g., diabetes or heart disease) needs to be tested.
  • Insufficient cultural adaptability: Current trials in the U.S. and South Korea provide some insights, but further exploration is needed to adapt chatbots to more diverse cultural contexts.
  • Privacy concerns in information sharing: Further research is required to achieve data sharing while respecting children's privacy.
  • Broader professional needs: The study primarily focuses on feedback from psychologists and social workers; future work should include input from doctors and nurses.

Conclusion

This study demonstrates the significant potential of AI-driven chatbots in improving communication among children, parents, and healthcare providers. The design and testing process of ARCH indicates that by creating a safe, emotionally resonant, and culturally adaptive space for expression, chatbots can significantly enhance the interaction experience in patient care settings. The authors provide valuable references for future research and encourage exploring the potential of this technology in broader contexts.