Barriers and enablers for the deployment of large language model-based conversational robots for older adults: A protocol for a systematic review of qualitative studies

摘要

BACKGROUND: Artificial intelligence-powered conversational agents have immense potential to provide social companionship and support for older adults. However, the deployment of large language model (LLM)-based conversational robots for seniors faces various technical, user acceptance, and ethical challenges. OBJECTIVES: This systematic review aims to synthesize insights from prior qualitative studies to identify key factors that influence the real-world application of LLM-based conversational agents for the aging population. The review will inform the user-centered design of these technologies, policy discussions on their governance, and highlight research gaps. METHODS AND ANALYSIS: Eleven electronic databases will be searched for qualitative studies exploring stakeholder perspectives on using AI chatbots and robots to assist seniors. Two reviewers will independently screen studies, extract data, and appraise methodological quality using the JBI checklist. Thematic analysis will be conducted to identify major barriers and enablers, and confidence in review findings will be assessed using the GRADE-CERQual approach. The review will adhere to PRISMA-P and ENTREQ reporting guidelines to ensure transparency. DISCUSSION: Understanding and addressing obstacles to implementing LLM-powered conversational agents for older adults is crucial for leveraging this technology to support the well-being of the rapidly aging global population. This systematic review will provide timely insights to guide the responsible development and deployment of AI companions for seniors. TRIAL REGISTRATION: ClinicalTrials.gov CRD42024601264.