End-user perspectives on the use of personal health information for AI development and the use of AI in care: an interview study

INTRODUCTION: The rapid advancement of artificial intelligence (AI) in health care necessitates that decision-makers consider end-user views on secondary data use and the application of AI in clinical settings. AIM: The aim of this study was to explore the perspectives of patients and health professionals on the use of personal health information for the development of AI tools and the implementation of these tools into health care. METHODS: An observational, cross-sectional study involving semi-structured, scenario-based interviews with 51 participants; grouped as health professionals, patients with chronic, mental health, or rare conditions, and individuals undergoing routine screening. RESULTS: Participants from all groups found it acceptable to use patient health information for AI on the condition that the information was intended to serve the greater good. All groups emphasised the potential for AI to support health professional staff by streamlining processes and reducing administration time. All groups also agreed on the importance of maintaining patient trust in health services and emphasised transparency, appropriate use of data, responsible data security measures, obtaining patient consent, and strong governance. Although these thematic overlaps demonstrated broad end-user agreement, there were also notable group-level differences in the emphases within these overarching themes. DISCUSSION: Although this research has shown that patients and health professionals shared broadly aligned but nuanced views on the use of AI in health care, a one-size-fits-all approach to seeking social licence in this area is unlikely to be possible. Recommendations include ongoing and continued engagement with health professionals and consumers that reflects the diversity of interest-holders and perspectives involved.