Public support for sharing health data for artificial intelligence (AI) research is conditional, as revealed in a recent study conducted by the National Institute for Health Research (NIHR) and published in BMJ Digital Health & AI. The study highlights that clear public benefits, robust safeguards, and meaningful consent are essential for gaining public trust in the use of personal health data.
Lead author Rachel Kuo, a Doctoral Research Fellow at NIHR, emphasized the growing public consciousness surrounding AI’s role in healthcare. She noted that while innovation is rapid, it also necessitates access to large volumes of patient data, which raises significant concerns about confidentiality and security. Kuo stated, “Our aim was to understand how people think about sharing their data in the context of AI, and whether AI introduces particular fears or perceived benefits that shape those decisions.”
The research involved eight online focus groups with a total of 41 participants from various backgrounds across the UK. Participants discussed realistic scenarios involving health data sharing for AI, including university-led research and projects involving commercial companies. The discussions led to three key themes regarding public perceptions of health data sharing.
Conditional Support for Data Sharing
Participants expressed cautious support for health data sharing, emphasizing the importance of anonymization. While many acknowledged that some risk is inherent in data sharing, they demanded greater transparency regarding data protection measures and potential repercussions if things go awry. Trust in data users varied significantly; institutions such as universities and the National Health Service (NHS) were generally viewed positively, whereas commercial involvement raised skepticism. This skepticism diminished when participants perceived a clear link between commercial usage and potential patient benefits, provided that oversight was stringent.
Evaluating Risks and Benefits
The discussions revealed that individuals weigh the perceived risks against potential benefits when considering data sharing. Concerns about discrimination, misuse, and unknown future risks were counterbalanced by potential advantages such as improved healthcare, faster diagnoses, and contributions to future medical advancements. Participants expressed a strong sense of altruism, particularly those with long-term health conditions or previous positive experiences with medical research, often citing a desire to contribute to the greater good.
Informed consent emerged as a critical factor in fostering trust among participants. They sought information that was clear, specific, and relevant to the particular study, presented in an accessible format. Participants highlighted the importance of consent processes, advocating against requests made during vulnerable clinical moments. Suggestions included tailored approaches, options to opt out of certain data uses, “cooling-off” periods, and the ability to withdraw consent later.
The study’s methodology was robust, incorporating Patient and Public Involvement (PPI) contributors who helped shape research questions and conduct focus groups. This collaborative approach ensured that the research addressed public concerns and built confidence among participants.
PPI co-producer Rosie Hill remarked on the significance of the study, stating, “It is essential that we understand, in real time, how this area of technology and science is advancing.” Another co-producer, Judi Smith, noted that the focus groups provided valuable insights into how individuals assess the complexities of sharing their data. One participant voiced the conflicting emotions many feel, saying that when wearing her “person-hat,” she had reservations about sharing her data, particularly with commercial entities. However, donning her “patient-hat,” she expressed a willingness to share data if it could expedite treatment for her condition.
Kuo concluded that as healthcare systems increasingly rely on extensive data to develop and evaluate AI technologies, public trust is crucial. She emphasized that while individuals are open to data sharing, this willingness is contingent upon clear conditions, including transparency, strong governance, meaningful consent, and demonstrable public benefits. Understanding these expectations is vital for creating an ethical and sustainable framework for data-driven innovation in healthcare.
The findings of this study underscore the need for ongoing public engagement and dialogue as technology and science evolve in the healthcare landscape.