ARTIFICIAL INTELLIGENCE IN HEALTH SETTINGS: A THEMATIC ANALYSIS OF PATIENT AND PUBLIC PERSPECTIVES ON UTILITY AND CONCERNS
Author(s)
Nicolas Hall, BSc1, Bill Byrom, PhD2, Thomas Smith, BSc3, Jessica James, MSc1, Lina Eliasson, PhD1.
1Sprout Health Solutions Ltd, Pinner, United Kingdom, 2Signant Health, Nottingham, United Kingdom, 3Eltlenz Global, Lda, Lisbon, Portugal.
1Sprout Health Solutions Ltd, Pinner, United Kingdom, 2Signant Health, Nottingham, United Kingdom, 3Eltlenz Global, Lda, Lisbon, Portugal.
OBJECTIVES: This exploratory study aimed to understand patient and public perceptions of the utility of artificial intelligence (AI) across different healthcare and health research contexts, as well as associated concerns.
METHODS: A cross-sectional survey was conducted among adults in the UK, including individuals with and without chronic health conditions. Participants were presented with six hypothetical scenarios illustrating different use cases of AI in healthcare and health research contexts. These scenarios included applications such as information management, clinical decision support, and the use of synthetic data to support training and research. Quantitative data were analysed using descriptive statistics. Qualitative data, derived from participants’ free-text responses, were analysed using thematic analysis to understand key areas of perceived utility and concern.
RESULTS: Thematic analysis revealed several broad areas of perceived utility aligned with the presented scenarios. Participants highlighted the value of AI for gathering, organising, and sharing information (scenarios 1-3), as well as supporting data processing to aid clinical decision-making (scenarios 3-4). Additional areas of utility included applications in healthcare professional training (scenario 5a) and in advancing research and development (scenario 5b). Timesaving for both patients and healthcare professionals emerged as a consistent cross-cutting theme across scenarios.Alongside these benefits, key thematic areas of concern emerged. These included doubts about the reliability and accuracy of AI outputs, concerns regarding data privacy and security, a preference for human interaction over AI systems, and worries about accountability and responsibility when AI is used.
CONCLUSIONS: Overall, AI is perceived as offering meaningful utility across a range of healthcare and health research applications, though the nature of this utility varies by use case. While patient and public attitudes are generally supportive, important concerns remain. Addressing issues related to reliability, data governance, human involvement, and accountability will be essential to support responsible adoption and sustained trust in AI-enabled healthcare systems.
METHODS: A cross-sectional survey was conducted among adults in the UK, including individuals with and without chronic health conditions. Participants were presented with six hypothetical scenarios illustrating different use cases of AI in healthcare and health research contexts. These scenarios included applications such as information management, clinical decision support, and the use of synthetic data to support training and research. Quantitative data were analysed using descriptive statistics. Qualitative data, derived from participants’ free-text responses, were analysed using thematic analysis to understand key areas of perceived utility and concern.
RESULTS: Thematic analysis revealed several broad areas of perceived utility aligned with the presented scenarios. Participants highlighted the value of AI for gathering, organising, and sharing information (scenarios 1-3), as well as supporting data processing to aid clinical decision-making (scenarios 3-4). Additional areas of utility included applications in healthcare professional training (scenario 5a) and in advancing research and development (scenario 5b). Timesaving for both patients and healthcare professionals emerged as a consistent cross-cutting theme across scenarios.Alongside these benefits, key thematic areas of concern emerged. These included doubts about the reliability and accuracy of AI outputs, concerns regarding data privacy and security, a preference for human interaction over AI systems, and worries about accountability and responsibility when AI is used.
CONCLUSIONS: Overall, AI is perceived as offering meaningful utility across a range of healthcare and health research applications, though the nature of this utility varies by use case. While patient and public attitudes are generally supportive, important concerns remain. Addressing issues related to reliability, data governance, human involvement, and accountability will be essential to support responsible adoption and sustained trust in AI-enabled healthcare systems.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
PCR268
Topic
Health Service Delivery & Process of Care, Organizational Practices, Patient-Centered Research
Topic Subcategory
Patient Engagement
Disease
No Additional Disease & Conditions/Specialized Treatment Areas