AI in health care

dlvr.it

Barriers and Facilitators to Health Care AI Adoption Among Those Living in Wales and Working in Health Care in Wales: Online Survey

Background: NHS Wales routinely collects patient-reported outcome measures, and these, together with other clinical data, offer an opportunity to design machine learning (ML) technologies that could advance the implementation of prudent health care principles (a health care strategy encouraged by the Welsh Government). However, the wide adoption of such technologies is not only dependent on the development of technically well-performing ML algorithms but also on end-user barriers and facilitators. Objective: This study aimed to identify potential end-user (patient and health care professional) barriers and facilitators to the use of ML in health care decision-making in Wales. The study’s objective was to provide actionable information for those who are developing and implementing ML technologies in health care, rather than contributing to the theoretical advance of technology implementation frameworks. Methods: An online survey using Microsoft Forms was conducted. It was open to anyone who was 16 years or older and lived in Wales (member of the public criterion) or was a registered health care professional working in Wales and participating in treatment or therapy decision-making (health care professional criterion). The anonymous survey was open from December 4, 2024, to March 4, 2025. The survey used single-choice, ranking, and free-text questions, which were phrased differently for both eligibility groups. Data analysis was based on the respondent-selected eligibility criterion and self-declared general attitude toward health care artificial intelligence (AI; generally supportive, opposed, or uncertain), using descriptive and inferential statistics, as well as a summary of free-text responses. Results: A total of 309 respondents filled out the survey, 179 selecting the member of the public criterion and 130 selecting the health care professional criterion. Among them, 209 self-identified as having a generally supportive attitude toward health care AI, 31 as generally being opposed to health care AI, and 69 as being uncertain. Overall, respondents placed a large emphasis on the presence of evidence for the technology’s effectiveness and humans being in control of the health care process, even if this meant that care processes were not as fast as they could be with a higher degree of automation. Those with a negative attitude toward AI placed more emphasis on human autonomy than other respondent groups. Conclusions: Those developing and implementing health care AI technologies should develop an unbiased evidence base for the effectiveness of their technologies, using transparent methodologies, and continue their evaluation when the technology is in place. Moreover, implementation should not decrease patient-clinician contact but automate specific tasks only and maintain a human in the loop.