Health Foundation explores opportunities and challenges of automation

The Health Foundation, working with the University of Oxford, has published research into the opportunities and challenges for automation and artificial intelligence in healthcare.

The research study explored the potential of automation in primary care, the increasing number of areas for automation, how AI and robotic technologies are being applied to clinical and administrative tasks, the challenges for making automation work on the front line, what automation might mean for the future of work in healthcare, and how the NHS can get this agenda right for the long term.

One of the aims of the report is to help “increase understanding among policymakers, organisation and system leaders and front- line staff” and bridge the gap between policy and practice. It also aims to explore “the human side of healthcare as well as the technical, to understand what its social and relational dimensions mean for how we should go about improving services.”

The research draws on an online YouGov survey of over 4,000 UK adults and over 1,000 NHS staff, and provides analysis into the challenges and potential presented by automation and AI in health.

In the survey, when participants were asked about the greatest challenges, 45% of NHS staff surveyed said they felt that patients might not accept these technologies or be suspicious of them. 39% felt that staff shortages or inadequate equipment might make it difficult to use these technologies properly.

Respondents viewed analysis of images and test results as the biggest opportunity presented by automation and AI in health care (40%), followed by risk prediction and screening (28%), use of robots in surgery (28%), and demand and capacity management, scheduling and rostering (28%.) Robotic carers and assistants received the lowest response (2%.)

The report then moves on to explore the main perceived benefit of automation, which was said “to make things more efficient and free up doctors and nurses to care for patients.”

The main perceived risk of automation and AI in healthcare for the public and NHS staff was it becoming more ‘impersonal’ with less human contact (51% for NHS staff surveyed, 45% for the public), followed by health care professionals not questioning the decisions computers make, creating risks to patient safety (42% for NHS staff, 44% for the public.) Nearly a third felt that it would be hard to know who’s accountable when things go wrong (29% for NHS staff, 32% for the public.)

Among the public, those who said they had heard, seen or read ‘a lot’ or ‘a fair amount’ about automation and AI in healthcare felt much more positive than negative about the use of these technologies, by 70% to 26%, while those who answered ‘not very much’ or ‘nothing at all’ felt more negative than positive, by 41% to 35%.

Will Warburton, Director of Improvement at the Health Foundation and co-author of the report, commented: ‘The NHS is facing unprecedented challenges, with the pandemic having placed an already strained system under severe financial and workforce pressures. We need to radically change the way that health care is delivered and the NHS must boost its ability to identify promising technologies to help achieve this. However, it is critical that automation technologies are co-designed with patients and staff and used in ways that support empathy, dignity and compassion and do not undermine them. Embedding new technology successfully in health care settings can also require significant training, testing and staff time so it is crucial that Government funds “the change” not just “the tech”.

‘There is also a very real risk that automation and AI could widen health inequalities. We know that digital inclusion poses particular challenges given inequalities in access to technology and digital literacy. The upcoming AI strategy must ensure that new technologies introduced to our healthcare system are designed with the most vulnerable in mind and ideally can help “level up” on health. Automation and AI have great potential in healthcare but they must be accessible for all.”

To read the paper, please click here.