1.4 Context

Artificial intelligence (AI) technologies are the latest innovation to transform healthcare.

Artificial intelligence (AI) technologies are the latest innovation in a series of digital technologies that have been transforming the delivery of healthcare. The broader efforts to enable the adoption of change and innovation in health and care settings (such as the digital transformation pathways outlined in the NHS Long Term Plan)⁶ are important foundations for the adoption of AI technologies.

The UK has published an ambitious strategy to remain a science and AI superpower in the next 10 years.⁷ The NHS is supporting this objective by developing a strategy for AI in health and care.⁸ Building on the work of the NHS AI Lab, the strategy’s vision is to enable the safe scaling of proven and fair AI technologies that deliver better outcomes for the UK population.

Existing levels of familiarity and experience in AI technologies

International surveys have shown that most healthcare workers lack direct experience with AI technologies.⁹ In the UK, a 2020 survey of over 1,000 NHS staff by the Health Foundation² found that three-quarters of respondents have heard, seen or read ‘not very much’ or ‘nothing at all’ about automation and AI.

The survey’s respondents were split between being positive or negative overall about the use of AI in healthcare, with some notable insights.

• Those more familiar with AI technologies tended to be more positive towards these technologies.
• Medical and dental staff respondents felt more positive about AI than nurses and midwives who in turn felt more positive than health care assistants.
• Assistive applications of AI, like image analysis and screening, were perceived as a bigger opportunity than autonomous forms of AI like robotic care assistants.

Respondents perceived the main benefits of automation and AI would be to improve efficiency and free up time to care for patients, a key objective identified in the Topol Review. The biggest identified risks of using automation and AI involved healthcare becoming more impersonal and healthcare workers failing to question any suggestions or decisions proposed by these systems. The biggest identified challenges were patients potentially not accepting these technologies and being suspicious of them, and staff shortages or inadequate equipment for implementation and use.

While most of the Health Foundation survey’s respondents anticipated that AI would improve their quality of work, a notable subset viewed AI as a threat to their jobs and professional status. This fear of AI replacing human jobs has been noted in several other research studies (particularly in relation to administrative positions and in radiology and pathology) along with other concerns like data governance, cyber security, patient safety, and fairness. ^{5, 10, 11}

The healthcare workers interviewed for this research shared similar views about the levels of experience and attitudes towards AI technologies within their settings, noting that most related activities involve a ‘small pool of experts’ employed by the NHS who are often self-taught. They noted that exposure to AI technologies varies across professions and roles, with radiology, dermatology, ophthalmology, and pathology leading in this space (driven by the need for, and availability of, AI-enabled machine vision technologies). 

However, interest in and understanding of how AI technologies can assist ongoing clinical and administrative challenges and improve patient outcomes is growing. Interviewees discussed current work in their settings on how to coordinate or develop data sets to train algorithms, investigations around static and dynamic AI models, and concerns about bias in AI systems. They also highlighted perceived challenges in ensuring the ongoing robust performance of AI systems and in understanding their impact on clinical safety and real-world patient outcomes.  

The growing deployment of AI in health and care settings is highlighted also in HEE’s AI Roadmap report,³ which surveyed over 200 AI technologies nearing or ready for market. The survey found that over 20 per cent of these technologies are estimated to be ready for large scale deployment within 2022, and an additional 40 per cent in the next 3 years. The most affected workforce groups identified in the survey included radiologists, general practitioners, workers in non-clinical administration, diagnostic radiographers, and cardiologists.

References

⁶ NHS. NHS Long Term Plan: Digital transformation. NHS England. https://www.longtermplan.nhs.uk/areas-of-work/digital-transformation/. Published 2019. Accessed February 28, 2022.

⁷ National AI Strategy - GOV.UK. https://www.gov.uk/government/publications/national-ai-strategy. Accessed February 28, 2022.

⁸ The National Strategy for AI in Health and Social Care - NHS AI Lab programmes - NHS Transformation Directorate. https://www.nhsx.nhs.uk/ai-lab/ai-lab-programmes/the-national-strategy-for-ai-in-health-and-social-care/. Accessed February 28, 2022.

⁹ Sinha S, Al Huraimel K. Transforming Healthcare with AI. In: Reimagining Businesses with AI; 2020:33-54. doi:10.1002/9781119709183.ch3

² Hardie T, Horton T, Willis M, Warburton W. Switched on. How Do We Get the Best out of Automation and AI in Health Care? 2021. doi:10.37829/HF-2021-I03

⁵ Topol E. The Topol Review: Preparing the Healthcare Workforce to Deliver the Digital Future. 2019. https://topol.hee.nhs.uk/the-topol-review/. Accessed February 28, 2022.

¹⁰ Lu X, Keane PA, Denniston AK. Time to regenerate: the doctor in the age of artificial intelligence. J R Soc Med. 2018;111(4):113-116. doi:10.1177/0141076818762648

¹¹ How to build trust with Trusts on artificial intelligence - Med-Tech Innovation. https://www.med-technews.com/medtech-insights/ai-in-healthcare-insights/how-to-build-trust-with-trusts-on-artificial-intelligence_1/. Accessed February 28, 2022.

3 AI Roadmap report and interactive dashboard - Health Education England. https://www.hee.nhs.uk/our-work/dart-ed/ai-roadmap. Accessed February 28, 2022.