NHSX has published guidelines for the health service on the use of AI, emphasising the need for the workforce to develop the appropriate skills.
Titled Artificial intelligence: how to get it right, the document from the Government’s policy unit for digital and data in healthcare includes a section on the challenges facing the NHS workforce in using the technology.
Other sections take in an overview of AI, its governance, the data that it uses and ‘what good looks like’ and how to monitor the impact of the technology.
It says a primary focus will be on expertise in critical thinking at senior levels to evaluate the potential for AI to provide benefits in any given situation. This will include a technical understanding of requirements for the evidence base and the computing resource needed.
In addition, NHS staff will need support in understanding the ethical considerations, and there will be a need for analytical skills to interpret complex information. The latter requires an appreciation of how AI works, the ability to design reproducible studies using open source analytics, and an understanding of the protocols and metadata needed.
The document points to work being done by the Alan Turing Institute, Health Data Research UK and the Health Foundation in laying the ground for these skills, and says the effort reflects elements of the NHS Interim People Plan.
Its section on what ‘good AI’ looks like highlights a handful of uses, including precision medicine, processing genome data, image recognition and providing operational efficiencies.
Along with this comes the need to monitor the implementation of AI, likely to be a challenge as it will be shaped heavily by the context. It could require a stronger analytical capability in the NHS to get the best results, says the document.
The ideas for a governance framework rely heavily on the Code of Conduct for Data Driven Health and Care, first published in September of last year, which NHSX says has now been recognised internationally as a source of guidance.
The section on clarifying data access and protection emphasises the importance of navigating the regulations, understanding patient data, and developing appropriate data sharing agreements and commercial models.
It also points NHSX’s publication in July of a data framework with five principles to maximise the benefits for patients and public in sharing data.
IT industry association techUK has applauded the guidance. Its associate director for health and social care, Ben Moody, commented: “techUK is delighted to see the NHS recognising the contribution that AI can make to better patient outcomes.
“As the report shows, AI is not a far-fetched technology of the future but an embedded part of NHS technology that is providing real benefit today. We welcome the focus on ensuring that the public and staff of the NHS have confidence that AI can be used safely and effectively so that we can fully realise the benefits that AI will bring in the coming years.
“The NHSX AI Lab is a significant investment and provides a great opportunity to bring together the world’s best health tech companies, clinicians and researchers to help the NHS to prove and spread solutions at pace.”
Image from GOV.UK, Open Government Licence v3.0