AI is being deployed in the NHS to improve healthcare delivery across a broad range of activities
Amongst the British population, there is a large contingent of people who believe that the drive towards health tech is an underhand scheme to privatise the NHS. The suspicion has been exacerbated by the coinciding of rapid progress in health tech over the past 10 years with a period of Conservative government, and the present Health Secretary’s strong propounding of it.
On the third and final day at CogX, the Health Stage was dedicated to discussing healthcare and the NHS in a future enhanced by technology. One session, Innovation and the NHS, brought together four examples of how artificial intelligence is improving the delivery of healthcarein hugely disparate ways, from diagnosis to supply chain management.
Nicholas Watkins, associate director for research and development at the NHS Blood & Transfusion service, and Andy Gray, CEO of Kortical, demonstrated how a project funded by a grant from Innovate UK was already delivering a 15% reduction in waste in the high value platelet distribution service. Platelets have a shelf-life of just 7 days, but demand from each hospital fluctuates dramatically from day to day. The use of AI can reduce both waste, and ad hoc logistical costs.
At Alder Hey hospital for children, inpatient experience is a critical component of care. Emma Hughes, head of innovation, and her team have been developing a virtual assistant to handle pre- and post-op basic questions. The project that began in 2016 is a collaboration with Hartree using IBM Watson.
Pearse Keane brought in DeepMind to Moorfields Eye Hospital to assist with triage following OCT scans. Ophthalmology is the leading outpatient specialty in the NHS. Together they developed a system for faster, more accurate diagnosis of 50 retinal diseases from OCT scans, which proved to be at least as good as leading diagnostic experts. The next step is to develop robust clinical validation.
Finally, Jacqueline Moxon, EMRAD project manager at Nottingham Healthcare NHS Trust, is in the early stages of a collaboration with three partners to use AI to cope with a severe understaffing problem. She forecasts that in three years’ time, AI will provide a second opinion in breast screening, and that it will be the key means of screening within six.
It is important that the opponents of this technological progress are persuaded of the good that is coming from it. A government might put the breaks on this progress if it determines that it will help it to win votes. Such a government might redirect spending largely, or entirely, to pay and recruitment. These are areas of need, and are highlighted in the NHS Long Term Plan, but withdrawing from investment in innovation will undermine progress in these personnel challenges, and reduce the ability of the health service to serve the British people.
At the same time, new models of clinical commissioning are beginning to emerge. There is a shift towards collaboration, and away from procurement. More progress is needed, and part of the outcome will likely be in the sharing of financial gains, so that partnerships that are funded by the UK taxpayer, enabled by UK medical expertise, using data from British patients, can return funds to the NHS for further investment.