Can technology and AI ‘save the NHS’? A look at the main party manifestos

With the NHS under major pressure from a pincer-like combination of escalating demand and persistent workforce shortages, there are widespread hopes that technology and artificial intelligence (AI) can come to the rescue. These hopes are woven through the main party manifestos on the NHS, all of which set out high aspirations to reap the benefits of digital, data and AI. And they’re not wrong to hope: effectively implemented and used, technology and AI offer significant potential for improving care quality and efficiency.

What’s missing, though, is any kind of route map or plan for how to get there, or for how the NHS will realise these benefits in practice. For whoever is appointed Secretary of State for Health and Social Care next week, drawing up such a plan must be a priority. Below, we explore what it will need to include.

What uses of health care technology stand out in the main party manifestos?

Given that the NHS is one of the public’s top concerns, it is no surprise that all parties devote significant sections of their manifestos to it. And in addition to headline promises on increasing funding, recruiting staff and cutting waits, the main party manifestos all contain important commitments on tech-enabled care.

Two themes stand out. The first is using technology to improve or even reshape how care is delivered and experienced. For example, both Labour and the Conservatives emphasise greater use of the NHS App to allow patients to view their health records and manage appointments and prescriptions. This improves the ‘front door’ (making entry into NHS services easier), as well as potentially enabling people to be more engaged in managing their health. The manifestos also allude to the necessary shift towards more community-based care, in which technology has an important role to play too – for example, the Liberal Democrats’ commitment to expanding virtual wards and investing in technologies that allow people to be treated closer to home.

A second theme is using technology to free up staff time and improve productivity. The Conservatives promise to ‘use AI to free up doctors’ and nurses’ time’, building on the recent Budget announcement of £3.4bn for productivity-improving technology in the NHS, while the Liberal Democrats commit to ‘replacing old, slow computers to free up clinicians’ time’ – to name just two examples. Commitments like these are a welcome recognition of the potential for technology to support NHS staff, and highlight that greater investment is needed to improve the digital maturity of the NHS. Yet capitalising on this potential will require politicians and policymakers to focus much more on something they have traditionally neglected: technologies that can help with administrative and operational tasks. Because while cutting-edge clinical uses of technology often dominate the headlines (think robotic surgery or AI imaging), non-clinical uses tend to have the potential for much wider application across the NHS workforce (to help with note-taking, communication and scheduling, for instance). Such operational uses of technology also tend to carry less clinical risk, and can therefore usually be implemented more quickly than clinical technologies. NHS England has started to explore more non-clinical uses of technology – for example through the Collaboration Charter with the Incubator for Artificial Intelligence – but greater political backing and investment will be necessary to maximise the impact of these opportunities. 

What’s missing from the manifestos on health care technology?

Ensuring that technology actually frees up time is easier said than done, however, and relies on far more than the technology itself. It needs to be effectively implemented and used – and this requires not just the right underpinning infrastructure, skills and culture, but also (often) the redesign of roles, processes and ways of working. And even when all that is done, productivity benefits will only emerge if freed-up time is used as intended. Our recent research shows this shouldn’t be taken for granted, with clinical staff having multiple calls on any freed-up time (including admin, management, training, research, quality improvement and reducing overtime). To get it right, skilled change management is essential to agree how freed up time will be used and proactively repurpose such time.

And here’s the potential gap when we look at the different parties’ plans on health technology – whether in manifestos or other statements. There is a lot about the ‘kit’ itself – replacing outdated computers, replacing ageing radiotherapy machines, doubling the number of scanners, etc – which of course is all essential. But the NHS will need more than just new kit: it will need the right training, infrastructure and support.

To illustrate the point, in a recent Health Foundation survey, clinical staff told us the biggest barriers they face in using technologies effectively include insufficient access to IT expertise, poor connectivity and lack of implementation support. And there is still further investment needed to get the basic digital infrastructure right – the focus of a Health Foundation funded research project currently underway. So the next government will have to tackle these pressing issues and fund ‘the change’, not just ‘the tech’. 

This all speaks to a broader challenge: to deliver on plans for tech-enabled service transformation, the new government will need to be confident there is an effective approach in place for overseeing and managing change – getting the ‘how’ right, as well as the ‘what’. As our colleague Penny Pereira recently pointed out, how quickly and successfully any new government can implement its ambitions will be determined by the capability of local organisations to introduce service changes.

In her recent speech to NHS ConfedExpo, NHS England Chief Executive Amanda Pritchard highlighted some key components of the required approach. One is strengthening NHS management, given the critical role managers play in facilitating and overseeing the adoption of innovation. Another is building improvement capability and developing communities of practice and improvement collaboratives – learning from the work of the NHS Modernisation Agency in the 2000s, and more recently the Q Community.

There are other important components too: capitalising on the NHS’s position as a national system to build consensus, coordinate reform and achieve change at scale; not distracting staff, providers and systems with unnecessary structural or administrative change; and incorporating implementation support and rapid evaluation into service transformation programmes to help teams adapt and embed new approaches to care.

For years, consecutive governments have poured billions into health and life sciences research. But for the UK to realise the ambition of becoming a global health tech ‘superpower’, the next government will need to pay much more attention – in terms of funding, capability and incentives – to the application and implementation of tech in health care. 

Finally, even when technological innovations and new approaches to care have been adopted, there will be a need for ongoing work to optimise their performance. Indeed, it’s likely that many gains from technology over the next few years won’t come from new technologies at all, but from the optimisation of existing technologies.

An obvious example is electronic health records (EHRs). In a recent Health Foundation survey, clinical staff named EHRs as one of the technologies most likely to deliver time savings over the next 5 years. But we also heard frustration that EHRs are not yet being used effectively or to their full potential. With the vast majority of NHS trusts now having EHRs, attention will soon need to shift beyond the current focus on take-up, towards realising the longer term potential of this technology – which, as the US experience shows, can take several years. This could include, for example, building further functionality into EHRs, such as adding wider datasets or integrating machine learning to enable risk prediction or decision support.

The 2016 Wachter review generated a step-change in the adoption of EHRs in the NHS. The next government should consider if a further review is now needed, to ensure the health service is in a position to realise the benefits of EHRs, and be ready for future opportunities.

How does artificial intelligence feature in manifestos?

What about artificial intelligence? Such has been the excitement about the power of the latest generative AI models, such as ChatGPT-4, that it’s hard to find a discussion of health technology that doesn’t include the potential of AI. And, sure enough, AI features in the main party manifestos.

But what’s conspicuously absent is any kind of route map towards AI being adopted and used at scale in the health service.

That’s not to say there aren’t important pockets of work on AI already underway within the NHS across England, and valuable national programmes like the MHRA’s Software and AI as a Medical Device Change Programme or the NHS AI Lab’s AI and Digital Regulations Service. But these isolated initiatives currently do not exceed the sum of their parts because of the lack of an overarching strategy for AI in health care.

In the Health Foundation’s recent engagement with expert stakeholders from the health service, academia and industry, we found broad agreement that progress in developing and deploying AI within the NHS is currently being hampered by the absence of an overarching strategy and agreed focus, as well as a lack of coordination and collaboration across national agencies and organisations. 

What’s more, the huge pressures facing the NHS make finding a way forward on AI especially urgent. If the health service is to withstand these pressures, it must have the ability to harness the transformative potential of AI to improve care quality and efficiency, and support patients and staff.

That’s why the Health Foundation is calling on the next government to prioritise the development of a strategy for AI in health care. 

Towards a comprehensive strategy for AI in health care

There are multifaceted challenges that such a strategy must address. These include building public and staff consensus around AI in health care, making the NHS’s data and digital infrastructure fit for purpose, supporting the demonstration, testing and spread of AI tools, creating clear and consistent regulation, and ensuring the right workforce skills and capabilities.

A strategy will need to combine supporting local innovation with setting out a small number of high-level priorities for AI use, to focus scarce attention and resources in a way that can have the biggest impact. And it should be developed under the guiding principle of responsibility – ensuring the use of AI is not only legal and ethical but also works for the greater social good.

Hearteningly, the manifestos do highlight individual steps in some of these areas – such as Labour’s proposal to create a new regulatory innovation office and introduce binding regulation on companies developing the most powerful AI models. The key to a successful strategy will be bringing together action on all these fronts in a coordinated way, providing clarity for the public, health care staff, the NHS and industry.

Technology and AI have significant potential to help make the NHS more sustainable, support staff and improve patient experience. But unless the next government pays far more attention to their effective implementation and optimisation, rather than headline-grabbing announcements about new kit, the chances of technology and AI ‘saving the NHS’ will unfortunately be slim.