Automation and digital technology in the NHS of the future

Tell us about your new role at NHSX. What are you most excited about?

I’m currently the digital health and AI lead. We’re still figuring out titles and roles within NHSX, but over the last two years I’ve been at NHS England where I’ve focused on standards and evidence in the digital health and AI space. Quite a bit of the work we’ve done is around creating an ecosystem for the safe, ethical and effective development and deployment of data-driven technologies.

We are seeing some really great technologies being developed and we are working closely with innovators, academia, tech companies and many others to understand the needs of the workforce and what is being developed that can help address some of the issues on the front line.

We’ve come quite far in the last few years and I’m really excited about how we now take the leap to help scale and create proper funding structures for these technologies, which is a focus that it’s been great to hear Simon Stevens talking about recently.

What kind of benefits do you think automation holds for patients and staff in the NHS?  

Automation is often most useful for back-end tasks, taking out the bits of someone’s job that are slightly boring and repetitive. In imaging, for example, a lot of time is spent measuring things – the size of a nodule or the distance between x and y. If you can automate those bits that frees up lots of time for the radiologist to spend on the more nuanced and complicated parts of their job. In King’s College Hospital this technology is already being used for maternity ultrasounds, with the computer helping to identify parts of the foetus and take measurements. 

Another example, which is starting to be used in the US, is using voice recognition technology to automatically input notes during a consultation. You can see how this might save time and reduce admin for the clinician, but it also might help the interaction feel different for the patient as the clinician can be more engaged with them rather than busy typing up notes. Eric Topol talks about it quite a bit in his review, along with lots of other examples. 

How do we make sure automation technologies are useful for staff and safe and effective for patients? What role does NHSX have in this?

We need to help create standards and frameworks and give people the support they need in implementation and to enable the market to flourish – that’s already a big part of my job. As set out in the Industrial Strategy Challenge Fund, we’re creating a dedicated team of commercial and legal experts to help support this, in collaboration with partners such as the Office of Life Sciences. 

We need to align some of the standards already out there, for example around how computer systems talk to each other and exchange information. Building APIs to enable this will also be part of our role. API stands for 'application programming interface', but basically sets out the rules for connecting with a particular website or system. 

We’ll also be looking at data standards, working with organisations like the Professional Records Standards Body to define best practice for things like discharge summaries or clinic letters, so that we can start to collect some of that information in a more coherent way. 

Of course, we need to involve the workforce in all of this. We’re currently doing a piece of work with the royal colleges to understand the needs of the workforce, and we need to do something similar with the public, talking to patients about what they feel is most important, and engaging with the charities developing their own work about how it can also align with the needs of the health system. Some horizon scanning work is already underway through the Accelerated Access Collaborative around this.

What do you think will be the biggest challenge for making automation technologies work on the front line across the NHS?

Encouraging the workforce to trust and support the technologies introduced will be key. But ultimately if something works and makes someone’s job easier people are usually happy to adopt it. 

Some people might worry that their jobs will be replaced by clever new technology. But the reality is that we already don’t have enough people, so we’re unlikely to replace whole roles with machines. We might be able to remove that repetitive and mundane part of a job and free up time to do something different though, so some people’s jobs might be altered. For example, if you work as an appointment booker, and we start automating appointment booking, maybe we could optimise your role so that you spend more time helping patients to navigate their care, or you handle things when the computer says no – things that take that human element. 

What we always say in the NHS is we’re giving you a ‘digital first’ option, but it’s not a ‘digital only’ option. While we’re trying to optimise the system as much as we can, we don’t want to eliminate the human factor. 

Of course, the other big challenge is that we have all these great ideas for automation but not the tech and hardware in place to roll it out. We can talk about all this whizzy stuff, but if we can’t get the fundamentals right we might as well not bother. So there’s a huge amount of work going on to build the technical architecture needed in the NHS. 

What are your hopes for technologies like automation and other data-driven tech over the next five years?

My colleague Jess Morley commented the other day that there’s so much hype right now, but in five years’ time we’ll probably only just be realising some of the things people are talking about today. Things like automation in imaging take a lot of time to embed, and five years really isn’t that long. Maybe there won’t be that much that’s different but things might be happening at scale and we’ll certainly have better frameworks in place to allow the new stuff to plug in. That doesn’t sound all that exciting but it’s the bit that needs to happen in order for things to really take off long-term. 

Who knows beyond that? Would we have predicted 10 years ago that we’d be splicing genes now? What I would say is that I hope in 5-10 years’ time we’ve learned from past mistakes and managed to bring more people along with us. We need to challenge the model more, both through the technologies we’re developing, but also in who we enable to develop things. We can’t just rely on outside suppliers to come in, and we need to empower our workforce to be more confident with technology. Because that sort of approach permeates through the system and means we won’t leave anyone behind. 

Coming together as NHSX is already allowing us to work better across silos, encouraging sharing, and showcasing the good stuff no matter whose brand is behind it. Matthew Gould, who will be NHSX CEO from 1 July, has already made it clear: ‘let’s not recreate something if it’s already out there, let’s try and enhance it’. I couldn’t agree more. 

This content originally featured in our email newsletter, which explores perspectives and expert opinion on a different health or health care topic each month.