Public dissatisfaction with the NHS is at a historic high.
A Nuffield Trust poll earlier this year found that just one in five people were satisfied with NHS services – the lowest since records began in 1983.
It isn’t difficult to see why: one in 11 people are on NHS waiting lists with around 1.2 million people using NHS services every single day.
And with 45% of 18-25 year olds using private general practitioners (GPs) – the age group with the least disposable income in society – this tells us everything we need to know – the NHS is not meeting people where they are.
The NHS was founded in 1948 amid the aftermath of war. More than 75 years on, it’s time to reimagine it for the world we currently live in.
Wes Streeting’s newly announced 10-year plan to modernise and futureproof the NHS, with artificial intelligence (AI) at the heart of it – signals a long overdue shift towards future-facing healthcare.
AI has a critical role to play in mitigating these problems. From smarter scheduling systems to predictive analytics that reduce did not attends (DNAs) and short notice cancellations. This technology can help unlock latent capacity and help frontline staff deliver more equitable, efficient, and human-centred care.
We explore how deploying AI into existing NHS systems can maximise resources, improve patient outcomes, and help address the growing disparity between those who can access care and those who can’t.
The hidden cost of missed appointments
Missed appointments are one of the NHS’s biggest but most avoidable costs. Around 8 million people miss their appointments each year, with every missed appointment costing the NHS an average of £165. Short-notice cancellations add further pressure, affecting around 4 million appointments.
While most people can book flights or meetings online with ease, NHS appointment systems remain outdated and inflexible.
AI can identify patients at risk of missing their appointments or those likely to cancel under short notice and, crucially, offers tailored solutions to increase attendance. By using the platform, an NHS trust in Essex saw a 23% reduction in no-shows.
Why is this important?
Missing appointments has serious consequences. If you have a long term health condition and a mental health condition and you miss just two GP appointments in a year, your chances of dying in that year go up 8 times.
If that was a genetic marker, we’d be spending billions on developing a drug to address that.
Predicting attendance and personalising approach
The platform is built using two complementary AI agents.
The first is a scheduling agent, refined over seven years, using anonymised non-personal data to understand and forecast attendance based on a wide range of operational and behavioural signals. This agent helps hospitals book the right number of patients.
Once that prediction is made, the second AI agent determines the most effective next step. It decides whether to follow up with a call, text or offer transport- and more importantly – how to personalise the wording and time of that outreach.
If a patient lives in an area with limited transport options, the agent may proactively suggest a carshare. Helen, for example, is disabled and finds it difficult to get to her appointments, especially living in a rural area with limited access to transport. Using this platform, Helen is able to get an Uber booked to make sure she gets there and back safely.
The platform is constantly evolving and learning from previous interactions with patients; what worked and what didn’t. It refines its approach over time to improve effectiveness. Healthcare providers benefit from increased clinical capacity, reduced administrative burden, and better patient outcomes.
When outreach alone is not enough, the scheduling agent enables dynamic scheduling. If someone cancels or is likely to no-show, the platform can identify and offer the open slot to another patient in need.
This kind of dynamic responsiveness is already showing promise. Clinics using personalised, AI-guided outreach are seeing fewer missed appointments. The change came not from increasing the number of reminders, but from making each one hyperpersonalised and effective, enabling the 6 million people waiting for care to fit into the 12 million empty slots.
Addressing barriers to care
Technology alone can’t solve every issue. But it can be built to work around real-world constraints that people face every single day.
We spend time speaking directly with patients to understand the barriers they face. For some, it’s financial – the cost of transport or time off work. For others, it’s physical – a lack of mobility or support. These challenges have a cumulative effect on health outcomes.
One major opportunity lies in how non-emergency transport is managed. The NHS currently spends around £460 million a year on 11 million journeys. By partnering with providers like Uber, we can repurpose that funding to ensure that a third of people eligible are able to get to their appointments safely, versus taking an expensive ambulance.
Taking action
In healthcare, AI is often praised for its ability to flag risk. But prediction without action is not enough.
AI must operate as a dynamic, continuously learning tool – one that listens to patients, responds to real-world constraints, and personalises outreach in real time to improve access. Real progress comes from closing the loop between insight and intervention – turning predictions into actions that make a measurable difference in patient care.
The government’s renewed focus on digital transformation is a vital step forward, AI adoption will enable clinicians to deliver more human-centred care – something that has become increasingly difficult in an under-resourced and overstretched system. As such, we have a greater opportunity to address the key issues that sit firmly within the DNA of the NHS: waiting lists, healthcare inequalities, financial stability and making the transition of moving to an automated outpatient booking process.
