Inside a healthcare revolution - how the NHS wants to use technology to keep you out of its hospitals

A nurse at Addenbrookes record a patient's vital signs on a portable electronic device - David Rose
A nurse at Addenbrookes record a patient's vital signs on a portable electronic device - David Rose

NHS at 70: Article banner

On Christmas Day last year, Brian Thomas* walked into A&E at Addenbrooke’s Hospital in Cambridge. He was a bit groggy but his wife, Jackie, wasn’t really concerned. She knew he’d been out late at the pub the night before, and thought he was just hungover.

Waiting for them was a hospital operating on what is traditionally the most poorly-staffed day of the year. With rotas filled by inexperienced nurses and junior doctors, it’s known even within the NHS as the worst time to get sick.

The triage nurse began to take Brian’s vital signs. According to Sian Coggle, Consultant in Acute Medicine and Infectious Diseases at Addenbrooke’s, in many hospitals he might well have been parked on a seat.

“The nurse would probably have thought, he’s been at the pub, he’s had a few too many, he’s a bit dehydrated.”

Instead, something else happened. As information about Brian’s case was plugged, detail by detail, into the electronic health record (EHR) system at the hospital, a yellow box popped up on the screen. The software had assembled the information and come to its own assessment.

“The box effectively said, ‘This person is really sick, have you considered sepsis?” says Coggle.

The triage nurse, prompted by the computer alert, summoned a doctor. Meanwhile, the system had prepared a list of all the tests needed, which could be commissioned with the click of a button. “No need to rush around filling in bits of paper, or looking for the paper drug chart,” says Coggle.

When the doctor arrived, the software even provided up to date guidelines about which antibiotics to prescribe, depending on the area of the body the infection seemed to be coming from.

It turned out that Brian Thomas was not hungover. He had Bacteraemic meningitis - sepsis, in other words. “Because our system flagged it up the team knew they had to get him up to ‘resus’, and get antibiotics into him in timely fashion,” says Coggle. “Whereas before…”

She trails off. The evidence is all too clear about what might have happened before. Research demonstrates that for every hour that antibiotics are delayed in sepsis cases, mortality increases by 8 per cent. It is a time-critical condition, every bit as much as cardiac arrest or stroke.

Sepsis Treatment Target Addenbrooke's

Without the Trust’s EHR, there is every chance that Brian Thomas would have been left for hours on a plastic chair in A&E, possibly to die on Christmas Day. Instead he was home two days later, and suffered no long term consequences.

Just two years earlier, that outcome would almost certainly have been very different. In 2015, Dr Coggle began looking at how quickly sepsis was being addressed across Cambridge University Hospital’s NHS Foundation Trust, which comprises Addenbrooke’s Hospital and The Rosie women’s and maternity hospital.  

Initially it was hard, because there was no data. When she began to collate that data, however, she was horrified to find that only 11 per cent of sepsis cases were meeting the recommended 1-hour target for the administration of lifesaving antibiotics after diagnosis. “It was pretty dire,” she says. “Sometimes it would be up to four hours. We realised we had a problem. Then we realised we could use Epic to prompt people.”

Epic is the electronic health record system that may well have saved Brian Thomas’ life. It went live across Addenbrooke’s and The Rosie hospitals at 2am on October 26 2014, taking the Trust all-but paperless literally overnight. Such was the scale of the cultural revolution that the moment remains etched on the memories of most of the 10,000 regular staff at the Trust’s  hospitals.

More than 5,000 old computers were chucked out. 7,250 new ones were installed. “Rover” devices (actually customised iPods) which could scan barcodes on every patient wristband became ubiquitous.

And the system to which all these bits of hardware connected allowed real-time access to real-time information - from scans to heart beat monitors - anywhere. “If I need to prescribe a drug, I can securely log on and do so, even from the Maldives if necessary,” says Dr Gareth Corbett, consultant gastroenterologist.  

Dr Afzal Chaudhry, Chief Clinical Information Officer at Addenbrooke’s, is also part of the renal transplant service there.

Dr Jag Ahluwalia (left) and Afzal Chaudhry, Chief Clinical Information Officer - Credit: David Rose/David Rose
Dr Jag Ahluwalia (left) and Afzal Chaudhry, Chief Clinical Information Officer Credit: David Rose/David Rose

“It means when I get called up at 4am because an organ has become available, I can log on without delay and have the patient’s entire record in front of me.”

The system can also be programmed so that, when certain parameters are met, it issues alerts, like the sepsis flag to Brian Thomas’s triage nurse, which was set up by Coggle and the Trust’s eHospital digital team. Others include the Paediatric Early Warning Score, PEWS. “It allows us to escalate serious cases much quicker,” says Hannah Nunn, senior sister on the paediatric ward. “We get them up the chain quicker. That makes us much safer. It can save lives.”

From that “dire” figure of 11 per cent, almost 80 per cent of patients diagnosed with sepsis at Addenbrooke’s are now being given antibiotics within 1 hour; 95 per cent get the right drugs within 90 minutes.

“This is not an IT programme that sits in the basement,” says Dr Jag Ahluwalia, Director of Digital at the Trust. “This is just how a 21st century hospital has to function. What is the data telling us? How can we help our clinicians to make use of that to treat people effectively, to save money and, ultimately, to save lives.”

The implications of the technological revolution at Addenbrooke’s and The Rosie have been dramatic not just for patients, but for staff too. Many took up the new system with gusto. Some resisted.  But all were affected, and sometimes for prosaic, rather than technologically sophisticated reasons.

“The biggest change, bar none, was the end of handwriting,” says Helen Balsdon, who has become the Trust’s first Chief Nursing Information Officer (CNIO). “My nurse colleagues could actually read doctors’ notes because they were in type on a screen.”

CNIO Helen Balsdon
CNIO Helen Balsdon

Balsdon moves with the brisk, relentless energy and purposefulness of senior nurses in the movies. As she strides the length of the hospital (to walk end-to-end in Addenbrooke’s is to cover a mile and a half of corridors) she enumerates the other benefits digitisation has brought.

“Second most important was charts. We don’t have to go looking for paper drug charts now. Without that, how can you give the patient the right drugs or the right care? Third is the RRT.”

The Rapid Response Team (RRT) operate on much the same way as the automated sepsis alerts. If the vital signs of a patient anywhere on site become a concern the system flags the case to a dedicated unit of 13 nurses, led by a consultant and registrar. They visit the patient’s bedside within 30 minutes, down from 90 minutes pre-digitisation and, if necessary, can move the patient to the Intensive Care Unit (ICU).

“It means that the expertise of the ICU can be spread throughout the Trust’s hospitals,” says Dr Vilas Navapurkar, consultant in intensive care.

“We’re getting to the sickest of the sick sooner. Between 2014-2017 there’s been a 50 per cent decrease in peri-arrest - those about to suffer cardiac arrest - through the automated capture of physiological data. That’s very, very good.”

Today’s access to complete patient data at any place, any time, allows staff at every level of the hierarchy to make clinical decisions. “It empowers us nurses,” says Lisa Wood, from the RRT. “We’re not just chasing doctors to be told what to do. We can alert them, then begin treatment while we wait.”

Having every bit of information at their fingertips means relatively junior staff, working alone, in the middle of the night, can feel confident about making even the most significant decision of all - when not to resuscitate a dying patient. “The DNR [the Do Not Resuscitate order], and all the other information is there to see in the patient’s electronic health record,” says Navapurkar.

Consultants Sian Coggle and Vilas Navapurkar - Credit: David Rose/David Rose
Consultants Sian Coggle and Vilas Navapurkar Credit: David Rose/David Rose

The cost of this Big Bang digitalisation at the Trust – the Epic EHR software, new computing equipment, handheld and mobile devices, IT infrastructure and support - is thought to be around £200m over 10 years. But there are also savings.

Thanks to the Rovers scanning medication barcodes and checking them again patients wristband barcodes, there has been a dramatic reduction in errors administering drugs. This and allergy alerts in the system prevent adverse reactions, saving many hundreds of “bed days” each year, and millions of pounds in the process.

Millions more are saved in staff time by the fact that clinicians can concentrate on the vital signs and ventilator data being automatically fed in to the system from ICU beds and operating theatres - rather than having to record them all manually.

Such is the extent of the change that patients have at times got grown frustrated with doctors fiddling away at screens. “When they see us on the Rovers they sometimes think we are just on our phones,” says Hannah Nunn. “Apart from that they are so used to technology in their lives, they assume that every hospital is like this.”

That is far from the case. There are more than 200 hospitals in the UK of which only about 10 per cent have computer systems on a par with Cambridge University Hospitals.

The reality is that though medicine is synonymous with some of the world’s most celebrated technological breakthroughs, from the discovery of DNA to heart transplants, the NHS is synonymous with what is probably this country’s most dysfunctional implementation and roll out of technology, ever.

The disasters of the past

The National Programme for Information Technology (NPfIT) was a decade-long attempt to digitise the whole of NHS secondary care - that is, hospitals (GPs are considered “primary care”). Ordained by central government and imposed upon an often resistant NHS staff, it was an utter disaster. By the time it was shelved in 2011, it had cost more than £12 billion. Parliament’s Public Accounts Committee described it as among the “worst and most expensive contracting fiascos in public sector history”.

Even amid the ashes of the NPfIT however, there was a realisation within government that hospitals could not simply languish in the age of the paper and pencil as the rest of the world went digital. So a National Information Board (NIB) was created to plot the way forward.

Shortly afterwards, in 2013, 15 so-called Academic Health Science Networks (AHSNs) were formed around the country, in an attempt to speed the path of innovative procedures, devices and technologies into the NHS’ Byzantine organisational structure.  

NHS at 70 - Spending as a proportion of GDP

For while it may be called the National Health Service, there is precious little that is National about it. Various levels of the service - GPs, ambulances, hospitals, social care - are barely coordinated and have little shared access to information. Paramedics arriving at a traffic accident, for example, often have to treat the injured “blind” - without any access to medical notes, which can mean that the administration of potentially lifesaving drugs actually triggers an allergic reaction.

“I have seen visitors from abroad genuinely terrified by the lack of data,” says Ben Collins of the King’s Fund.

“They see old people admitted to hospital and doctors with no way of getting access to their medical history, with no way of contacting their GP because the record lists no name. So they can’t find anything or anybody to give them any information about the patient in front of them.”

Improving this pitiful contact between hospitals and the nation’s 7,500 or so GP practices could not be more important. Unlike hospitals, GPs have successfully run patient EHRs for many years now. Consequently, it is increasingly the GP practice that is the critical repository of the details of any patient’s treatment “cradle to grave”. And as they increase in importance, practices - which are effectively small private businesses - are also growing in influence, teaming up into chains responsible for thousands, if not tens of thousands, of patients.

But if GPs are tending to band together, hospitals have rarely been less coordinated. They are operated by separate trusts, of which there are more than 200, and are paid for the care they provide by Clinical Commissioning Groups (CCGs), also numbering more than 200.

In all, the NHS budget is £126bn per year - more than 7 per cent of GDP - and the service employs more than 100,000 doctors; 300,000 nurses and health visitors; 22,000 midwives; and 20,000 ambulance staff. Some 130,000 more work on research, and there are 40,000 managers. “It’s the most incomprehensible system imaginable,” says Collins. “We have a phenomenally complicated system. It’s hellishly difficult to work out just who is in charge.”

NHS at 70 - How staffing has increased

In October 2014, a year after the AHSNs were set up, NHS England published its plan for the next half decade, called the The Five Year Forward View which, in its very opening paragraphs, made clear that “the NHS needs to adapt to take advantage of the opportunities that science and technology offer”.

A month later the NIB reported, issuing the Government’s digital health strategy: Personalised Health and Care 2020: Using data and technology to transform outcomes for patients and citizens. Three years after the disaster of NPfIT, tech transformation of the NHS was back on the political agenda.

Patients cost the NHS more as they get older

With the effort to drag the NHS into the information age given new impetus, a national advisory group on “health information technology” was set up in late 2015, chaired by the eminent American doctor, Robert Wachter.  He soon made it clear that while the NPfIT’s aims of digitising the NHS were not going to be jettisoned, its methods were.

Instead of top down injunctions issued from Cabinet level, Wachter advocated finding local trusts, like CU at Addenbrookes, which were doing well, then rolling out their expertise and experience to others.  The so-called “Global Digital Exemplar” trusts were born - a mere 12 to begin with, now expanded to 16.

Wachter, like almost any doctor you talk to in the NHS, understood that what was at stake was more than a few shiny screens. He recognised that as our population ages and patients thus enter the healthcare system with more complicated problems, the existing NHS way of doing things was no longer sustainable. Not only would patients die unnecessarily, but costs would rocket.

“In order for the NHS to continue to provide a high level of healthcare at an affordable cost, it simply must modernise and transform,” he noted. “This will involve enormous changes in culture, structure, governance, workforce and training.”

The NHS’ burden is increasing - particularly from elderly people

Since then it has also become clear that the prize on offer for pushing through such change is equally enormous. A recent report by McKinsey estimated that digitisation of records alone could save up to 11.5 percent of health expenditure - that’s £14.5bn a year.

The Big Data revolution

Money is just the start. For digitisation of patient health records means data. Vast troves of it. The NHS, if it can digitise patient records and treatment data across primary and secondary care, will be sitting on what Harry Evans, also at the Kings Fund, calls “the single richest dataset on medical health anywhere in the world”.

Computer analysis of such data would almost certainly lead to pioneering and cost-saving treatments. The world’s most celebrated Artificial Intelligence company, Deepmind, which was founded and is based in Britain but is now owned by Google, is already using its algorithms to trawl retinal scans from Moorfields’ Eye Hospital and spot disease that even the most experienced human consultant might miss.

“There’s really a lot going on in data analytics,” says Dr Shaun O’Hanlon, chief medical officer for EMIS, a leading provider of EHR software for GP practices. “We’ve got 40m patient records on EMIS, you can analyse that database to create risk scores for diabetes, coronary disease...”

Moorfields Eye Hospital
Moorfields is home to Deepmind's new healthcare venture

Such information, he stresses, is “suitably anonymised” and patients are “given the option to opt out”. For like everyone involved in handling patient data, he is mindful of the dangers as well as the opportunities it offers. For example the last great effort to harness “Big Data” for the public good -  called - became another NHS tech fiasco, all despite the backing of nearly everyone in the medical profession.

The culprit this time was not the IT, but the PR. With little public consultation, the proposal that their records were to be shared and analysed was effectively sprung on an unsuspecting population.

The backlash was inevitable. In 2016, when almost a million people who had chosen to opt out discovered that their information was still being shared anyway, was scrapped, two years after its abortive launch.

Even so, few today doubt that, in one form or another, Big Data will be back to play a transformative role in the NHS. It is not just eye scans where algorithms are besting humans. From ECG analysis to drug delivery, myriad start-ups claim that their machines will deliver far greater accuracy in the near future than humans can now - saving billions in the process.

From Hospital to Home Care

Yet the biggest savings of all will come not from treating people in different ways, but not having to treat them at all. Increasingly, the NHS is focusing its time and technological effort on keeping people out of its waiting rooms, wards and operating theatres altogether.

Given that a place on an NHS bed costs about £350 per day, it’s easy to understand why. Rooms at the Dorchester cost about the same.

Part of the movement towards prevention will involve capitalising on the fashion for consumer health tech devices, like FitBits, which track users’ exercise regimes and produce data that can already be logged on hospital software systems like Epic at Addenbrooke's. Other, more sophisticated devices are beginning to allow constant monitoring from home of patients who previously had to go to hospital for outpatient assessments, often unnecessarily.

Asthma inhaler
Are the days of the 'dumb' inhaler numbered?

These are things like smart spoons, whose handles can record the tremors of Parkinsons’ sufferers and track the deterioration of their hand control. Smart inhalers, for asthma, are logging air quality when users take a puff, and better calibrating doses. Aggregated, that data may provide new insights into what triggers an attack.

At a simpler level, the NHS now recommends some smartphone Apps which it thinks effective and thus save it money. One is for insomnia.  Others are being looked at which allow patients with mental health problems to track their mood. Such self-reported data can then be assessed for patterns which might predict if they are at risk of an “event” - as attempted suicides are euphemistically known.

According to Eric Topol, an American geneticist who is currently leading a review of technology in the NHS for Health Secretary Jeremy Hunt, the combination of apps, wearables, and sensors will combine in the next few years to move healthcare “out of the hospital and into the home”.

"People won’t need to be in a hospital,” he says. “Unless they’re very acutely ill in the intensive care unit, or just coming out of the operating room, most people will be perfectly well at home with sensors which offer continuous, real time monitoring of blood pressure, heart rate and the like - better thn hospital but with a much lower expense factor.”

The consequences for hospitals and those who work in them will, says Topol, be profound.

“Just imagine the implications on the [NHS] workforce over time, because with [home sensors] then you don’t need the staff of a hospital to support patients. This is a very significant thing. It’s so expensive to support hospitals, it’s so labour intensive and so much of that can be taken over by technology and it will be.”

Topol’s is a vision in which the need for the giant “hotel hospitals” of today, where large numbers of often elderly patients frequently spend many days tended to by huge numbers of staff, simply disappear. Instead of relentlessly needing ever more beds, he thinks, the pressure will go the other way.

“Over the course of the next decade there will be a big trend towards at home monitoring,” he says. “So we will see a major reduction in the need for hospital beds and facilities as we know them today. I can't see ICUs disappearing but those other wards..."

This is a radical concept, in which patients use cutting edge technology keep themselves healthier for longer and then, when they do fall sick, are monitored at home for as long as possible. It is future in which the biggest costs of the NHS - hospitals and the staff in them - are all but eliminated. But while such cost savings will certainly be politcally alluring, prioritising prevention will require a fundamental change in the economy of the NHS: how treatment is paid for.

A new payment model needed

At the moment, CCGs, which control the purse strings, pay per procedure. So innovations which allow hospitals to perform more procedures, quicker, are welcomed. “But who in the NHS is going to pay for a new diagnostic piece of kit,” says Michael Branagan Harris, a business consultant who specialises in helping private companies navigate the procurement complexities of the health system.

“Currently patients can’t get devices that monitor them at home on the NHS, because there’s no mechanism to pay. Doctors can’t prescribe such devices like drugs. And hospitals won’t pay because the ‘treatment’ is not happening at the hospital, it’s happening at home.”

Faced with such bureaucratic conflicts, the government freely admits that getting innovation into the NHS is “unnecessarily complex”. Today, even if new devices meet  the stringent guidelines for acceptance there is still no central path to roll them out across the NHS. Instead, manufacturers must become travelling salesmen, travelling door-to-door to every trust procurement department, cutting individual deals - and different prices - with each.

“Currently it takes 15 years to get an innovation from first clinical trial to widespread use,” says Mike Hannay, Chair of the AHSN network. “We have to speed that up.”

A host of initiatives, with names like the Innovation and Technology Payment, or the Accelerated Access Pathway (AAP), have been established to attempt to hack a clearer, quicker path to doing so. And though only a handful of innovations are selected each year, slowly, a new model is emerging. The NHS Innovation Accelerator, for example, provides funding to medical entrepreneurs primarily emerging from the NHS. “We are trying to become an accelerator for in-house talent,” says Hannay. “We might not have $10bn companies today, but one day we might.”

Stakes in such companies, and the profits they generate, might one day provide an alternative source of funding for the NHS. Even now, some money is trickling in. “We’ve worked with a simple messaging system for school nurses in Leicestershire,” says Hannay. “That’s generating revenue which is coming back into the NHS trust.” The result? Government is now calling for “significant improvement” in the “commercial capability” of the NHS to smooth the way for more “win-win deals with innovators”.

Ali Parsa, CEO Babylon Health
Ali Parsa, CEO Babylon Health

One tech company that has made it is Babylon Health, which is now paid by the NHS for its GP-at-Hand service - essentially remote video GP consultations done via a computer screen. Ali Parsa, Babylon’s excitable CEO, promises not only that patients who use GP-at-Hand get to see a doctor within 2 hours, but also that 40 per cent who call in are so reassured by its automated, AI-powered, pre-consultation triage, that they realise they don’t need to see a doctor after all.

What remains to be seen, though, is whether GP-at-hand and services like it will be adopted across the nation. Both doctors and patients can be hard to convince of the merits of change. “Some will never adapt,” says Parsa, ruefully.

Innovation, good. Iteration, bad

But roll-out is the critical factor. Because whether it is cutting edge big-data analysis by Deepmind, tele-heath like GP-at-hand, Smartphone apps like DrDoctor or wearable devices like Fit-bits - there is no shortage of medical innovation in Britain. Health, after all, is a vast market which is guaranteed to grow.

What’s in short supply is a way of getting the best new technologies used throughout the whole of England’s fragmented, un-national National Health Service.

“There are pockets of excellence. But it’s really hard is to get the whole NHS to do it,” says Mike Hannay. Part of that is because there is no money in dissemination. The NHS currently spends more than £1bn a year on research, but only a few million on implementation.

Mustafa Suleyman, Deepmind Co-founder
Mustafa Suleyman, Deepmind Co-founder

“Look at any innovative organisation like Apple or Microsoft - they spend more on adoption than R&D,” he adds. “They know it’s critical. Why do we think we’re any different?

“We’ve got an ageing population which is growing fast. And no government, of whichever colour, has the money to keep spending the way we do to treat that ever-ageing population as we do now. Society just can’t afford it. So the most valuable thing we can do is keep patients out of hospitals, out of GP surgeries, and help individuals keep themselves healthy at home.

Then, when we have to treat them we have to find new, more efficient ways of doing so. And the key to both those things is the spread and adoption of innovation and technology across the NHS. That’s the way to a sustainable NHS for the next 70 years. In fact without that, there soon won’t be an NHS at all.

*names changed