Worker pointing at a monitor
Picture of health: In trusts that have implemented the Oxevision system, 94 per cent of staff say it has helped improve patient safety

It is a moment seared on Prince Ade-Odunlade’s memory. When he was a senior manager at an NHS trust in England’s Midlands, a psychiatric patient died after staff failed to notice an abrupt deterioration in the man’s physical health. At the subsequent inquest, Ade-Odunlade promised the patient’s family he would do his utmost to prevent a similar tragedy occurring in future. “I felt this sense of loss with the family,” he recalls. “And my promise is that I will do everything in my capacity to . . . minimise, or completely stop, inpatient deaths.”

His staff identified a potential solution in a piece of technology invented by Oxehealth, an Oxford university spinout founded a decade ago. Its Oxevision monitoring system uses infrared-sensitive cameras placed in patients’ rooms to measure pulse and breathing rates remotely. Alerts are sent when potentially high-risk activity is detected — for example, if a patient spends a prolonged time in the bathroom. Ade-Odunlade, who is now chief operating officer at Derbyshire NHS Foundation Trust — which has also adopted the software — believes it may have saved his patient’s life, as it could have tipped off staff about the changes in vital signs.

Oxevision is now being used in half of all mental health trusts in England, and its enthusiastic embrace by the NHS is explained partly by statistics about the number of mental health inpatients who come to harm. Unexpected or unintended events caused the death of one psychiatric hospital inpatient every 2,700 occupied bed days, compared with one death every 26,000 days for acute or general hospital inpatients. “While there are many reasons for that difference, broadly speaking, it is an order of magnitude more dangerous to be a mental health inpatient than a general hospital inpatient,” notes Hugh Lloyd-Jukes, Oxehealth’s chief executive.

However, 94 per cent of staff in trusts that have implemented the Oxevision system say it has helped improve patient safety and 87 per cent say it has contributed to better care for patients, according to surveys initiated by the company and carried out by the trusts. At the same time, 70 per cent of patients report a greater sense of wellbeing.

But it was not a foregone conclusion that the technology would be deployed in the mental health field, says Professor Lionel Tarassenko, who led the team that devised it as a founder-director of Oxehealth. The company was spun out of Oxford’s Institute of Biomedical Engineering, which he established in 2008.

About 12 years ago, he became aware of research by leading institutions, such as Massachusetts Institute of Technology, which demonstrated it was possible to use a camera to track vital signs, such as heart or breathing rates. “Every cardiac pulse, every heartbeat — there’s a tiny change in colour in all of the vessels in your face as the fresh blood comes in with each beat, and these cameras are good enough to pick it up,” he explains.

“I thought, if I leave [the breakthrough] within the academic world, then yes, we’ll refine it, we’ll do lots of interesting things — but how do I actually discover where this can best be used for healthcare applications?” Having already been involved in three university spinouts, he decided that “one good way to test the market is to get some seed funding . . . and test ideas, because the market effectively will tell you, at least commercially, where the applications are”.

He secured about £1mn from IP Group, a UK intellectual property business that invests in technology start-ups, to set up the company. The die was cast following an encounter between one of his former PhD students, Oliver Gibson, now the company’s director of research, and a medical student friend who happened to be undertaking a psychiatry rotation at Broadmoor, the UK’s oldest high-security psychiatric hospital.

As Gibson described the invention to her over dinner, his friend remarked that a video camera that could monitor vital signs would be “ideal” for her patients, Tarassenko recalls. “We started talking to Broadmoor, we got their clinicians on board, we did a pilot study, and the rest is history,” he says.

Tarassenko emphasises that it is hospital staff across the NHS who have driven Oxevision’s adoption, happy to be able to focus their energies on other aspects of their jobs in the knowledge that there are constant “eyes on the patient”.

Underlining its potentially transformative impact, he recalls the case of a patient in an acute psychiatric hospital who had been violent on admission and had been put in an isolation room and sedated. However, the urgency of the situation meant staff did not have time to check the patient’s notes and were unaware that he was being medicated for severe epilepsy. He went into cardiac arrest as the sedative interacted with the epilepsy drug.

“His heart rate went to zero, his breathing rate went to zero. That was picked up by the video camera and [staff] went straight in and brought him back . . . If it had not been picked up, it’s very likely that person would have died in the isolation room,” Tarassenko says.

A tablet showing the graphic interrface of Oxehealth software
Progress reports: Oxehealth’s software can reduce the need for frequent physical checks on patients, the company says

While the NHS overall is having to contend with acute staff shortages, not least in mental health, Oxehealth executives are at pains to emphasise that their product is not a substitute for trained personnel. Lloyd-Jukes notes that Oxevision is “an assistive technology that is helping doctors and nurses focus on what they are trained to do. We aren’t a diagnostic technology . . . We provide a risk management and care planning tool that gives additional insights and expands the constellation of data available to the clinical practitioner. They exercise judgment on all of the information they have in front of them.”

The technology can save time and money. Lloyd-Jukes says health economists at the University of York have found that “a typical NHS trust, if they spend a pound [on the technology], can expect to make more than a pound back in cash savings”. He adds that “providers who have incorporated Oxevision into their system of care have reported fewer injuries, leading to less cash [being spent] on agency staff . . . In the context of a very stretched NHS workforce, probably two-thirds of the total benefit comes in increased time to care.”

Richard Morrow, assistant director of public and physical health at Derbyshire, acknowledges that some healthcare staff raised questions about its scope when the trust first decided to adopt the technology, which was after a very similar incident to the patient death experienced by Ade-Odunlade. However, the team quickly realised that it was a supplement to patient care, not a replacement, he adds.

Maria O’Brien, chief nurse at Central and North West London NHS Foundation Trust, says that since the technology was installed 18 months ago, staff have seen a reduction in patient falls of up to 40 per cent in areas where older people are cared for, as well as those receiving acute psychiatric care. There has also been a 29 per cent decrease in assaults in its psychiatric intensive care units.

Since the trust had also made other improvements, it is hard to state definitively that Oxevision is behind this improvement, “but I’m pretty sure it had a significant impact”, O’Brien says.

The response from patients has been positive, too, she adds. Staff do not have to make as many physical checks, so patients’ sleep is less disturbed, for example.

Tarassenko points out that a requirement to carry out checks every 15 minutes on patients through the night adds up to a total of 100 observations per patient. “We can reduce that to one — 99 checks have become superfluous because you get [the information] directly from our software,” he says.

The company says identifiable patient data are deleted, by default after 24 hours, although, on rare occasions, it can be retained by clinicians beyond that period should an incident, such as a patient fall, require a review. Most trusts, Tarassenko says, “run the implicit informed consent model” under which patients are given information sheets and briefings that explain that the Oxevision system is being used in the hospital and they may discuss opting out of it with their psychiatrist.

Privacy concerns have, nevertheless, reared their head. Oxehealth acknowledges that a “service user group” at Camden and Islington NHS Foundation Trust in London raised concerns over the way consent was obtained by the trust when the technology was piloted there.

The trust is no longer using the technology. The National Mental Health and Learning Disability Nurse Directors Forum subsequently worked with a cross-section of organisations, including representatives of users and carers, to develop national guidance on how to implement an informed consent regime for “vision-based patient monitoring systems”, the company says. This has now been published. O’Brien, who was a member of the forum, says its aim was to distil, then disseminate, best practice.

In the past three years, Oxehealth has entered the Swedish market and recently signed its first contract in the US. However, it has yet to turn a profit, choosing instead to plough the money it makes into fresh research.

For example, it is developing a new Oxevision device that, if successful, will allow clinicians to monitor the quality of patients’ sleep — “a hugely important factor in diagnosing and monitoring psychiatric conditions”, Lloyd-Jukes says.

He adds: “We are still investing for growth . . . I see a route to operating profit over the next three or four years.”

For now, Oxevision is continuing to garner satisfied customers. Morrow enthuses “There isn’t really anything else like it . . . in the market that provides us with a level of robust data and a robust patient safety intervention. It’s quite unique.”

Copyright The Financial Times Limited 2024. All rights reserved.
Reuse this content (opens in new window) CommentsJump to comments section

Follow the topics in this article