Oxevision camera monitoring on NHS mental health wards

Oxevision is a wall-mounted infrared camera unit made by Oxehealth, now called LIO. National Health Service (NHS) mental health trusts, the public bodies that run NHS hospitals, place it in patients' bedrooms. It measures pulse and breathing without contact and sends color-coded alerts to nurses when, for example, a patient leaves the room.

How it works

Oxehealth Ltd began as a spin-out of Oxford University biomedical engineering. Its camera flags a patient leaving the room, staying in the bathroom for more than three minutes, or lingering at the doorway for more than three minutes. Nurses watch the alerts on monitors at the ward base, the nurses' station on the ward.

The ombudsman is the independent official who rules on NHS complaints in England. Under the procedure the ombudsman examined, it records pixelated video during red alerts. Staff can call up a clear image for ten seconds to check vital signs. That raw, unblurred footage is kept for twenty-four hours and then deleted automatically. It records no sound.

Each trust uses it under its own local procedure. Essex Partnership University NHS Foundation Trust was an early adopter. Its executive operational committee recommended roll-out in 2020, after a trial on four wards. Observations are the scheduled checks nurses make on each patient in person. One-to-one observation means a nurse stays with one patient. That trust now keeps patients' observation records electronically inside the Oxevision system.

How widely it is used

Oxevision is in live use across a disputed share of England's NHS mental health trusts. The vendor claims 50 percent. Legal submissions put it at about 40 percent, roughly 20 of 50 trusts. The National Survivor User Network, a network of mental health service users and survivors, counts 25 active trusts. As of mid-2026 there has been no national suspension.

Who controlled the evidence

Oxford Health NHS Foundation Trust evaluated the tool internally in February 2023. The campaign group Stop Oxevision obtained that evaluation under a freedom-of-information request. It shared it with Novara Media in January 2026. The report has not been published in full.

The evaluation recorded that the vendor had written the trust's business case. It recorded that the vendor asked survey questions the report called leading. It also recorded that the vendor guarded its own evaluation work on the tool, known as Early Insight, as the bedrock of its sales and marketing drive. The report concluded that successful business relationships are built on trust, and that this was not evident.

Oxford Health spent about 2.4 million pounds on the system between November 2021 and September 2024. Its estimated staffing savings were about 709 thousand pounds.

The case file argues that the check meant to be independent of the vendor, the word on whether the tool works, was in the seller's hands. No independent error or false-alert rate for the alerts has been published.

What the ombudsman found

On 27 March 2026 the Parliamentary and Health Service Ombudsman partly upheld a complaint, case C-2118934. A former inpatient, called Miss B in the report, brought it against Essex Partnership University NHS Foundation Trust.

During her admission in 2021 and 2022, the trust did not seek her consent for Oxevision or revisit it. It did not switch the camera off when she asked. It gave her no information or resources about it, and kept no record of how staff used it in her care.

The ombudsman found the surveillance had made her feelings of anxiety, vulnerability, and fear worse. She changed her behavior to try to protect herself from the technology. It ordered the trust to pay her 925 pounds and send a letter of acknowledgment. The trust must also produce an action plan and re-review its operating procedure.

One part of the complaint did not succeed. On balance, the ombudsman found staff had continued in-person observations. It checked observation charts, a nurse adviser's review, and door key-card data. So the case is not a finding that the camera replaced human care.

What the procedure said about consent

The trust's earlier procedure told staff to tell patients that Oxevision is part of normal practice and that consent for its use will not be required.

The ombudsman found even the revised 2025 procedure departs from the standards of the General Medical Council, the UK's regulator of doctors. It also departs from NHS England's February 2025 principles for using digital technologies in mental health inpatient care. A nurse in charge or a doctor can override a refusal from a capacitous patient, meaning one judged able to make the decision. They need the approval of the multidisciplinary team, the staff of different professions caring for the patient, and must judge it clinically safe. So the trust may keep filming a capacitous patient who has refused.

Separate reporting describes patients told that images would be blurred when clear footage existed. Others were told the cameras were mandatory trust policy when no such policy existed. A peer-reviewed critique reports staff reassuring patients that the camera was not recording them all the time.

The inquiry and the rename

The Lampard Inquiry is a statutory inquiry into the deaths of more than 2,000 people under Essex mental health inpatient care. It opened on 9 September 2024. It postponed its public Oxevision evidence and took Stop Oxevision's evidence privately on 14 May 2025. It heard evidence from the Essex trust and the vendor in October 2025.

At the inquiry, the trust admitted there is the possibility of misuse across all trusts using the tool. That includes malicious staff access to live bedroom feeds. The vendor's head of patient safety accepted that round-the-clock filming could be a very significant invasion of privacy.

In August 2025, during the inquiry, Oxehealth renamed itself LIO, or LIO Health. The National Survivor User Network reports that the parent company, and possibly the product, were renamed. The oxehealth.com address now redirects to liohealth.com.

The data-protection complaint

On 18 May 2026 the law firm Bindmans, acting for Stop Oxevision, asked the Information Commissioner's Office for an expedited investigation. That office is the UK's data-protection regulator. The request covered consent and transparency, and opaque data arrangements between trusts and the vendor. It also covered how pixelated video and derived data are classified, and retention and reuse without a clear lawful basis.

In June 2026 Novara Media reported that the office had opened an investigation. It quoted a Bindmans solicitor saying there is a clear legal risk that other NHS trusts using the tool in similar ways could face comparable complaints.

What each side says about safety

The vendor claims reductions of 44 percent in self-harm, 48 percent in falls, 37 percent in assaults, and 71 percent in one-to-one observations. It also cites counts of incidents detected on the Essex trust's wards. These are marketing figures.

A campaign-linked meta-analysis found no statistically significant association between Oxevision and four of five measures of patient safety. Novara Media reports that a 2024 review found insufficient evidence for the promised safety and cost benefits. A peer-reviewed critique argues the flagship acceptability study's conclusions did not match its own participants' data on privacy.

Campaigners and bereaved families have linked the technology to at least four inpatient deaths. That link comes from journalism and campaign reporting. It is not an adjudicated finding that the technology caused the deaths.

Novara Media also reports that the vendor plans to roll out a whole-hospital patient-monitoring platform in 2026.

Where the facts come from

The main sources are the ombudsman's final report and Harriet Williamson's reporting for Novara Media in January and June 2026. Others are Bindmans' press releases from 2025 and 2026, and a 2025 statement by the National Survivor User Network. The list also includes a Lampard Inquiry update, the vendor's website, a 2026 critique in BMC Psychiatry by Porter and Edwards, and Stop Oxevision's resources page.

What this network is drawn from

This network follows the pattern the case file documents. It is not a reconstruction of the actual device. It shows the monitor, the bedroom video, the nurses, trust clinical governance, the observation record, and the vendor's platform. It also shows three checks. Patients are outside the network, and it computes no clinical, distress, or safety outcome for anyone.

What the available tools can and cannot address

A failure pathway is a link between two parts of the network where a mistake made by one part can be passed on to the other. A tool closes a pathway when mistakes stop passing along it. The work along it may go on.

This case has a budget of 9 units. Understand the system is not on offer here, so every tool costs its listed price. Explore (No Targets) sets no targets. The other three levels all ask for the network's mistakes to be contained, meaning corrected rather than building on each other. They also ask that the monitor still adds to the work.

Lingering effects is the Lab setting in which some effects stay after their cause is gone. It starts on under Explore (No Targets) and Service Targets Only, where you can switch it off. Service and Safety Targets keeps it off, and All Governance Targets keeps it on. With it on, three tools work at reduced strength: Peer sharing rules, Store less data, and Vet connections. Vet connections then closes no pathway here.

Under Service Targets Only, one tool meets the targets on its own: Mark AI-written records, for 2 units, or 3 at its stronger setting. It stops mistakes passing along Monitor reads the bedroom video, Alert history read at handover, and Record data used in the vendor's evaluation. With lingering effects on, 224 different sets of tools meet the targets within the budget, or 588 counting stronger settings. With lingering effects off, Vet connections alone also meets them, for 3 units.

Under Service and Safety Targets, you must also close every failure pathway, keep the ward from falling behind, and stop the Privacy gauge draining. Eight pathways are open before any tool is used. The first four are Monitor reads the bedroom video, Alerts shown to nurses, Vendor reports given to governance, and Readings and alerts logged in the record. The other four are Nurses record observations, Alert history read at handover, Record data used in the vendor's evaluation, and Video and data held by the vendor.

The cheapest way costs 7 units: Escalate checks, Mark AI-written records, and Store less data. Six different sets of tools meet the targets, or nine counting stronger settings, all within 9 units. Every one includes Escalate checks and Mark AI-written records. All Governance Targets also sets a carbon target and asks the monitor to add more to the work. The same sets meet those targets too.

The sources describe no one running two checks on this network: Independent evaluation of the monitor, and Audit trail of monitor use. No tool on offer adds either. So the targets can be met while no fully independent evaluation of Oxevision exists.

More is not better here. Using every tool at its strongest setting costs 36 units, four times the budget. It contains the mistakes and closes every failure pathway. It still meets the targets at no level, because the monitor then adds too little to the work.