An insight into an innovative imaging device that uses artificial intelligence to provide safer resection of brain tumours.
At a new, state-of-the-art specialist and critical care centre in South Wales, an advanced patient monitoring solution is supporting care for some of the most critically ill patients – throughout the pandemic and beyond, as well as contributing to new research insights. Delivered during the COVID crisis, the speed and scale of the project enabled the hospital to scale-up ICU capacity, in the face of a surge of patients with complex needs. Louise Frampton reports.
As the number of seriously ill patients with COVID-19 steadily increased throughout 2020, a decision was made to open Gwent’s flagship ‘super critical care centre’ months ahead of schedule. In an impressive feat of construction, the completion of the new Grange University Hospital was advanced at pace to provide much-needed extra ICU capacity to cope with a surge of critically ill patients during the second wave of the COVID-19 pandemic.
The state-of-the art Grange University Hospital is equipped to treat the most seriously ill patients, or those with significant injuries, and is a designated trauma unit. With 560-beds, a 24-hour acute assessment unit, emergency department and helicopter pad, ‘The Grange’ is one of the biggest NHS construction projects Wales has ever seen – costing around £350m to build.
The project was part of the Aneurin Bevan University Health Board’s strategy to allow other hospitals across Gwent to concentrate on providing routine diagnosis and treatment without having to give priority to emergency patients. The aim was to improve the flow of patients throughout the NHS system. However, as the health system experienced severe demand for intensive care beds, the urgent need for extra capacity became apparent.
“Other healthcare providers had set up field hospitals, such as requisitioned sports centres and rugby stadiums, to allocate the projected overspill of patients during the pandemic, but this was never an option for us,” commented Dr. David Hepburn, an intensive care consultant at the Grange.
“The building was in progress and it was due to be officially opened in March/ April 2021, so we prioritised finishing this as a matter of urgency and it was ready to move into by November 2020. The developer, Laing O’Rourke, pulled out all the stops to complete the hospital as quickly as possible.”
Dr. Hepburn explained that the plan was to replace the two existing ICUs, which included an eight-bed unit at Neville Hall (a district general hospital in Abergavenny), and a 16-bedded unit at the Royal Gwent (a local general hospital in the city of Newport). These would be merged into a single 30-bedded ICU with the latest monitoring technologies supporting complex patient care. Each patient would have their own single room, reducing the risk of healthcareassociated infection, with views outside – which would make a significant difference to patients’ wellbeing.
“We made it a priority to get in there as quickly as possible. Having been through the first COVID-19 wave, we knew that a lot of our problems would be solved by being on one site,” he explained. “Instead of running two units in parallel, we could group all the staff and patients in one unit, which would mean the pool of doctors and nurses would be greater and our resources would not have to be duplicated across two hospitals.
“Moving into ‘the Grange’ was a gamechanger in terms of being able to deal with a large volume of COVID patients in the second wave. The design of the unit is such that we could surge to 60 patients without leaving the footprint of the ICU.”
Before the Grange opened, the ICUs “broke their banks”, in the first wave, and ended up taking over adjacent wards. However, the design of the Grange meant that in the event of a surge of patients, it would be possible to double up patients in each room, as the rooms were so large.
“This gave us double the capacity if we needed it, going forward. This made a very big difference, when we hit our peak of getting on for 50 patients, as we were still able to do this within the confines of the ICU,” commented Dr. Hepburn.
He pointed out that it was important that the final project was finished but not ‘rushed’ – everything had to be future proofed, including the patient monitoring technologies. These needed to provide accurate data for clinical decision-making and ensure patient safety, both during the COVID crisis and beyond, but they also needed to meet future digital objectives – they had to be able to support advanced applications using artificial intelligence, automated vital signs capture (as part of the vision for a paperless hospital), as well as being capable of facilitating potential new concepts of healthcare delivery, in the future
The solution
As Gwent is a pilot site for the implementation of the ASCOM eICU system across Wales, a main priority for the patient monitoring technology was the ability to output data in a way that the ASCOM system could interrogate and understand.
As part of government plans to accelerate digital improvements for critical care, the NHS Wales Informatics Service has signed a £13 million contract with ASCOM. The seven-year agreement provides a technical platform to develop the all-Wales Intensive Care Information System. Funding is being provided by the Welsh Government, Health Boards and the Critical Care Trauma Network.
The aim is to transform critical care by automating the collection of data from monitors and devices used to support patients with life-threatening illness or injuries. With increasing pressure on intensive care services, reducing the administrative burden on NHS Wales staff will free-up more time for patient care.
At the Grange, observations are currently taken down by the nurses and recorded on paper, at the end of the bed. Ventilatory parameters, peek airway pressures, peak levels, oxygen concentration, I:E ratio, and other data from the ventilators are recorded on the sheets along with the patient’s vital signs measurements. Staff also record arterial blood gasses, point of care testing, urine output and fluid input, so they can calculate the patient’s fluid balance.
“Taking down this information and making calculations is an immense task for nurses and amounts to around three hours per day of work, at present. With the new ASCOM system, the data will be populated automatically, which will free up staff to focus on hands-on nursing tasks rather than noting down vitals. It will make a significant difference once the ASCOM system is fully operational,” said Dr. Hepburn.
“As the patient data will be automatically populated, transcription errors will also be eliminated. If there are any events, you will be able to look back and interrogate the data. The ASCOM system will also capture results and allow you to identify whether the patient is at risk of kidney injury, for example, or suggest a different dose range for a drug because their kidney function is impaired.
“It will perform some of the tasks normally performed by the pharmacist and prescribing will also be electronic. If two drugs have the potential to interact with each other the machine will flag this up. This has been available to GPs for many years but is only now becoming part of hospital practice. We hope this will reduce drug costs and prescribing errors.”
The hospital hopes to have more and more parameters captured automatically; even urometers will be linked to the system – the fluid will be weighed and the data ‘bluetoothed’ to the central monitoring system, so that urine output will be captured accurately. Smart infusion pumps will also link into the system.
“Having all of the systems integrated into one platform, with the patient monitors providing an integral part of the functionality, will be a step in the right direction,” commented Dr. Hepburn
In the long term, integration with NHS Wales’ systems will enable intensive care staff to:
- Support national audit and research needs
- Record patient assessments electronically
- Manage prescriptions and drug administration at the bedside
- Connect with bedside equipment to record vital signs and fluid balance
- Calculate a patient’s acuity scores
- Better manage infection control
- Manage daily care plans
- Create reports on results and department objectives
- Support national audit and research needs
With this advanced digital vision at the heart of the project, it was important that the patient monitoring technology was compatible with ASCOM’s platform. This was a key factor in the decision to install Mindray’s technology.
“We had dealt with Mindray before, for both monitoring and ultrasound, and found them very adaptable. They often acted on feedback to unlock and develop further functionality, so we knew that if we wanted a particular derivation or data, their software could offer this flexibility. The monitors have clear, wipe-clean screens and the technology is easy to use and intuitive. In addition, the company was also able offer UK-based service and support.
“It was important that the supplier could re act quickly and be on hand to troubleshoot, rather than rely on service departments located overseas – particularly during the pandemic. During the first wave, the hospital found that some companies experienced supply chain challenges and prioritised local need. Therefore, security of supply with a UK-based service team was a high priority at the procurement stage,” commented Dr. Hepburn.
With COVID pressures quickly escalating, the roll-out of the monitoring technology required a fast response. Mindray supplied, installed and commissioned 350 patient monitors, as well as the connectivity systems required, in just six weeks – a significant undertaking. The installation included TM80 wearable telemetry devices (for monitoring while transferring patients), BeneVision N1 monitors, which plug into Mindray’s larger N12, N15 and N17 high acuity bedside monitors, as well as the N22 monitor, which is designed to allow the screen to be adaptedfrom portrait to landscape format
The BeneVision patient monitors provide seamless integration with other bedside devices, such as ventilators, anaesthesia systems, and infusion pumps, through the BeneLink module, while Mindray’s eGateway system and central monitoring server enable the distribution of data between the bedside monitors and workstations. The eGateway will also provide connectivity with ASCOM and integration into the electronic patient record (EPR) system, as the Grange transitions to paperless workflows.
The solution now provides continuous patient monitoring from admission to discharge, including during transport. With the Mindray system, the monitor can follow the patient throughout their care journey.
When the patient is moved to another area within the hospital, the module simply unplugs from the side of the monitor and can be used as a transport monitor. It can then be ‘plugged’ into the host monitor at the bedside, at the new location. There are no cumbersome leads to disconnect and reconnect, or clean between patients, and the patient is continuously monitored – ensuring seamless data and patient safety at all times.
“The system captures every heartbeat,” commented Dr. Hepburn. “From the minute a patient arrives in the ICU, they have a basic six-way monitoring data set, which includes ECG electrodes and oxygen levels. Anyone that is intubated or on a ventilator will have end-tidal carbon dioxide (ETCO2) measured (capnography).
“We also capture data on invasive and non-invasive blood pressure, and temperature. Most patients have central venous access, so the ability to transduce the central line in terms of checking the positioning is very important. However, we also use advanced monitoring, including PiCCO cardiac output monitors, which work through analysis of pulse contour and thermodilution curves. Therefore, we needed the monitors to be capable of outputting special parameters. This also included the need to be compatible with outputs such as pulse pressure variation and bispectral index measurements,” Dr. Hepburn explained.
The quality of the data is also extremely important. The patient monitors from Mindray feature CrozFusion technology, which combines and analyses SpO2 waveforms and ECG to help reduce false arrhythmia alarm notifications and increase the accuracy of heart rate and pulse rate.
“There can be a lot of interference; if you don’t get the filtering right, ECGs can frequently flash up errors or tell you that the patient’s heart isn’t beating. When patients have poor perfusion, have cold hands or a lot of fluid oedema, you can lose the biological signals, but the Mindray technology offers good filtering – it amplifies these signals and rejects any artefacts. It means we can rely on the data and it will give an accurate insight into what is going on with the patient,” commented Dr. Hepburn.
He added that some older styles of monitors, particularly in the case of SATS probes, can be affected by skin colour or tone, resulting in less reliable readings.
“Many are calibrated on Caucasian skin so tend to give erroneous signals in patients from the BAME community. We need to rely on the monitors for all of our patients. This is very important. With the Mindray technology, the SATS probe is very reliable. It also has a forehead module, so if data isn’t picked up from the patient’s fingers, a module can be put on their forehead. When we compare this data to our invasive blood gas measurements, we have found this to provide an accurate representation,” he continued.
The design of the Mindray monitors also enables data to be shared and viewed from a ‘slave’ monitor. This provides visibility at all times.
“If you need to leave the bedspace to attend to a patient next door and assist a colleague, you can bring up the monitoring data from the adjacent bed space. This means you can keep an eye on the patient while you are helping with another. I haven’t seen this function elsewhere.
“There are other thoughtful features for end-of-life situations. If the patient’s family is having some quiet time with the patient, it can be distressing for the family members to see the monitor as the patient’s vital signs decrease. We can put the monitors on to ‘privacy mode’ which means it goes blank; there are no alarms, so it is more peaceful and unobtrusive
“However, the ‘slave’ function allows the data to be sent to a monitor in another room to allow the clinician to know what is happening physiologically and identify the appropriate time to come in and support the family, and let them know their loved one has passed away. If you can de-medicalise end-of-life care and turn off the alarms and numbers on the screen, the experience is much more patient centred and dignified,” Dr. Hepburn explained.
Big data and trend analysis
The patient monitoring technology also needed to support the hospital’s future development of artificial intelligence-driven improvements in daily care – from early detection of deterioration in patients with sepsis or acute kidney injury, to pioneering research to further current understanding of the predictors of mortality in patients with COVID-19
“The advantage of the big data sets that we capture is that we can use them in research. If you have ECG data for patients in the ICU, over a period of time, you can feed this to an AI system. Trend analysis could facilitate pre-recognition of events and changes in the patient’s status, including deterioration.
“For example, AI could help recognise that patients have a number of ectopic beats of their heart before they go into atrial fibrillation. We may be able to target patients that are at risk and intervene to stop them from going into atrial fibrillation. We have used this type of data set already during the COVID surge,” Dr. Hepburn explained.
“NEWS scores are well ratified in identifying the deteriorating patient and this is already supported by the monitors, but there may also be more subtle signs and markers that AI can identify from huge data sets, such as in the case of sepsis.”
He pointed out that Professor Tamas Szakmany, a consultant at the Grange, has been working with colleagues from Public Health England and Cardiff and Nottingham Trent universities to use big data captured in the ICU to further ‘predictive medicine’
The researchers have already investigated blood results captured on patients in the ICU to derive a risk score, by feeding this data into an AI system. This research has identified molecular “signatures” in the immune component of the blood which indicate how patients in intensive care with sepsis, septic shock and systemic inflammatory response syndrome are likely to respond.
Sepsis is a life-threatening condition – requiring admission to intensive care – and occurs when the immune system overreacts to an infection and starts to damage the body’s tissues and organs. It is a major healthcare problem in the UK and accounts for a quarter of intensive care admissions in the UK.
Despite this, there is a lack of knowledge of the immune processes involved in sepsis or the clinically relevant molecules at play. This would help clinicians to distinguish between sepsis and systemic inflammatory response syndrome (SIRS) – which are very similar – to improve patient management through more appropriate treatment as well as help to identify potential new therapies.
The work means that, for the first time, clinicians could be able to test and preemptively manage and treat patients based on their immune profiles. The team believe this approach could also be applied to COVID-19, as it manifests as a sepsis-like disease in more severe cases.
The future ICU
In the future, big data and patient monitoring technologies could also pave the way for new models of care, according to Dr. Hepburn. The systems installed at the Grange have the functionality to provide the foundations for increased remote monitoring of patients, to enable greater oversight of care across healthcare boundaries by experienced consultants. The ‘Tele-ICU’ is an area of research currently being explored by Professor Szakmany.
“In the future, there is the potential to cover a number of ICUs remotely, a model already used in the US. There could be one consultant coming on to their shift at night, covering five smaller units, with central access to the patient data and cameras by
the beds. This could enable a consultant located in a time zone where it is day, to oversee a shift at a location where it is night. This could reduce fatigue and make better use of limited resources and skill sets. Hands-on nursing care and competent registrars will still be required on site, but this model could facilitate greater access to senior clinical decision-making without the consultant having to be physically present at the bedside,” explained Dr. Hepburn.
“Intensive care is a very data driven specialty. A lot of the decision making is based on huge data sets. To have this data at your fingertips will enable this decision making and facilitate new models of working in the future. It will make the job much more sustainable – you could be on call from home and it will also solve the problem of the workforce crisis. There aren’t enough intensive care consultants, and this isn’t going to change for quite some time
“Remote monitoring and the ‘Tele-ICU’ won’t replace the face-to-face human touch, but it could allow us to expand the number of ICU beds that are covered by dedicated intensive care consultants without having to expand the workforce. Potentially, you could have a very small ICU in Aberystwyth looked after by the University Hospital of Wales. This model works in some countries already and there is some evidence to suggest that this reduces length of stay and mortality.
“The more experienced the centre and staff overseeing the care of patients the better the outcomes tend to be. There is a move towards bigger and bigger ICUs but, if you use the remote model, you wouldn’t have to close the smaller units. They could receive the same level of care as the larger hospital.”
Ultimately, the technologies now in place the Grange will enable teams to work seamlessly across organisational boundaries, to share good practice and to rapidly implement change driven by evidence.
“There are lots of interesting possibilities,” Dr. Hepburn concluded.
