Scientists are using artificial intelligence to predict when people will fall ill from acute kidney injury (AKI), one of the biggest NHS killers.
Known as the ‘silent killer’, AKI is often diagnosed late and is hard to predict. It involves sudden damage or decreased blood flow to the kidneys that is often treatable.
Without rapid treatment, patients can die, end up on dialysis or need a transplant.
The condition contributes to nearly 20% of all hospital admissions, accounts for around 100,000 deaths every year in the UK, and costs the NHS £1.2 billion annually.
The new AI system from DeepMind Health can analyse up to 600,000 data points – such as blood tests, heart rate and blood pressure – and calculate whether someone will develop AKI up to 48 hours in advance.
The health and technology company’s deep learning algorithm was applied retrospectively to records of more than 700,000 patients from the US Department of Veterans Affairs – the largest integrated healthcare system in the United States.
It was able to detect 55.8% of all inpatient episodes of AKI and 90.2% of all acute kidney injuries that required subsequent administration of dialysis.
They are hoping to pilot the technology in UK hospitals within the next 12-18 months.
The researchers believe it could be extended to help detect other deadly conditions, including sepsis, which can be difficult to spot and was responsible for more than 350,344 hospital admissions in 2017/18.
Other areas the technology could be deployed in include helping medical staff predict a patient’s length of stay, their risk of readmission when discharged, their risk of falling and the risk of picking up a hospital-acquired infection.
Dr Dom King, the health lead for DeepMind Health, said: ‘This progress represents potentially a very significant change in how medicine is practised and care is delivered.
‘A lot of care at the moment is very reactive and this represents the potential to re-move the needle to proactive, preventative care. And that requires a lot of careful thought.’
Dr King, previously a general surgeon, continued: ‘When you spend a few years practising you develop a kind of sixth sense about patients who are getting unwell – people will say ‘their vital signs are normal and their blood tests are normal’, but you just know that there’s something not quite right.
‘And the current alerts are very simple and rules-based and don’t really pick up the subtlety of those patients at the earlier signs of deterioration, so it really is mindblowing for me as a doctor that in some way these AI systems are almost doing what an expert physician does, which is to look at not one or two factors but to look at thousands of factors, from what time of the year it is to what part of the hospital the patient is in, and all of these things contribute in some way to their risk score.
‘I think we’ve seen some real progress in the last couple of years in AI as applied to medical imaging, but this I think is potentially more impactful because it affects many more patients and it informs many more clinicians.’
The research was published in Nature on Wednesday.
The team have also been trialling a mobile app which sends experts breaking news-style alerts to help quicker diagnosis of AKI.
Use of the Streams app at the Royal Free Hospital has led to AKI cases being detected in 14 minutes or less – a process which might otherwise have taken many hours, a UCL-led evaluation found.
Patients with AKI deteriorate rapidly and clinicians face challenges to detect it quickly due to imitations of current NHS technology and reliance on manual observations.
Experts believe up to one in three deaths from AKI may be preventable if clinicians are able to intervene earlier and more effectively.
The app has also been found to reduce NHS costs by around £2,000 per hospital patient – from £11,772 to £9,761 for a patient with AKI.
And the detection rate rose from 87.6% to 96.7% in emergency cases.
Dr Chris Streather, Royal Free London chief medical officer and deputy chief executive, said the results were “incredibly encouraging”.
He added: ‘Digital technology is the way forward for the NHS. In the same way as we can receive transport and weather alerts on our mobile devices, doctors and nurses should benefit from tools which put potentially life-saving information directly into their hands.
‘In the coming months, we will be introducing the app to clinicians at Barnet Hospital as well as exploring the potential to develop solutions for other life-threatening conditions, like sepsis.’
Paul Leeson, Professor of Cardiovascular Medicine, University of Oxford, said: ‘This is important work in which the team have overcome several technical challenges to show it is possible to successfully apply AI to large scale electronic health records.
‘The AI was able to identify over half the patients who went on to develop kidney problems during the next 48 hours.
‘Trials are still needed to test whether this early warning is useful to doctors to improve patient care, without causing too many false alarms, or missing patients that the AI also overlooked. However, this is another strong example of how AI appears to have the potential to augment delivery of healthcare.’
Source: Read Full Article