Acute kidney injury (AKI), formerly known as acute kidney failure, is a sudden loss of kidney function. It is
caused by a number of medical problems including infections of the bloodstream, dehydration, drug toxicity, blockages of the urinary tract (from for example kidney stones) and specific renal diseases such as glomerulonephritis.
AKI affects up to 20% of acute hospital admissions and is associated with a number of life threatening complications like excess acidity of the blood and high potassium levels.
It is known that if a patient suffers an AKI, then they are likely to spend more time in hospital and have less chance of surviving emergency hospitalisation. Around half the patients who die in hospital in the Royal Free have evidence of an AKI prior to death.
NHS England estimates that 100,000 deaths per annum are associated with AKI and the healthcare costs of this problem in the NHS exceeded £1 billion.

If we can intervene in patients with early signs of an AKI then we may be able to save lives
AKI is an important cause of chronic kidney disease (CKD) and CKD progression. A single, severe AKI episode increases the risk of developing CKD by 30 times and the risk of needing chronic dialysis by 10 times. 30% of patients need dialysis for AKI in the Royal Free renal unit end up requiring longterm dialysis.
If we can intervene as soon as possible in patients with early signs of an AKI, then we may be able to save lives, get people home earlier and hopefully prevent many people ending up with CKD and on dialysis.
The challenge of AKI is that it can affect so many patients, from a young mother with preeclampsia on the labour ward, to an elderly patient in hospital who had had a stroke. These patients may be located across the whole hospital and may be undergoing treatment for many other things. Most of our inpatients have
regular blood tests and several thousand tests of kidney function are carried out each week across our hospitals. One in five of these patients having blood tests will have a change in kidney function.
So how do we ensure that when there is an AKI, patients get the right treatment at the right time? To help crack this problem the Royal Free has approached DeepMind Health.

DeepMind is a British artificial intelligence research company
The Royal Free is working with DeepMind to develop clinical applications that could improve patient safety and clinical care. DeepMind has worked with Royal Free staff to develop its first app and it is focused on the management of AKI called Streams.
So what does Streams do and how will it help AKI patients?

Streams monitors a patient’s kidney function by taking a flow of data from our laboratory systems. It applies a nationally mandated ‘algorithm’ which has been developed by NHS England to detect significant changes in kidney function that can indicate that a patient may be developing AKI. When AKI is detected, the application notifies a clinician and sends them important tests results and details of their relevant diagnosis. This means they can rapidly decide whether the patient needs help, what treatment they require and ensure they are appropriately managed, wherever they are in the hospital.
Chris Laing explains why Streams is so important for his work
Streams will give us the opportunity to monitor, in real time, on a smartphone, the kidney function of patients having blood tests across our hospitals and help us to intervene early to prevent their AKI getting worse. We believe this will help save lives, prevent longterm kidney damage and get patients safely home earlier.
We are delighted that a collaborator of this calibre, backed by one of the largest technology companies in the world, is working on AKI in the Royal Free as its first healthcare project. The project was originally announced in February 2017 and received considerable national interest. It involved the handling of some patient data and understandably there was concern to ensure this was done safely and securely.
How patient data is handled safely and securely
We understand that patients need to be confident that their health data is secure. We took rigorous measures to protect the security of patient data and the project did not involve any research nor transfer any ownership of patients’ data to Deep-Mind. Like many other clinical software systems, Streams processes encrypted data and presents relevant information to licensed clinicians who are authorised to review the results and must securely log in. DeepMind can’t do anything with this data other than serve clinical care under conditions defined by the Trust and our clinical teams. We actually believe new standards for healthcare data security and privacy are being set.
Streams is an improved care pathway involving a renal specialist team who will use the application to be piloted from November 2017 onwards and this new pathway will undergo a full evaluation with our UCL colleagues.
DeepMind is an artificial intelligence (AI) company, so can we use AI to treat AKI?
Currently Streams uses no AI but there is definitely potential for doing this type of research in future under the appropriate regulatory approvals. Ultimately we may be able to predict in advance whether patients may get AKI before it happens, help diagnose the underlying cause of kidney failure or identify patients who may subsequently get CKD and then have a live application so we can use this information to support clinicians.”
More information on DeepMind Health and the Streams application is available here
The Royal Free London Trust Q and A on the project is available here
By Chris Laing, Consultant Nephrologist an Associate Medical Director, Royal Free London