Clinical audit is a quality improvement process aiming to find a solution to the problem or to make the health system more efficient and safe. The continuous cycle of audits and re-audits ensures that the hospital improves its delivery of care on a regular basis.
This comes at a huge cost.
An audit needs to be completed manually by a large number of staff involved in entering the data, analysing it, reporting and implementing its findings. The hours spent on compiling and audit are hours lost to the frontline health service and that puts an extra burden on a cash-strapped health service.
A helping hand of a computer.
What if we could use a super smart machine to take the burden off our NHS? Three young scientists from the University of Bristol (UK) developed a new Artificial Intelligence (AI)-based algorithm that can achieve just that.
The findings of their simulation experiment have just been published in the International Journal of Surgery earlier this month. In this hot off the press article, they proved that the algorithm makes the process quicker, cheaper and pain-free.
How does it work?
Frideswide, which the algorithm is called, takes over the heavy burden of the data crunching by scanning thousands of clinical letters, scan reports, blood tests, referral letters and discharge summaries. It then puts it together in an easy-to-follow network and works out the patterns within individual patient’s journeys.
In a fast-paced clinical environment, with a stack of patient files piling up on the clinician’s desk, it is almost impossible to go through all the data across all specialities, especially if the first admission dates back years ago.
The algorithm ensures that a clinician can see the whole picture and can quickly tease out the relevant pieces of information, and spend the saved time on delivering a better, more informed care.
Improving the care, one patient at a time.
The algorithm also suggests different ways the health service can be improved. You can set it up to work out savings, lower the number of invasive tests or improve the speed or accuracy of the diagnosis.
Will it replace a doctor with a machine?
“Absolutely not. Frideswide is there to help the clinician. It takes away the tireless task of digging through the salt mines of data and gives you the answer quickly and accurately.”
– says Max Brzezicki, the first author of the paper.
Currently within the NHS, the principles of machine learning AI are used to identify common presentations, like diabetic eye disease or degeneration of the brain, where the algorithm aims to be better than the doctor reading the scan or interpreting the test result.
Frideswide, however, is there to help the doctor or a nurse and reduce their paperwork, so that they can spend more time at the bedside.
The future of research
The article describes a simulation experiment where the algorithm was used. Now, the researchers want to use it to deliver a better care in real people. A trial is currently underway to work out the efficiency of Frideswide in Southmead Hospital, Bristol.
You can read more about the algorithm in the paper: Frideswide – An Artificial Intelligence Deep Learning Algorithm for Audits and Quality Improvement in the Neurosurgical Practice