February 1, 2024  Print

Australian researchers have developed a world first test which aims to identify which COVID-19 patients will need urgent, lifesaving, medical treatment.

The test, developed by teams of researchers from The Westmead Institute for Medical Research (WIMR), the University of Sydney Infectious Diseases Institute (Sydney ID) and the University of Queensland, needs a few drops of blood to predict, with near-perfect accuracy, which COVID-19 patients will develop potentially deadly secondary bacterial infections, such as pneumonia.

Secondary bacterial infection is the number one reason people deteriorate after COVID-19, and this usually occurs in the first seven days to four weeks after diagnosis.

The study that led to the development of this test has been published today in leading journal, Lancet Microbe.

Lead author of the study, WIMR’s Dr Maryam Shojaei said, “The current method to detect secondary bacterial infection is unreliable and time-consuming.  It takes up to three days as it requires examining microbiological culture (namely, growing the bacteria on a petri-dish) in a hospital laboratory.

“This new method overcomes these limitations. It can provide front-line doctors with time-critical information within a matter of hours and with a much higher accuracy (more than 95%).”

Dr Shojaei explained, “It works by eavesdropping on the immune system. It picks up the signal released by the body’s immune system on the first sign of a bacterial infection.

“By picking up this early warning signal, the test can work faster than the conventional microbiological culture method. This gives doctors a better chance to get on top of patient’s infection before it overwhelms them and potentially kills them.  It is a great example of Precision Medicine at work.”

The secret behind the test’s accuracy lies in the use of advanced artificial intelligence (AI) in its development.

Dr Meagan Carney from the University of Queensland who is first author of the study explains that the AI algorithm is the secret that underpins the test performance.

“We deploy machine learning, which is a form of AI, to look for signatures hidden in the millions of copies of RNA (sister copies of DNA) circulating in the patient’s blood,” said Dr Carney.

The study represents another successful Australian story in the battle against the COVID-19 pandemic. Originally founded by Dr Benjamin Tang in early 2020, the study assembled teams of scientists and doctors from across the globe (Australia, Southeast Asia, Europe, Middle East and USA) to develop cutting-edge technology to better diagnose COVID-19 patients. Now in its fourth year, the consortium has already published several landmark scientific papers.

Dr Shojaei added, “We hope this study will lead to the development of a simple blood test that can help during future pandemics with other viral infections. This type of test will be extremely valuable for the triage of patients when hospitals are running near or beyond capacity.