Skip links and keyboard navigation

Queensland clinicians testing new tool to help detect sepsis

Queensland’s health experts are testing a new artificial intelligence (AI) tool designed to help detect patients with sepsis, an illness that threatens the lives of up to 21,000 Queenslanders a year.

Sepsis can be triggered by any infection. It could be as simple as an infection from a mosquito bite, wound or the flu that leads to sepsis, a condition caused by an abnormal response of the body to an infection that damages organs and tissues. This can lead to organ failure as your body goes into septic shock.

Queensland Health’s Deputy Director General Clinical Excellence Queensland, Dr Jillann Farmer, said because sepsis can be difficult to diagnose, testing was underway to identify new methods to aid detection.

“There’s no single symptom of sepsis and there’s no diagnostic test. It can initially look like the flu, gastro, or a typical urinary, skin, or chest infection,” Dr Farmer said.

“Symptoms can vary from person to person and are different for adults and children, and may include feeling very hot or cold, aching muscles, fast breathing, or vomiting.

“If sepsis is identified and treated quickly, patients are more likely to have better outcomes. This is why we are keen to look at new ways to identify sepsis quickly and improve the health of Queenslanders,” she said.

Queensland Health and Townsville Hospital and Health Service are now testing cutting edge technology in the form of artificial intelligence algorithms.

Townsville University Hospital and Clinical Excellence Queensland are collaborating with the Clinical Excellence Commission New South Wales to use data logged in the Integrated Electronic Medical Record (ieMR) in digital hospitals to help clinicians with the early identification of sepsis in patients.

Four hospitals in Sydney, including the Westmead Hospital, were the first in Australia to deploy the artificial intelligence tool.
Townsville University Hospital was chosen as the Queensland site to support the testing of the artificial intelligence tool due to a special interest by local clinicians in the illness.

Adult Sepsis Clinical Co-Lead for Queensland Health, Dr Paul Lane, from Townsville University Hospital, said the testing involved developing a predictive algorithm for Queensland to scan the information inputted into ieMR.

“The algorithm will use natural language processing technology to examine the notes of clinicians, looking for keywords and health indicators that may predict the onset of sepsis,” Dr Lane said.

“Quick treatment relies on people being aware of and recognising the symptoms. If there is a risk of sepsis, an alert will be sent to the treating team, enabling treatment to be implemented quickly and providing enhanced opportunities to improve patient outcomes.

“The working group at Townsville University Hospital is excited to be a part of the project and look forward to seeing the testing results,” he said.

Testing of the artificial intelligence tool is due to conclude late 2021 and the project team will then consider the results.

For the signs and symptoms of sepsis visit:

Last updated: 24 March 2021