Scientists at CSIRO, Australia’s national scientific organization, have led research to develop a machine learning tool that will give health professionals early warning of a deteriorating patient condition.

The study, published in Nature: Scientific Reports, in collaboration with Princess Alexandra Hospital and Metro South Health, demonstrates that early warning deterioration alerts can be set to monitor patients two to eight hours before they are triggered by current clinical standards. showed.

Dr. Sankalp Khanna, a CSIRO scientist, says medical professionals use data contained in electronic medical records (EMRs) to detect when a patient’s vital signs, such as blood pressure and temperature, reach danger zones, causing decline in the patient. He said he was able to predict the timing.

The large amount of data in EMR creates the potential for better patient care. For example, information from data can be used by medical staff to make decisions that prevent patients from becoming worse due to adverse events or acute illnesses. Until recently, patient data was still not available electronically at some hospitals, limiting their ability to develop and leverage digital tools.

“Until now, there has been no way to use all the EMR data to predict patient health. This new tool has the potential to transform the day-to-day functioning of the healthcare system,” said Dr. Khanna. said.

“When we applied the tool to a test cohort of 18,648 patient records, we found that the tool was able to We have achieved 100% sensitivity.

“Our scientists have the expertise to transform data into useful information to guide clinical choices. The new tool also sets out reasons for caution that can guide intervention choices.” .

“Alert alerts medical staff when a patient is at risk of deterioration that could lead to death, cardiac arrest, or unplanned admission to the ICU. can be notified.

“Clinical decision support tools like these are preemptive solutions that can provide medical staff with the opportunity to intervene early to prevent patient outcomes,” he said.

Dr. David Cook, Intensive Care Unit Staff Specialist in the Intensive Care Unit at Princess Alexandra Hospital, said this work is a truly useful and workable method for managing unanticipated patient deterioration across a large hospital. says there is.

“It’s done without duplication of process, nor does it interfere with the established best-practice systems used to recognize sick and deteriorating ward patients,” Dr. Cook said.

CSIRO scientists are currently discussing clinical trials with partners to investigate how alerts work and how they can be best implemented in clinical workflows.