Back to Articles
Machine Learning Considered to Address South Australia's Ramping Crisis

The Canberra Times

SKIPPED

Details

Date Published
10 Jan 2024
Priority Score
2
Australian
Yes
Created
8 Mar 2025, 02:41 pm

Authors (1)

Description

Machine learning technology has been floated as a potential solution for South Australia's ramping crisis after a review...

Summary

The article reports on the potential use of machine learning to improve resource allocation and address ambulance ramping in South Australia, as recommended by an independent review conducted by Dr. Bill Griggs and Professor Keith McNeil. The review was prompted by allegations of inappropriate prioritization of ambulance patients, although it found no evidence of such practices. The state government plans to implement the review's recommendations, which include leveraging AI for better resource management across departments. This suggests an innovative application of AI in public health policy, though it primarily focuses on resource efficiency rather than directly addressing catastrophic AI risks.

Body

Machine learning technology has been floated as a potential solution for South Australia's ramping crisis after a review found no evidence clinicians were being pressured to treat ambulance patients ahead of sicker patients in waiting rooms. The state government commissioned the independent report in December after Dr David Pope, an emergency specialist doctor and president of the South Australian Salaried Medical Officers Association, alleged health bureaucrats were intimidating hospital staff to inappropriately prioritise ambulance patients to reduce ramping statistics. The ambulance union then made a counter-claim in January that waiting room patients were being prioritised over ambulances after a 54-year-old man known as Eddie died after waiting more than 10 hours for an ambulance. But the report's authors, Dr Bill Griggs and Professor Keith McNeil, on Thursday said they could find no evidence to support either claim after an "exhaustive" examination. Emergency room data did show a trend for non-ambulance arrivals to be seen quicker than those conveyed by ambulance, but there was no evidence this was a result of inappropriate staff directions. "We found absolutely no evidence that practices had been enacted, or that data was being manipulated, to present a more palatable view of the ambulance ramping situation," the report said. Premier Peter Malinauskas said the government would implement all five of the review's recommendations, including strengthening the words of relevant policies and conducting a review into organisational safety culture. Dr Griggs said communicating across departments was holding the system back. He suggested using artificial intelligence to better allocate resources across information "silos". "There is a potential here to use machine learning, there's potential here to use people who have a very good understanding of flow and logistics," Dr Griggs said. He pointed to the airline industry as a model for how data could be used to more effectively manage staff and resources. While Dr Pope criticised the review's scope for being too narrow, he applauded its recommendations. "It's pleasing that the reviewers have come up with these and hopefully this allows us to go forward and fix a lot of the issues that are arising," Dr Pope told reporters. The Ambulance Employees Association also welcomed the review's findings and vowed to work collaboratively to implement its recommendations. But opposition health spokeswoman Ashton Hurn said it was an "absolute joke" that Dr Griggs and Prof McNeil only spoke to about 15 clinicians face-to-face as part of their consultation. "The doctor's union had a survey which came out on Monday that interviewed 50 people," she said. "What we know is that 90 per cent of doctors who responded to that survey indicated that they had witnessed this pressure. "I don't think this gives the average person listening at home any confidence there's no pressure within our health system." Australian Associated Press