How this aged care provider uses AI to prevent hospital admissions

IRT (the Illawarra Retirement Trust) has developed an AI falls predictor system, which uses data about aged-care clients - such as medications, sleeping habits, their health and their physical environment - to assess their risk of a fall.

The client and IRT can then use that information to put in place strategies to prevent a fall, or prevent injury if they do fall, and subsequently keep them independent and living in their own home for as long as possible.

Falls have the potential to cause life-changing injury in older people.

IRT chief executive officer Patrick Reid said people who suffered serious injury were often hospitalised for a long time and went home with decreased mobility, which affected their physical strength and their social and community engagement.

A serious fall can also lead to death: falls contribute to the deaths of 1200 older people in NSW each year.

Using 100 tested mathematical models, the IRT team looked at how to identify those home care clients at risk of a fall and over time refined the system, which returned a list of people who might need help.

IRT home care clinical and quality adviser Raelene Lewis - who Mr Reid said came up with the idea originally - then began to look at what could be put in place to prevent a fall.

"It could have been medication interventions, it could have been environmental interventions like getting rid of rugs, making the floor less slippery, getting ramps instead of stairs - all the usual things that you can do within home care," Mr Reid said, or strategies to improve physical fitness and strength.

He said the predictor "worked really well", with marked improvements in the rate of falls.

Ms Lewis said the proportion of clients who had a fall each month dropped from 7 per cent to 4 per cent, which meant about 30 people each month were saved from having a fall when they otherwise might have.

"For those individuals that's a huge difference, because we know when older people fall they're more likely to be hospitalised, they're more likely to have poorer health outcomes and some, particularly if they have a fracture or a long-term hospitalisation, it's more likely to result in their death," she said.

Ms Lewis said the system empowered customers to make choices about their care and services they received, and how they used their home care funding.

IRT's executive general manager of IT, John Vohradsky, said a unique quality of the project was that it took a proactive rather than reactive approach to a problem.

"All technology to date in aged care has been reactive... So the fall's happened, how do we now have an outcome?" Mr Vohradsky said.

"But to be able to predict it and get ahead of the fall is what the difference is."

It was also a hidden technology, he said, as opposed to something that was intrusive.

Ms Lewis said the system had also had a positive impact on staff, boosting their knowledge and ability to offer solutions for clients.

The system has collected two IT awards, claiming not only the 2024 iTnews Award for best health project, but also Australian Technology Project of the Year, which encompasses all industries.

IRT is looking to expand the falls predictor to residential aged care, and is working on similar tools for wound care and medications.

The organisation has also recently won an Innovation and Technology Across Care excellence award for a project within residential aged care that alerts nurses when sensors on walls detect a person has had a fall.