Top Menu

New system to use predictive analytics to improve care

RMIT and Telstra Health partner to improve aged care data analytics (clockwise from left) Sarah Evans, Michael Donnelly, Professor Lawrence Cavedon, Larissa Briedis and Professor Irene Hudson

A research and vendor collaboration is developing software to predict the deterioration of aged care residents’ health to improve care planning and reduce avoidable hospital transfers.

The clinical decision support software is being developed for Telstra Health’s residential aged care software suite through a $1 million partnership announced on Tuesday between RMIT University, Telstra Health and the Digital Health Cooperative Research Centre.

The software will analyse clinical data of aged care residents for signs of deteriorating health to help providers plan and direct staffing and clinical resources more effectively and reduce unnecessary emergency hospitalisations.

Once developed, the predictive algorithms will be embedded directly into Telstra Health’s software, which is being used for more than 55,000 residents across Australia, said Larissa Briedis, senior strategy specialist at Telstra Health.

“The project will leverage data from existing residential aged care providers using Telstra Health’s Clinical and Care Management software to develop algorithms to provide advanced indications of deteriorating resident condition,” Ms Briedis told Australian Ageing Agenda.

“Other complementary data sources may be identified and leveraged in the course of the project.”

The research aims to identify clinically validated measures to reliably predict deterioration, she said.

“This will enable care staff to adjust the care of the residents accordingly together with reducing avoidable emergency hospitalisations,” Ms Briedis said.

In addition to keeping more aged care residents out of avoidable emergency care, the project has enormous potential to provide earlier indications when residents are approaching end of life, said Digital Health CRC CEO Dr Victor Pantano.

“The earlier we can ascertain that an aged care resident is approaching end of life, the earlier we can enact their advance care plan and honour their preferences – an important process for the aged care resident, their carers and families, and the aged care provider,” Dr Pantano said.

Researchers working with gerontologists, aged care staff

Clinical decision support software to predict deterioration is already used in acute care settings, but this will be the first for aged care.

The research team will work with gerontologists and aged care staff to interpret historical data and develop new predictive analytics techniques, as well as adapting existing decision support methods from the acute care sector, said Lawrence Cavedon, a professor of computer science at RMIT.

“Researchers will work closely with clinicians to understand reliable signs of patient deterioration, how this might be identified from recorded data, and to manage any related ethical issues,” Professor Cavedon said.

The new algorithms will first be tested using historical data then applied to current data in a trial setting before ultimately being integrated into Telstra Health’s software.

Telstra Health has begun approaching providers using its residential aged care software to participate in the research and trial, Ms Briedis said.

“Telstra Health and RMIT will announce participating providers as they come on board,” she said.

Analysis with participating aged care providers is expected to commence in early 2020.

Comment below to have your say on this story

Subscribe to Australian Ageing Agenda magazine and sign up to the AAA newsletter

Tags: , , , , , , ,

, , , , , , ,

One Response to New system to use predictive analytics to improve care

  1. Ian MacDonald October 24, 2019 at 9:51 am #

    As wonderful as this sounds, I hate to be devil’s advocate. If this is a good as the promoters suggest, what happens to these skills ” gerontologists and aged care staff interpret historical data and “ clinicians {who} understand reliable signs of patient deterioration,..” As these are skills that make these people capable to work in this field, it still remains that data is collected by humans about human condition.
    Observation will never go away. Where some can correlate well and have this amazing capacity – learnt over years of study and observation well .. electronics dispenses with this skill set. This particular task degenerates to a lesser employee on less pay(no offence to those people – I am one). Extrapolating that further, this reduces the need of aged care providers to pay people who have sacrificed their time and family life and gone without an adequate income to pay the mortgage to get these, once necessary abilities. I think it’s part of a race to the bottom.

    Sure, use the technology, I’m not saying not to. In effect though, the analysis of a deteriorating condition is unique to each human body and therefore demands a unique, individualised approach to each person – not a blanket algorithm.

Leave a Reply