"Revolutionizing Aged Care: How AI Predicts Health Risks Before They Happen"

By examining patient data, an AI-powered application created by RMIT University and Telstra Health forecasts health risks like mortality, depression, and falls.

Nearly 500,000 elderly Australians rely on the aged care system as their lifeline. 456,000 Australians were using aged care services as of June 30, 2023, according to the AIHW's GEN Aged Care Data, with 258,000 receiving home care help and 193,000 in residential care. However, this support system is under tremendous pressure to provide high-quality treatment in response to the rising demand. A novel AI-powered tool to forecast health hazards has been introduced by Telstra Health, RMIT University, and the Digital Health Cooperative Research Centre in order to tackle these issues head-on. This tool will revolutionise the way elderly care facilities monitor and manage the well-being of its residents.

According to Dr. Tabinda Sarwar, the project lead and data scientist at RMIT University, the tool is a much-needed improvement for a system that is already overburdened. According to Dr. Sarwar, "the tool can automatically monitor both structured and free-text electronic patient records for 36 evidence-based indicators of deterioration." "A comprehensive system to support nursing staff and improve resident care outcomes is provided by these indicators, which are further linked to predicting various health risks."

Nursing staff members are in charge of the daily needs and health monitoring of the senior residents at aged care facilities. Staff members have to manage several residents with various health concerns, which results in a heavy task. It is not the best idea to provide a manual screening tool in light of this current burden. RMIT and Telstra Health collaborated to create a data-driven platform that can forecast unfavourable health events in addition to tracking residents' health conditions.

Research Australia has named this digital health solution the 2024 Digital & Data Health Innovation Award winner. The crew is extremely grateful for this distinction. "It is an acknowledgement and accomplishment of improving the work of nursing staff and, consequently, the quality of life for senior citizens in assisted living facilities," Dr. Sarwar says.

Data gathered from aged care facilities on a daily basis is used by the tool. "This tool was developed based on the routine note-taking and health-related information recorded by nursing staff," Dr. Tabinda Sarwar continues. Through the use of natural language processing (NLP) tools, the program analyses this data to find early indicators of decline and creates prediction alerts for a range of health hazards. Progress notes, geriatric evaluations, and observation charts are essential data sources that guarantee a thorough approach to health monitoring. "We were able to anticipate indications of decline by applying sophisticated data analysis and machine learning methods to the daily information that was gathered," Dr. Sarwar says. "This includes risks like depression, falls, and even death, as demonstrated by the data that was extracted."

The central infrastructure for the project was provided by Telstra Health's Clinical Manager system, which is implemented at more than 360 sites around Australia. The partnership also involved assistance from the Digital Health Cooperative Research Centre (CRC) and feedback from elderly care nursing staff. Dr. Sarwar points out that RMIT contributed researchers and technological professionals to build digital tools and solutions, while Telstra Health made access to nursing staff and aged care facilities available. This initiative would not have been feasible if the Digital Health CRC had not played a crucial role. The Digital Health CRC, which brought together academia and industry, demonstrated the enormous potential of teamwork in utilising technology to tackle difficult health issues.

It wasn't easy to come up with a universal answer for various elderly care institutions. The instrument needed to be extremely flexible because nursing personnel in various homes brought to light particular difficulties. Accordingly, the most challenging aspect of the project was coming up with a solution that could deal with more general problems and significantly affect a wider population, she continues. In order to verify the tool's usability, the team carried out a separate investigation to make sure the created solution is easy to use and that nursing staff can readily adopt it.

We used machine learning and statistical methods to verify and assess the tool's accuracy and performance, which were crucial for guaranteeing its clinical viability. Additionally, machine learning models were used to anticipate deterioration, underscoring the importance of data mining and machine learning to the project's success.Telstra Health is now using the tool, and deployment is ongoing. Many senior living facilities have already indicated interest in implementing the technique, according to Dr. Tabinda Sarwar. The tool's rights are owned by Telstra Health, so any ambitions to increase its capacity to identify more health hazards will rely on their long-term plans.