Predicting resident deterioration and acute care needs in aged care

Project title: Predicting resident deterioration and acute care needs in aged care

Project dates: ​30/9/2019 – 31/12/2022

Grants and funding: ​Digital Health Cooperative Research Centre and Telstra Health

Description

The goal for this project is to better equip aged care workers to identify deteriorating residents, so that appropriate care and resources can be provided.  This project will produce an algorithm for use in Telstra Health’s aged care software that will:

a. Identify deteriorating residents so that clinical and other resources can be adjusted accordingly, and to reduce avoidable emergency hospitalisations; and

b. Identify deteriorating residents approaching end of life, so that goals of care can be reviewed as specified in the Advance Care Directive (if completed), in collaboration with the resident and carers; and 

c. leverage/adapt existing algorithms and decisions support methods from the acute sector.

Research strategy

This project will leverage from Telstra Health’s residential aged care system and data collected by it to develop algorithms to provide advanced indication of deteriorating condition. Advanced Data Mining and Analytics methods will be developed and applied to support automated decision-making processes for early detection of conditions potentially requiring acute care intervention.

Rationale

With an ageing population, there are an increasing number of people receiving aged care. Additionally, an increasing number of those who receive care in a residential facility require acute care, and therefore higher needs for medical and other health services.  There is also an increasing community expectation to receive care in the home where possible, and for aged care to be of a high standard. 

As a result of these shifts, there is increasing pressure on aged care providers to deliver and demonstrate quality care. This is a challenge for many care providers, in a setting that has been de-medicalised for some years, and where clinical information is often gathered manually and in formats that are difficult to interrogate.   

Key people

• Lawrence Cavedon
• Irene Hudson
• Xiuzhen (Jenny) Zhang
• Jeffrey Chan
• Vural Aksakalli