About the data
This year’s datathon will use data from Uniting NSW.ACT, one of Australia’s largest not-for-profit providers of aged care and community services. The data provides an opportunity to explore how analytics and machine learning can help identify emerging risks and support better outcomes for older people.
The context
Uniting supports almost 8,000 older people across Residential Aged Care homes in NSW and the ACT. The quality and safety of aged care services are overseen through Australia’s national aged care regulatory framework, including monitoring and assessment by the Aged Care Quality and Safety Commission.
Residential aged care homes are assessed using quality, compliance, staffing and experience measures. Together, these measures provide important insights into the quality of care and the experiences of people living in aged care.
The Star Rating System
Star Ratings help older people, their families and carers understand and compare the quality of residential aged care homes. Ratings are published through the Australian Government’s My Aged Care service.
- Compliance (30%): Reflects a home’s performance against regulatory requirements and relevant compliance decisions.
- Residents’ Experience (33%): Reflects feedback from people living in aged care about the care and services they receive.
- Staffing (22%): Considers care delivered by registered nurses, enrolled nurses and personal care workers.
- Quality Measures (15%): Uses clinical indicators such as pressure injuries, restrictive practices, unplanned weight loss, falls and medication management.
Together, these measures provide a broader picture of the quality and safety of care.
The challenge
How can we use data to identify the early signals that a Residential Aged Care home may need additional support?
Uniting’s Risk-Based Support Model (RBSM) is an early-warning and decision-support framework that brings together indicators across five areas: Clinical, Customer, Quality, Operational and People. These indicators provide a broader view of where a home may require additional attention or support.
Currently, each indicator contributes equally to the overall risk assessment. However, not every indicator may have the same relationship with future quality outcomes. Some may provide stronger early signals, while areas containing more indicators can have greater influence simply because they contain more measures.
Your challenge
Using historical data, your team will explore evidence-based and explainable approaches to improve how RBSM indicators are weighted.
- Identify which indicators provide the strongest early signals of changes in future quality outcomes.
- Assess whether some indicators should have greater influence than others.
- Explore how indicators can be combined into a more meaningful overall risk assessment.
- Test whether these relationships remain reliable across different homes and over time.
- Keep the resulting approach simple, transparent and explainable.
The objective is not simply to build the most accurate predictive model. The goal is to develop a practical and explainable approach that can identify emerging risks earlier, support better prioritisation and help Residential Aged Care homes receive the right support sooner.