Development and validation of a frailty index based on Australian Aged Care Assessment Program data

Area of Science:

• Gerontology
• Public Health
• Biostatistics

Background:

• Frailty is a key determinant of health outcomes in older adults.
• Existing frailty measures may not be readily applicable to large-scale administrative datasets.
• Developing a robust frailty index from existing aged care data is crucial for research and planning.

Purpose of the Study:

• To develop and validate a frailty index using data from the Australian aged care eligibility assessment program.
• To assess the predictive validity of the developed frailty index for mortality and permanent residential aged care entry.

Main Methods:

• Retrospective cohort study utilizing the Registry of Senior Australians (ROSA) dataset.
• Analysis of 903,996 non-Indigenous Australians aged 65+ assessed between 2003-2013.
• Development of a 44-item frailty index and evaluation of its predictive accuracy using Cox proportional hazard models.

Main Results:

• The mean frailty index score was 0.20, with an increasing proportion of older adults exceeding a score of 0.20 over time.
• Higher frailty index scores significantly predicted increased risk of death and permanent residential aged care entry.
• The frailty index demonstrated good predictive validity (AUC 0.63-0.64) for one-year mortality and aged care entry.

Conclusions:

• A validated frailty index can be derived from Australian aged care eligibility assessment data.
• This index holds potential for use in aged care research within Australia.
• Further investigation is needed to determine its utility in aged care planning and policy development.