A newly released feasibility study from the Australian Institute of Health and Welfare (AIHW) has uncovered a newfound capacity to predict the prevalence of Australians experiencing the earliest stages of dementia.
Using data algorithms to assess Australian healthcare data, the study found that Medicare items associated with geriatrician attendances, imaging of the head, and some pathology had strong predictive value of a later dementia diagnosis, especially where a patient has a combination of these items in their recent medical history.
When deployed, the models were found to have an accuracy of 80 per cent, accurately identifying 25,000 cases of early-stage dementia.
It is a step forward in improving government services relating to the early stage of the condition.
Up until now, rates of dementia in the country could only be confirmed once a patient recorded dementia-related prescription medication use, hospital admission or death.
Maree McCabe AM, chief executive officer of Dementia Australia, welcomes the research initiative, telling Aged Care News that this improvement in data analytics will allow great improvement of policies and support services.
“For Dementia Australia, up-to-date data is invaluable,” she says.
“It ensures we have a correct view of the scale of dementia in Australia and gives us a better sense of the real impact it is having on so many Australians, from across all backgrounds.”
The AIHW notes in the report that further work will be done to tighten the accuracy of the machine-learning based technology, for the purpose of:
• exploring further individuals who are misclassified due to data gaps (such as missing hospital information, alternative services streams such as Aboriginal Health Service use)
• assessing similarity between dementia diagnosis pathways and other conditions
• identifying individuals who did not access the PBS during the period of time assessed but are captured in other data sets in later years.