Predicting early dementia using Medicare claims

The article was originally posted on LinkedIn by Fleur de Crespigny, Dementia Unit Head, AIHW. 

I’m pleased to announce that today my team has released a new feasibility study on predicting early dementia using Medicare claims, linked to other health and aged care data within the National Integrated Health Services Information Analysis Asset (NIHSI-AA).

Linked data provides an excellent source of information for the ongoing monitoring of dementia in Australia. This piece of work is geared towards building a better understanding of how many people there are with newly diagnosed dementia in the community and improving how we measure prevalence and incidence. The study builds on our current body of work, including our recent report Dementia in Australia, to help fill known data gaps.

A lack of available diagnosis information in primary and secondary care data means that early dementia is difficult to capture in dementia prevalence estimates.

Currently, a diagnosis of dementia can be identified in hospital admitted patient care, mortality, aged care and Pharmaceutical Benefits Scheme (PBS) data. However, these sources tend to be best at identifying people with later stages of dementia, as disease progression leads to more frequent contact with various parts of the health and aged care systems.

This study takes a step towards completing the picture of dementia in Australia by enabling earlier identification of dementia in linked data using Medicare Benefits Schedule (MBS) items.

Importantly, it demonstrates the feasibility of training algorithms to recognise dementia cases from MBS service history in the two years leading up to being prescribed and dispensed dementia-specific medications for the first time. For instance, geriatrician services, brain scans and chemical pathology all emerged as being highly associated with a recent dementia diagnosis.

Congratulations to my team for delivering such an important resource, particularly Lisa Irvine and Bronwyn Wyatt who lead this work. Linking administrative data is helping overcome data limitations associated with determining the prevalence of dementia, and will assist us to deliver a cutting edge dementia monitoring program over the longer term as well as produce targeted, policy relevant reports on specific health and welfare issues for people with dementia.

To see more of the work we are doing on dementia, please explore some of the recent reports produced by my unit at the links below:

Previous article