Assigning demographic variables and region of residence
Age and sex
The cohort included people aged 30 and over. Younger people were out of scope for this study because the number of people aged under 30 living with dementia (including childhood dementias) would be too small to analyse by smaller geographies, and people are likely to use different services.
Age at the first service event in 2019 was calculated from the "age at event zero" variable in the NIHSI patient demography file (95% of the cohort had a service event in January 2019).
Sex was derived in the following order of priority: from the NIHSI patient demography file, or the last service event in 2019, or the latest service event in the NIHSI.
Geographical region of residence
This report presents information on people living with dementia based on where they lived, not where they received services. People are likely to have used services outside of their region of residence.
To ensure comparability across regions, data were only analysed for people living in states/territories with hospital data in the NIHSI 2020–21: NSW, Victoria, Queensland, South Australia, Tasmania and the ACT. Similarly, people who only had a dementia record in a private hospital service were excluded because private hospital data were not consistently available for all states/territories.
The smallest geography in the NIHSI is Statistical Area level 2 (SA2) (ABS 2016). Region of residence was based on a person’s latest recorded SA2 region or postcode in the NIHSI, either in the analysis year of 2019 or, if they had died, their death record in January 2020–December 2021.
People living with dementia may need to change their residence to be nearer to family and carers, or to receive health and/or aged care services. The SA2 regions and postcodes of residential aged care facilities were not available in the NIHSI, so residence was taken from other sources. It is known that people often do not update their address in the Medicare system (used for PBS and MBS data), so these sources were considered less accurate than deaths or hospitals data.
When a person had records in more than one dataset, the latest recorded SA2 region or postcode was taken using the following order of priority:
- National Death Index (for those in the cohort that died between January 2020 and December 2021)
- Hospital data (NHMD and NNAPEDC) (2019)
- Pharmaceutical Benefits Scheme (2019)
- Medicare Benefits Schedule (2019)
- Postcodes were only used if there was no record of SA2 region. Postcodes were allocated to the SA2 region where the largest percentage of the population lived.
Data are presented at four levels of geography:
Data are only presented for people living in NSW, Victoria, Queensland, South Australia, Tasmania and the ACT. Western Australia and the Northern Territory were excluded because their hospital data were not available in the NIHSI. SA2s with state code ‘Other territories’ (Christmas Island, Cocos (Keeling) Islands, Jervis Bay and Norfolk Island) were excluded from analysis due to small counts and privacy concerns.
SA4 regions are 76 geographic areas covering Australia, with boundaries defined by the ABS (2016). This geography was chosen as the smallest area that had adequate counts of people living with dementia for data analysis. Data were not able to be released from the NIHSI at both SA4 and Primary Health Network level due to data confidentiality concerns with small areas where their boundaries do not align.
Based on the 2016 Remoteness Area Structure within the Australian Statistical Geography Standard (ABS 2018a). The 5 Remoteness Areas are based on a measure of relative access to services: Major Cities, Inner Regional, Outer Regional, Remote and Very Remote Australia. In this report, results for Remote and Very remote Australia were aggregated to ensure confidentiality and quality.
Remoteness areas do not necessarily correspond with SA2 boundaries and one SA2 can cover several remoteness areas. In this report, SA2s were allocated to the remoteness area where the largest percentage of the population lived, using the population distribution as of 30 June 2016.
Based on the 2016 Socio-Economic Indexes for Areas (SEIFA) Index of Relative Socio-economic Disadvantage (IRSD) (ABS 2018b). The IRSD ranks statistical areas from the most disadvantaged area (lowest quintile) to the least disadvantaged area (highest quintile), based on the relative socioeconomic conditions at an area or 'neighbourhood' level, not at an individual level. People were assigned an IRSD quintile based on the SA2 of their residential address. The small number of people living in an SA2 with no corresponding SEIFA score were excluded from the socioeconomic area analyses.
The majority (71%) of people in the dementia study cohort lived in Major cities, which are socioeconomically diverse. Furthermore, socioeconomic effects can be compounded by access issues for people living in rural and remote areas. For this reason, data are also presented for IRSD quintiles within Major cities and Outside major cities (Inner Regional, Outer Regional, Remote and Very Remote Australia).
ABS (Australian Bureau of Statistics) (2016) Australian Statistical Geography Standard (ASGS): Volume 1—Main structure and greater capital city statistical areas, July 2016, ABS, Australian Government, accessed 30 September 2022.
ABS (2018a) Australian Statistical Geography Standard: Volume 5 – Remoteness Structure, July 2016, ABS, Australian Government, accessed 30 September 2022.
ABS (2018b) Census of Population and Housing: Socio-Economic Indexes for Areas (SEIFA), Australia, 2016, ABS, Australian Government, accessed 30 September 2022.