Health outcomes in aged care residents with dementia during the COVID-19 pandemic

The methods used in this report were adapted from methods described in the Technical notes of the report, Hospitalisations and deaths following COVID-19, 2020–2022: a linked data analysis (AIHW 2025b). Methods specific to this report are described below.

Scope of the COVID-19 Register

While community COVID-19 cases were tested and reported nationally, the additional responsibilities applied to aged care providers prompted more stringent surveillance in aged care facilities. For example, when Rapid Antigen Tests (RATs) were made available to residential aged care facilities in high-risk areas in August 2021, provider responsibilities included contacting the Public Health Unit to inform them of any positive RAT (Department of Health, Disability and Ageing 2021).

For more information, see Scope of the COVID-19 Register.

Considerations for interpreting report findings

For more information, see Considerations for interpreting report findings.

Sex

The Medicare Consumer Directory (MCD) in the COVID-19 Register was used as the primary data source for determining sex (recorded as 'male', 'female' or ‘missing’), followed by the Australian Immunisation Register (recorded as 'male' or 'female'), followed by the National Notifiable Diseases Surveillance System (recorded as 'male', 'female', ‘another term’, ‘not stated/inadequately described’ or ‘missing’). It is not known if the people completing the original administrative records interpreted ‘sex' to mean sex at birth or gender identity. 

Due to small numbers, results for people whose sex was recorded as ‘another term’ are not available.

This report uses the terms 'men and women' to mean 'male and female', but it should be noted that some people may not identify with these terms.

Place of residence

Place of residence data were primarily based on a person’s latest SA2 information in the Medicare Consumer Directory in the COVID-19 Register. It should be noted that the remoteness area or socioeconomic area of a person’s residential aged care facility may not be the same as the area/s they previously lived in.

Confidence intervals

To aid the interpretation of findings, 95% confidence intervals were included in hospitalisation and mortality data to show the range of uncertainty around each estimate (that is, that 95 out of 100 times the property being estimated will fall between the upper and lower values). A wider confidence interval band means greater uncertainty. In this report, when confidence intervals of two different estimates do not overlap, the difference between estimates is likely to be statistically significant. 

Data suppression 

Some data in this report have been suppressed or aggregated to avoid potential disclosure of an individual’s personal information. Suppressed data are indicated by missing data in figures or ‘n.p.’ (not published) in data tables.

The COVID-19 Register requires counts under 6 to be suppressed (or aggregated) to avoid disclosure of personal information. This complies with the principles of personal information disclosure in the Privacy Act 1988, the Five Safes Framework and the Guidelines for the Disclosure of Secondary Use Health Information for Statistical Reporting, Research and Analysis. This is also helpful to ensure analysis is based on counts that are large enough to make inferences from. Where groups have low numbers, there may be reliability issues when applying findings from the small group to a broader population.

Defining the study cohort

Identifying aged care residents with a COVID-19 diagnosis

This report focused on people aged 50 and over who had a COVID-19 diagnosis during a permanent or respite residential aged care episode between January 2020 and December 2022. 

Analysis was conducted for 2 time periods: pre-Omicron (25 January 2020 to 14 December 2021) and Omicron (15 December 2021 to 30 June 2022). 

Rather than looking at all COVID-19 diagnoses among people living in residential aged care, this report focused on hospitalisation and mortality outcomes after a person’s latest COVID-19 diagnosis while living in residential aged care up to the end of 2022. 

• During the pre-Omicron period, there were very few aged care residents who had more than one COVID-19 diagnosis recorded (counts were too low to publish).
• During the Omicron period, 6.4% of residents (3,610 people) with a dementia record, and 5.7% of residents (1,800 people) with no dementia record had more than one COVID-19 diagnosis recorded (Table S-A2).

The date of a resident’s COVID-19 diagnosis was used to assign them to a COVID-19 period, depending on the type of analysis. For the profile of residents with a COVID-19 diagnosis and COVID-related hospitalisation analyses, residents could be assigned to one or both COVID-19 periods. For example, a person who had a COVID-19 diagnosis during the pre-Omicron period and another during the Omicron period was included in analysis for both periods. For the mortality analysis, because there is only one record of death, the date of a resident’s last recorded COVID-19 diagnosis was used to assign them to a COVID-19 period.

For more information, see Identifying people with COVID-19 for methods used to construct the COVID-19 cohort.

Dementia records in the COVID-19 Register

Aged care residents were identified as having a dementia record if they had one or more of the following in the COVID-19 Register between 1 July 2010 and 31 December 2022:

• dispensed a dementia-specific medication through the Pharmaceutical Benefits Scheme (PBS, including Repatriation Pharmaceutical Benefits Schedule (RPBS) information)
• hospital admission with a principal, additional or supplementary code diagnosis of dementia
• emergency department presentation with a principal or additional diagnosis of dementia
• a dementia record in datasets within the National Aged Care Data Clearinghouse (NACDC):
  • Aged Care Funding Instrument (ACFI) assessment with a mental or behavioural diagnosis of dementia
  • National Screening and Assessment Form (NSAF) where dementia was recorded as a health condition impacting care needs
  • Dementia and Cognition Supplement for home care
• National Disability Insurance Scheme (NDIS) approved plan with a primary disability code of dementia
• dementia recorded as an underlying or additional cause of death up to 31 December 2022.

See Dementia in Australia Data sources for detailed information about codes used for analysis in each of these data sets.

As dementia identification is based on records in available administrative data sets, the ‘dementia record’ group does not include all people who were living with dementia in residential aged care. Conversely, ‘people with no dementia record’ is likely to include residents who either had undiagnosed dementia or had a dementia diagnosis that was not recorded in any of the available data sets.

There is emerging evidence that SARS-CoV-2 infection may result in cognitive decline in some cases (Quan et al. 2023). As there are currently no administrative data sets that include information on dementia diagnosis date, people were assigned to a dementia status group based on whether their first dementia record occurred prior to, or on the same day as, a COVID-19 diagnosis. If a person’s first dementia record was after a COVID-19 diagnosis within the study period, that person was excluded from analysis. This rule applied to the analyses based on a resident’s last COVID-19 diagnosis (hospitalisations and mortality after a COVID-19 diagnosis, COVID-19 vaccination and the use of antivirals). For the analyses based on all aged care residents (MBS, PBS and COVID-19 diagnoses by month and year), if a person had any record of dementia during the analysis period they were assigned to the dementia group.

People living in rural and remote areas are less likely to be identified as having dementia as they often access specialised aged care services (AIHW 2025c) for which data are not available in the COVID-19 Register (such as the Multi-Purpose Services Program and the National Aboriginal and Torres Strait Islander Flexible Aged Care Program). It should also be noted that dementia is often underreported on death records (AIHW 2025a; Xu et al. 2022).

Analysis methods

Measures of health outcomes

As this report focused on people in residential aged care, the analysis did not use Estimated Resident Population (ERP) data (which are often used in other reporting) to derive percentages. Rather, this report used more specific denominators that better reflected the cohorts of interest. The measures used to report health outcomes in this report are summarised in Table T1. 

All denominators were stratified by dementia status as appropriate, however the total residential aged care population was used as the denominator to calculate the percentage of residents dispensed a dementia-specific medication, rather than just people with a dementia record, as dispensing of these medications is often the first record in the linked data of a person having dementia.

Medicare Benefits Schedule data

The Medicare Benefits Schedule (MBS) data collection contains claims data for Medicare services subsidised by the Australian Government. This includes services provided by community-based health professionals, including general practitioners (GPs), medical specialists and allied health professionals.

People may receive primary care services in ways that are not captured in MBS data, including services delivered under Department of Veterans’ Affairs arrangements; services provided to public patients in hospitals; services provided under a state-funded service; dementia specific services such as the Dementia Behaviour Management Advisory Service or Severe Behaviour Response Teams; and services provided by a salaried GP or any other salaried medical officer arrangement. In-hospital MBS attendances were excluded from all analyses.

MBS outputs are by calendar year quarters rather than COVID-19 periods. Data from 2019 were included to show trends prior to the pandemic.

If a person had any record of dementia during the analysis period they were assigned to the dementia group.

A cut-off of 1 January 2010 was applied as the earliest date of entry for aged care episodes that were ongoing. Aged care episodes in Western Australia were excluded.

MBS attendances were calculated based on the 'number_of_services_provided' variable in the COVID-19 Register.

Attendances were only included if the ‘date_of_service’ was during a residential aged care episode (that is, >= entry date and <= exit date).

Attendances were counted by distinct values for all available variables: aihw_ppn_a, date_of_service, benefit_paid, broad_type_of_service, bulk_billing_flag, date_of_processing, fee_charged.

Codes used are shown in Table T2.

*Counts of telehealth attendances may not capture services where telehealth is allowed under the item descriptor but not included in the MBS Group or Subgroup name.

Pharmaceutical Benefits Scheme data

The Pharmaceutical Benefits Scheme (PBS) database contains medications eligible for government subsidy dispensed under the PBS. The PBS database also includes medications supplied under the Repatriation Pharmaceutical Benefits Scheme (RPBS, available for eligible veterans, war widows/widowers and their dependants). The PBS database does not contain data on the dispensing of privately prescribed medications, medications to public hospital in-patients and over-the-counter medications.

PBS outputs are by calendar year quarters rather than COVID-19 periods. Data from 2019 were included to show trends prior to the pandemic.

If a person had any record of dementia during the analysis period they were assigned to the dementia group. 

A cut-off of 1 January 2010 was applied as the earliest date of entry for aged care episodes that were ongoing. Aged care episodes in Western Australia were excluded.

The number of scripts were calculated based on the ‘number_of_scripts_dispensed' variable in the COVID-19 Register.

Scripts were only included if the ‘date_of_supply’ was during a residential aged care episode (that is, >= entry date and <= exit date).

Scripts were counted by distinct values for all available variables: date_of_supply item_code benefit_amount number_of_scripts_dispensed repeat_prescription_indicator patient_contribution_amount quantity_supplied_reg24_adj.

Anatomical Therapeutic Chemical (ATC) Classification

PBS items were mapped to the Anatomical Therapeutic Chemical (ATC) Classification, a classification system for medicines maintained by the World Health Organization. The ATC classification groups medicines according to the body organ or system on which they act, and their therapeutic and chemical characteristics. Medicines are given an ATC classification in the Schedule of Pharmaceutical Benefits according to their main therapeutic use in Australia as registered with Therapeutic Goods Administration and listed on the PBS. Codes used are shown in Table T3.

Hospitalisation data

This report focused on hospitalisation outcomes after a person’s latest COVID-19 diagnosis while living in residential aged care, by COVID-19 period, up to 30 June 2022. This approach was taken to better understand outcomes arising during this hospitalisation, including death. As previously noted, most residents only had a single infection in each period (Table S-A2).

This analysis did not look at hospitalisations with an acute respiratory code without a COVID-19 code.

The number of residents with a hospital admission involving continuous ventilatory support was excluded from this report due to low counts.

For more information, see Hospitalisations.

Mortality data

This report focused on mortality outcomes after a person’s latest COVID-19 diagnosis, by COVID-19 period and overall (total) while living in residential aged care up to 31 December 2022. This approach was taken to better understand the factors surrounding the most recent COVID-19 diagnosis which preceded death, including vaccine status, antiviral use, and care needs inferred from the Aged Care Funding Instrument (ACFI).

Previous COVID-19 diagnoses were not included in this analysis; therefore, this report presents the “percentage of residents with a COVID-19 diagnosis who died due to COVID-19”, using the number of aged care residents with at least one COVID-19 diagnosis as the denominator. This is different to a case fatality rate, which uses the total number of COVID-19 diagnoses as the denominator.

For more information, see Deaths.

Aged Care Funding Instrument data

Between March 2008 and September 2022, the Aged Care Funding Instrument (ACFI) was used a tool used to allocate funding in residential aged care and included data on day-to-day care needs, categorised as 'nil', 'low', 'medium', or 'high' based on responses to 12 questions across 3 domains: Activities of daily living, Cognition and behaviour and Complex health care. When assessed for funding purposes, people with high care ratings in a domain had more severe needs and require extensive assistance and care in that domain, whereas those with a low care rating had less severe needs. For more information, see How are care needs assessed using the Aged Care Funding Instrument.

This analysis explored hospitalisation and mortality outcomes among residents who had an ACFI assessment in the linked data dated on or before the date of their latest COVID-19 diagnosis. ACFI levels (see Table T4) were taken from the latest ACFI assessment available and were aggregated where necessary to ensure data confidentiality.

COVID-19 vaccination

For more information, see COVID-19 vaccination.

COVID-19 antiviral medications

For more information, see COVID-19 antiviral prescriptions.