3: Dementia prevalence and incidence
Monitoring the number of people living with dementia (prevalence) and the number of new cases of dementia each year (incidence) is required for policy development and service planning for health and aged care systems. Monitoring incidence over time indicates if the rate of diagnoses is increasing, decreasing or stable and whether certain population groups are more affected than others. However, this is difficult given the challenges of obtaining a timely diagnosis and lack of national general practitioner (GP) and specialist data with diagnostic information. See Timeliness of dementia diagnosis and Dementia diagnosis and management for more information.
Addressing the data gap and improving data
There is no single authoritative data source for deriving dementia prevalence in Australia, leading to varying estimates of how many people in Australia have dementia. Estimating dementia incidence is difficult because dementia is a gradual evolution of signs and symptoms rather than an acute event, and incidence requires information such as date of diagnosis and whether a diagnosis was made close to symptom onset (AIHW 2023a). There is no comprehensive source of information on dementia diagnosis date, meaning that dementia incidence is currently unmeasurable. National data on dementia incidence would also facilitate the estimation of dementia survival rates in Australia, which are currently not well understood. Accurate information on the number of Australians living with, and dying from, dementia in Australia is critical for monitoring trends and informing dementia policy and service planning (AIHW 2023a). Better data are also needed for the prevalence of different dementia types (for example, Alzheimer’s disease, vascular dementia) and severities, as well as people with mild cognitive impairment.
The Australian Dementia Network (ADNeT) Clinical Quality Registry, with its goal to register all Australians newly diagnosed with either dementia or mild cognitive impairment (ADNeT 2023), may be a future source of data that can contribute to estimates of dementia incidence at the population level. The ADNet Clinical Quality Registry is described in further detail at Dementia diagnosis and management.
Existing epidemiological studies on healthy ageing, especially when linked with existing administrative data that enable identification of people living with dementia, could also inform better estimates of dementia prevalence and incidence. For example, the Sax Institute’s Analysis of Population Traits and Risk Factors (ADAPTOR) study already links longitudinal data from over 200,000 participants from the 45 and Up Study with data on their use of hospitals, GPs, health services and prescription medication. Adding cognitive measures to existing studies could improve measures of dementia prevalence and incidence in Australia. There are challenges with this approach, as many existing surveys use volunteers which are not representative of the general population, where dementia risks and prevalence may be greater in those not participating in voluntary surveys (Brayne and Moffitt 2022).
Bringing together disparate data sources through data linkage would also help overcome some of the limitations associated with estimating the prevalence of dementia in Australia. Linked administrative data sets are currently being used to monitor dementia, and methodologies could be developed and refined using linked data to model estimates of dementia prevalence in Australia. However, data linkage is dependent on the capture of dementia and the quality of information provided in the individual data sets. At present, improvements are needed in available data sets and more types of data linked together. For dementia, available national data often capture people who have used a health, aged care or other type of support service, or people who died with a record of dementia, which skews towards capturing people who access services and/or who have more advanced dementia. Therefore, improvements in existing population-based studies and national surveys, or the development of new studies are needed to inform the estimation of dementia prevalence and mild cognitive impairment in Australia.
Proposed data improvement activities
There are 7 activities proposed to improve available data on dementia prevalence and incidence in the population. Some activities are directly focused on improving dementia prevalence and incidence data overall, while others are focused on improving or developing new data. Further, some activities are dependent on the completion of other activities in this plan.
Activities to improve dementia prevalence and incidence data include:
- incorporate dementia data in enduring linked data assets to meet the needs of dementia monitoring
- increase coverage of existing data for national dementia incidence monitoring
- investigate approaches to estimate the national prevalence of dementia
- assess new self-reported dementia data collected in the 2021 Census
- estimate dementia prevalence in priority population groups
- develop a Regional Insights portal of dementia data for local areas
- assess impact of dementia coding changes in ICD-11 when incorporated in the Australian health system.
These activities may involve single projects, or multiple projects to enable monitoring of trends over time, incorporate improvements in data or to focus on a specific population group. Each activity provides information on the intended outcome, priority rating, level of investment required, timeframe for completion of the activity and who is responsible for undertaking the activity.
Activities aimed at improving dementia diagnosis data in primary and secondary health care data would also improve dementia prevalence and incidence data. See Dementia diagnosis and management for additional activities.
There have been substantial developments and progress made with enduring linked data assets in recent years. The Australian Institute of Health and Welfare (AIHW) is aiming to expand an enduring linked data asset, which would allow approved projects to access data sets within the linked data asset using streamlined governance arrangements and pre-existing linkages of widely used health and aged care administrative data.
This activity involves the inclusion of dementia-focused data sources (such as data from Dementia Support Australia and the ADNet Clinical Quality Registry) in existing enduring linked data assets for dementia research and monitoring. Once the linked data asset is established and data governance and approvals are in place, dementia analysis would draw on data sets from the linked data asset including mortality data, hospital data, Pharmaceutical Benefits Scheme (PBS), Medicare Benefits Schedule (MBS), aged care, and any other relevant data sets which would improve dementia monitoring. Other relevant data sets could include data from the Department of Social Services Data On Multiple INdividual Occurrences (DOMINO), National Disability Insurance Scheme (NDIS), and Department of Veterans’ Affairs (DVA). This would allow for monitoring of people with dementia accessing welfare and disability support services, and services offered by DVA.
The enduring linked data asset could improve available data on a range of priority topics identified in the National Centre for Monitoring Dementia (NCMD) work plan, including culturally and linguistically diverse (CALD) status (for example, country of birth at a minimum), First Nations people, dementia among veterans, younger onset dementia, childhood dementia and rarer causes of dementia.
This activity relies heavily on improvements in existing national data infrastructure and governance arrangements, and ideally would be made available to all researchers.
Outcomes: Better diagnosis, prevalence, cost, services, transitions to aged care and outcomes data for the Australian national population and by priority groups.
Priority: High
| Alignment | Foundational | Dependency | Priority Group Data | Priority score (Maximum 10) |
|---|---|---|---|---|
| 4 | 3 | 0 | 1 | 8 |
- Alignment: highly important for multiple data gaps including National Dementia Action Plan (NDAP) objectives 4, 5, 6 and 7
- Foundational data: the linked data set would allow a breadth of dementia monitoring and research
- Dependencies: relies on linked data asset funding, infrastructure to support data linkage and appropriate governance, access and approvals from data custodians
- Priority group reporting: likely – would allow improved data on priority groups
Level of investment: Medium–High
The level of investment is dependent on the number of data sets to be linked and the resourcing involved to maintain an enduring linked data set. Work to include a single data source into an enduring linked data set would require medium investment. However, in total this is a high investment activity.
Timeframe: Short term for incorporation of single data set (<2 years); Long term (6+ years) to meet needs of dementia monitoring
Work to include a single data source could commence now and be completed in the short term, however, incorporating multiple data sources to suit dementia monitoring needs and maintaining an enduring linked data set are long-term activities.
Responsible stakeholder: AIHW NCMD
This activity involves investigating potential existing sources of dementia incidence data (such as the ADNeT Clinical Quality Registry, the ADAPTOR study and other epidemiological studies) and determining what work could be undertaken to estimate what coverage is required and how could this be achieved. This activity is focused on reviewing available data from existing studies and registries and recommends activities to expand sources to achieve nationally representative dementia incidence data. Following outcomes from this investigation, this plan would be updated to incorporate new activities.
Activities aimed at improving dementia diagnosis data in primary and secondary health care data would also improve dementia incidence data (see Dementia diagnosis and management for additional activities).
Outcome: Recommendations on activities to expand existing data sources to achieve nationally representative dementia incidence data
Priority: High
| Alignment | Foundational | Dependency | Priority Group Data | Priority score (Maximum 10) |
|---|---|---|---|---|
| 3 | 2 | 1 | 1 | 7 |
- Alignment: highly important to inform incidence monitoring
- Foundational data: needed to inform better incidence data
- Dependencies: relies on continued funding for existing registries and studies
- Priority group reporting: likely – however, this is dependent on data on priority groups available in existing sources
Level of investment: Low
The level of investment to undertaken initial work to investigate potential sources and what coverage is required to improve national estimates of dementia incidence is low. Work to expand existing data to improve capture of dementia incidence and monitor this on an ongoing basis would be a high investment activity.
Timeframe: Short term (<2 years)
Responsible stakeholder: AIHW NCMD; academic researchers
This activity involves a systematic investigation of different options (and combinations) to better estimate the prevalence of dementia. This would involve a review of international and national surveys and population-based studies used to estimate dementia prevalence and determine if these are appropriate for the current Australian context.
This activity would examine pre-existing nationally representative surveys, such as the ABS National Health Survey, and large population-based studies. The need for a new Australian population-based study, like the longitudinal Cognitive Function and Ageing Studies in the UK, would also be assessed if there are limitations with expanding or augmenting existing national surveys or population-based studies.
This assessment should consider and propose methodologies to sample participants and undertake cognitive screening to permit robust national estimates of the prevalence of dementia and mild cognitive impairment by key demographics. This activity should also consider priority populations and provide recommendations for collecting data in key priority groups. Completion of this activity would result in recommendations for estimating national prevalence of dementia through cognitive screening in population-based studies.
This investigation should evaluate the benefits of better prevalence data for reporting, policy and programs, and the anticipated benefits of screening or other methods to identify persons with dementia in a population that does not have a dementia diagnosis. Participants in such a study would need to be informed about the potential for a diagnosis of dementia according to the tool being used, and the risks of overdiagnosis harms associated with this (such as misdiagnosis) and consider any psychological harms of a diagnosis provided without a program of support offered in conjunction.
Following this investigation, this plan would be updated to incorporate new activities to implement the recommendations. Suitable data sources identified in this activity could also be used to compare and validate existing and emerging prevalence estimates from national enduring linked data sets.
Outcome: Recommendations to develop national estimates of dementia in Australia through population-based studies
Priority: High
| Alignment | Foundational | Dependency | Priority Group Data | Priority score (Maximum 10) |
|---|---|---|---|---|
| 3 | 2 | 2 | 1 | 8 |
- Alignment: highly important to improve prevalence data
- Foundational data: prevalence data
- Dependencies: investigation provides a pre-requisite to inform subsequent data collection
- Priority group reporting: likely – this study would inform data collection methods for priority groups
Level of investment: Low
The level of investment to develop a strategy for capturing national dementia prevalence through screening tests in population-based studies is low. Subsequent work to implement the activity would require high investment.
Timeframe Short term (<2 years)
Responsible stakeholder: AIHW NCMD in collaboration with input from academic researchers and experts in population-based studies.
The 2021 Census included a question on long-term health conditions including dementia for the first time. A current project under the NCMD aims to compare self-reported dementia records in the 2021 Census with dementia records in other data sets available in the Multi-Agency Data Integration Product (MADIP).
While there is potential that the 2021 Census data could capture people with mild to moderate dementia who have not yet had contact with hospital or aged care services (and thus be a valuable source for capturing additional cases), there is also potential that the public are not willing to share health information in the Census and lead to under-reporting. Analysis of the current Census data on dementia and other chronic conditions will inform the value of the Census data and any corrections to estimates required and the value for advocating for the inclusion of dementia as a long-term health condition in subsequent Censuses.
Outcome: Inform the use of Census data for monitoring dementia and estimating dementia prevalence in Australia; provide evidence to support the inclusion of a question on dementia in future Censuses
Priority: High
| Alignment | Foundational | Dependency | Priority Group Data | Priority score (Maximum 10) |
|---|---|---|---|---|
| 3 | 2 | 1 | 1 | 7 |
- Alignment: highly important gap
- Foundational data: may inform data methods and associations with self-reporting, and may allow comparisons with subsequent Censuses if this question is repeated
- Dependencies: data custodian and ethics approvals, repeating Census question
- Priority group reporting: likely
Level of investment: Low
Timeframe: Short term (<2 years)
Responsible stakeholder: AIHW NCMD (this activity is currently underway)
This activity involves establishing methods and implementing studies to estimate dementia prevalence in priority populations. This may involve undertaking new dementia prevalence studies, expanding existing studies to improve the capture of dementia or using linked data to estimate dementia in priority population groups (if considered appropriate). The data collection methods would need to be designed to be culturally appropriate, recognising that different cultural attitudes around dementia will require different approaches. Activities should be designed to maximise participation and data collection. For this reason, the activity would involve distinct projects for specific population groups.
Outcome: Improved methods and data to estimate prevalence of dementia among priority groups
Priority: High
| Alignment | Foundational | Dependency | Priority Group Data | Priority score (Maximum 10) |
|---|---|---|---|---|
| 3 | 2 | 2 | 1 | 8 |
- Alignment: highly important data gap
- Foundational data: may inform prevalence data
- Dependencies: minimal
- Priority group reporting: likely
Level of investment: Low–Medium
The level of investment is dependent on the type of activity being undertaken. Development of new methodologies using existing data would require low investment, but new studies would require more investment.
Timeframe: Short term (<2 years) to Medium term (2–6 years)
Responsible stakeholder: Academic researchers; organisations with experience in collecting data among priority population groups; AIHW (for linked data projects).
A Regional Insights portal for dementia would bring together dementia statistics across a range of topics at a local level, allowing users to easily access and compare dementia statistics for geographic areas. This would include estimates for dementia prevalence of dementia, health and aged care service use and other key statistics. This would be a similar design to the Regional Insights for Indigenous Communities (RIFIC) website. This activity could build on the current project exploring the geographical variation in health service use by people with dementia currently underway under the NCMD. This activity supports the National Health Reform Agreement commitment to regular reporting on services in regional, rural, and remote communities.
Consideration would be needed on how to best report dementia statistics across a range of topics at a local level and the quality of the data by smaller geographies, and among priority population groups. For example, for data on dementia among First Nations people this could be the incorporation of dementia data into the existing RIFIC website. This would support Priority Reform Four: Shared access to data and information at a regional level in the Closing the Gap Implementation Plan 2023. However, sufficient data needs to be available at the local level for this population group.
Outcome: Easy access to local geographic and comparator dementia statistics
Priority: Medium
| Alignment | Foundational | Dependency | Priority Group Data | Priority score (Maximum 10) |
|---|---|---|---|---|
| 2 | 2 | 0 | 1 | 5 |
- Alignment: moderately important data gap
- Foundational data: prevalence data at smaller geographies
- Dependencies: highly dependent on developments in availability of primary health care data to report on local geographies for data such as dementia prevalence, services, workforce
- Priority group reporting: likely – improves data on dementia in rural and remote populations and potentially on priority populations
Level of investment: Medium–High
Investment in the development and usability of a web-based portal is required. Maintenance of this website over several years would lead to this activity overall being a high investment activity.
Timeframe: Medium term (2–6 years)
Responsible stakeholder: AIHW NCMD; academic researchers; organisations with experience in collecting data among priority population groups.
The International Classification of Diseases (ICD) is a global health classification system and forms the basis for the recording of health data and production of health-related statistics. The ICD is revised every 10–20 years to ensure its currency and utility. ICD-11 represents a significant advance on the 10th revision of ICD, currently in use in Australia and internationally. In addition to updated scientific content, ICD-11 has been developed for use in electronic environments and is linked to other relevant classifications and terminologies (AIHW 2023b). The Australian ICD-11 Task Force was established in 2022 to develop a roadmap of activities regarding the implementation of the ICD-11 in Australia (AIHW 2023b).
This activity will monitor the implementation of ICD-11 in Australia and assess its impact on dementia coding across a range of contexts, including morbidity and mortality coding, as well as adaptations for clinical use in primary care, specialty care and research.
Outcome: Known impact of coding changes for dementia in ICD-11 and more rapid adaptations to existing routine monitoring of dementia using ICD coded data
Priority: Medium
| Alignment | Foundational | Dependency | Priority Group Data | Priority score (Maximum 10) |
|---|---|---|---|---|
| 2 | 3 | 0 | 1 | 6 |
- Alignment: moderately important data gap
- Foundational data: yes (interoperability)
- Dependencies: highly dependent on when and if the ICD-11 is implemented in Australia
- Priority group reporting: likely
Level of investment: Low
Timeframe: Short term (<2 years)
This activity could only commence after the ICD-11 has been implemented in Australia.
Responsible stakeholder: AIHW NCMD.
ADNeT (Australian Dementia Network) (2023) The Australia Dementia Network Clinical Quality Registry, ADNeT website, accessed 24 May 2023.
AIHW (Australian Institute of Health and Welfare) (2023a) Dementia in Australia, AIHW, Australian Government, accessed 1 August 2023.
AIHW (2023b) International Classification of Diseases revision, AIHW website, Australian Government, accessed 31 July 2023.
Brayne C and Moffitt TE (2022) ‘The limitations of large-scale volunteer databases to address inequalities and global challenges in health and aging’, Nature Aging, 775–783, doi:10.1038/s43587-022-00277.