This page presents information on the number and characteristics of dementia-related episodes of admitted patient care (referred to as 'hospitalisations') from the National Hospital Morbidity Database, including hospitalisations due to dementia in 2021–22 by:
- age and sex
- sex between 2012–13 and 2021–22
- the type of dementia recorded
- how the hospitalisation ended and whether palliative and other end of life care was provided
- geographic and socioeconomic areas.
Each hospitalisation is assigned a principal diagnosis (the main reason for being admitted to hospital) and can also be assigned one or more additional diagnoses (conditions that impact the provision of care but are not the main reason for being admitted to hospital). This page largely focuses on hospitalisations with a principal diagnosis of dementia, or 'hospitalisations due to dementia'. See Hospitalisations due to dementia versus hospitalisations with dementia for information on hospitalisations with dementia recorded as an additional diagnosis.
See Box 9.1 for key terms and considerations for hospitalisations for people with dementia and the Technical notes for further information on hospitals data.
This section aims to provide a comprehensive picture of the impact of dementia on Australia’s hospital systems. The following terms are used to distinguish dementia hospitalisations:
Hospitalisations due to dementia are hospitalisations where dementia was recorded as a principal diagnosis (the main reason for admission).
Hospitalisations with dementia are hospitalisations where dementia was recorded as a principal diagnosis and/or an additional diagnosis (where dementia impacted the hospitalisations but was not the main reason for admission).
It is important to note that the statistics presented here cannot be considered a full count of hospitalisations among people with dementia for various reasons including:
- Hospital records only include conditions that were significant in terms of treatment, investigations needed and resources used during the ‘episode of care’. This means that hospitalisations among people with mild dementia may be under-recorded because the early stages of dementia are less likely to affect the care provided in hospitals.
- Since 2015 dementia may be recorded in hospitals data using ‘supplementary codes’ rather than as an additional diagnosis (especially when dementia was not strongly impacting the care received by the patient) (see Box 9.3 for more information). Supplementary codes are newly available in the data, further exploration of these codes is currently underway.
Further, specific dementia types may be misclassified or simply attributed to Unspecified dementia by medical professionals in the hospital setting (Crowther et al. 2017). As a result, the number of hospitalisations for a specific dementia type may not be accurate. The most reliable data are likely to be for the most common types of dementia. Caution should be taken when interpreting hospital statistics by dementia type.
In 2021–22, there were just over 11.6 million hospitalisations in Australia. Of these, about 25,700 were hospitalisations due to dementia, which is equivalent to 2 out of every 1,000 hospitalisations in Australia.
Table 9.1 shows that of the hospitalisations due to dementia in 2021–22:
- there were more hospitalisations for men than women (13,300 hospitalisations compared to 12,300 hospitalisations)
- men also had a higher age-standardised rate of hospitalisations than women (49.7 hospitalisations for men and 36.2 hospitalisations for women, per 10,000 population aged 60 and over)
- the average length of stay was 14.8 days – this was 5 times higher than the average length of stay of 2.7 days for all hospitalisations (AIHW 2023a)
- the average length of stay was longer for men (15.9 days) than women (13.6 days).
Number of hospitalisations
Age-standardised rate (number per 10,000)
Average length of stay, days per hospitalisation
Total number of bed days
Note: Age-standardised rates, which apply to people aged 60 and over, were standardised to the Australian population as at 30 June 2001 and are expressed per 10,000 population.
Source: AIHW analysis of National Hospital Morbidity Database.
With increasing age, hospitalisations due to dementia increased in number but decreased in length
There were distinct patterns in hospitalisations due to dementia in 2021–22 by age and sex (Figure 9.1):
- the number of hospitalisations for both men and women increased with age up to 80–84 years, then decreased in the older age groups
- the rate of hospitalisations increased for both men and women up to age 90–94, before decreasing among those aged 95 and over
- people with younger onset dementia (aged under 65) had a greater average length of stay (31 days) than older people – the average length of stay decreased with increasing age, to 11 days for people hospitalised due to dementia aged 95 and over. This trend was observed among both men and women.
Figure 9.1: Overview of hospitalisations due to dementia in 2021–22, by age and sex
This figure shows various measures of hospitalisations due to dementia in Australia by sex and age, including: the estimated number of hospitalisations due to dementia, rate of hospitalisations (per 10,000 population), bed days and average length of stay (days).
All rates have been age-standardised to adjust for population differences, providing a more accurate reflection of time trends. Figure 9.2 shows the number and age-standardised rate of hospitalisations due to dementia between 2012–13 and 2021–22:
- the rate of hospitalisations increased gradually each year from 2012–13 to 2014–15 (5.4% annual average increase) and from 2015–16 to 2021–22 (1.0% annual average increase)
- between 2014–15 and 2015–16 there was a large increase in both the number of hospitalisations (24% increase) and the age-standardised rate (22% increase). This increase is likely due to changes in coding for rehabilitation which came into effect in 2015–16 (see Box 9.2 for more information)
- although the numbers of hospitalisations were similar throughout the period, men consistently had higher rates of hospitalisations than women.
Figure 9.2: Hospitalisations due to dementia between 2012–13 and 2021–22: number and age-standardised rates
Two line graphs showing the estimated number and age-standardised rates of hospitalisations due to dementia in Australia by sex between 2012–13 and 2021–22.
Box 9.2: Coding changes impacting hospitalisations due to dementia over time
With many national health data collections, the primary purpose is to support the administrative objectives of the collecting organisation rather than to monitor disease. Factors such as changes in coding practices may affect how well conditions are recorded in the data.
The period 2014–15 to 2015–16 saw a substantial increase in hospitalisations due to dementia (where dementia was the principal diagnosis). In July 2015, a new coding standard ACS2104 Rehabilitation was introduced into the ICD-10-AM. Changes that may have led to an increase in admissions with dementia coded as the principal diagnosis include:
- prior to July 2015, patients who were admitted to hospital specifically for rehabilitation were assigned a principal diagnosis of Z50.9 Care involving use of rehabilitation procedure, unspecified, and the condition which led to the need for rehabilitation (such as dementia) was assigned as an additional diagnosis
- in July 2015, this coding standard was revised so that the principal diagnosis should be assigned to the condition which led to the need for rehabilitation, and Z50.9 Care involving use of rehabilitation procedure, unspecified should be assigned as an additional diagnosis.
When changes such as these are implemented in short time periods, distinct steps up or down in numbers and rates are expected to be observed in administrative data.
In 2021–22, the most common dementia types recorded in hospitalisations due to dementia were:
- Alzheimer’s disease (29%)
- Other dementia (20%)
- Unspecified dementia (16%)
- Vascular Dementia (12%) (Figure 9.3).
There were differences by sex in the dementia types recorded in hospitalisations due to dementia. Compared to men, there was a slightly higher proportion of hospitalisations among women due to Alzheimer’s disease (30% of women compared with 28% of men) and Unspecified dementia (18% of women compared with 14% of men). There was a higher proportion of men with hospitalisations due to Vascular dementia than women (14% of men compared with 9.0% of women).
Although not a distinct type of dementia, delirium superimposed on dementia occurs when a person with pre-existing dementia experiences delirium, and is a common cause of hospitalisations among people with dementia. In 2021–22, 15% of hospitalisations due to dementia were for Delirium superimposed on dementia (Table S9.3).
As a person may have multiple types of dementia, it is possible to have more than one type of dementia recorded among hospitalisations with dementia. While the majority of dementia hospitalisations had only one type of dementia recorded, almost 10% (about 8,700 hospitalisations) had 2 or more types recorded.
Figure 9.3: Hospitalisations due to dementia in 2021–22: percentage, by type of dementia and sex
This figure shows the percentage of hospitalisations due to dementia in Australia by type of dementia and sex in 2021–22. The most common dementia type recorded was Alzheimer’s disease.
In 2021–22, 40% of all hospitalisations due to dementia ended in the patient being discharged home (Figure 9.4). This proportion was considerably less than seen for all hospitalisations in 2021–22, where 93% ended in the patient being discharged home (AIHW 2023b).
Of all dementia hospitalisations, 5.2% of hospitalisations were discharged to their usual place of residence which was a residential aged care facility. 14% were transferred to a residential aged care facility which was not their usual place of residence. Of those discharged or transferred to a residential aged care facility, 2 in 3 patients were aged over 80 (Table S9.9).
Around 24% of hospitalisations due to dementia ended with a ‘Statistical discharge’, which is when the intent of care changed during their stay in hospital (for example, from acute care to rehabilitation or palliative care). A further 9.2% ended with discharge or transfer to another acute hospital and 5.2% ended with the death of the patient.
Figure 9.4: All hospitalisations and hospitalisations due to dementia in 2021–22: percentage, by mode of separation
A bar graph showing the percentage of all hospitalisations, and hospitalisations due to dementia in Australia by mode of separation in.
6% of all hospitalisations due to dementia involved palliative and other end-of-life care
In 2021–22, there were over 900 palliative care-related hospitalisations due to dementia, accounting for 5.8% of all hospitalisations due to dementia (3.6% for palliative care and 2.2% for other end-of-life care) (Figure 9.5a). Palliative care-related hospitalisations include hospitalisations with a care type of palliative care, in which specialist palliative care is provided, as well as hospitalisations where a palliative care diagnosis was recorded, but care was not necessarily delivered by a palliative care specialist (referred to as other end-of-life care). See Palliative care services in Australia for more detail.
Compared to women, men tended to have higher rates of hospitalisations due to dementia involving palliative care or other end-of-life care, and this held across most age groups; for both men and women, the rate of hospitalisations increased with age (Figure 9.5b).
Figure 9.5: Palliative care and other end-of-life care hospitalisations due to dementia in 2021–22, by age and sex: (a) number and (b) age-specific rate
This figure has two bar graphs showing the estimated number and age-specific rates of palliative care and other end-of-life care hospitalisations due to dementia in Australia by sex and age in 2021–22.
Figure 9.6 shows the age-standardised rate of hospitalisations due to dementia by state and territory, remoteness areas and socioeconomic areas for 2021–22.
In 2021–22, after adjusting for population differences, the age-standardised rate of hospitalisations due to dementia for people aged 60 and over:
- varied across states and territories – ranging from 34 hospitalisations per 10,000 population in ACT to 56 hospitalisations per 10,000 population in Queensland
- varied by remoteness – the rate of hospitalisations due to dementia was highest in Major cities (45 hospitalisations per 10,000 population) and lowest in Remote areas (36 hospitalisations per 10,000 population)
- was lowest among people living in the highest socioeconomic areas (37 per 10,000 population) and highest among people living in the lowest socioeconomic areas (47 per 10,000 population). The other socioeconomic areas had similar rates of hospitalisation (ranging between 42 and 44 per 10,000 population).
Socioeconomic and demographic factors such as income, education, employment, geographic access to health and social support services, as well as belonging to a culturally and linguistically diverse (CALD) community, may impact hospitalisations with dementia. These and other barriers may affect this group's ease of access to hospital services. Refer to Understanding Dementia among First Nations people and Dementia in priority groups for more information.
Figure 9.6: Hospitalisations due to dementia in 2021–22: age-standardised rate, by state and territory, remoteness area and socioeconomic area
A bar graph showing the age-standardised rates of hospitalisations due to dementia by state or territory, remoteness areas and socioeconomic areas in 2020–21.
Previous sections have presented hospitalisations due to dementia (that is, when dementia was recorded as the principal diagnosis), but understanding hospitalisations with dementia (that is all hospitalisations with a record of dementia, whether as the principal and/or an additional diagnosis) provides important insights on the wide-ranging conditions that can lead people living with dementia to use hospital services.
In 2021–22, there were just under 94,500 hospitalisations with dementia (47,000 men and 47,500 women), which is equivalent to 8 out of every 1,000 hospitalisations in Australia (AIHW 2023a).
Trends over time in hospitalisations with dementia were different to trends in hospitalisations due to dementia (Figure 9.7). The age-standardised rate of hospitalisations with dementia were fairly stable between 2012–13 and 2014–15 (0.5% annual average increase) and decreased slightly between 2015–16 and 2021–22 (2.2% annual average decrease). Between 2014–15 and 2015–16, there was a decrease in both the number (8.1% decrease) and the age-standardised rate (10.6% decrease) of hospitalisations. By comparison, the age-standardised rate of hospitalisations due to dementia increased between 2014–15 and 2015–16 (as shown in the Hospitalisations due to dementia increased over time section).
Supplementary codes were introduced into hospital administration data in 2015 to record chronic conditions that are part of a patient’s current health status, but don’t meet the inclusion criteria as a principal or additional diagnosis. When accounting for the supplementary codes, hospitalisations with dementia rise to 197,000 people (95,000 men and 102,000 women). This is equivalent to almost 2 out of every 100 hospitalisations, and is more than double the estimate of hospitalisations with dementia based on principal and additional diagnoses alone.
Refer to Box 9.3 for more information on how changes to the way dementia has been recorded in hospital data over this period may have contributed to the decrease in hospitalisations with dementia.
Figure 9.7: Hospitalisations with dementia between 2012–13 and 2021–22 by sex: (a) number and (b) age-standardised rate
Two line graphs showing the estimated number and age-standardised rates of hospitalisations with dementia in Australia by sex between 2012–13 and 2021–22.
Box 9.3: Potential factors impacting hospitalisations with dementia over time
In July 2015, 29 supplementary codes for chronic conditions (including dementia) and a new Australian Coding Standard were implemented in existing classification systems and coding rules for assigning specific codes to diagnoses found in medical records (ACCD 2015). As a result, from July 2015 onwards dementia can be recorded in hospital data using a supplementary code when it does not meet the criteria to be recorded as a principal or additional diagnosis.
It is not yet understood whether the introduction of supplementary codes has impacted recording of chronic conditions as additional diagnoses in hospitals data. It may have resulted in less recording of dementia as an additional diagnosis, especially in cases where dementia did not directly impact the care provided to the patient. The exclusion of supplementary codes may explain the sudden decrease in the rate of hospitalisations with dementia between 2014–15 and 2015–16, and subsequently, the slightly declining rates to 2021–22 (Figure 9.7).
The quality of the supplementary codes is still being assessed but these will be factored into analysis in the near future to better quantify hospitalisations with dementia.
What were the most common principal diagnoses when dementia was an additional diagnosis?
In 2021–22 there were almost 77,500 hospitalisations where dementia was recorded as an additional diagnosis. Examining the most common principal diagnoses recorded for these hospitalisations provides insights on the diverse reasons why people with dementia require hospital services. The most common principal diagnoses were:
- Delirium, not induced by alcohol and other psychoactive substances, ICD-10-AM codes F05.0–9 (7,000 hospitalisations or 9% of hospitalisations where dementia was an additional diagnosis)
- Problems related to medical facilities and other health care, ICD-10-AM codes Z75.0–9 (7,300 hospitalisations or 9.4% of hospitalisations where dementia was an additional diagnosis).
- these hospitalisations are where patients are required to stay in hospital longer to rehabilitate or are waiting to transfer to new accommodation. These hospitalisations may be counted as a separate hospitalisation rather than a continuation of a hospitalisation from an acute episode (such as a fracture)
- Fracture of the femur, ICD-10-AM codes S72.0–9 (4,700 hospitalisations or 6%)
- Other disorders of urinary system, ICD-10-AM code N39X (2,800 hospitalisations or 3.6%) (Figure 9.8).
Other common principal diagnoses recorded for these hospitalisations included respiratory diseases, cardiovascular diseases and sepsis.
Falls (ICD-10-AM codes W00–W19) accounted for 97% of all hospitalisations due to fractures of the femur, and 95% of hospitalisations due to fractures of the lumbar spine and pelvis, where dementia was an additional diagnosis. Women had twice as many hospitalisations than men for a fracture of the femur and fracture of the lumbar spine and pelvis, with an additional diagnosis of dementia (Table S9.10).
Figure 9.8: Most common principal diagnoses where dementia was an additional diagnosis in 2021–22, by sex
A bar graph showing the number of hospitalisations for the most common principal diagnoses when dementia was recorded as an additional diagnosis, by sex in 2021–22.
ACCD (The Australian Consortium for Classification Development) (2015) Reference to Changes for ICD-10-AM/ACHI/ACS 9th edn, Independent Hospital Pricing Authority, Australian Government, accessed 17 August 2022.
AIHW (Australian Institute of Health and Welfare) (2023a) Admitted patient care 2021-22, AIHW, Australian Government, accessed 19 June 2023
AIHW (2023b) Palliative care services in Australia, AIHW, Australian Government, accessed 20 June 2023.
Crowther GCE, Bennett MI & Holmes JD (2017) 'How well are the diagnosis and symptoms of dementia recorded in older patients admitted to hospital?', Age and Ageing, 46:1, doi: https://doi.org/10.1093/ageing/afw169.