Comparing ABS long-term health conditions data sources

• The Australian Bureau of Statistics (ABS) has several sources of long-term health conditions data which provide data to inform health policy and assist with health service planning.
• The 2021 Census asks a single, long-term health conditions question which allows for the analysis of long-term health conditions data at more detailed geographic and sub-population levels than ABS health surveys can support, and across a range of socio-economic and demographic dimensions.
• The National Health Survey (NHS) is designed to provide detailed data about the health status of the population, including long-term health conditions and health risk factors, and collects demographic and socio-economic characteristics to allow for cross-classification. It is the recommended data source for national and state/territory level long-term health condition prevalence rates, except for mental disorders.
• The National Aboriginal and Torres Strait Islander Health Survey (NATSIHS) is designed to obtain national benchmark information on a range of health-related issues including long-term health conditions, disability, lifestyle factors, physical health, and use of health services. It enables the monitoring of trends in the health of the Aboriginal and Torres Strait Islander population over time.
• The National Study of Mental Health and Wellbeing (NSMHW) is the recommended source for prevalence data for mental disorders as it uses diagnostic criteria rather than relying on respondents self-reporting that they have a mental health condition. Data can be cross-classified with data on service use and other mental health-related topics.
• Long-term health conditions data will be impacted by aspects unique to the Census or the survey in which it is collected, such as the scope, size and characteristics of the sample, and collection methodology.

In recent years, there has been increasing demand by government and private organisations for the regular collection of long-term health conditions data to:

• inform health policy
• assist with health service planning
• target programs and resources to assist in the prevention and treatment of chronic disease
• monitor change and report against key health indicator frameworks.

The ABS has several sources of long-term health conditions data. This article aims to:

• summarise the data sources in terms of their purpose, collection methods and concepts collected
• outline the advantages and limitations of each data source
• provide guidance to data users to enable informed use of the available data
• help data users select the most appropriate long-term health conditions data source to meet their analysis needs.

The 2021 Census asked a question for the first time about whether people had a long-term health condition. The following ABS health surveys also collect information about people’s long-term health conditions:

• National Health Survey (NHS)/National Aboriginal and Torres Strait Islander Health Survey (NATSIHS)
• National Study of Mental Health and Wellbeing (NSMHW)
• Patient Experience Survey (PEx).

There are other household surveys that collect information about people’s long-term health conditions. They include the General Social Survey (GSS) and the National Aboriginal and Torres Strait Islander Social Survey (NATSISS). However, for the purposes of this article we will focus on the Census and the ABS health surveys listed above.

It is not uncommon for Census and other ABS surveys to collect information on the same topic (e.g. need for assistance, labour force participation, education), nor is it uncommon for the counts of the Census to not align precisely with other ABS surveys due to differences in collection method. It is important to understand the differences in each source of information, and the best way to use this data.

The table in the section titled Comparison and information sources for long term health conditions data summarises each collection, the scope, what the collection is best for, the geographic levels each collection contains, and other information about each collection.

The Census provides a contemporary picture of Australian society every five years. As the Census is collected from the entire population in Australia, it is a critical source of information for understanding small areas and small population groups.

The 2021 Census includes a new health topic to capture data about the proportion of Australians reporting selected long-term health conditions. This allows for the analysis of long-term health conditions data at more detailed geographic and sub-population levels than ABS health surveys can support. The Census also collects information about a range of characteristics of people enabling analysis across a range of socio-economic and demographic dimensions.

The Census included a single, long-term health conditions (LTHP) question which asked people of all ages if they have been told by a doctor or nurse that they have any of these long-term health conditions:

• arthritis
• asthma
• cancer (including remission)
• dementia (including Alzheimer’s)
• diabetes (excluding gestational diabetes)
• heart disease (including heart attack or angina)
• kidney disease
• lung condition (including COPD or emphysema)
• mental health condition (including depression or anxiety)
• stroke
• any other long-term health condition(s).

The ten selected health conditions were chosen based on outcomes of stakeholder consultation, how common they are, and how the data collected can inform services, policy, and research.

Respondents were asked to include health conditions that have lasted, or are expected to last, for six months or more and to include health conditions that:

• may recur from time to time
• are controlled by medication
• are in remission.

For more information, see Census dictionary: Health.

The Long-term health conditions question in the Census aims to identify people who have a current, diagnosed (i.e. told by a doctor or nurse) long-term health condition (i.e. has lasted or is expected to last for six months or more).

The main advantages of Census as a source of long-term health conditions data are:

• it includes the total Australian population, as opposed to ABS health surveys which are based on a sample of the total population. This means the data can be used to look at the proportion of the population with long-term health conditions among people in smaller geographic regions or in small population groups, which is not possible from surveys
• the data can be cross-classified with other data collected in the Census
• it includes people in non-private dwellings such as hospitals and nursing homes which are out of scope for most ABS health surveys.

The main limitations to the Census as a source of long-term health conditions data are:

• the data was collected for the first time in the 2021 Census so there is currently no time series data available
• the long-term health conditions data collected are based on a single question. ABS health surveys such as the NHS and NATSIHS have a detailed set of questions to capture prevalence of these conditions more accurately
• data relies on the respondent self-reporting health conditions, or in some cases reporting health conditions of other people in the household, in most cases without an interviewer present to help
• data is only available for the listed conditions above
• under-reporting may occur for some conditions compared with surveys. This is due to self-reporting and reporting on behalf of other members of the household, as well as potential sensitivities about individual health conditions.

The NHS is a large sample survey. It is designed to provide detailed data about the health status of the population, including long-term health conditions and health risk factors such as smoking, Body Mass Index, diet, exercise, and alcohol consumption. It collects demographic and socio-economic characteristics to allow for cross-classification.

The NHS is the recommended data source for national and state/territory level long-term health condition prevalence rates, except for mental disorders. The National Study of Mental Health and Wellbeing (NSMHW) is the recommended source for information on mental disorders.

The NATSIHS is a large sample survey, similar to the NHS. It is designed to obtain national benchmark information on a range of health-related issues including long-term health conditions, disability, lifestyle factors, physical health, and use of health services. It enables the monitoring of trends in the health of the Aboriginal and Torres Strait Islander population over time.

The NHS and NATSIHS collect data on a broad range of health conditions a person may have, with the primary focus on current and long-term conditions.

Some conditions of particular policy importance are asked in individual modules to ensure high quality, detailed results for these conditions. These are:

• asthma
• arthritis
• cancer
• cardiovascular disease
• diabetes and high sugar levels
• kidney disease
• osteoporosis
• mental, behavioural, and cognitive conditions.

The NHS and NATSIHS also collect information on sight and hearing and other long-term conditions, as well as self-assessed health and disabilities.

There are some differences in the way NATSIHS long-term health conditions data is collected compared with the NHS. For example, to take account of language and cultural differences, the collection method and survey questions sometimes vary. In remote areas some questions are reworded to enhance a person’s ability to understand concepts or some topics are excluded if they are considered problematic to collect or not applicable. The NATSIHS only collects four mental health conditions, while the NHS collects over 20.

In the specific conditions modules in the NHS and NATSIHS, respondents are asked whether they have been diagnosed with the condition (‘ever told’). Respondents are also asked whether the condition is current and long-term, except where an assumption is made (e.g. cancers are considered to be long-term if they are identified as current).

Each reported condition is then classified into the following condition status output categories:

1. Ever told has condition, still current and long-term
2. Ever told has condition, still current but not long-term
3. Ever told has condition, not current
4. Not known or not ever told, but condition current and long-term (used where respondents are not asked about diagnosis)
5. Never told, not current or long-term (used for respondents without a condition).

The prevalence of a condition is usually reported based on categories 1 and 4 above (where applicable in NHS), i.e. that the condition is current and long-term. Diabetes and some heart, stroke and vascular conditions are considered to be long-term and current, regardless of the condition status reported at interview.

Current long-term conditions are defined as medical conditions (illnesses, injuries, or disabilities) which are current at the time of interview and which have lasted at least six months, or which the respondent expects to last for six months or more, including:

• long-term conditions from which only infrequent attacks may occur, e.g. asthma
• long-term conditions which may be under control, for example, through the continuing use of medication
• conditions which, although present, may not be generally considered ‘illness’ because they are not necessarily debilitating, e.g. reduced sight
• long-term or permanent impairments or disabilities.

For further detail regarding collection methodology and interpretation specific to each condition, see the appropriate topic section in 4363.0 - National Health Survey: Users' Guide, 2017-18  or National Aboriginal and Torres Strait Islander Health Survey - Methodology - 2018-19.

The NHS and NATSIHS use a conditions classification which is based on the 10th Revision of the International Classification of Diseases (ICD-10).

The classification considers:

• the types of long-term conditions more commonly reported in a population-based survey and for which reliable estimates could be produced
• the types of conditions or groups of conditions known to be of particular interest to data users
• the variability of the descriptions of conditions provided by respondents.

For most conditions, interviewers can select the appropriate condition from either category responses for questions or from a computer-based coder.

Results are generally not available classified to the most detailed ICD-10 condition level as the NHS and NATSIHS are sample surveys and there are not enough observations to support reliable estimates at that level of detail.

The main advantages of the NHS and NATSIHS as sources of long-term health conditions data are:

• the conditions can be cross-classified with other health-specific items collected in the surveys such as health risk factors and health service use
• these surveys collect more detailed data relating to health conditions, e.g. asking detailed questions can help to clarify whether a condition is current or long-term
• the surveys collect more detail about other reported long-term health conditions which can also be published if supported by the sample and of sufficient quality
• the survey questions are administered by trained interviewers which improves the quality
• the surveys provide time series data and enable analysis of trends over time.

The main limitations of the NHS and NATSIHS as sources of long-term health conditions data are:

• The NHS is a nationally representative sample survey, and the NATSIHS is a sample survey of the Aboriginal and Torres Strait Islander population, rather than the total Australian population. This means that the NHS and NATSIHS samples do not generally support reliable output at lower geographical levels or for specific sub-populations of interest (e.g. country of birth). Modelled small area estimates from the NHS and NATSIHS have been produced previously, although these aren’t as reliable as direct estimates.
• Data relies on the respondent self-reporting health conditions, or in some cases reporting health conditions of other people in the household in the case of proxy interviews. However, this impacts the data to a lesser extent in these surveys, compared with the Census, given the conditions questions asked are more detailed and they are interviewer administered.
• The scope of the NHS and NATSIHS excludes certain parts of the population of interest such as people living in non-private dwellings like hospitals and nursing homes. This may affect estimates of the number of people with some long-term health conditions (for example, conditions which may require periods of hospitalisation or long-term care).

The National Study of Mental Health and Wellbeing (NSMHW) is a component of the Intergenerational Health and Mental Health Study (IHMHS). The NSMHW aims to help understand the mental health of Australians, including their use of services and their social and economic circumstances.

The NSMHW provides an evidence base for planning and supporting national mental health services and related mental health and suicide prevention initiatives in Australia. 

The NSMHW provides information on:

• the prevalence and severity of mental disorders in Australia
• demographic, social and health characteristics of people experiencing mental disorders
• access to services and the impacts of these services.

The NSMHW includes modules from the World Health Organization’s World Mental Health Composite International Diagnostic Interview (WMH-CIDI 3.0), covering a range of affective disorders, anxiety disorders and substance use disorders. Some of the standard international modules have been tailored to an Australian context. 

The NSMHW also includes a single, long-term health conditions question to provide information on comorbidity of mental disorders with physical health conditions.

The NSMHW uses a structured interview in which trained ABS interviewers ask respondents about their experiences of symptoms of mental ill health throughout their lifetime. Responses are assessed to determine whether a person’s symptoms meet diagnostic criteria for having a mental disorder in their lifetime and in the 12 months prior to the study interview. Assessment of mental disorders is based on definitions and criteria of the World Health Organization's International Classification of Diseases, Tenth Revision (ICD-10).

The main advantages of the NSMHW as a source of mental health data are:

• it is the recommended source for prevalence data for mental disorders as it uses diagnostic criteria rather than relying on respondents self-reporting that they have a mental health condition
• it provides prevalence data on people’s experiences of mental ill health across their lifetime and in the last 12 months
• data can be cross-classified with data on service use and other mental health-related topics
• data are comparable internationally.

The main limitations of the NSMHW as a source of mental health data are:

• it is a nationally representative sample survey rather than being asked of the total Australian population
• it provides prevalence data on affective, anxiety and substance use disorders only, and not other mental disorders such as psychosis
• the age scope is limited to those aged 16 to 85 years
• the scope of the NSMHW does not include some populations such as people living in non-private dwellings or in very remote parts of Australia
• estimates are not available separately for Aboriginal and Torres Strait Islander Australians.

The short module to collect long-term health conditions is a single question which has been included in several ABS household surveys. These include the Patient Experience Survey (PEx), the National Study of Mental Health and Wellbeing (NSMHW), the General Social Survey (GSS), and the National Aboriginal and Torres Strait Islander Social Survey (NATSISS). This section covers the PEx as an example of this single question approach in ABS surveys.

The PEx collects a single, long-term health conditions question which asks people aged 15 years and over if they have any of the following conditions that have lasted, or are likely to last, 6 months or more:

• arthritis or osteoporosis
• asthma
• cancer
• diabetes
• heart or circulatory condition
• mental health condition, including depression or anxiety
• long term injury
• any other long-term health condition.

Multiple responses are allowed. The interviewer reads out each category and pauses after each one for a ‘yes’ or ‘no’ response.

The long-term health conditions question in the PEx aims to identify people aged 15 years and over who have a condition that is current and long-term (i.e. has lasted, or is likely to last, 6 months or more).

This module can provide useful information about the characteristics of people with long-term health conditions relative to those without, or about those with a particular long-term health condition when cross-classified with other survey items. For example, long-term health conditions data can be cross-classified with survey items such as a person’s self-assessed health status, whether the person has seen a general practitioner for their own health in the last 12 months, or other socio-economic and demographic variables (e.g. age, sex or Socio-Economic Indexes for Areas (SEIFA) – Index of Relative Socio-Economic Disadvantage).

As the data is based on a single question rather than a detailed set of questions as is collected in the NHS and NATSIHS, it is not as reliable and therefore is not recommended for reporting long-term health conditions prevalence rates. Surveys using the question also exclude certain parts of the population of interest such as people in hospitals and nursing homes.

The table below presents the reported rates for each of the long-term health conditions from the Census 2021, NHS 2020-21 and PEx 2020-21 (as an example of the short module to collect long-term health conditions).

Long-term health conditions data will be impacted by aspects unique to the Census or the survey in which it is collected, such as the scope, size and characteristics of the sample, and collection methodology. The tables below provide an overview of these for Census and the ABS health surveys. For detailed information see the explanatory material for each collection, linked in the last table.