Australia is an ethnically diverse nation. In 2020, an estimated 3 in 10 (30%, or 7.7 million) people living in Australia were born overseas (ABS 2021a). According to the 2016 Census of Population and Housing almost half (49%) of Australians had been born overseas or had one or both parents born overseas (ABS 2017a).
People from some culturally and linguistically diverse (CALD) backgrounds can face greater challenges when navigating the health-care system than people who do not identify as CALD. These can include language and cultural barriers, such as not knowing where to seek help or how to access services. Understanding patterns of disease within CALD populations is important to being able to address the health needs of the CALD population in Australia.
What are the challenges in reporting on the CALD population
Reporting on the health needs of CALD populations in Australia is complex and challenging. The term CALD can have multiple definitions, and includes aspects such as a person’s country of birth, their ancestry, where their parents were born, what language/s they speak and their religious affiliation. There can also be large differences within CALD groups; (for example, those born in the same country may not identify with the same culture or speak the same language). Many data collections do not collect any information on CALD, or collect information on one aspect only. This is inadequate for identifying all people from CALD backgrounds as often a range of information is required.
How can linked data be used to report on the health of the CALD population?
Using linked data can provide a solution to some of these challenges. By combining different sources of information, it is possible to tell a rich story of a person’s demographic profile and interactions with various services. The Multi-Agency Data Integration Project (MADIP) combines information from data sets such as the Census, National Health Surveys (NHS), Medicare Benefits Schedule, prescription medicines (from the Pharmaceutical Benefits Scheme) and deaths registrations (ABS 2021c). Because the data are linked at the level of the individual, information from one data set (for example country of birth from the Census), can be used to supplement information in data sets that do not collect this information.
What did the example analyses show?
This report uses linked Census, death registration and NHS data to explore 3 commonly reported health outcomes in conjunction with the range of CALD information collected in these data. The CALD variables available were: country of birth of person, country of birth of parents, language spoken at home, proficiency in spoken English, religious affiliation, ancestry and year of arrival in Australia. Two of the health outcomes selected – self-reported health status and the proportion with a chronic condition – were from the NHS, and the third – all-cause mortality – from deaths registrations.
There was some evidence that the selected health outcomes were influenced by migration patterns. In general, newer migrants tend to be younger and from Asian countries such as China and India. In contrast, those from European countries such as Greece and Italy tend to be older and have arrived earlier. Additionally, new migrants are often subject to strict health screening requirements prior to entry.
Among the groups who could be considered as CALD:
- those who were born in Asian countries, who spoke Asian languages, and who identified with Asian ancestries generally had better health than the non-CALD Australian population, as measured by the likelihood of having a chronic condition and mortality rates
- those born in European countries, who spoke European languages and identified with European ancestries generally had the highest age-standardised proportions with a chronic condition and all-cause mortality rates
- even when adjusted for age, migrants who had been living in Australia for longer tended to have higher mortality rates than recent migrants.
Overall, variations in the selected health outcomes were observed for all of the CALD variables investigated in this report when data could be presented at the most granular level (such as for individual countries of birth or language). Where data were aggregated these differences were reduced, and when presented in binary CALD versus non-CALD form, the CALD population often had better outcomes. However, presenting results at the most granular level presents challenges, even for large data collections.
What are some key considerations when using linked data for CALD populations?
Both the linkage rate and linkage quality are important when using linked data, as both can introduce bias into reported outcomes. The linkage rate and linkage quality did vary by CALD group for the data collections used in this report. Where linkage rates are low for a particular group, the data may not be representative of this group and generalisability is reduced. Additionally, analysis based on linked data can underestimate the true prevalence or rate of a health outcome, particularly for CALD populations with low linkage rates.
However, linkage has an advantage in that, where variables are present in both data collections, missing data can be supplemented by the data in other collections.
Overall, linked data provides the opportunity to explore the health outcomes and service use using a range, or combination, of CALD variables. This allows the flexibility to tailor the definition of ‘CALD’ used to the health outcome of interest.
It is important to note, however, that the health of CALD populations is a product of many factors, including environmental, economic, genetic and socio-cultural factors in their home country and Australia, as well as their migration experience – many of which are unable to be captured consistently in data.
The Multi-Agency Data Integration Project
Country of birth of person
Country of birth of parents
Language spoken at home and proficiency in spoken English
Year of arrival in Australia
Combining CALD variables and defining CALD
Appendix A: Data sources
Appendix B: Methods
End matter: Acknowledgements; Abbreviations; Symbols; Glossaries; References