Using linked Census data to categorise the population by ethnicity and enhance understanding of ethnic inequalities in health in Australia

Ethnicity is not collected routinely in Australian health data but the proxy of ancestry is collected in the Census which is now linked to health data. We aimed to develop an approach to using ancestry data to examine health disparities by ethnicity. We established an expert and community panel to inform our work. We analysed 2016 Census data linked to death registrations for 2016-2021 in 20 million people and compared how different approaches to categorising ethnicity impacted on detection of mortality inequalities between groups. We developed the following approach: (1) Creation of an Anglo-Celtic ethnic group which included shifting those identifying as Australian or New Zealander only out of the Oceania group where they are currently grouped with Indigenous Peoples; (2) For those reporting two ancestries: prioritise ethnic minority over national identities (e.g. Australian) and for the remainder create multiethnic groups based on the combined ancestries; (3) First analyse data at the maximum level of granularity, but then collapse up granularity where no disparities are identified to simplify results and messaging. We demonstrated that these approaches improved identification of mortality inequalities in Māori and Pasifika. Inequalities were either hidden or underestimated when country of birth was used rather than ethnicity.