A groundbreaking study led by Population Data Science at Swansea University has resulted in the creation of a comprehensive longitudinal harmonised national ethnicity spine for Wales. Leveraging 22 years of anonymised health and administrative data from over 4.2 million individuals, researchers have developed a robust research ready-data asset (RRDA) to address the urgent need for improved ethnicity data in healthcare research.
The COVID-19 pandemic highlighted health disparities among different ethnic groups, emphasising the importance of accurate and accessible ethnicity data. The new ethnicity spine harmonises multiple data sources, ensuring consistency and reliability for researchers.
The study, spanning 22 years from 2000 to 2021, utilised data from 26 different sources within the Secure Anonymised Information Linkage (SAIL) Databank, encompassing 46 million ethnicity records for over 4.2 million individuals in Wales, UK. Researchers aimed to establish a comprehensive and research-ready data asset that harmonises, de-duplicates, and makes ethnicity data accessible for healthcare research. Between the 10-year gap of census data which is often the only resource for ethnicity data, the work of this research showed the value of comprehensive approaches for capturing and curating information through routinely collected data.
Using four distinct selection methods (latest, mode, weighted mode, and composite), the study demonstrated the effectiveness of categorising ethnic groups across various demographic factors such as age, sex, deprivation, health board, and residential location.
The results, published in BMJ Open, align closely with the Office for National Statistics (ONS) 2011 census data, confirming the spine’s accuracy. With a record completeness rate of over 95% in 2021, the RRDA has been utilised in numerous studies and has contributed to various outputs completed and published through the One Wales collaboration – demonstrating its value as a tool for researchers aiming to understand and address ethnic health inequalities.
Professor Ashley Akbari, co-lead author of the study, said: “The development of the population-scale individual-level linked ethnicity spine provides a robust maintainable resource for researchers in the One Wales project, as well as other users and projects who access data available within the SAIL Databank to utilise ethnic group information for research in Wales, and provides a reproducible method and template which can easily be deployed in other trusted research environments in the UK and beyond”.
Fatemeh Torabi, co-lead author of the study, said: “Our work enables research on sub-populations, which is key to promoting diversity and inclusion in primary data assets for research.
This work paves the potential for more effective and impactful research and policy, enabling the development of more targeted interventions that promote equitable outcomes for all.”
This work was completed as part of the One Wales collaboration response to the COVID-19 pandemic, supported and funded by the Medical Research Council, Health Data Research UK, Economic and Social Research Council (ESRC), UK Research and Innovation (UKRI), ADR Wales and the Wales COVID-19 Evidence Centre.