Article Text
Abstract
Background There are striking ethnic variations in health status, outcomes, and health service utilisation. Quantifying and understanding this variation is essential to identify the health needs of ethnically diverse populations and assess the extent to which healthcare is equitable. Incomplete ethnic coding in routine data collection in Scotland undermines these efforts.
Methods We examined ethnic variations in cardiovascular, cancer, maternal and child, and mental health using data linkage methods to bring together data on self-defined ethnicity and health outcomes. Personal identifiers available in both data sets were used to link the encrypted unique record identifiers (census number and Community Health Index number) using exact and probability matching to create a look-up file containing only corresponding encrypted record identifiers. These methods ensured that no personal information was shared between participating organisations. The resulting anonymised database included information on around 4.6 million people with 7 years' follow-up data.
Results Important ethnic differences in all health areas were observed. The White Scottish population had higher cardiovascular risks than Other White British, but Pakistani populations had the highest risks and Chinese the lowest. Breast cancer screening non-attendance was substantially lower in Pakistani women than White Scottish. White Scottish populations had higher hospital admission rates for mental health problems than most other groups, with the exception of African / Black women, and the Mixed groups.
Conclusions These findings contribute substantially to developing ethnicity and health research internationally and provide data that NHS Scotland can use to highlight ethnic inequalities and inform action to address them.