Aims: To explore issues in the collection and analysis of ethnicity data in primary care, and discuss the implications of this for health services research.
Methods: Data routinely collected by 25 Wellington Independent Practice Association (WIPA) general practices in 2001.
Results: Practices varied in the level of ethnicity data coverage achieved, ranging from less than 10% to greater than 90% of patients. Combining practice data with National Health Index (NHI) ethnicity data increased coverage to at least 70% for registered patients at 23 of 25 practices. There were differences between the data collected in general practices and ethnicity recorded on NHI. However, with the exception of the 'Other' category, this disagreement was not systematic. Practices had lower proportions of patients with ethnicity recorded as 'Other' (not further specified) than NHI ethnicity records.
Conclusions: The study demonstrated that it is possible to collect quality ethnicity data in general practices, although there are challenges. Merging practice-collected ethnicity data with NHI ethnicity data increases coverage. However, there is some mismatch between data sources. The findings support the need for standardised, consistent approaches to ethnicity data collection and analysis, as well as systems and policies that facilitate the collection of high quality data.