For Debate: Collecting ethnic group data for inpatients: is it useful?

Since April last year British hospitals have had to collect data on the ethnic group of every patient. Patients must categorise their own ethnic group using the categories of the 1991 census. Claire Hilton argues that such data collection is unhelpful and will not meet its stated aim of helping to provide equitable access to NHS services for all ethnic groups. Self categorisation is unreliable, and the number of ethnic groups available is inadequate, in particular for the white population. Additional data on the size of each ethnic group in the local population and on rates of illness in particular groups are necessary to assess whether services are being provided equitably. Moreover, ethnicity is a multidimensional concept, and information on patients' language and cultural and religious practices, as well as their place of origin, is necessary if services are to be culturally sensitive.

The Department of Health's report Ethnicity and Health: A Guide for the NHS outlines the considerable evidence of inequalities in the availability and use of health services among minority ethnic groups.1 Since April 1995 collection of data on ethnic groups has been mandatory for all hospital inpatients, with the aim of providing “equitable access to NHS services for all ethnic groups.”2 The definitions of ethnic groups are, however, mainly geographical, whereas ethnicity is a multidimensional concept. The limited data that hospitals are now collecting are inadequate to meet the need to develop equitable access—and, by lulling hospitals into a false sense that they are doing something, may even undermine that aim.