The potential for measuring ethnicity and health in a multicultural milieu - the case of type 2 diabetes in Australia

<b><span style="font-size: 12px;">Objective</span></b><span style="font-size: 12px;"><br />Ethnicity influences health in many ways. For example, type 2 </span><span style="font-size: 12px;">diabetes (T2DM) is disproportionately prevalent among certain ethnic groups. </span><span style="font-size: 12px;">Assessing ethnicity is difficult, and numerous proxy measures are used to </span><span style="font-size: 12px;">capture its various components. Australian guidelines specify a set of </span><span style="font-size: 12px;">variables for measuring ethnicity, and how such parameters should be </span><span style="font-size: 12px;">categorised. Using T2DM data collections as an illustrative example, this </span><span style="font-size: 12px;">study sought to examine how ethnicity is measured in Australian health </span><span style="font-size: 12px;">databases and, by comparing current practice with Australia’s existing </span><span style="font-size: 12px;">benchmark recommendations, to identify potential areas for improvement of </span><span style="font-size: 12px;">the health data landscape.<br /></span><div><br /><b>Design</b><br />We identified databases containing information from which ethnic group-specific estimates of T2DM burden may be gleaned. For each database, details regarding ethnicity variables were extracted, and compared with the Australian guidelines. <b><br /><br />Results</b><br />Data collection instruments for 32 relevant databases were reviewed. Birthplace was recorded in 27 databases (84%), but mode of birthplace assessment varied. Indigenous status was commonly recorded (78%, n=25), but only nine databases recorded other aspects of self-perceived race/ethnicity. Of 28 survey/audit databases, 14 accommodated linguistic preferences other than English, and 11 either excluded non-English speakers or those for whom a translator was not available, or only offered questionnaires in English.</div><div><b><br />Conclusions</b><br />Considerable variation exists in the measurement of ethnicity in Australian health data- sets. While various markers of ethnicity provide complementary information about the ethnic profile within a data-set, nonuniform measurement renders comparison between data-sets difficult. A standardised approach is necessary, and identifying the ethnicity variables that are particularly relevant to the health sector is warranted. Including self identified ethnicity in Australia’s set of recommended indicators and as a core component of the national census should be considered. Globalisation and <span style="font-size: 12px;">increasing migration mean that these findings have implications internationally, </span>including for multi-ethnic countries throughout North America and Europe.</div>