Methodology to identify Iranian immigrants for epidemiological studies.

ABSTRACT

Determining ethnic differences in cancer patterns using administrative databases is often a methodological challenge for information on ethnicity or place of birth is commonly lacking. This paper describes the approach we used to identify Iranians residing in British Columbia (BC), Canada and who were registered within the BC Cancer Registry. A listing of common Iranian surnames and given names was generated from two sources a residential telephone book (with a high density of Iranians) and a provincial breast cancer screening program (which allowed for the selection of women born in Iran). Surnames and given names were reviewed manually and the Iranian names were identified and coded as 'highly probable' and 'probable' Iranian. A name directory was then created and linked with the BC Cancer Registry to identify Iranian cancer cases. Using this method, 1729 surnames and 737 given names were selected from the telephone book, and 1881 surnames and 757 given names from the screening program. The majority of these names were coded as 'highly probable' Iranian (98% and 96% for surnames and given names, respectively). 12% of surnames and 10% of given names were common to both sources. A listing of the most common Iranian surnames and given names is provided. In conclusion, in the absence of other ethnicity data, surnames and given names can be very helpful to identify persons of specific ethnicities when these ethnic groups have distinctive names.