[Improved opportunities for the identification of people with a migrant background for mortality research using the example of Bremen]

Background: 19.6% of Germany's population has a "migrant" background. Comprehensive epidemiological research on health and health development of this large, heterogeneous and increasingly important population group in Germany is still deficient. There is a lack of results on mortality and morbidity, particularly concerning chronic diseases and disease processes.

Objective: The aim of this paper is to combine and to compare already applied methods with new methodological approaches for determining the vital status and the mortality of immigrants from Turkey and the former Soviet Union.

Methods: For this purpose we used data from the state of Bremen (666 709 residents, last update 2010). We examined 2 methodological aspects: (i) possibilities for identifying immigrant background in the data of residents' registration office with different methods (onomastic, toponomastic, etc.) and (ii) opportunities for record linkage of the obtained data with the Bremen mortality index.

Results: Immigrants from Turkey and the former Soviet Union were successfully identified in databases of the residents' registration office by a combination of different methods. The combination of different methodological approaches proved to be considerably better than using one method only. Through the application of a name-based algorithm we found that Turkish immigrants comprise 6.9% of the total population living in Bremen. By combining the variables "citizenship" and "country of birth" the total population proportion of immigrants from the former Soviet Union was found to be 5%. We also identified the deceased immigrant population in Bremen. The information obtained from residents' registration office could be successfully linked by death register number with the data of the Bremen mortality index. This information can be used in further detailed mortality analyses.

Conclusion: The results of this analysis show the existing opportunities to consider the heterogeneity of the German population in mortality research, especially by means of combination of different methods to identify the immigrant background.