[The use of routine data to study rehabilitative care in populations with a migrational background: potential and limitations].

OBJECTIVE: Little is known about the rehabilitative care of populations with a migrational background. Using the assessment of rehabilitation effectiveness as an example, we outline the potentials and limitations routine data from social security carriers has for this research field.
METHODS: We draw on an 80% random sample of all completed medical rehabilitations in the year 2006 funded by the German Statutory Pension Insurance Scheme (n=634 529). The assessment of rehabilitation effectiveness was based on the occupational performance at the time the rehabilitation was completed. Migrational background was defined by nationality. By means of logistic regression adjusted for socio-economic/-demographic and health-related variables, we examine whether rehabilitation effectiveness differs between German and non-German nationals. RESULTS AND DISCUSSION: Non-German nationals have a higher chance of completing medical rehabilitation with a lower occupational performance - irrespective of differences in socio-economic/-demographic and health-related variables. Odds ratios for a lower rehabilitation effectiveness were 1.23 [95%-CI=1.16;1.30] for Turkish nationals and 1.47 [95%-CI=1.38;1.56] for persons from the former Yugoslavia. Different aspects limit the validity of the analysis. (1) By using nationality, only a selection of all persons with a migrational background can be identified. (2) Important covariates cannot be considered. (3) The assessment of rehabilitation effectiveness is inaccurate, because valid data on occupational performance prior to rehabilitation is missing.
CONCLUSION: Similar to routine data from other social security carriers, data from the German Statutory Pension Insurance Scheme has limitations for health services research on populations with a migrational background. Solutions comprise the application of computer-aided procedures for an accurate operationalisation of migrational background and the use of survey data for a valid assessment of rehabilitation effectiveness. © Georg Thieme Verlag KG Stuttgart · New York.