[Correlation between Area-Level Sociodemographic Patterns and Estimates of Need for Medical Care].

OBJECTIVE

Besides the known determinants age, sex, and morbidity, regional sociodemographic factors may be important for the level of health care needs. This study proposes a possible characterisation of area-level sociodemographic patterns and proves their association to variation in morbidity, mortality and health service utilisation.

METHODS

We used the 412 counties of Germany as unit of analysis. To describe area-level sociodemographic patterns a factor analysis was conducted on a set of 27 indicators from official statistics. Two factors were retained and rotated according to the Varimax criteria, which explained 34.2 and 33.0% of variance. Mortality, utilisation of inpatient health services as well as parameters of the nationwide outpatient claims data [relative risk score (RRS) and the level of health care use (LB)] served as determinants of need for medical care and were correlated to the extracted factors.

RESULTS

Factor 1 describes regional disparities in socio-economic and health status variables and is called socio-economic health index (SGX). Factor 2 characterises the spatial distribution of interregional migration and household size and is called urbanity index (UX). There was a strong positive correlation between SGX and RRS (r=0.77), mortality (r=0.68 and r=0.78 for overall and premature death rate) and inpatient health care use (r=0.62). UX was not correlated with RRS but weakly inversely correlated with inpatient health care use (r=-0.28). Both SGX and UX were significantly correlated to the level of outpatient health care use (r=0.39 and r=0.40). Stratification of LB by type of practitioner revealed that SGX was mainly associated with health care provided by general practitioners while UX was linked to health care provided by specialists, particularly psychotherapists.

CONCLUSIONS

The extracted factors to describe area-level sociodemographic patterns showed distinct correlations to indicators for medical care use. While SGX was mainly associated with overall morbidity, UX showed consistent relations with specific medical care needs, which may be linked to urban living conditions. Therefore, UX may refer to need for care independently from overall morbidity on the one hand and to structural specifics in health care services on the other hand. The meaning of SGX and UX needs to be further investigated taking additional determining factors into account.