gms | German Medical Science

Introduction: Perfluorooctanoic Acid (PFOA) is suspected to have various adverse effects on human health, in particular on foetal development. In 2006, elevated PFOA concentrations in drinking water have been observed in several areas of North Rhine-Westphalia (NRW), Germany.

The PerSpat-Project aims to integrate spatial statistics and environmental health and was established a) to evaluate the applicability of different statistical approaches for spatial realignment and b) to investigate associations among PFOA in drinking water and birth outcome. We here report on the additional integration of census data to improve the spatial realignment between both datasets. Analyses are based on two secondary datasets of different spatial and temporal structures: the state birth registry is a postal code dependent dataset and includes various biometric, medical and social variables of more than 1.5 million births from 2003 to 2014 in hospitals in NRW. The drinking water database is a compilation of a state monitoring programme and self-investigated information on PFOA-concentrations in water supply areas from 2006 to 2017.

Additionally, spatial data of residents in NRW have been obtained from the 2011 European Union Census (https://www.zensus2011.de), which includes a registry based data collection of all inhabitants in Germany.

In order to assign exposure data (drinking water concentrations) to individuals (birth cohort) different spatial resolutions (drinking water supply areas and postal code areas) have to be aligned.

Methods: Spatial assignment of both datasets is conducted using residential data from the 2011 European Union Census. Residential data are joined to a national grid consisting of squares with 100 m side length. Geographical intersection of both, water supply and postal code areas, with the national grid is computed. Weighted by the number of residents, the proportion of every water supply area related to postal code areas is calculated.

Results: 336 of 864 postal code areas in NRW are completely located within one water supply area, while the majority of postal code areas intersects with two or more water supply areas.

When census data is integrated, additional 335 postal code areas can be assigned to a water supply area (based on assumption that at least 97 % of the population should be provided by a single water utility).

Based on census data 671 of 864 postal code areas are served with an amount of at least 97% by one of 190 water supply areas and could therefore definitely be assigned to it. For the 193 remaining postal code areas, population-related prediction of the relative contribution of multiple water supply areas can be distinctly enhanced.

Conclusion: The usage of census data significantly improved geographical realignment of postal code and water supply areas. More than 75 % of NRW’s postal code areas could be aligned to a single water supply area. Additionally the weighting of water supply areas in the remaining postal code areas may be used in further statistical models of realignment.

The authors declare that they have no competing interests.

The authors declare that a positive ethics committee vote has been obtained.