[A cancer mortality atlas on a small geographic scale: procedure, validity and possibilities for its use]

Cancer mortality in two counties (Dithmarschen and Nordfriesland) of the Federal State of Schleswig-Holstein (Germany) was assessed by analysing the death certificates for the period from 1980 to 1991. Age-adjusted cancer mortality rates were calculated for the more common cancers for counties as well as for smaller sub-scales. The procedure in respect of data base, data protection, data flow, calculation of rates and mapping of cancer mortality rates on small geographic scales is described and discussed. The data collected by us was compared with the data provided by the offices of statistic affairs on a county scale (only for the more common cancers). The comparison was done for each gender and each county, respectively. The agreement was high (> 95%) for 16 of 22 compared items, and below 90% for two items only. The validity of the data, therefore, was considered sufficient for epidemiological studies. As an example, results regarding female breast cancer rates are shown. The possibilities for the utilization of cancer registries on small geographic scales are discussed: Depending on the incidence of the occurrence of different cancers, cancer registries should provide data on multiple regional scales. A more flexible use of the possibilities of the assessment of regional cancer rates is imperative to utilise the very expensive and time-consuming instruments of analytical epidemiological studies of the causes of cancer more effectively.