Cancer incidence around an oil refinery as an example of a small area study based on map coordinates

The aim of the study was to create a rapid method for estimating cancer risk in areas defined by exact map coordinates and to test it using as an example incidence of leukemia near an oil refinery. The method can be used to investigate possible local excesses of cancer of suspected environmental origin in Finland. Map coordinates with an accuracy of 10 m for the place of residence of each Finn were obtained from national registers. Based on this data set, numbers of inhabitants and expected number of cancer cases by sex and age in squares of 500 x 500 m were calculated for all of Finland. Observed number of cancer cases in each square were obtained based on record linkage of the map coordinates with the Finnish Cancer Registry. The ratio of observed and expected number of cancer cases was modeled using Poisson regression. The example analysis included all 23 leukemia and 531 any cancer cases registered in an area around an oil refinery in 1983-1986. There was no significant association between distance from the oil refinery and risk of leukemia or any cancer. The method proved fast and efficient in comparing areal differences in cancer incidence in Finland. In cases of environmental concern, geographical analyses of existing registers is a rapid method to perform first analyses when evaluating the need for further studies.