Bayesian focused clustering for a case-control study on lung cancer in Trieste

摘要

The relationship between four putative sources of environmental pollution (incinerator, shipyard, iron foundry and city center) and lung cancer risk for men in Trieste (Italy), is investigated using a Bayesian framework by a case-control study. In the analysis information on smoking habits and exposure to occupational carcinogens are taken into account to adjust for known risk factor as potential confounders. The models are based on distances between subject place of residence and the different sources of environmental pollution, as a proxi for exposure. Models enable estimation of the risk gradient and directional effects separately for each putative source. We found that risk of lung cancer is highly related to the city center and incinerator sources. However, as the models appeared to be sensitive to modelling choices, any point analysis should be provided with careful sensibility analysis.