Radiation epidemiology: concept, methodology and statistical resources

As radiation exposure is the main point of interest in radiation epidemiology, epidemiologists try to relate the risk of diseases (mainly the risk of cancer) to the different levels and patterns of humankind exposure to radiation. Statistics as a branch of mathematics is able to prove associations and infer causality. As many researches are object of methodological limits, mainly those related to both the insufficient size of the sample and descriptive analysis as well as the choice of methods and variables, this paper aims at describing firstly the main kinds of epidemiological studies. Secondly, it relates distributions and summary measures (central tendency measures, measures of dispersion and normal distributions) and hypothesis tests as well necessary for each study. It also discusses the most appropriate statistical resource to the epidemiological evaluation. Finally, the main aim of this study is both to elaborate a systematic review of the researches that have been already done in Brazil since 2000, focusing on the effects caused by the occupational exposures to ionizing radiation in order to establish positive associations between them and to analyze the risk related to the workers health. This paper has as its basis the Reports in Public Health (Public Health Books-CSP) from which several studies about the exposure effects to ionizing radiation and referred kinds of cancer (e.g.: leukemia, skin cancer, thyroid gland cancer and bone cancer) have been taken as object of analysis. The relevance of this study lies in the most applied methods of risk to establish positive associations in ionizing radiation, in the relation between workers' workplace and his health. (author)