Fuzzy approach to admission for pneumonia and air pollution

Abstract

The effects of air pollution on human health have been released and relates primarily to the respiratory and circulatory systems. Studies in Brazil especially in large cities indicate exposure to the pollutants particulate matter, sulfur dioxide, ozone, carbon monoxide and oxides of nitrogen as related to hospital admissions for respiratory diseases like asthma and pneumonia. The statistical techniques used for this identification are usually the logistic regression and Poisson regression (generalized linear models or additive). These models estimate the chance or risk of hospitalization or death for these respiratory diseases. A new way of analysis of epidemiological data has been gaining ground in research on public health. Is the approach or fuzzy or diffuse nebula. Widely used in automation and control, recently have emerged articles using this new approach. As for pollution and its effects, there is no citation in the national literature of this kind of study. Thus we propose a fuzzy approach to associate exposure to environmental pollutants and hospital admissions for respiratory diseases in Sao Jose dos Campos, Brazil. It will be a study with computer simulation using Matlab tools for fuzzy logic. The data are taken from the hospital and the pollutants Datasus Cetesb. Input functions will be built with the pollutants particulate matter, sulfur dioxide and ozone, and temperature and relative humidity and the output of the model is the number of admissions. Lags will be used because the effects of pollutants sometimes are not noticed on the same day, but on subsequent days. The output of the fuzzy model will be correlated with the actual data using the Pearson correlation. (AU)