Generación de un modelo espacial de riesgo de enfermedades respiratorias crónicas a partir de datos de calidad de aire en la ciudad de Quito entre los años 2013 a 2017.

Abstract

Air pollution and adverse effects on public health have received considerable attention in the city of Quito. The study of the spatial risk model (MER) was designed to evaluate the probability of acquiring chronic respiratory diseases (CKD) in the population of the metropolitan district of Quito (DMQ), and how these responses are related to environmental pollution. The binary logistic regression model (MRLB) and the Bayesian logistic regression model (MRLIB) are used to treat the effects of the population's attention and the evaluation of the most important factors through which the factors Air quality, weather data and satellite images. 21 variables were taken into account for the logistic regressions, from which, using correlations and diagnostic tests, 11 variables were selected to evaluate the behavior of the dependent variable (number of patients). The final model tries to predict the risk of presenting CKD in the DMQ plots and observe the health status of the population, through the risk probability maps (MPR). The model with the highest prediction of adjustment was the MRLIB, since the Bayesian inference allows obtaining more accurate predictions based on the MCMC methods.