Abstract
Background
To propose a new method for including the cumulative mid-term effects of air pollution in the traditional Poisson regression model and compare the temperature-related mortality risk estimates, before and after including air pollution data.