Data Sources and Methods for the Air Health Indicator
The Air Health Indicator (AHI) is based on two temporal functions: annual ozone (O3) and fine particulate matter (PM2.5) concentrations and annual mortality risks at the national level. The annual air pollutant concentrations were obtained from the National Air Pollution Surveillance (NAPS) Network.
To identify the cardiopulmonary mortality-air pollution association and establish annual mortality risks, a Bayesian 2-level hierarchical model was applied. First, the city-specific risk for each city (census division) was estimated. Annual city-specific risks of cardiopulmonary mortality due to O3 or PM2.5 were estimated by a generalized Poisson model for each census division. To handle potential overdispersion with daily mortality counts and to account for confounders, a generalized additive over-dispersed Poisson regression model was applied. Second, the national risks for each year were estimated by pooling over the city-specific risks within region and between regions simultaneously. Finally, testing was done of the resulting data for annual national attributable risks (due to O3 or PM2.5) to determine whether a time trend could be detected.
The annual cardiopulmonary mortality risk estimates showed no upward or downward trend. It was therefore possible to assert that the annual mortality risk was constant over the period analysed. The average of 18 annual national risk estimates for O3 (0.011 per 10 parts per billion (ppb) for 1990 to 2007) and the average of eight annual national risk estimates for PM2.5 (0.025 per 10 ppb for 2001 to 2007) were used. It was assumed that this constancy persisted and each of these averages was projected for the years 2008 to 2010, due to the lack of mortality data to develop risks estimates for these years.
The overall annual risk was derived from the product of annual air pollutant concentrations and the average of annual mortality risk estimates of O3 and PM2.5. Finally, a time-trend analysis was also performed on the AHI using the Sen’s test, a non-parametric linear trend test. A trend was detected at the 95% confidence interval for the O3 AHI but not for PM2.5.
- Date Modified: