Developing an Air Quality Health Index for Cape Town, South Africa – A time-series study

Abstract

Background: South Africa currently uses a single-pollutant based air quality index (AQI) to communicate the air quality levels. However, this does not fully capture the risk associated with air pollution related illness as the index is reported using the highest concentration level of all the pollutants. Therefore, a revision of the AQI using a multipollutant approach could better reflect the combined health effect associated with multipollutant exposure. Methods: Using time-series analysis we derived the air pollution coefficient, daily concentration values for PM10, NO2, SO2 and O3 were regressed against total all-cause mortality in a Quasi-Poisson generalized additive model using data from Cape Town, 2006 – 2015. We estimated the short-term association after controlling for confounders, long-term trends and seasonality. Thereafter, we calculated the exponential function of the product of the coefficients from each single-pollutant model and the daily pollutant concentration – these calculated values were multiplied by 10 and further divided by the maximum daily percent excess mortality to scale the daily values for the AQHI. We provide preliminary findings, as the analysis is still ongoing. Results: The coefficients obtained per 10 µg/m3 were 0.007: PM10, 0.014: NO2, 0.009: SO2 and 0.017: O3. These values will be used to develop the multi-pollutant health based air quality index. The lag structure for the pollutant and cause-specific mortality models also showed significant evidence of mortality displacement for PM10 on both cardiovascular and respiratory mortalities from lag5–10. Conclusions: This study serves as a framework that can be adopted and validated by the department of environmental affairs in South Africa. The presence of harvesting supports the need for an updated AQHI. It further merits consideration for review and update of the existing AQI. Thus, improving health risk communication of ambient air pollution for the general population and frail people. Keywords: AQHI, mortality, multipollutant, harvesting