The combined effects of heat events and poor air quality conditions can severely affect population health. A novel correlational method was developed to assess the impact of the short-term variations of environmental variables (air pollutants and ambient conditions) on community health responses (mortalities and emergency department visits). A multi-dimensional clustering approach was proposed by combining hierarchical and k-means clustering to promote flexibility and robustness to improve the correlation procedure. The study focused on the health records of the elderly population and people diagnosed with cardiorespiratory causes. The study investigated multiple health records on different levels of investigation: total, elderly, cause-based, and elderly cause-based records. The developed method was validated by investigating the short-term impact of ambient air temperature, relative humidity, ground-level ozone, and fine particulate matter on the health records during hot and warm seasons in the municipalities of Mississauga and Brampton, Peel Region, Ontario, Canada for 15 years. The analysis confirmed the association between moderate levels of environmental variables and increased short-term daily total deaths and emergency department visits, while the elderly sector showed higher vulnerability to environmental changes. Furthermore, the association with extreme heat conditions and poor air quality levels was affirmed with cause-based mortalities and emergency visits; the correlation was strongest with elderly cause-based health records. Findings confirm that cardiorespiratory patients, especially elderly people, were at the greatest risk of poor environmental conditions.
Keywords: Air quality; Clustering analysis; Emergency visits; Extreme heat; Greater toronto area; Greenbelt; Heat wave; Mortality.
Copyright © 2022 Elsevier Inc. All rights reserved.