Assessing health impacts of the December 2013 Ice storm in Ontario, Canada

Abstract Background Ice, or freezing rain storms have the potential to affect human health and disrupt normal functioning of a community. The purpose of this study was to assess acute health impacts of an ice storm that occurred in December 2013 in Toronto, Ontario, Canada. Methods Data on emergency department visits were obtained from the National Ambulatory Care Reporting System. Rates of visits in Toronto during the storm period (December 21, 2013 – January 1, 2014) were compared to rates occurring on the same dates in the previous five years (historical comparison) and compared to those in a major unaffected city, Ottawa, Ontario (geographic comparison). Overall visits and rates for three categories of interest (cardiac conditions, environmental causes and injuries) were assessed. Rate ratios were calculated using Poisson regression with population counts as an offset. Absolute counts of carbon monoxide poisoning were compared descriptively in a sub-analysis. Results During the 2013 storm period, there were 34 549 visits to EDs in Toronto (12.46 per 1000 population) compared with 10 794 visits in Ottawa (11.55 per 1000 population). When considering year and geography separately, rates of several types of ED visits were higher in the storm year than in previous years in both Toronto and Ottawa. Considering year and geography together, rates in the storm year were higher for overall ED visits (RR: 1.10, 95 % CI: 1.09-1.11) and for visits due to environmental causes (RR: 2.52, 95 % CI: 2.21-2.87) compared to previous years regardless of city. For injuries, visit rates were higher in the storm year in both Toronto and Ottawa, but the increase in Toronto was significantly greater than the increase in Ottawa, indicating a significant interaction between geography and year (RR: 1.23, 95 % CI: 1.16-1.30). Conclusions This suggests that the main health impact of the 2013 Ice Storm was an increase in ED visits for injuries, while other increases could have been due to severe weather across Ontario at that time. This study is one of the first to use a population-level database and regression modeling of emergency visit codes to identify acute impacts resulting from ice storms