Heat-related illness (HRI) is of growing public health importance, particularly with climate change and an anticipated increased frequency of heat waves. A syndromic surveillance system for HRI could provide new information on the population impact of excessive heat and thus be of value for public health planning. This study describes the association between 911 medical dispatch calls for HRI and temperature in Toronto, Ontario during the summers of 2002-2005.
A combination of methodological approaches was used to understand both the temporal trend and spatial pattern in the relationship between 911 medical dispatch data and temperature. A case definition for HRI was developed using clinical and empirical assessments. Generalized Additive Models (GAM) and Zero inflated Poisson regression (ZIP) were used to determine the association between 911 calls and mean and maximum temperature. The validity of the HRI case definition was investigated by making comparisons with emergency department visits for HRI. Descriptive, aberration detection, and cross-correlation methods were applied to explore the timing and volume of HRI calls in relation to these visits, and the declaration of heat alerts. Finally, the existence of neighbourhood level spatial variation in 911 calls for HRI was analyzed using geospatial methods.
This is the first study to demonstrate an association between daily 911 medical dispatch calls specifically for HRI and temperature. On average, 911 calls for HRI increased up to a maximum of 36% (p<.0001) (median 29%) for each 1°C increase in temperature. The temporal trend of 911 calls for HRI was similar to emergency department visits for HRI and heat alerts, improving confidence in the validity of this data source. Heterogeneity in the spatial pattern of calls across neighbourhoods was also apparent, with recreational areas near the waterfront demonstrating the highest percentage increase in calls.
Monitoring 911 medical dispatch data for HRI could assist public health units carrying out both temporal and geospatial surveillance, particularly in areas where synoptic based mortality prediction algorithms are not being utilized. This previously untapped data source should be further explored for its applications in understanding the relationship between heat and human health and more appropriately targeting public health interventions.
| Identifer | oai:union.ndltd.org:TORONTO/oai:tspace.library.utoronto.ca:1807/17296 |
| Date | 26 February 2009 |
| Creators | Bassil, Katherine |
| Contributors | Cole, Donald |
| Source Sets | University of Toronto |
| Language | en_ca |
| Detected Language | English |
| Type | Thesis |
| Format | 1621246 bytes, application/pdf |
Page generated in 0.0018 seconds