|

Bicycling injuries in Ontario 2002-2017: descriptive analyses of hospital visit and fatality data

Author(s): Harris, Watson, Branion-Calles, Rosella

Slidedeck Presentation:

Slide deck link

Abstract:

Background:

Many studies of injuries to bicyclists rely on police reported data. However, bicyclist crashes and incidents are some of the least likely to be police reported among all road injuries. Records held in population health data offer the potential for deeper understanding the total burden of bicycling injuries.

Aims:

To conduct descriptive analyses of Ontario population health records of bicyclist injuries resulting in emergency department (ED) visits, hospitalization, and death.

Methods:

We accessed Ontario data in the following databases at the Institute for Clinical and Evaluative Sciences (ICES): the Discharge Abstract Database (DAD) for hospitalizations, the National Ambulatory Care Reporting System (NACRS) for ED visits, and Vital Statistics records for deaths. Bicyclist injuries were identified through International Classification of Diseases (ICD) 10 codes in the V100 range. Due to data lags, the included years were 2003 to 2017.

Results:

Between 2002 and 2017, there were an average of 23,221 ED visits per year (73.3% male); 1,249 hospitalizations per year (75.9% male); and 25 deaths per year (87.4% male) due to bicycling injury in Ontario. We noted a trend of an increasing contribution of on in-traffic bicycling crashes that did not involve a motor vehicle over the observation period in both ED visits (increasing from 8.6% in 2002 to 31.1% in 2017) and hospitalizations (increasing from 10.1% in 2002 to 33.9% in 2017).

Discussion:

Interpretation of count data is complicated by application of appropriate denominators. We discuss application of Canadian Community Health Survey (CCHS) data as a potential source of data on the number of cycling trips taken during the observation period. The increasing contribution of in-traffic crashes not involving a motor vehicle could reflect increasing proportions of single bicycle crashes in Ontario, but may be an artifact of providers’ application of ICD coding. The risk of double counting in databases must be considered. Future avenues and limitations of health care utilization data to study bicycling injury will be explored.

Conclusions:

Health care and population health databases offer a complementary set of records to explore the burden of bicycling injury and evaluate the impact of safety interventions at a population level.