Skills

R Programming

Statistics

Spatial Analysis

Injury Epidemiology

blog

As a part of a University level GIS course I taught I needed to generate some raster data for a lab exercise on map algebra. I wrote a function to “rasterize” polyline data I had on hand, that is, to convert the discrete polyline data to a continuous density surface. This type of operation used in things like land use regression models or in metrics of walkability or bikeability, which require all input data to be continuous surfaces.

To produce kernel density estimates (KDE) of point processes in a linear network: \[\lambda(z)= \sum_{i=1}^{n} \frac{1}{\tau} k(\frac{d_{iz}}{\tau})y_i\] Using the Quartic function: \[\lambda(z)= \sum_{i=1}^{n} \frac{1}{\tau}(\frac{3}{\pi}(1-\frac{d_{iz}^2}{\tau^2}))y_i\] Where, \(\lambda\)(z) = density at location z; \(\tau\) is bandwidth linear network distance; \(k\) is the kernel function, typically a function of the ratio of \(d_{iz}\) to \(\tau\); \(d_{iz}\) is the linear network distance from event \(i\) to location \(z\). I wanted to implement a network-based KDE in R based on the algorithm outlined in Xie & Yan (2008).

Publications

More than 50% of detected pedestrian injuries were falls in the street or roadway, which may be missed if users of administrative data rely solely on pedestrian external cause codes for detection. Although on-street falls were more common in individuals aged 65 and older, they contributed to the burden of injury requiring emergency department visits across all age groups. Additionally, an increasing percentage of bicycling injuries were classified as ‘in-traffic’ but not resulting from motor vehicle collisions over the 15-year observation period Healthcare records offer a means to detect these injuries that are likely overlooked by police reporting.

We observed significant inequities in transportation injury risk requiring hospitalization, with higher incidence rates among low-income and Indigenous populations for injuries involving motor vehicle occupants, pedestrians, and cyclists

Crowdsource bicycling incident reports across 9 Canadian cities suggest dangerous passes are the most commonly reported bicycling incidents, with the highest ratio of near misses to collisions (9:1) while incident types leading to highest proportion of injury were with motor vehicles turning left.

Canadian Community Health Survey captures more bicycling/walking than the census. Across data sources, walking more common among women than men. Men had higher risk of a fatality than women for bicycling and walking. Both data sources have key limitations for measuring bicycling and walking. Implementing a national household travel survey should be a priority in Canada.

A tutorial for using R as a GIS with a specific focus on applied research in transportation safety

We estimated installation of fully separated cycle tracks along one Toronto corridor would prevent approximately 152.9 injuries and 0.9 fatalities over a 10-year period.

People bicycling for leisure were more likely to be younger, male, higher income, and identify as white. Few bicyclists commuted by bike.

Out-of-home commute modes declined during COVID-19, with increases in telework. Commuting by public transit was most strongly associated with change in commute mode to avoid COVID-19 risk. Among pre-COVID-19 transit commuters, 18.2% continue to rely on transit, and personal motor vehicle use is more common (13.0%) than walking (3.4%) or cycling (2.9%).

E-scooters used more for transport than recreation, potentially filling a niche. Also viewed as convenient, faster, and better in hot weather than walking. Non-white non-riders significantly more likely to intend to try e-scooters. E-scooters disproportionately replace walking and bicycling for all trip types. Women significantly more likely to cite safety-related barriers to e-scooter use.

Data from 8 North American show the odds of reporting a bicycling crash were lower in cities that had existing PBSPs (Boston, Montreal, Toronto).

Data from 7 European cities show higher crash risks for less frequent cyclists, men, those who perceive cycling to not be well regarded in their neighbourhood, and those who live in areas of very high building density.

In our case study we found that measuring bicycling once, resulted in a larger sample with better representation of sociodemographic groups, but different estimates of long-term bicycling behaviour.

We found bicyclists with greater spatial access to bicycling specific infrastructure had a higher likelihood of perceiving bicycling to be safe.

Official safety data underreport the burden of cycling safety incidents. In Vancouver, British Columbia only 1 in 8 cycling incidents were reported to insurance claim. The burden of cycling safety incidents is much greater than official data sources suggest. Innovations are needed for cycling safety surveillance.

Crowdsourced collision data have potential to fill in gaps in reports to official collision sources and that crowdsourced near-miss reporting may be influenced by perceptions of risk.

We explored a method for adjusting estimates of per capita alcohol consumption for tourist influence in 26 census divisions (CD) in British Columbia, Canada. The adjustments for tourist influence decreased consumption estimates by approximately 2% provincially and between 1% and 16%, regionally.

We examined how different radon concentration thresholds were associated with patterns of regional radon vulnerability, estimated areas and populations at risk, and lung cancer mortality trends in BC. Lowering the threshold from its current guideline value of 200 to 50 Bq m−3 resulted in better classification accuracy, a 2.5-fold increase in the relatively small population at risk, and persistent separation in lung cancer mortality trends between areas of high and low vulnerability.

Here, we present a novel method for predictive radon mapping that is broadly applicable to regions throughout the world.