Route infrastructure and bicycling injury risk Results from the Bicyclists Injuries and the Cycling Environment (BICE) Study Anne Harris, Conor Reynolds,Meghan Winters, Peter Cripton, Mary Chipman, Michael Cusimano, Shelina Babul, Jeff Brubacher, Steve Friedman, Garth Hunte, Hui Shen, Kay Teschke
safety concerns deter cycling Survey of 1400 cyclists & potential cyclists in Metro Vancouver, top deterrents all related to safety: ice & snow on route car, bus & truck traffic vehicles driving faster than 50 km/h glass or debris on route motorists who don t know how to drive safely near bikes risk of injury from car-bike collisions
differences in cycling injury rates Europe v. North America Netherlands Fatalities per 100 million km Injuries per 10 million km Denmark Germany United Kingdom Canada - BC United States 0 10 20 30 40 [data sources: International - Pucher & Buehler Transport Reviews 2008;28:495-528 BC - Motor Vehicle Branch, 2005 to 2007, TransLink s 2008 Trip Diary Survey, Census 2006]
why the differences? Safety in numbers? Route design?
challenges to studying route infrastructure Comparing locations -> are you sure the number of users travelling through these locations is comparable? http://www.toronto.ca/transportation/publications/brochures/2010_ped.pdf Comparing injured to noninjured individuals -> are you sure any risk differences can be attributed to route design?
Bicyclists Injuries & the Cycling Environment
Bicyclists Injuries & the Cycling Environment Utilitarian cycling Infrastructural characteristics Case-crossover design
participating cities Vancouver 2 participating hospitals 0.6 million people rain in winter, temperate summer lots of hills 26 km of bike lanes & paths per 100,000 population 3.7% of trips by bike Toronto 3 participating hospitals 2.5 million people snow in winter, heat in summer mostly flat 11 km of bike lanes & paths per 100,000 population 1.7% of trips by bike 10
study overview Cyclist to Emergency Department cyclist to emergency department interview site observations data analysis
interview to map route & choose control sites
observations of injury & control sites injury site control site 1 control site 2
Comparisons cumulated over all person-trips, using conditional likelihood method in Proc Logistic case-crossover design features Control sites randomly selected from injury trip: controlling for exposure to risk, i.e., distance ridden on each route type Sites observed by researchers blinded to site status (injury or control): preventing observation bias Comparisons made within a person-trip: controlling for personal & trip characteristics Comparisons cumulated over all person-trips, using conditional logistic regression
results
participants & trips Toronto 273 Vancouver 417 } 690 male 59% 19 to 39 years old 62% income > $50,000 56% cycle > 52 times/year 88% wore helmet 69% wore high viz clothes 33% trip < 5 km 68% weekday, daylight 77% commute 42% other transport 32%
injury circumstances Car Bus or truck Vehicle door Streetcar or train track Other cycling surface Infrastructure Cyclist Pedestrian Dog Other collision % of injury events 0% 5% 10% 15% 20% 25% Collisions n=497 Avoiding motor vehicle collision Avoiding other collision Loss of balance Bike malfunction Braking too hard Item caught in wheel Cornering Clipped into toe clips Other fall Motor vehicle involved, n=331 No Motor vehicle involved, n=347 Falls n=181
major street sidewalk cycle track comparison of 15 route types main focus of study
relative risks by route type decreased risk toy 10 1/100 1/10 1/2 1 Major streets with parked cars no bike infrastructure shared lane bike lane Major streets, no parked cars no bike infrastructure shared lane bike lane Local streets no bike infrastructure bike route bike route with traffic diverters bike route with traffic slowing Separated from traffic sidewalk multiuse path, paved multiuse path, unpaved bike only path cycle track
relative risks by route type decreased risk toy 10 1/100 1/10 1/2 1 Major streets with parked cars no bike infrastructure shared lane bike lane Major streets, no parked cars no bike infrastructure shared lane bike lane Local streets no bike infrastructure bike route bike route with traffic diverters bike route with traffic slowing Separated from traffic sidewalk multiuse path, paved multiuse path, unpaved bike only path cycle track
relative risks by route type decreased risk toy 10 1/100 1/10 1/2 1 Major streets with parked cars no bike infrastructure shared lane bike lane Major streets, no parked cars no bike infrastructure shared lane bike lane Local streets no bike infrastructure bike route bike route with traffic diverters bike route with traffic slowing Separated from traffic sidewalk multiuse path, paved multiuse path, unpaved bike only path cycle track
relative risks by route type decreased risk toy 10 1/100 1/10 1/2 1 Major streets with parked cars no bike infrastructure shared lane bike lane Major streets, no parked cars no bike infrastructure shared lane bike lane Local streets no bike infrastructure bike route bike route with traffic diverters bike route with traffic slowing Separated from traffic sidewalk multiuse path, paved multiuse path, unpaved bike only path cycle track
relative risks by route type decreased risk toy 10 1/100 1/10 1/2 1 Major streets with parked cars no bike infrastructure shared lane bike lane Major streets, no parked cars no bike infrastructure shared lane bike lane Local streets no bike infrastructure bike route bike route with traffic diverters bike route with traffic slowing Separated from traffic sidewalk multiuse path, paved multiuse path, unpaved bike only path cycle track
relative risks by route type decreased risk toy 10 1/100 1/10 1/2 1 Major streets with parked cars no bike infrastructure shared lane bike lane Major streets, no parked cars no bike infrastructure shared lane bike lane Local streets no bike infrastructure bike route bike route with traffic diverters bike route with traffic slowing Separated from traffic sidewalk multiuse path, paved multiuse path, unpaved bike only path cycle track
relative risks of other significant features increased risk 1/10 1 2 3 10 Grade Flat Uphill Downhill Streetcar tracks No Yes Construction No Yes
intersection type grade routes meeting travel direction intersection characteristics analysis of 210 injuries at intersections with 272 intersection control sites
intersection risks Lower risk Higher risk route types meeting at intersection
intersection risks Lower risk Higher risk intersection type
intersection risks Lower risk Higher risk intersection type
intersection risks Lower risk Higher risk motor vehicle speed
intersection risks Lower risk Higher risk cyclist travel direction
intersection risks Lower risk Higher risk route grade
limitations & strengths Extremely severe and mildest injuries not included. all injury studies focus on defined categories of injuries here, those who attended emergency department within 24 hours Not possible to test many route designs available elsewhere: multiple types of cycle tracks innovative intersection designs More route designs tested than in prior studies, objectively measured.
key themes reduce obstacles/obstructions streetcar tracks construction traffic circles increase traffic separation/protection cycle tracks local streets with traffic diverter
safe routes also preferred routes?
cycling in cities suvey bike only paths 85% likely to choose paved multi-use paths 77% likely to choose route preferences: top 5 of 16 unpaved multi-use paths 71% likely to choose cycle tracks 71% likely to choose local street bike routes with traffic calming 65% likely to choose
Route Preference route preference vs. safety High Preference Bike only path Multiuse path paved Multiuse path unpaved Local street bike route & traffic diverters Local street bike route Cycle track Major street bike lane, no parked cars Local street Major street shared lane, no parked cars Major street bike lane, parked cars Major street shared lane, parked cars Major street with parked cars Major street no parked cars Low Preference Less safe More safe Route Safety
Route Preference route preference vs. safety High Preference Bike only path Multiuse path paved Multiuse path unpaved Local street bike route & traffic diverters Local street bike route Cycle track Major street bike lane, no parked cars Local street Major street shared lane, no parked cars Major street bike lane, parked cars Major street shared lane, parked cars Major streets: shared lanes, parked cars, no bike facilities Major street with parked cars Major street no parked cars Low Preference Less safe More safe Route Safety
Route Preference route preference vs. safety High Preference Mixed-use facilities Bike only path Multiuse path paved Multiuse path unpaved Local street bike route & traffic diverters Local street bike route Cycle track Major street bike lane, no parked cars Local street Major street shared lane, no parked cars Major street bike lane, parked cars Major street shared lane, parked cars Major street with parked cars Major street no parked cars Low Preference Less safe More safe Route Safety
Route Preference route preference vs. safety High Preference Multiuse path paved Multiuse path unpaved Bike only path Local street bike route Bike-specific facilities Local street bike route & traffic diverters Cycle track Major street bike lane, no parked cars Local street Major street shared lane, no parked cars Major street bike lane, parked cars Major street shared lane, parked cars Major street with parked cars Major street no parked cars Low Preference Less safe More safe Route Safety
thanks to everyone, especially study participants and... Vancouver study team Melody Monro Evan Beaupre Niki Blakely Jill Dalton Martin Kang Theresa Frendo Jack Becker David Hay Peter Stary Toronto study team Lee Vernich Vartouji Jazmaji Kevin McCurley Andrew Thomas Doug Chisholm Nancy Smith Lea Fred Sztabinski David Tomlinson Barbara Wentworth Funders Funders
cyclingincities.spph.ubc.ca Study background Interview and site observation forms Peer-reviewed publications Reports & brochures My email: anne.harris@ryerson.ca 42
bike signage driveways # of marked lanes distance visible along route presence of intersections other features studied these not significant
streetcar tracks... OR = 3.04 a special risk in Toronto almost one-third of crashes interactions with cars important, because many crashes begin with avoidance manoeuvres
downhill grades... RR = 2.32 a special risk in Vancouver Sometimes compounded with difficult route features: limited sight lines traffic circles speed bumps
behaviour & environment There are (those) who believe that human error is the prime culprit in the majority of cases; who find... that (the individual) was momentarily careless and inattentive and thus injured himself... It (is important) to note that even human error when it occurs is harmful only because it is taking place in an environment with some hazards. After all, mining produces some fifty times the injury toll that banks do, not because miners tend to be fifty times as careless as do bank clerks, but simply because the mining environment is much less forgiving to the inevitable human error as and when it occurs. [source: Weiler PC. Protecting the Worker from Disability: Challenges for the Eighties. Report submitted to Russell H. Ramsay. Toronto: Ontario Minister of Labour, April, 1983 ]
number of fatalities traffic fatalities in Canada driving, walking, cycling 3500 3000 3105 Drivers & Passengers Pedestrians Bicyclists 2500 2000 1500 1605 1000 500 0 586 125 1988 1998 2008 307 41 [data source: Transport Canada]
fatality & injury rates per person-trip fatality & injury rates per 100 million person-trips (note log scale) 0.1 1 10 100 1000 10000 Bus passengers (US) 0.4 161 Fatalities Injuries Motor vehicle drivers & passengers (BC) 9.6 713 Pedestrians (BC) 14.7 392 Cyclists (BC) 13.8 1,398 Motorcyclists (US) 537 10,337 [data sources: US - Beck et al., 2007; BC - Motor Vehicle Branch, 2005 to 2007, TransLinks 2008 Trip Diary Survey, Census 2006]
Fatality & injury rates per distance travelled (BC) fatality & injury rates per distance travelled 30 25 Fatalities per 100 million km Injuries per 10 million km 20 15 10 5 0 Drivers & passengers Pedestrians Cyclists [data sources: BC Motor Vehicle Branch, 2005 to 2007, TransLink s 2008 Trip Diary Survey, Census 2006]
limitations Most severe and mildest injuries not included all injury studies focus on defined categories of injuries here, those who attended emergency department within 24 hours Not possible to test many route designs available in Europe: multiple types of cycle tracks innovative intersection designs But more route designs tested than in other studies to date, all objectively measured.
Cycling injuries vs. health
transportation & illness
risks vs. benefits of cycling authors location benefits & risks taken into account ratio of benefit : risk British Medical Association, 1992 United Kingdom physical activity traffic crashes 20 : 1 lives saved vs. lost Woodcock et al., 2009 London, England physical activity population air pollution traffic crashes 49 : 1 lives saved vs. lost 15 : 1 DALYs saved vs. lost Johan de Hartog et al., 2010 Netherlands physical activity traffic crashes individual air pollution 9 : 1 lives saved vs. lost Rojas-Rueda et al., 2011 Barcelona, Spain physical activity traffic crashes individual air pollution 96 : 1 lives saved vs. lost Rabl & de Nazelle, 2012 Europe physical activity population air pollution traffic crashes individual air pollution 19 : 1 Euros saved vs. lost
risks vs. benefits of cycling in spite of the hostile environment in which most cyclists currently ride, the benefits in terms of health promotion & longevity far outweigh the loss of life years in injury on the roads [source: British Medical Association. Cycling: Towards Health and Safety. Oxford: Oxford University Press, 1992]
why the differences? Best evidence: safety in numbers [source: Jacobsen. Injury Prevention 2003;9:205-9]