To the Editor: Cycling as a means of commuting to and from work has increased significantly in the Washington, DC, area over the past two decades. Cycling is viewed as a healthy way to exercise and as an environmentally responsible mode of transportation.1–3 Washington, DC, is favored by a generally mild climate, and the District and many of its neighboring jurisdictions have pursued a policy of making their communities more “bike friendly.”4 With more than 10,000 staff members working in DC, the World Bank Group (WBG) is one of the largest employers in the metropolitan area. An increasing proportion of its staff commutes to work by bike. There are few studies on accident rates of cyclists and even fewer adjust accident rate with risk exposure—the time spent or distance covered while cycling.1–3 The purpose of this study was to investigate commuting patterns and risk-adjusted accident rates of World Bank bike commuters.
Materials and Methods
Study participants were identified from lists of people who had registered for a free bike-parking place in a WBG garage and those who were included in the distribution list of the WBG Bike Club. In June 2008, e-mail invitations were sent to the identified bike commuters, and a notice of the survey was posted on the World Bank’s internal “Kiosk” bulletin board. Once the survey was underway, weekly reminders were sent to nonrespondents, and an additional announcement was posted on the Kiosk before the final week of the survey.
Survey questions include demographic information (age, gender, and country of origin), biking equipment, commuting patterns (frequency, distance, duration, use of a bike path, and use of World Bank facilities such as racks, lockers), experience in collision and noncollision accidents in the past 3 years, and health consequences. A collision accident was defined as a direct hit with a car, another vehicle, or a pedestrian, and a noncollision accident was defined as a fall or other accident not involving a vehicle or a pedestrian. Respondents were also asked to describe the circumstances of their most recent accident in free-form responses.
A total of 1047 unique e-mail addresses were identified from the combined lists (see “Methods”). During the 4-week survey period, 564 (54%) valid responses were received. Survey respondents are slightly younger than World Bank staff overall and much more likely to be men (65% compared with 46% in the general World Bank population).
Characteristics of Bike Commuters
Survey respondents were largely commuters. Eighty-five percent of respondents said they used their bikes to commute to work and, of those, 90% said they commuted only by bike. More than half of respondents reported using a bike path for part of their commute. Twenty percent reported commuting for less than a year, and 73% have commuted for less than 6 years. Recent commuters tended to be more evenly divided between men and women, whereas “senior” commuters (more than 10 years experience) were more likely to be men (10% vs 6% senior female commuters; Table 1).
The median length of a one-way commute was close to 5 miles with a median commuting time of 25 minutes; the average length was 5.75 miles and average time was 28 minutes. The longest reported ride was 14.5 miles, and the longest commuting time was 90 minutes. Over half reported living in zip codes within the District of Columbia, slightly more than a quarter lived in Montgomery County, Maryland, 8% in Arlington County, Virginia, and most of the remaining in Alexandria and Fairfax County, Virginia.
Respondents were also asked to report the number of days on which they commuted in the last month and last year. The average number of days reported in the previous month (May 2008) was 12.9 days, which is higher than the average for the year (115.1, ie, 9.6 days per month). Because riders who make long commutes may be less likely to ride regularly, the total miles commuted in a year were estimated by cross-tabulating reported riding days with commuting distance. The average two-way commuting distance for all commuters in the last year was 1335 miles in a year. In total, survey respondents rode 610,000 miles in their commuting trips in the preceding year.
Accident Experience in the Preceding 3 Years
In the past 3 years, 66% of respondents reported no collision and 52% reported no noncollision accidents (Table 2). More noncollision accidents (247 reported by 161 respondents) than collision accidents (121 reported by 92 respondents) were reported. Of those who reported collision accidents, 79% had just one accident in the preceding 3 years. Accident rates for women and men were similar, although women experienced slightly fewer noncollision accidents than men.
Based on reported accidents, commuting frequency, and distance traveled, WBG commuters experienced collision accidents in 0.030% of their trips or 300 per million trips. Their accident rate per mile traveled was 0.0072% or 72 per million miles (1 collision accident per 13,888 miles). Noncollision accidents occurred at a rate of 574 per million trips and 137 per million miles traveled (1 noncollision accident per 7299 miles).
Circumstances of Accidents and Injury
Eighty-six participants provided descriptions of collisions involving other vehicles, pedestrians, and animals. Seventy-two percent of collision accidents happened with a motorized vehicle. The most common cause was a “right hook,” in which an overtaking vehicle makes a right turn, striking the cyclist or causing him or her to fall. Less than one third of WBG commuters reported injuries associated with accidents, and 12% received treatment as a result.
Noncollision accidents were more common than collision ones, and reported injury rates were also higher. Half of those reporting noncollision accidents said that they had been injured, and 19% received some form of treatment. Reported causes of noncollision accidents included obstacles in the road, such as potholes or debris, slippery surfaces, and mechanical failures.
Association Between Accidents and Demographics
Although not statistically significant, results indicate that being more than 50 years old, men, or traveling on a bike path was associated with greater risk of accident compared with younger, female cyclists, and those on the roadway riders (Table 3).
Because most available statistics on cycling accidents do not include good measures of the underlying exposure (miles or time traveled or trips made), it is difficult to compare the experience of Bank commuters with a standard reference. A recent survey of cyclists in the Delaware River Valley5 found that 15% of all respondents reported a collision with an object or animal; 6% with another bicyclist; 4% with a pedestrian; and 14% with a motor vehicle. Eighty-three percent reported no need for emergency department treatment. The study provides no exposure information specifically for commuters.
A 1975 study by Jerrold Kaplan6 surveyed members of the League of American Bicyclists. The accident rate of World Bank staff is higher than that found in any state by him. Accident rates (all causes) ranged from 166 per million miles in Oregon to 87 in California. The accident rate in Maryland was 130, the District of Columbia was 125, and Virginia was 107. The accident experience of Kaplan’s sample may not be fully comparable with those of World Bank commuters. Over half of the reported miles ridden by his sample were for recreation or touring, and only 22% were commuting to work or school. Nor is it clear that the definition of accident used by Kaplan was the same as that used in this study. If seeking treatment is taken as the definition of a serious accident, then Bank commuters experienced approximately 34 serious injuries from all types of accidents per million miles ridden, which places them in the middle of the range of serious accident rates reported by Kaplan6 and better than the rates reported for the District (41), Virginia (40), or Maryland (50).
Comparable statistics for other modes of transport are hard to find. Most sources of automobile accident statistics focus on fatalities.7,8 The National Highway Transportation Safety Administration’s Fatality Analysis and Reporting System reported 1.37 fatal accidents per 100 million miles traveled in 2007.7 In 2003, Transport Canada reported 8.9 fatalities and 711 injuries per billion vehicle kilometers.9 These rates correspond to 0.014 fatalities and 1.15 injuries per million vehicle miles. Nevertheless, these data are probably not directly comparable with the self-reported accidents of World Bank cyclists. Car accident statistics are based on reportable accidents, which generally involve injuries or significant property damage. Only a fraction of the accidents involving World Bank cyclists were ever reported to the police or involved medical treatment. If we take the filing of a police report as the standard for a “reportable” accident, then the comparable accident rates for Bank commuters are approximately 18 collisions and 12 noncollision accidents per million miles. Based on reported injuries (which may not be comparable with car accident injuries), the accident rate was 90 collisions and 287 noncollisions per million miles.
Although not statistically significant, our study results indicate that those traveling on the bike path incurred greater risk of accident than those on the roadway riders. Similar results were found in a study of bicycle-motor vehicle collisions in the city of Palo Alto, California.10 They found that bicyclists on a sidewalk or bicycle path incurred a 1.8-times greater risk than those in a roadway. Wrong-way cycling on a road or on a sidewalk greatly increased cyclists’ risk of accidents. In contrast, those riding on a roadway in the same direction as adjacent traffic had no increase in risk relative to the study average. In the roadway, most accidents were likely to occur at intersections, not as a result of overtaking vehicles or sideswipes. Kaplan’s study6 also indicated that accident rates on bike paths were more than twice the rate for riding in major streets.
For all its health and environmental benefits, riding a bike exposes cyclists to the risk of serious injury. The survey data show that the rate of injury to Bank staff resulting from collision and noncollision accidents was low, but not negligible. Had the same staff members used their next preferred alternative mode of transportation, they would have experienced fewer accidents and injuries. We cannot assess the costs and benefits of a commuter’s choice—the risk of injury versus the salutary effects of exercise, being crowded into buses and trains versus exhilaration of riding in fresh air, the frustration of traffic grid lock versus the ability to ride on pleasant bike paths. The findings of our study will help bikers to make an educated choice whether to commute by bike based on the risk profile. We must assume that those who have decided to ride their bikes have made such an evaluation and decided that the benefits outweigh the costs. Nevertheless, all other things being equal, a reduction in accident rates would improve the quality of cyclists’ lives and reduce social costs.
We thank the support of the World Bank Health Services and General Services Departments. The views expressed here are those of the authors and may not reflect those of the World Bank, its Executive Directors or the countries they represent.
The study was fully funded by the World Bank Group.
Eric V. Swanson, PhD
Jasminka Goldoni Laestadius, MD, PhD
Anne Gaelle Selod, MD
Jian Ye, MD, PhD
The World Bank
Lennart Dimberg, MD, PhD
Institute of Medicine
Department of Primary Health Care
University of Gothenburg
1.Morris JN. Exercise in the prevention of coronary heart disease: today’s best buy in public health. Med Sci Sports Exerc
2.de Geus B, Van Hoof E, Aerts I, Meeusen R. Cycling to work: influence on indexes of health in untrained men and women in Flanders. Coronary heart disease and quality of life. Scand J Med Sci Sports
3.Oja P, Vuori I, Paronen O. Daily walking and cycling to work: their utility as health-enhancing physical activity. Patient Educ Couns
4.Bicycle Friendly Communities program. Available at: http://www.bikeleague.org/programs/bicyclefriendlyamerica/communities/
. Accessed June 15, 2008.
5.Delaware Valley Regional Planning Commission. Bicycling in the Delaware Valley 2005: Use, Safety, and Demographics. 2007. Available at: www.DVRPC.org/reports/07050.pdf
. Accessed June 15, 2008.
6.Jerrold K. Characteristics of the Regular Adult Bicycle User. 1975. Available at: http://www.bikexprt.com/research/kaplan/index.htm
. Accessed June 15, 2008.
7.Fatality Analysis Reporting System. National Fatality Statistics. Available at: http://www-fars.nhtsa.dot.gov/Main/index.aspx
. Accessed June 15, 2008.
8.Evans AW. Estimating transport fatality risk from past accident data. Accid Anal Prev
9.Canadian Motor Vehicle Traffic Collision Statistics. 2003. Available at: http://www.tc.gc.ca/roadsafety/tp/tp3322/2003/page5.htm
. Accessed June 15, 2008.
10.Wachtel A, Lewiston D. Risk factors for bicycle-motor vehicle collisions at intersections. ITE J
Readers are invited to submit letters for publication in this department. Submit letters online athttp://joem.edmgr.com. Choose “Submit New Manuscript.” A signed copyright assignment and financial disclosure form must be submitted with the letter. Form available atwww.joem.orgunder Author & Reviewer information.