G Model
AAP-2600; No. of Pages 11
ARTICLE IN PRESS
Accident Analysis and Prevention xxx (2011) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Accident Analysis and Prevention
journal homepage: www.elsevier.com/locate/aap
A prospective cohort study on minor accidents involving commuter cyclists in
Belgium
Bas de Geus a , Grégory Vandenbulcke b , Luc Int Panis c,d , Isabelle Thomas b , Bart Degraeuwe c ,
Elke Cumps a , Joris Aertsens c , Rudi Torfs c , Romain Meeusen a,∗
a
Department of Human Physiology & Sports Medicine, Faculteit LK, Vrije Universiteit Brussel, Belgium
C.O.R.E. and Department of Geography, Université catholique de Louvain (UCL), Belgium
c
Flemish Institute for Technological Research (VITO), Boeretang 200, 2400 Mol, Belgium
d
Transportation Research Institute (IMOB), University of Hasselt, Belgium
b
a r t i c l e
i n f o
Article history:
Received 10 May 2011
Received in revised form
14 September 2011
Accepted 27 September 2011
Keywords:
Utilitarian cycling
Incidence
Incidence rate
Prospective study
Exposure
SHAPES
a b s t r a c t
The purpose of this study is to gain insight into bicycle accidents. Bicycle accident data and weekly
exposure data were prospectively collected for one year to calculate the incidence rate (IR) of bicycle
accidents. An accident was included if it occurred during utilitarian cycling, resulting in an acute injury
with corporal damage. If an accident occurred, a detailed questionnaire was filled out to collect detailed
information about its circumstances and consequences.
A sample of 1087 regular (≥2 cycling trips to work a week) adult (40 ± 10 years) cyclists was analyzed.
Over the 1-year follow-up period, 20,107 weeks were covered, accumulating 1,474,978 cycled kilometers.
Sixty-two participants were involved in 70 bicycle accidents, of which 68 were classified as ‘minor’. The
overall IR for the 70 accidents was 0.324 per 1000 trips (95% CI 0.248–0.400), 0.896 per 1000 h (95%
CI 0.686–1.106) and 0.047 per 1000 km (95% CI 0.036–0.059) of exposure. Brussels-capital region is the
region with the highest IR (0.086; 95% CI 0.054–0.118), with a significantly (P < 0.05) higher IR compared
to Flanders (0.037; 95% CI 0.025–0.050). Injuries were mainly caused by ‘slipping’ (35%) or ‘collision with
a car’ (19%). The accidents caused abrasions (42%) and bruises (27%) to the lower (45%) and upper limbs
(41%). Police, hospital emergency department or insurance companies were involved in only 7%, 10% and
30% of the cases, respectively. It is noteworthy that 37% of the participants indicated that they could have
avoided the accident.
In order to decrease the number of accidents, measures should be taken to keep cycling surfaces clean
and decrease the number of obstacles on bicycle infrastructure. Roads and intersections need to be built
so that the collisions between cars and bicycles are decreased to a minimum. Car drivers and cyclists
should pay more attention towards each other. Underreporting of minor bicycle accidents in Belgium
is confirmed, and is higher than expected. Reliable accident statistics, taking into account exposure, are
needed to decide which road safety measures are the most effective. The ‘safety in numbers’ principle is
also applicable for minor bicycle accidents.
© 2011 Elsevier Ltd. All rights reserved.
1. Introduction
Cycling is recognized as an excellent way of being physically active and maintaining good health (Oja et al., 1998, 2011;
Hendriksen et al., 2000; de Geus et al., 2008, 2009). Furthermore, increased active transport (walking and cycling) could
have a substantial role in meeting targets for urban air quality,
∗ Corresponding author at: Department of Human Physiology & Sports Medicine,
Faculteit LK, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium.
Tel.: +32 2 6292222; fax: +32 2 6292876.
E-mail address: [email protected] (R. Meeusen).
greenhouse-gas emissions, and could result in major public-health
benefits (Int Panis et al., 2004; Woodcock et al., 2007).
While cycling brings many personal, environmental and societal
benefits, important barriers to cycling exist. These include fear of
crime/vandalism, bad weather, social pressure, hills and slopes and
long commuting distances (e.g. Pucher et al., 1999; Rietveld and
Daniel, 2004; Gatersleben and Appleton, 2007; Parkin et al., 2008).
Other important barrier to cycling are concerns about traffic safety,
with women fearing accidents more than men (Vuori et al., 1994;
Byrnes et al., 1999; Garrard et al., 2008; Tin Tin et al., 2010), lack
of adequate infrastructure (Pucher et al., 1999; Parkin et al., 2007;
Vandenbulcke et al., 2009) and exposure to air pollution (Int Panis
et al., 2010; Jacobs et al., 2010). The balance between risks and
0001-4575/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aap.2011.09.045
Please cite this article in press as: de Geus, B., et al., A prospective cohort study on minor accidents involving commuter cyclists in Belgium.
Accid. Anal. Prev. (2011), doi:10.1016/j.aap.2011.09.045
G Model
AAP-2600; No. of Pages 11
ARTICLE IN PRESS
B. de Geus et al. / Accident Analysis and Prevention xxx (2011) xxx–xxx
2
benefits is a topic of ongoing debate (de Hartog et al., 2010; Int
Panis, 2011).
The modern traffic system is primarily designed for motorized
vehicles and often fails to make provision for other road users.
Pedestrians and cyclists incur higher crash risks than motorists
(in particular car drivers) in terms of accidents per distance covered (Pucher and Dijkstra, 2000; van Boggelen et al., 2005; BRSI,
2009; Elvik, 2009; Tin Tin et al., 2010). In Norway, the risk of injury,
expressed as fatalities per kilometer while cycling is about 7.5 times
higher than for car drivers (Pucher and Dijkstra, 2000; Elvik, 2009).
In the Netherlands, for all age groups, about 5.5 times more fatal
injuries are recorded per kilometer travelled by bicycle than by car
(CBS, 2008). A study in Portland (US) (Hoffman et al., 2010) reported
that nearly one in five bicycle commuters will experience an event
leading to injury in any given year, regardless of gender, age, body
mass index (BMI), or cycling skill level.
Most statistics on bicycle accidents result from retrospective
surveys (Jacobson, 2003). A weakness of a retrospective study
design is the selection and recall bias, resulting in the fact that especially the most serious injuries will be remembered or registered. It
is well known that most road accident statistics strongly underestimate the total number of cycling accidents meaning that only the
‘tip of the iceberg’ is investigated (Dhillon et al., 2001; De Mol and
Lammar, 2006; Hoffman et al., 2010), particularly when there is no
hospitalization and the cyclist is the only party involved (Veisten
et al., 2007; Vandenbulcke et al., 2009). A comparison of hospital
admissions related to cycling accidents and police registrations of
such accidents show that the latter register only 50% of the total
number of cycling accidents in Europe (De Mol and Lammar, 2006)
and only 10% in the US (Stutts et al., 1990; Pucher and Dijkstra,
2000). Several authors have estimated that in Belgium only about
15% of the cycling accidents are officially reported (Doom and
Derweduwen, 2005; De Mol and Lammar, 2006; BRSI, 2006).
A shortcoming of the current literature and another weakness
specific to retrospective study designs is that no precise recording
of exposure data (bicycle usage, i.e. travel time, distance, frequency) is available nor is it possible to record such data precisely
(Aultman-Hall and Hall, 1998). Yet the recording of exposure data
is essential for the calculation of the incidence rate (IR). Data on
the numerator (accidents) and denominator (exposure) recorded
separately are inadequate to determine an IR, making comparisons between countries or regions within one country difficult
(Jacobsen, 2003).
For the implementation of safety measures and the assessment
of injury costs, a complete and accurate recording of minor and
major accidents and the recording of the cycling exposure is essential. So far, only Aertsens et al. (2010) and Hoffman et al. (2010)
have published results of a cohort study of minor bicycle accidents
and detailed exposure data simultaneously.
The purpose of this study was therefore to record bicycle accidents, and to gain insight into exposure data during one year,
using a prospective study design. The combination of accident and
exposure data allows us to calculate the incidence rate of bicycle
accidents.
2. Methodology
2.1. Study design
This study is part of the SHAPES project (Systematic analysis of
Health risks and physical Activity associated with cycling PoliciES)
which is at the crossroads of health, transport and air pollution
research. The aim of the SHAPES project is to analyze the benefits
and risks (voluntary and involuntary) of cycling and to advise policy makers in order to facilitate the implementation of integrated
policies related to cycling for transport. Within the SHAPES project
exposure to traffic exhaust and accidents during cycling were considered as risks.
For this part of the SHAPES project, an online data registration
system was developed to get insight into bicycle accidents and factors related to their occurrence in regular commuter cyclists in
Belgium. Exposure (bicycle usage) was recorded simultaneously in
a prospective way.
For this study only utilitarian cycling was taken into account.
Utilitarian cycling was defined as commuting to or from work or
cycling for other transport purposes. ‘Sports-related’ cycling, e.g.
cycling for competition or cycling that is not done in traffic, was
excluded.
This web- and e-mail-based data collection system was incorporated into the website of the SHAPES project: www.shapes-ssd.be.
Access to the online data collection system was not restricted, so
that anyone could access the website and participate so long as the
inclusion and exclusion criteria were met. The following inclusion
criteria were checked: (1) age between 18 and 65 years; (2) having a paid job based outside the home; (3) cycling to work at least
2 times a week in the preceding year; (4) living in Belgium. After
entering their e-mail address, an automatically generated e-mail
was sent to the tentative participant. In this first e-mail, information related to the purpose of the study was given. The main dataset
for this study was collected between March 10th 2008 and March
16th 2009.
Those who fulfilled the inclusion criteria received a second automatically generated e-mail with a request to fill out the General
Questionnaire (GQ). Together with this e-mail they also received
the first weekly travel diary (TD). The last question of the TD asked
whether an accident had occurred during the past 7 days. If an
accident occurred, the participant got access to the Prospective
Questionnaire (PQ).
2.2. Study area
Belgium is a small and highly urbanized European country covering approximately 30,000 km2 and has approximately 11 million
inhabitants. At the European level (EU 15), Belgium is ranked fourth,
with a bicycle share of 2.42% (in traveller-km/person/yr), compared
to Germany: 2.47%; Denmark: 5.48%; the Netherlands: 6.66% (EU,
2003; Rietveld and Daniel, 2004; Vandenbulcke et al., 2011).
Belgium is subdivided into three institutional regions: the
Brussels-capital region (BCR, central – officially Dutch-French bilingual), the Flemish region (north, Dutch-speaking) and the Walloon
region (south, French-speaking). Although Belgium is divided into
three institutional regions, traffic legislation is the same, including
the mandatory use of lights and reflectors. The use of bicycle helmet or other protective measurements for commuting purposes
is non-mandatory. Nevertheless, the regions have a certain freedom to adopt different flanking policies. In Flanders traffic calming
measures are more frequent, traffic speeds and speed limits for
motorized traffic on secondary roads are lower and more bicycle
infrastructure is available (Vandenbulcke et al., 2009).
The BCR (more than one million inhabitants) is a particular
region as it consists of a single condensed urbanized built-up
area. Data from the Belgian ‘National household survey’ showed
a clear-cut north-south division in terms of bicycle use, infrastructure, pro-bicycle policies, flat vs. hilly terrain, and attitude towards
cycling (for details, see Vandenbulcke et al., 2009, 2011). In Flanders, 12% of all trips to work are made by bicycle, compared to 2%
in Wallonia and 1% in BCR (NIS, 2001). For commuting to work this
seems to stay the same in Flanders throughout the years, as indicated by data collected in 2007–2008 (OVG, 2008). For BCR and
Wallonia a relative increase in commuting trips by bicycle from
1.2% to 1.5% of all home-work trips was estimated between 2005
Please cite this article in press as: de Geus, B., et al., A prospective cohort study on minor accidents involving commuter cyclists in Belgium.
Accid. Anal. Prev. (2011), doi:10.1016/j.aap.2011.09.045
G Model
AAP-2600; No. of Pages 11
ARTICLE IN PRESS
B. de Geus et al. / Accident Analysis and Prevention xxx (2011) xxx–xxx
and 2008 (FOD, 2010). A report from Pro Velo shows a 294% increase
in the number of cyclists between 2000 and 2008 in BCR (Pro Velo,
2010).
2.3. Questionnaires – travel diary
The General Questionnaire (GQ) is inspired by the Belgian
National household survey on mobility (NIS, 2001) and recent literature (e.g. Kim et al., 2007). In this GQ, traffic-related aspects are
questioned, e.g. travel frequency, presence of bicycle paths on the
road to work, and postal code associated with the places of residence and work. The GQ was also used to collect self-reported
demographic variables: gender, age, height, weight, level of education, job category, perceived health, and living situation.
The Prospective Questionnaire (PQ) is inspired by existing
national official registration systems for traffic accidents (FPS Economy), recent literature (e.g. Brooks and Fuller, 2007; Richter et al.,
2007) and the BLITS® online registration system for sports injuries
(Cumps et al., 2007). The PQ is designed to collect detailed information on the (i) context and circumstances of the accident (e.g.
purpose of the trip, weather conditions, time of day and visibility, type of road, road conditions, traffic situation), (ii) cause of the
accident and injury, (iii) presence and cause of possible physical
injuries, (iv) type of injury (e.g. affected body part(s), injury type),
(v) protective and preventive measures taken at the time of the
accident, (vi) presence of material damage, (vii) medical care, (viii)
reporting by police, insurance company, hospital, and (ix) possibility of avoiding the accident.
In order to obtain a detailed overview of the exposure, participants filled out travel diaries (TD) of their travel frequency,
time spent cycling and distance travelled over the past 7 days.
The same TD has been used in other studies (de Geus et al., 2008,
2009). All participants that fulfilled the inclusion criteria received
an automatically generated e-mail every week at the same moment
(Monday at 2 AM) inviting them to fill out one unique TD.
2.4. Incidence, incidence rate
Incidence was defined as the number (N) of injuries during the
1-year follow-up period. Incidence rates (IR) and corresponding
95% confidence intervals (95% CI) were calculated as the number
of injuries reported per (i) 1000 trips; (ii) 1000 h; and (iii) 1000 km
of exposure.
Exposure data were limited to trips, hours and kilometers
cycled by every participant. The accident and exposure data were
assigned to gender and region of the place of residence (Section
3.2). Incidence rates for the various types of road infrastructure,
environmental characteristics (urban vs. rural), cause of the accident, and medical consequences could therefore not be calculated
(Sections 3.3–3.6).
2.5. Accident and injury definition
Inclusion criteria for the registration of an accident and injury
were: (i) accident during utilitarian cycling; (ii) acute accident
and/or injury; (iii) with corporal damage; (iv) injury more serious
than a muscle cramp or bruise.
Accidents were categorized as ‘minor’ or ‘major’ bicycle accident
according to the definition used by the Belgian National Institute
for Statistics. ‘Minor’ is defined as accidents where the person is
hospitalized for less than 24 h. ‘Major’ is defined as hospitalization
of more than 24 h.
3
3. Results and discussion
3.1. Participants
After one year, 1849 persons left their e-mail address on the
server (Table A, Appendix). Of these, 646 either did not open the first
e-mail or did not meet the inclusion criteria. Only those (N = 1187;
68% men) who filled out more than one TD were retained for the
data analyses presented in this paper. Thirty-six percent of those
retained for the data analysis lived in the BCR, 49% in Flanders and
15% in Wallonia.
Descriptive data of the injured participants are reported in
Table B (Appendix) and compared with those participants not
involved in an accident (SHAPES participants). Table B also
includes data from a representative sample of the Belgian cycling
population, collected within the framework of the national socioeconomic census carried out in 2001 by the National Institute for
Statistics (NIS, 2001).
The SHAPES study population consists of highly educated (88%),
middle aged individuals (40 ± 10 years of age), with only 2.4% blue
collar workers. Almost 72% of the study population indicated that
they live with a partner. They have a mean body mass index of 23.2
(±3.1) and perceive their health as being ‘good’ (49%) to ‘very good’
(43%). No significant differences for demographic variables were
found between those not reporting any accident and those who did
report one or more accident(s). Significantly (P < 0.05) more participants in the SHAPES study are men, have a higher education and
less participants are blue collar workers in comparison with the NIS
(2001) data.
3.2. Exposure–incidence–incidence rate
3.2.1. Overall incidence rate
For a better understanding of ‘how dangerous’ cycling is and
to make the link between health and sustainable travel, exposurebased fatality rates are needed (Christie et al., 2007). Only when
we understand the differences in how much people cycle, and then
express the number of injuries per unit of exposure, can we measure how ‘safe’ and health-enhancing these activities are, and how
specific policies contribute to improved safety.
During the 1-year follow-up period a total of 20,107 weeksworth of travel diaries were collected. During this period, the
participants cycled 214,644 trips, 78,099 h and 1,474,978 km.
A total of 70 bicycle accidents resulting in an acute body injury
were reported by 62 participants. One participant sustained three
injuries and six participants sustained two injuries. Of the 62
injured participants, 44 were men and 18 were women.
The overall incidence rate (IR) in our study is 0.324 per 1000
trips (95% CI 0.248–0.400), 0.896 per 1000 h (95% CI 0.686–1.106)
and 0.047 per 1000 km (95% CI 0.036–0.059) of exposure. In other
words, 1 accident occurred every 3066 trips, 1116 h or 21,071 km
cycled. In one of the few prospective cohort studies (Hoffman et al.,
2010), conducted in the Portland (US) metropolitan area, where
bicycle accidents and exposure were measured simultaneously,
164 cyclists reported 192 traumatic events and 49 cyclists reported
50 serious traumatic events. In their study, the IR of traumatic
events and serious traumatic events was 15.0 (95% CI 13.2–17.5)
and 3.9 (95% CI 2.9–5.1) per 100,000 miles commuted, respectively.
These results are very similar, as the IR in our study was 7.5 (95%
CI 5.8–9.4) per 100,000 miles.
As part of the SHAPES study, Vandenbulcke et al. (2009) analyzed
the road accident statistics between 2002 and 2005 (annual data
from NIS) in a population of cyclists aged between 18 and 65 years,
including mostly major accidents (accidents that required hospital
treatment) and fatal injuries. Time spent cycling to and from work
was used as the denominator. They showed that in Belgium the
Please cite this article in press as: de Geus, B., et al., A prospective cohort study on minor accidents involving commuter cyclists in Belgium.
Accid. Anal. Prev. (2011), doi:10.1016/j.aap.2011.09.045
G Model
AAP-2600; No. of Pages 11
ARTICLE IN PRESS
B. de Geus et al. / Accident Analysis and Prevention xxx (2011) xxx–xxx
4
Fig. 1. Incidence rate (95% CI) per 1000 km for the total study period.
risk of being seriously injured or killed is 0.069 per 100,000 min
cycled compared to 1.493 minor accidents per 100,000 min in this
study. This difference in IR is probably due to the severity of the
recorded accidents. In an epidemiologic study conducted in New
Zealand (Tin Tin et al., 2010), using only on-road cycling accidents
leading to deaths and serious injuries (Abbreviated Injury Scale ≥ 3)
0.052 injuries occurred per 100,000 min spent cycling.
Nine accidents occurred in the month of June, followed by seven
accidents in December, seven in January, seven in February and
seven in April. The absolute number of accidents per week and the
distance cycled during that same week were used to calculate the
IR per week (Fig. 1). Although participants cycled more kilometers
during spring (440,830 km), than in winter (295,695 km), the IR was
not significantly different between the 4 seasons.
The IR in the weeks with snow or icy roads was 0.099 (95%
CI 0.053–0.145) and in weeks without snow or icy roads the IR
was 0.048 (0.036–0.060). In the study of Hoffman et al. (2010),
December and January were the months with the highest IR.
Continuous maintenance of the bicycle infrastructure is therefore
considered to be one of the most important measures to decrease
the number of traumatic events during this period (Hoffman et al.,
2010).
3.2.2. Incidence rate per region
Data from the National Institute for Statistics (NIS, 2009) show
that in Flanders the absolute number of fatal accidents decreased
from 122 to 76 between 2000 and 2008, whereas in BCR only 3 fatal
accidents were recorded during the same period. Major accidents,
involving a hospitalization of more than 24 h, stayed approximately
the same over time in BCR, Flanders and Wallonia. The number of
minor bicycle accidents (hospitalization of less than 24 h) increased
between 2000 and 2008, by +125% in BCR, +25% in Flanders and +2%
in Wallonia. The large increase in accidents in the BCR could be the
result of the large increase in bicycle usage. As mentioned before,
a report from Pro Velo shows a 294% increase of the number of
cyclists between 2000 and 2008 (Pro Velo, 2010).
Data from this study, counting mostly minor accidents, showed
a higher incidence in Flanders, followed by BCR and then Wallonia (Table 1). These results could make us wrongly conclude that
cycling in Flanders is unsafe compared to BCR and Wallonia. Brussels is the region with the highest IR, with a significantly (P < 0.05)
higher IR compared to Flanders.
The so called ‘safety in numbers’ principle (Jacobsen, 2003;
Robinson, 2005; Elvik, 2009) is based on official statistics, including mostly major and fatal injuries. Therefore, it is not known if
the ‘safety in numbers’ effect applies to non-reported accidents,
like minor accidents and accidents only involving a single vehicle
(Elvik, 2009).
Vandenbulcke et al. (2009) have shown that Flanders has the
highest incidence of fatal injuries and major accidents, but when
exposure was taken into account the risk of a cyclist being seriously injured or killed was lower compared to Wallonia. In Flanders,
the risk was spatially homogeneous and lower than the average for the whole of Belgium (0.069 per 100,000 min cycled). In
Wallonia, the casualty risks were much more varied: there was
a very low casualty risk (equal or close to zero) in the majority
of communes (due to the fact that very few if any cyclist was
seriously injured or killed). On the other hand, nearly 38% of communes had quite a high casualty risk (0.069–3.716 per 100,000
bicycle-minutes).
When looking at the regional level, the results from this study
and Vandenbulcke et al. (2009) are complementary, as the present
study calculates the incidence rate of minor and major accidents.
Both studies indicate that Flanders has the highest bicycle share,
the highest incidence, but the lowest IR for minor and major accidents. From the results of this study compared to those presented
in Vandenbulcke et al. (2009), it is found that in the BCR the IR
of minor accidents is high, but low for major and fatal accidents.
The BCR is highly urbanized making on the one hand the interaction between bicycles and cars more frequent (which potentially
increases the number of accidents), but on the other hand resulting in slowing down of motorized traffic (which would decrease
the severity of the injury if an accident occurs between a bicycle
and a car) caused by the large number of hurdles (e.g. traffic lights,
pedestrian crossings, congestion).
Our results demonstrate that the ‘safety in numbers’ principle
is also applicable for minor bicycle accidents. Possible explanations for the ‘safety in numbers’ effect are motorists driving more
carefully when they expect to see, and share the road with, proportionally more cyclists (Jacobsen, 2003) or being more likely to be a
cyclist themselves which in turn (may) induce a different driving
behavior (Vandenbulcke et al., 2009).
Table 1
Incidence, exposure and incidence rate per region.
Incidence
Number of injuries (N)
Exposure
Frequency (# of trips)
Time (h)
Distance (km)
Incidence rate (95% CI)
/1000 trips
/1000 h
/1000 km
Brussels-capital region
Flanders
Wallonia
28
34
8
64,982
20,153
325,210
116,262
45,190
909,033
22,920
8540
160,873
0.431 (0.271–0.590)
1.389 (0.875–1.904)
0.086 (0.054–0.118)
0.292 (0.194–0.391)
0.752 (0.499–1.005)
0.037 (0.025–0.050)
0.349 (0.107–0.591)
0.937 (0.288–1.586)
0.050 (0.015–0.084)
Values in Bold indicate a significant difference (P < 0.05).
Note: 511 (2.54%) travel diaries could not be attributed to a specific region.
Please cite this article in press as: de Geus, B., et al., A prospective cohort study on minor accidents involving commuter cyclists in Belgium.
Accid. Anal. Prev. (2011), doi:10.1016/j.aap.2011.09.045
ARTICLE IN PRESS
G Model
AAP-2600; No. of Pages 11
B. de Geus et al. / Accident Analysis and Prevention xxx (2011) xxx–xxx
Table 2
Incidence, exposure and incidence rate per gender.
Incidence
Number of injuries (N)
Exposure
Frequency (# of trips)
Time (h)
Distance (km)
Incidence rate (95% CI)
/1000 trips
/1000 h
/1000 km
5
Table 3
Most cited weather condition and state of the road at the time of the accident.
Men
Women
44
18
149,346
57,633
1,143,299
60,592
18,891
304,164
0.341 (0.248–0.435)
0.885 (0.642–1.128)
0.045 (0.032–0.057)
0.314 (0.173–0.455)
1.006 (0.554–1.458)
0.062 (0.034–0.091)
Note: 509 (2.53%) travel diaries could not be attributed to a specific gender.
As Bhatia and Wier (2011) indicate, caution should be taken in
the use of ‘safety in numbers’ in transportation policy, planning
dialogue and decision making. Assuming a causal chain leads from
numbers to more beneficial driver behavior is premature. In a country or city where cycling is not part of everyday transport culture
more safety measures should be taken in the first place in order to
create a safer environment for cyclists.
3.2.3. Incidence rate per gender
In our study 44 men and 18 women were involved in an accident. Men cycled more frequently, during a longer time and larger
distances compared to women and had more accidents during the
1-year follow-up period (Table 2). Although women have the highest IR per 1000 h and per 1000 km, differences between genders
were not statistically significant, probably due to the wide confidence intervals and insufficient power.
We conclude that men and women have the same ‘risk’ of having an accident, but also that women cycle less than men. In the
study of Tin Tin et al. (2010), men had a higher rate of collision
and other injuries compared to women, expressed as the number
of injuries per million hours spent cycling. In both studies the gender difference was non-significant as the confidence intervals are
wide and often overlap, probably caused by the insufficient study
power. In the Hoffman et al. (2010) study, no statistical difference
in gender, age, BMI, skill level, or prior traumatic events was found
between those commuters who experienced a traumatic or serious
traumatic event and those who did not.
3.3. Circumstances of the accident
Fifty seven (83%) accidents occurred during a trip to or from
work. Wednesday (29%), Monday (21%), and Thursday (21%)
are the days with the highest number of accidents. Fifty three
percent of the accidents took place during the morning peak
hours (06:45–09:15 AM) and 17% during the evening peak hours
(17:45–19:15 PM). Kim et al. (2007) also found that morning peak
hours (06:00–09:59 AM) are the most dangerous moment of the
day to cycle to work. They pointed out that this may be caused
by driver behavior. Both motorists and cyclists tend to drive and
cycle more aggressively during the busy morning commute and
such behavior can increase the likelihood of having an accident
(Kim et al., 2007). Other studies found that most crashes did occur
in the afternoon or early evening (Thompson and Rivara, 2001; Eid
et al., 2007; Loo and Tsui, 2010) or around noon (Rosenkranz and
Sheridan, 2003).
In the present study, most accidents occurred when it was not
raining and the surface of the road was dry (Table 3). In the Hoffman
et al. (2010) study, most traumatic events (57%) and serious traumatic events (64%) occurred during daylight on a clear/dry day (40%
and 54%, respectively). Poor roadway surface conditions were a factor in 21% of the traumatic events and 20% of the serious traumatic
Dry
Rain
Cold temperature (<0 ◦ C)
Dry
Wet surface with puddles
Glazed frost
Weather condition
Surface of the road
70.0
17.1
7.1
57.1
17.1
20.0
Values are a % of total.
events: tracks on the road, loose gravel, and steel plates were cited
most often.
In a detailed analysis based on police-reported accidents, using
all kinds of accident data (from fatal to possible or no injury)
inclement weather conditions (rain, snow, fog) was found to
increase the severity of injury (Kim et al., 2007). Although this
variable may capture the confounding effects of driver and cyclist
behavior, the authors believe this effect is largely due to increased
slipperiness which reduces both the vehicle’s and bicycle’s maneuverability and can lead to a more severe accident than if braking and
maneuvering were optimal. Reduced visibility can affect perception
times which can lead to a more severe accident since drivers and
cyclists have less time to maneuver and thereby possibly reduce
the force of the crash.
Vandenbulcke et al. (2009) demonstrated that it is important to
make a distinction between urban and rural environments when
analyzing accident statistics. It was shown that for major and
fatal accidents, the casualty rates for cyclists are higher in lessurbanized environments, while the reverse is true in urban areas
(Vandenbulcke et al., 2009). Participants were asked to report
whether the accident had occurred inside or outside the builtup area and if they perceived the traffic density as busy or calm.
Fifty-nine percent of the accidents took place inside the built-up
area while traffic was perceived as ‘calm’. Another 26% of the accidents occurred inside the built-up area with traffic perceived as
busy. Also, the location of the bicycle accidents were allocated to
the different commuter zones of the urban regions (i.e. city centre, agglomeration and urban fringe, which are defined on the basis
of functional and morphological criteria) (Luyten and Van Hecke,
2007). Overall, most of these accidents occurred in municipalities of
the urban agglomeration (41%) and in the city centre (30%), rather
than in the suburbs (13%) or outside the urban regions (8%).
In this study, a differentiation was made between (i) a ‘bicycle
path’ which is separated from traffic (e.g. by a hedge, an open space
or barrier), (ii) a ‘bicycle lane’ which is a portion of a roadway designated by striping and signing or (iii) on-road cycling (which may
include pavement markings for the preferential but not exclusive
use by cyclists).
Sixty-nine percent of the participants were cycling on-road,
while 21% were on a bicycle lane and 10% on a bicycle path at the
time of the accident. Nearly half (49%) of the accidents occurred
on a continuing street and 36% at an intersection. Table 4 further
divides the type of road (bicycle lane/path) in relation to the place
of the accident.
The safety while cycling on-road or off-road, on bicycle paths or
bicycle lanes has often been a point of debate. In Portland (US), the
Table 4
Bicycle path/lane and place of the accident.
Bicycle lane
Bicycle path
Public road without
any markings for
bicycles
Intersection
Continuing street
Other
1.4
2.9
31.4
8.6
18.6
21.5
0.0
0.0
15.7
Values are a % of total.
Please cite this article in press as: de Geus, B., et al., A prospective cohort study on minor accidents involving commuter cyclists in Belgium.
Accid. Anal. Prev. (2011), doi:10.1016/j.aap.2011.09.045
G Model
AAP-2600; No. of Pages 11
ARTICLE IN PRESS
B. de Geus et al. / Accident Analysis and Prevention xxx (2011) xxx–xxx
6
Table 5
Most cited causes of the accident and the injury.
Slipping
Collision with car
Collision with pedestrian
Collision with cyclist
Collision with road sign
Hindrance on the road (constructions)
Other
Table 6
Type of injury.
Accident
Injury
32.9
11.4
5.7
4.3
1.4
7.1
22.8
35.7
18.6
4.3
4.3
5.7
1.4
24.3
Values are a % of total.
greatest numbers of both traumatic and serious traumatic events
occurred on infrastructure, bike lanes/wide shoulders, and residential streets, with motorized vehicles (Hoffman et al., 2010). This
pattern may, at least in part, be due to a disproportionate amount
of cycling occurring on streets with bike lanes.
However, evidence is accumulating that purpose-built bicyclespecific facilities (bike paths) reduce crashes and injuries among
cyclists compared to cycling on-road (Pucher, 2001; Reynolds et al.,
2009; Dill, 2009). The Organisation for Economic Co-operation and
Development (OECD, 1998) reported that bicycle lanes and paths
are safer than no bicycle facilities between junctions, but that the
risk at junctions is increased with bicycle paths. Increasing the distance between the cyclist riding off-road on a bicycle path and the
motorized traffic decreases the visual contact between the cyclist
and the car driver, and may increase the accident risk at intersections/junctions. The discussion is even more complex when looking
at traffic safety on roundabouts. Studies (Daniels et al., 2008, 2009)
conducted in Flanders showed that the conversion of intersections
into roundabouts produced a significant increase (+27%) in the
number of accidents involving cyclists on or nearby roundabouts.
The increase was even higher for accidents involving fatal or serious
injuries (+41 to +46%). When inside built-up areas, the construction of roundabouts increased the number of minor accidents by
48% and the number of accidents causing fatal or serious injuries
by 77% (Daniels et al., 2008). In a subsequent analysis, it was found
that regarding all injury crashes with cyclists, roundabouts with
cycle lanes appeared to perform significantly worse compared to
mixed traffic, separate cycle paths, and grade-separated cycle paths
(Daniels et al., 2009).
One should keep in mind that in our study, and most other
studies, frequency of cycling on bicycle paths or bicycle lanes or
the number of times an intersection is crossed was not measured.
Therefore we do not make any statement on what would be the
safest or most risky bicycle infrastructure, a point also made by
Daniels et al. (2008, 2009) and Hoffman et al. (2010).
From the point of view of contact with traffic exhaust, bicycle
paths however need to be as far away from the motorized traffic
in order to decrease the exposure to air pollution (Int Panis et al.,
2010).
Number count
Abrasion
Bruise
Muscle torn
Bone fracture
Sprain
Cut
Burn
Concussion
Loss of consciousness at scene
In shock
Total
57
37
11
8
8
7
5
2
1
1
133
% of total number of
injury types
41.6
27.0
8.0
5.8
5.8
5.1
3.6
1.5
0.7
0.7
100
in New Zealand, using mainly on-road accidents resulting in deaths
and serious injuries (Abbreviated Injury Scale (AIS) ≥ 3). Tin Tin et
al. (2010) showed that the most common mechanisms of cycling
injuries were non-collision crashes (40% of all cases and 34% of
cases with serious injuries). Collisions with a car, pick-up truck or
van were the cause of the injury in 26% of all cases and 39% of cases
with serious injuries.
In the study of Hoffman et al. (2010), 29% of all traumatic events
and 48% of all serious traumatic events involved a motorized vehicle. An analysis of the severity of hospitalized patients found that
currently the greatest burden is cyclists involved in collisions with
motor vehicles (Chong et al., 2010). Thirty-two percent of the
cyclists with an Injury Severity Score (ISS) of >8 were involved in a
collision with a motor vehicle (Davidson, 2005).
Fatal and serious bicycle accident rates rise markedly with
higher speed limits of the motorized traffic (Stone and Broughton,
2003; Chong et al., 2010). Therefore, cyclist safety can be improved
by reducing bicycle and motor vehicle speed (Garder et al., 1998;
Fernandez de Cieza et al., 1999; Koike et al., 2003). The probability
of injury to cyclists and pedestrians varies with speed with a clear
threshold effect at 32.2 km/h (20 mph) (speed difference between
the motor vehicle and the cyclist or pedestrian), which supports the
commonly used 30 km/h speed limit in residential neighborhoods
(Kim et al., 2007). A study conducted in London showed that the
32.2 km/h zones are effective measures for reducing road injuries
and deaths (Grundy et al., 2009).
However, Chong et al. (2010) and Hoffman et al. (2010) indicate that injury from collisions between cyclists and pedestrians is
not inconsequential. In this study, ‘collision with pedestrians’ was
the third most cited cause of accident and injury. Therefore, moving cyclists to shared paths for cyclists and pedestrians needs to be
approached with caution. If the infrastructure does not allow a separation between cyclists and pedestrians, the speed limit for shared
bicycle–pedestrian pathways should be set at 10 km/h (6.2 mph) for
cyclists (Kleinberger et al., 1998; Eppinger et al., 1999, 2000).
3.5. Medical consequences
3.4. Cause of the accident and injury
A separate question dealt with the self-perceived cause of the
accident and the cause of the injury. For example, a cyclist has to
make a maneuver to avoid an opening door (cause of the accident)
and then comes into contact with a pedestrian and gets injured
(cause of the injury). The two most cited causes of accident and
injury were ‘slipping’ (36%) and ‘collision with car’ (19%) (Table 5).
According to the participants’ description of the accident, in 7 of
the 8 cases, the car hit the cyclist.
As shown in Table 5, non-collision events (‘slipping’) were the
most frequent causes of accidents and injuries, and collision with
a car the second most frequent cause. These results are in accordance with an epidemiologic study (Tin Tin et al., 2010) conducted
3.5.1. Type of the injury
In 28 participants (40%) only one type of injury was registered.
Most injuries were minor (see Section 3.5.3) but two participants
had a concussion, one other lost consciousness and another one
was in shock (Table 6).
3.5.2. Injured body parts
Of the 70 recorded accidents, 46 involved injuries to more than
1 body part. Of the 179 body parts that were injured, the knee was
most frequently hurt (20% of the cases) (Table 7).
As in other studies, the lower and upper limbs were the body
parts most often injured (Hoffman et al., 2010). Injuries of the head,
face, abdomen and spine were likely to be associated with a serious
Please cite this article in press as: de Geus, B., et al., A prospective cohort study on minor accidents involving commuter cyclists in Belgium.
Accid. Anal. Prev. (2011), doi:10.1016/j.aap.2011.09.045
ARTICLE IN PRESS
G Model
AAP-2600; No. of Pages 11
B. de Geus et al. / Accident Analysis and Prevention xxx (2011) xxx–xxx
Table 7
Injured body parts.
Table 8
Reported in official statistics.
Number count
Hip, knee and leg
Shoulder and arm
Head and neck
Back
Trunk (front side)
Total
7
81
74
19
4
1
179
% of total number of
injury types
45.3
41.3
10.6
2.2
0.6
100
traumatic event (Hoffman et al., 2010). However, compared to other
studies which are based on officially reported accidents, including
mostly hospital and police report data, head injuries in this study
were rare.
The most severe injuries seem to be the result of collisions with a
motor vehicle. In the study of Tin Tin et al. (2010), the most common
injuries in crashes involving a collision with a motor vehicle were
traumatic brain injuries (29% of the cases) and open wounds in
the head, face or neck (26%). In cycling injuries unrelated to motor
vehicle collision, traumatic brain injuries and open wounds in the
head, face or neck were present in only 15% and 17%, respectively
(Tin Tin et al., 2010).
3.5.3. Injury severity
The International Classification of Diseases (ICD-9-CM) Injury
Severity Score (ICISS) was used to measure the severity of injuries.
Each lesion was coded with ICD-9-CM and the corresponding exclusive Survival Risk Ratio (SRR) was assigned according to Osler et al.
(1996). A given ICD-9s SRR thus represents the likelihood that any
individual person will survive that particular ICD-9 injury. The ICISS
is defined as the product of all the SRRs for each of the individual
person’s injuries and scores range from 0 (death) to 1 (complete
recovery). For more details we refer to Osler et al. (1996) and
Rutledge et al. (1998).
The total number of participants with ICD-codes (SRR < 1) was
7. The total number of ICD-codes was 185 and the sum of ICISS
(expected number of deaths) was 0.085. The mean ICISS score was
0.988 ± 0.011, with a maximal score of 0.998 and a minimal score
of 0.970. According to the definition of Cryer and Langley (2006),
no participant had a ‘serious injury’ defined as an ICISS of ≤0.941
(a probability of death of at least 5.9%).
In this study two participants were hospitalized for more than
24 h. One of those two accidents was caused by ‘collision with a car’
and the other one was caused by ‘slipping’. In these two cases the
ICISS scores were not higher (0.995 and 0.983) than the ICISS scores
of the other participants with an ICD-code.
At the time of filling out the questionnaire, 70% reported that
the injury would not result in permanent damage. The remaining
30% indicated that it was not yet possible to indicate if the accident
had caused permanent damage.
No relation was found between the severity and the cause of
the accident. This result could be explained by the small number of
accidents with the ICISS code (N = 7) or by the fact that mostly minor
bicycle accidents were reported in this study. In a detailed crash
analysis study using police report data (Richter et al., 2007), it was
shown that lower injury severity (MAIS, ISS) occurred for crashes
in urban areas and for cyclists who used bicycle lanes, compared
with crashes in rural areas and in bicycle traffic lanes.
3.6. Reported accidents in official statistics
Table 8 represents the number of accidents that were officially
reported by police, hospitals or insurance companies. Only 7.1% of
the recorded accidents were reported in official police statistics.
From the two participants that were hospitalized for more than
Police
Hospital
Insurance
With official record
Without official record
No police intervention
Self-care
Ambulant
emergencies
No medical intervention
Yes
No
All injured
participants
Injured
participants with
an ICD-code
7.1
4.3
88.6
47.1
25.7
10.0
17.1
30.0
70.0
40.0
33.3
6.5
0.0
16.7
57.1
0.0
28.6
2.0
% of total within each category.
24 h, only one was reported in an official police report. In a second
analysis, we linked the severity of the injury (those with an ICDcode) with police, hospital and insurance data. These results clearly
show that as the severity of the injury increases, the likelihood of
being reported in official statistics increases.
These results confirm that collecting data through police, hospital and insurance companies will underestimate the actual number
of accidents. Secondly, more severe accidents are more likely to be
reported in official statistics leading to bias in the data. Recorded
bicycle accidents in official statistics in Belgium was estimated
to represent only about 15% of the total accidents (Doom and
Derweduwen, 2005). A case study in Belgium showed that only 20%
of the injured cyclists admitted to hospitals were also recorded by
the National Institute for Statistics using police report data (De Mol
and Lammar, 2006). In Norway, based on police reports, a comparison of the total estimates with the official records revealed that
the official statistics reported approximately 12% of the minor and
moderate bicycle injuries, 33% of serious injuries, and 71% of severe
injuries (Veisten et al., 2007).
3.7. Avoiding the accident
To the question “could you have avoided the accident”, 37%
of respondents said that they could have avoided the accident.
‘Imprudence’ from the cyclist itself was the cause of accident in
26% of cases. ‘Distraction’ was reported to be responsible for 11%
of accidents. Kim et al. (2007) concluded in their study on policereported accident data that cyclist fault is more closely correlated
with greater cyclist injury severity than driver fault.
Bicycle safety relies on infrastructure, conditions and individual behavior (Rietveld and Daniel, 2004). Risk-taking behavior is
a potential confounder in all observational studies of injury and
is difficult to measure reliably because of the likelihood that participants will provide socially desirable responses. Cyclists choose
risky or non-risky behaviors based on individual values, beliefs and
personality traits. Cyclists with different values and beliefs will
choose different actions under the same conditions because they
appraise the same situation differently. Knowledge of cyclists’ value
systems and perceptions of risk is important in designing bicycle
safety programs for adult bicycle riders (McCoy, 2002).
4. Impact of minor bicycle accidents
In Belgium in 2008, 7132 accidents recorded by official police
statistics were classified as “minor accidents” (hospitalization of
less than 24 h) (NIS, 2009). Based on our data, we can extrapolate
the incidence in our study to Belgium. Because only two accidents
required hospitalization for more than 24 h, the following estimation is based on 68 accidents. As only 5.9% of the minor cycling
accidents are officially reported, it is estimated that 120,881 victims
Please cite this article in press as: de Geus, B., et al., A prospective cohort study on minor accidents involving commuter cyclists in Belgium.
Accid. Anal. Prev. (2011), doi:10.1016/j.aap.2011.09.045
G Model
AAP-2600; No. of Pages 11
ARTICLE IN PRESS
B. de Geus et al. / Accident Analysis and Prevention xxx (2011) xxx–xxx
8
would have suffered from minor bicycle accidents in Belgium in
2008.
As part of the SHAPES project, Aertsens et al. (2010) calculated the average cost (material, medical and intangible) of a minor
bicycle accident to be D 841 (95% CI D 579–1205) or D 0.125 per kilometer cycled. The inclusion and exclusion criteria for ‘minor bicycle
accident’ were somewhat different than the ones used for this part
of the SHAPES project. In the Aertsens et al. (2010) study, accidents
with only material damage were also included and accidents that
required hospitalization of more than 24 h were excluded.
5. Study limitations
Some limitations of this study should be noted. Due to the nature
of the study, well-skilled cyclists, who cycle a greater distance and
time than occasional cyclists, are more likely to be captured in the
data. The small number of observations may be the reason why
the differences between genders or between the regions were not
significant. Yet another limitation of this study is that the study
population is self-selected.
actions should contain ‘soft’ (communication/education) and ‘hard’
(enforcement) methods and should be used simultaneously (Berg,
2006). Achieving immediacy in the delivery of a road safety campaign message, in terms of proximity to the target behavior, might
tend to increase campaign effect in the shorter term, and complement any long-term campaign effects achieved using mass media
delivery (Phillips et al., 2011).
Special emphasis should be paid to traffic education (Pucher and
Dijkstra, 2003; Tin Tin et al., 2010), particularly: (1) for specific age
groups for which the accident risk is higher; (2) in specific regions
where the accident risk is generally higher due to lower densities
and higher speed habits. Examples of measures are:
◦ improving driver training for motorists, in order to make them
more mindful/respectful of cyclists;
◦ increasing the (perceived) risk of being punished (following an
illegal/dangerous maneuvers or violations of the traffic regulations with respect to cyclists);
◦ informing all road users of the increased risk of accidents during
the morning and evening peak hours and during winter months
and bad weather conditions in general.
6. Conclusion and policy recommendations
6.1. Conclusion
In this study, a large cohort (N = 1187) of regular (≥2 cycling trips
to work/week) commuter cyclists, aged 18–65 years was prospectively followed during a 1-year period, during which exposure
and bicycle accidents were registered simultaneously. In total 62
cyclists experienced at least one injury. The registered accidents
were classified as ‘minor’ bicycle accidents in 97% of the cases,
resulting in abrasions and bruises of the upper and lower limbs. We
have demonstrated that a prospective design can provide accurate
data on cycling exposure and minor accidents, which allowed the
calculation of the incidence rate.
Underreporting of minor bicycle accidents in Belgium is confirmed as only 7% of the prospectively recorded bicycle accidents in
this study were reported in police statistics. The likelihood of being
reported in official statistics increases with increased severity of
the injury.
The incidence rate of minor bicycle accidents in Belgium is 0.896
per 1000 h (95% CI 0.686–1.106) cycled. There are indications that
injury ‘risk’ differs between regions and gender, but when taking
exposure into account these differences are no longer significant.
The study highlights that exposure has to be taken into account in
order to make reliable statements about the effectiveness of safety
measures. The ‘safety in numbers’ principle is also applicable for
minor bicycle accidents.
Cyclists involved in an accident cycle faster and cycle a greater
distance per week compared to those not involved in an accident.
‘Slipping’ and ‘collision with a car’ are the most cited causes of accidents. 37% of the cyclists indicate that the accidents could have been
avoided. Measures to improve the street environment like traffic
calming schemes, continued maintenance of bicycle infrastructure,
and speed limits in residential areas may contribute towards more
safety for all vulnerable road users. Emphasis should be put on
continuous cyclist and driver education through promotional and
educational campaigns on road safety.
6.2. Policy recommendations
When designing an appropriate policy with the aim of increasing traffic safety, in particular for the vulnerable road users like
cyclists, actions should be taken towards cyclists and drivers of
motorized vehicles as both parties can be the cause of injury. These
For cyclists in particular, more actions towards preventive and
protective approaches should be taken. For instance, preventive
measures (e.g. lights, reflectors, safety jackets, etc.) should be
taken/encouraged to make the cyclists more conspicuous in traffic and to decrease the risk of having an accident. Although this
study did not look at the efficiency of helmet use, the authors suggest to take protective measures, such as wearing a helmet, because
proper helmet use will decrease the severity of the (head) injuries
if a collision occurs. Commuters and schoolchildren (=future commuters) should be instructed on safe cycling practices, in order to
encourage them to cycle and to make them more aware of the risks
associated to a bicycle accident (e.g. with a lorry).
The bicycle network should be kept clean, in general and especially during winter (effective de-icing).
Official registration of minor and major bicycle accidents and
bicycle use must be improved to permit better targeted actions.
Ethical issues
The Vrije Universiteit Brussel ethical committee approved the
study.
Table A
Flow chart for the in- and exclusion of the participants and Prospective
Questionnaire.
Left e-mail on the server
Did not respond to the first e-mail
Exclusion criteria participants
Age? <18 and >65
Paid job outside home? No
Cycling frequency? <2×/week
Living in Belgium? No
Participants filling out GQ
Filled out more than 1 TD
Inclusion criteria PQ
Reported an accident? Yes
Correctly reported? Yes
Recreational cycling? No
Acute accident and/or injury? Yes
Corporal? Yes
Only bruise or cramp? No
1849
377
23
116
101
29
1203
1187
293
286
234
223
174
70
Values are absolute numbers of (tentative) participants.
GQ: General Questionnaire; PQ: Prospective Questionnaire; TD: travel diary.
Please cite this article in press as: de Geus, B., et al., A prospective cohort study on minor accidents involving commuter cyclists in Belgium.
Accid. Anal. Prev. (2011), doi:10.1016/j.aap.2011.09.045
ARTICLE IN PRESS
G Model
AAP-2600; No. of Pages 11
B. de Geus et al. / Accident Analysis and Prevention xxx (2011) xxx–xxx
9
Table B
Comparison between the whole group of SHAPES participants and the group involved in an accident.
NIS (2001)
Men + women
Age (mean ± SD)
38.8
?
Height (mean ± SD)
?
Weight
(mean ± SD)
BMI# (mean ± SD)
?
Education (% of total)
69.3
Lower (primary/secondary)
Higher
30.7
(postsecondary
education/university)
Work situation (% of total)* , §
?
Student (with a
paid job)
40.8
Employee
23.9
Civil servant
3.7
Freelance
Executive
0.7
Worker (blue
28.9
collar)
2.0
Other
Perceived health (% of total)
–
Very good
Good
–
Average
–
Poor
–
Living situation (% of total)
–
Without partner
–
With partner
SHAPES participants (N = 1187)
Men (55%)
Women (45%)
Injured participants (N = 62)
Men + women
Men (68%)
Women (32%)
Men + women
Men (73%)
Women (27%)
37.7 ± 9.7
167.3 ± 6.1
61.8 ± 8.4
37.7 ± 9.43
177.0 ± 9.5
72.0 ± 12.5
38.2 ± 9.0
181.5 ± 7.1
78.0 ± 9.3
36.5 ± 10.6
166.8 ± 5.8
58.1 ± 5.9
39.2
?
?
38.3
?
?
39.7 ± 10.2
175.7 ± 10.3
72.1 ± 12.3
40.7 ± 10.3
179.8 ± 6.6
76.9 ± 10.2
?
?
23.2 ± 3.1
23.8 ± 2.9
72.5
65.3
10.8
13.1
5.8
7.7
7.9
7.1
27.5
34.7
89.2
86.9
94.2
92.3
92.1
92.9
?
?
1.8
0.9
3.7
3.8
0.0
14.3
31.7
26.7
4.3
1.0
35.3
52.2
20.4
3.0
0.4
20.8
49.5
25.7
5.5
9.1
2.4
48.5
25.7
5.8
11.0
3.1
53.2
26.4
5.1
5.4
1.0
40.4
28.8
3.8
17.3
1.9
28.9
36.8
0.0
23.7
2.6
71.4
7.1
7.1
0.0
0.0
1.0
3.4
4.9
4.9
5.1
3.8
5.3
0.0
–
–
–
–
–
–
–
–
42.6
49.2
6.9
0.3
42.5
50.2
6.9
0.3
44.1
48.5
7.1
0.3
44.2
46.2
9.6
0.0
36.8
55.3
7.9
0.0
64.3
21.4
14.3
0.0
–
–
–
–
27.2
71.8
22.4
77.6
38.3
61.7
36.5
63.5
23.7
76.3
28.6
71.4
22.1 ± 2.8
22.8 ± 2.6
23.6 ± 2.2
20.9 ± 2.3
#
BMI: body mass index.
NIS (2001) – population of cyclists with a paid job outside their home (18–65 years).
*
Significant difference between NIS and SHAPES: (Chi2 ) P < 0.05.
§
Definitions not entirely in accordance with these of the census (NIS, 2001). To be interpreted with caution.
–: Definitions not in accordance with these of the Census (NIS, 2001). Cannot be used for comparison.
?: Data not available.
Competing interests
The authors declare that they have no competing interests with
respect to the results discussed in this paper.
Acknowledgements
This paper is part of the SHAPES project (Systematic analysis of
Health risks and physical Activity associated with cycling PoliciES)
which is at the crossroads of health, transport and air pollution
research. The aim of the SHAPES project is to analyze the benefits
and risks (voluntary and involuntary) of cycling and to advise policy makers in order to facilitate the implementation of integrated
policies in different domains related to cycling for transport.
The work reported in this paper was financed by the Belgian
Science Policy under the Science for Sustainable Development Program (Project N◦ : SD/HE/03). The authors wish to thank Nico Smets
and Hanny Willems for creating the online registration system and
database as well as all the participants for their willingness to
participate.
Appendix.
Tables A and B.
References
Aertsens, J., de Geus, B., Vandenbulcke, G., Degraeuwe, B., Broekx, S., De Nocker, L.,
Liekens, I., Mayeres, I., Meeusen, R., Thomas, I., Torfs, R., Willems, H., Int Panis,
L., 2010. Commuting by bike in Belgium, the costs of minor accidents. Accident
Analysis and Prevention 42, 2149–2157.
Aultman-Hall, L., Hall, F.L., 1998. Ottawa-Carleton commuter cyclist on- and
off-road incident rates. Accident Analysis and Prevention 30 (1), 29–
43.
Bhatia, R., Wier, M., 2011. Safety in numbers re-examined: can we make valid or
practical inferences from available evidence? Accident Analysis and Prevention
43, 235–240.
Brooks, J.H.M., Fuller, C.W., 2007. The influence of methodological issues on the
results and conclusions from epidemiological studies of sports injuries. Sports
Medicine 36 (6), 459–472.
Belgian Road Safety Institute (BRSI), 2006. Cycling accidents in an urban
context: three years (1998–2000) of casualties along the regional
roads of the Brussels capital region. Available online at: http://www.
bivv.be/dispatch.wcs?uri=716833556&action=viewStream& language=nl (in
French/Dutch).
Belgian Road Safety Institute (BRSI), 2009. Observatorium voor de Verkeersveiligheid. Themarapport fietsers Verkeersongevallen met fietsers
2000–2007. Available online at: http://bivvweb.ipower.be/Observ/NL/
fietsrapport NL def.pdf (in French/Dutch).
Berg, H.-Y., 2006. Reducing crashes and injuries among young drivers: what kind of
prevention should we be focusing on? Injury Prevention 12 (Suppl. I), i15–i18.
Byrnes, J.P., Miller, D.C., Schafer, W.D., 1999. Gender differences in risk taking: a
meta-analysis. Psychological Bulletin 125 (3), 367–383.
CBS (Statistics Netherlands), 2008. Available: http://www.cbs.nl/en-GB/menu/
thema/verkeer-vervoer/cijfers/default.htm.
Chong, S., Poulos, R., Olivier, J., Watson, W.L., Grzebieta, R., 2010. Relative injury
severity among vulnerable non-motorised road users: comparative analysis
of injury arising from bicycle–motor vehicle and bicycle–pedestrian collisions.
Accident Analysis and Prevention 42 (1), 290–296.
Christie, N., Cairns, S., Towner, E., Ward, H., 2007. How exposure information can
enhance our understanding of child traffic “death leagues”. Injury Prevention
13, 125–129.
Cryer, C., Langley, J.D., 2006. Developing valid indicators of injury incidence for all
injury. Injury Prevention 12, 202–207.
Cumps, E., Verhagen, E., Meeusen, R., 2007. Efficacy of a sports specific balance training programme on the incidence of ankle sprains in basketball. Journal of Sports
Science and Medicine 6, 212–219.
Daniels, S., Nuyts, E., Wets, G., 2008. The effects of roundabouts on traffic safety
for bicyclists: an observational study. Accident Analysis and Prevention 40,
518–526.
Please cite this article in press as: de Geus, B., et al., A prospective cohort study on minor accidents involving commuter cyclists in Belgium.
Accid. Anal. Prev. (2011), doi:10.1016/j.aap.2011.09.045
G Model
AAP-2600; No. of Pages 11
10
ARTICLE IN PRESS
B. de Geus et al. / Accident Analysis and Prevention xxx (2011) xxx–xxx
Daniels, S., Brijs, T., Nuyts, E., Wets, G., 2009. Injury crashes with bicyclists at
roundabouts: influence of some location characteristics and the design of cycle
facilities. Journal of Safety Research 40, 141–148.
Davidson, L.A., 2005. Epidemiology and outcome of bicycle injuries presenting to an
emergency department in the United Kingdom. European Journal of Emergency
Medicine 12, 24–29.
de Geus, B., Van Hoof, E., Aerts, I., Meeusen, R., 2008. Cycling to work: influence on
indexes of health in untrained men and women in Flanders. Coronary heart disease and quality of life. Scandinavian Journal of Medicine and Science in Sports
18, 498–510.
de Geus, B., Joncheere, J., Meeusen, R., 2009. Commuter cycling: effect on physical performance in untrained men and women in Flanders. Minimum dose to
improve indexes of fitness. Scandinavian Journal of Medicine and Science in
Sports 19, 179–187.
de Hartog, J.J., Boogaard, H., Nijland, H., Hoek, G., 2010. Do the health benefits of
cycling outweigh the risks? Environmental Health Perspectives 118, 1109–1116.
De Mol, J., Lammar, P., 2006. Half the road victims are not reported
in the statistics. Verkeersspecialist 130, 15–18, Available online at:
http://www.isaweb.be/bestanden/Artvsp130.pdf (in Dutch).
Dhillon, P.K., Lightstone, A.S., Peek-Asa, C., Kraus, J.F., 2001. Assessment of hospital and police ascertainment of automobile versus childhood pedestrian and
bicyclist collisions. Accident Analysis and Prevention 33, 529–537.
Dill, J., 2009. Bicycling for transportation and health: the role of infrastructure. Journal of Public Health Policy 30 (Suppl. 1), S95–S110.
Doom, R., Derweduwen, P., 2005. Optimalisation of road accident statistics and use
in urban safety management. Research contracts CP/02/392, CP/F1/391 (final
report) for Belgian Road Safety Institute (BRSI), Center for Sustainable Development (CSD) and the Belgian Science Policy (BSP) (in Dutch).
Eid, H.O., Bashir, M.M., Muhammed, O.Q., Abu-Zidan, F.M., 2007. Bicycle-related
injuries: a prospective study of 200 patients. Singapore Medical Journal 48 (10),
884–886.
Elvik, R., 2009. The non-linearity of risk and the promotion of environmentally
sustainable transport. Accident Analysis and Prevention 41, 849–855.
Eppinger, R., Sun, E., Bandak, F., Haffner, M., Khaewpong, N., Maltese, M., Kuppa, S.,
Nguyen, T., Takhounts, E., Tannous, R., Zhang, A., Saul, R., 1999. Development of
Improved Injury Criteria for the Assessment of Advanced Automotive Restraint
Systems II. NHTSA, Washington, DC.
Eppinger, R., Sun, E., Kuppa, S., Saul, R., 2000. Supplement: Development of Improved
Injury Criteria for the Assessment of Advanced Automotive Restraint Systems II.
NHTSA, Washington, DC.
EU (EuropeanUnion), 2003. Energy and transport in figures. European Commission, Directorate-General for Energy and Transport, Brussels Available online
at: /http://gasunie.eldoc.ub.rug.nl/root/2004/2999175/S.
Fernandez de Cieza, A.O., Ortiz Andino, J.C., Archilla, A.R., Gomez, A.M., Gonzalez,
C.G., Mengual, S., Morrall, J., 1999. Nonmotorized traffic accidents in San Juan,
Argentina. Transport Research Record 1695, 19–22.
FOD, 2010. Federale Overheidsdienst Mobiliteit, Mobiliteit en Verkeersveiligheid,
Diagnostiek woon-werkverkeer 2008, eindverslag. Nr 46. Available online at:
http://www.mobilit.fgov.be/data/mobil/RapportWWV 2008n.pdf.
Garder, P., Leden, L., Pulkkinen, U., 1998. Measuring the safety effect of raised bicycle
crossings using a new research methodology. Transport Research Record 1636,
64–70.
Garrard, J., Rose, G., Kai Lo, S., 2008. Promoting transportation cycling for women:
the role of bicycle infrastructure. Preventive Medicine 46, 55–59.
Gatersleben, B., Appleton, K.M., 2007. Contemplating cycling to work: Attitudes and
perceptions in different stages of change. Transportation Research Part A 41,
302–312.
Grundy, C., Steinbach, R., Edwards, P., Green, J., Armstrong, B., Wilkinson, P., 2009.
Effect of 20 mph traffic speed zones on road injuries in London, 1986–2006:
controlled interrupted time series analysis. BMJ 339, b4469.
Hendriksen, I.J.M., Zuiderveld, B., Kemper, H.C.G., Bezemer, P.D., 2000. Effect of commuter cycling on physical performance of male and female employees. Medicine
and Science in Sports and Exercise 32, 504–510.
Hoffman, M.R., Lambert, W.E., Peck, E.G., Mayberry, J.C., 2010. Bicycle commuter
injury prevention: it is time to focus on the environment. The Journal of Trauma,
Injury, Infection, and Critical Care 69 (5), 1112–1119.
Int Panis, L., De Nocker, L., Cornelis, E., Torfs, R., 2004. An uncertainty analysis of air
pollution externalities from road transport in Belgium in 2010. Science of the
Total Environment 334–335, 287–298.
Int Panis, L., de Geus, B., Vandenbulcke, G., Willems, H., Degraeuwe, B., Bleux, B.,
Mishra, V., Thomas, I., Meeusen, R., 2010. Exposure to particulate matter in traffic: a comparison of cyclists and car passengers. Atmospheric Environment 44,
2263–2270.
Int Panis, L., 2011. Cycling: health benefits and risks. Environmental Health Perspectives 119 (3).
Jacobs, L., Nawrot, T.S., de Geus, B., Meeusen, R., Degraeuwe, B., Bernard, A., Sughis,
M., Nemery, B., Int Panis, L., 2010. Subclinical responses in healthy cyclists briefly
exposed to traffic-related air pollution. Environmental Health 9, 64.
Jacobsen, P.L., 2003. Safety in numbers: more walkers and bicyclists, safer walking
and bicycling. Injury Prevention 9, 205–209.
Kim, J.K., Kim, S., Ulfarsson, G.F., Porrello, L.A., 2007. Bicyclist injury severities in bicycle–motor vehicle accidents. Accident Analysis and Prevention 39,
238–251.
Kleinberger, M., Sun, E., Eppinger, R., Kuppa, S., Saul, R., 1998. Development of
Improved Injury Criteria for the Assessment of Advanced Automotive Restraint
Systems. NHTSA, Washington, DC.
Koike, H., Morimoto, A., Kitazawa, A., 2003. Unevenness of intersection pavement
and bicycle safety. Transport Research Record 1846, 56–61.
Loo, B.P.Y., Tsui, K.L., 2010. Bicycle crash casualties in a highly motorized city. Accident Analysis and Prevention 42, 1902–1907.
Luyten, S., Van Hecke, E., 2007. De Belgische stadsgewesten 2001. Statistics Belgium
– Working Paper no. 14. FPS Economy (Direction for Statistics and Economic
Information), Brussels.
McCoy, C.A., 2002. Development and pilot testing of a bicycle safety questionnaire
for adult bicyclists. Public Health Nursing 19 (6), 440–450.
NIS (National Institute for Statistics), 2001. National house hold survey on mobility.
Realization and results. Research contract MD/13/036, FUNDP-GRT-LV-UIAIW-INS. Belgian Federal Science Policy Office, Brussels Available online at:
http://www.belspo.be/belspo/fedra/proj.asp?l=en&COD=MD/DD/18.
NIS (National Institute for Statistics), 2009. CEESE – ULB et Transport & Mobility Leuven – Indicateurs de transport géographiquement désagrégés pour
la Belgique, FOD Economie, Algemene directie Statistiek en Economische
Informatie.
OECD Organisation for Economic Co-operation and Development, 1998, Scientific
Expert Group on the Safety of Vulnerable Road Users (RS7). Safety of vulnerable
road users, DSTI/DOT/RTR/RS7(98)1/FINAL.
Oja, P., Vuori, I., Paronen, O., 1998. Daily walking and cycling to work: their utility
as health-enhancing physical activity. Patient Education and Counseling 33 (1
Suppl.), S87–S94.
Oja, P., Titze, S., Bauman, A., de Geus, B., Krenn, P., Reger-Nash, B., Kohlberger, T.,
2011. Health benefits of cycling; a systematic review. Scandinavian Journal of
Medicine & Science in Sports 21, 496–509.
Osler, T., Rutledge, R., Deis, J., Bedrick, E., 1996. ICISS: an international classification of disease-9 based injury severity score. Journal of Traumatology 41,
380–388.
OVG, 2008. Onderzoek VerplaatsingsGedrag Vlaanderen. Available online at:
http://www.mobielvlaanderen.be/ovg/ovg03.php?a=19&nav=10.
Parkin, J., Wardman, M., Page, M., 2007. Models of perceived cycling risk and route
acceptability. Accident Analysis and Prevention 39 (2), 364–371.
Parkin, J., Wardman, M., Page, M., 2008. Estimation of the determinants of bicycle
mode share for the journey to work using census data. Transportation 35 (1),
93–109.
Phillips, R.O., Ullebergb, P., Vaa, T., 2011. Meta-analysis of the effect of
road safety campaigns on accidents. Accident Analysis and Prevention 43,
1204–1218.
Pro Velo, Observatoire du vélo: Observations 1998–2010, 2010. Available online at:
http://www.provelo.org/IMG/pdf/Observatoire velo RBC rapport final 2010.
pdf (in French).
Pucher, J., Komanoff, C., Schimek, P., 1999. Bicycle renaissance in North America: recent trends and alternative policies to promote bicycling. Transportation
Research Part A 33 (7/8), 625–654.
Pucher, J., Dijkstra, L., 2000. Making walking and cycling safer: lessons from Europe.
Transportation Quarterly 54 (3), 25–50.
Pucher, J., 2001. Cycling safety on bikeways vs. roads. Transportation Quarterly 55
(4), 11–99.
Pucher, J., Dijkstra, L., 2003. Promoting safe walking and cycling to improve public
health: lessons from the Netherlands and Germany. American Journal of Public
Health 93 (9), 1509–1516.
Reynolds, C.C., Harris, M.A., Teschke, K., Cripton, P.A., Winters, M., 2009. The impact
of transportation infrastructure on bicycling injuries and crashes: a review of
the literature. Environmental Health 8, 47.
Richter, M., Otte, D., Haasper, C., Knobloch, K., Probst, C., Westhoff, J., Sommer,
K., Krettek, C., 2007. The current injury situation of bicyclists—a medical and
technical crash analysis. Journal of Trauma 62, 1118–1122.
Robinson, R.L., 2005. Safety in numbers in Australia: more walkers and bicyclists,
safer walking and bicycling. Health Promotion Journal of Australia 16 (1),
47–51.
Rietveld, P., Daniel, V., 2004. Determinants of bicycle use: do municipal policies
matter? Transportation Research Part A 38, 531–550.
Rosenkranz, K.M., Sheridan, R.L., 2003. Trauma to adult bicyclists: a growing problem
in the urban environment. Injury: International Journal of the Care of the Injured
34, 825–829.
Rutledge, R., Osler, T., Emery, S., Kromhout-Schiro, S., 1998. The end of the Injury
Severity Score (ISS) and the Trauma and Injury Severity Score (TRISS): ICISS,
an International Classification of diseases, Ninth Revision based prediction tool,
outperforms both ISS and TRISS as predictors of trauma patient survival, hospital
charges, and hospital length of stay. Journal of Trauma 44, 41–49.
Stone, M., Broughton, J., 2003. Getting off your bike: cycling accidents in Great Britain
in 1990–1999. Accident Analysis and Prevention 35 (4), 549–556.
Stutts, J.C., Williamson, J.E., Whitley, T., Sheldon, F.C., 1990. Bicycle accidents and
injuries: A pilot study comparing hospital- and police-reported data. Accident
Analysis and Prevention 22 (1), 67–78.
Thompson, M.J., Rivara, F.P., 2001. Bicycle-related injuries. American Family Physician 63 (10), 2007–2014.
Tin Tin, S., Woodward, A., Ameratunga, S., 2010. Injuries to pedal cyclists on New
Zealand roads, 1988–2007. BMC Public Health 10, 655.
van Boggelen, O., Janssen, R., Everaars, E., 2005. Effect toename fietsaandeel
op de verkeersveiligheid. Adviesdienst Verkeer en Vervoer, Rotterdam, the
Netherlands (in Dutch).
Vandenbulcke, G., Thomas, I., de Geus, B., Degraeuwe, B., Torfs, R., Meeusen, R., Int
Panis, L., 2009. Mapping bicycle use and the risk of accidents for commuters who
cycle to work in Belgium. Transport Policy 16, 77–87.
Please cite this article in press as: de Geus, B., et al., A prospective cohort study on minor accidents involving commuter cyclists in Belgium.
Accid. Anal. Prev. (2011), doi:10.1016/j.aap.2011.09.045
G Model
AAP-2600; No. of Pages 11
ARTICLE IN PRESS
B. de Geus et al. / Accident Analysis and Prevention xxx (2011) xxx–xxx
Vandenbulcke, G., Dujardin, C., Thomas, I., de Geus, B., Degraeuwe, B., Meeusen,
M., Int Panis, L., 2011. Cycle commuting in Belgium: spatial determinants and
‘re-cycling’ strategies. Transportation Research Part A 45, 118–137.
Veisten, K., Saelensminde, K., Alvaer, K., Bjørnskau, T., Elvik, R., Schistad, T., Ytterstad,
B., 2007. Total costs of bicycle injuries in Norway: correcting injury figures and
indicating data needs. Accident Analysis and Prevention 39 (6), 1162–1169.
11
Vuori, I.M., Oja, P., Paronen, O., 1994. Physically active commuting to work – testing
its potential for exercise promotion. Medicine and Science in Sports and Exercise
26 (7), 844–850.
Woodcock, J., Banister, D., Edwards, P., Prentice, A.M., Roberts, I., 2007. Energy and
transport. Lancet 370 (9592), 1078–1088.
Please cite this article in press as: de Geus, B., et al., A prospective cohort study on minor accidents involving commuter cyclists in Belgium.
Accid. Anal. Prev. (2011), doi:10.1016/j.aap.2011.09.045