Public Health Agency of Canada / Agence de santé publique du Canada
Skip all navigation -accesskey z Skip to sidemenu -accesskey x Skip to main menu -accesskey m Skip all navigation -accesskey z
Français Contact Us Help Search Canada Site
PHAC Home Centres Publications Guidelines A-Z Index
Child Health Adult Health Seniors Health Surveillance Health Canada

Volume 17, No.2 -1997

 [Table of Contents] 


Public Health Agency of Canada (PHAC)

Bicycle-related Injuries in Children: A Study in Two Ontario Emergency Departments, 1994
Hardy J Schwartz and Robert J Brison

Objective: To identify risk factors associated with bicycling injuries, and to further define the magnitude of this problem. Design: Information from the database of the Kingston and Region Injury Surveillance Program (KRISP)-the Kingston reporting centre for the Canadian Hospitals Injury Reporting and Prevention Program (CHIRPP)-was used to identify all bicycling injuries in children who were treated at participating emergency departments in 1994. A follow-up mail survey was then used to obtain further information about the incidents. Setting: Two southeastern Ontario emergency departments that serve a regional population of 125,000 people. Focus Population: Children and youth aged 0-19. Results: The KRISP database documented 348 bicycling injuries in this population. Mail surveys were sent to 280 subjects and completed by 163 of these children or their parents. Our descriptions of the circumstances of the injury events and of the most frequent resulting injuries are similar to patterns in previously published reports. Sixty percent of the children who returned surveys indicated they were injured on bicycles they had been riding for less than one year, even though most of these children were experienced bicycle riders. We estimated that there were 9.4 injury events per 1000 children in the Kingston region. Conclusions: Bicycling injuries remain a significant cause of childhood morbidity and mortality. We found that many children were injured on bicycles with which they were not familiar. Consequently, we propose as a preventive measure that children be targeted at the time of purchasing a new bicycle for helmet purchase, bicycle fitting and enrollment in safety courses.

Key words: Bicycling; Ontario; pediatrics; wounds and injuries


There were 212 bicycle-related deaths in Ontario recognized by the coroner's office between 1986 and 1991.1 More than half of all bicycling fatalities involve children under the age of 20, with head injury being the most common cause of death or hospital admission.1-7 Consequently, bicycling injuries have become a major focus of pediatric public health education and prevention.

The high proportions of fatalities and hospital admissions due to head trauma make helmet compliance an important target of intervention. However, helmet use among children continues to be, at best, poor.8,9 In addition, helmets provide no protection against fractures and soft-tissue damage to extremities, which have been well documented as the most common bicycling injuries.4,6,7,10,11 Hence, additional study is required to identify measures of prevention and protection in order to decrease the morbidity, mortality and health care costs associated with this popular activity.

This study was undertaken to provide a population-based descriptive analysis of all bicycle-related injuries that presented over a one-year period to the only two emergency departments serving the 125,000 people in the Kingston, Ontario, region (the Kingston General and Hotel Dieu hospitals). We identified the factors contributing to the bicycling incidents, described the frequencies of injuries with respect to type, location and severity, and assessed characteristics of the bicycle and the cyclist, including safety practices before and after the injury event.


We used data from the Kingston and Region Injury Surveillance Program (KRISP), the southeastern Ontario regional reporting centre for the Canadian Hospitals Injury Reporting and Prevention Program (CHIRPP), which has been previously described.12 CHIRPP collects and codes information on the circumstances surrounding the injury occurrence using a questionnaire completed on a voluntary basis by patients (or family/friends). KRISP identifies 100% of all injured persons who report to its participating hospitals. Of these patients, 85% agree to complete the CHIRPP questionnaire, and about 85% of these respondents note on the form that they will agree to further telephone follow-up.

From the KRISP database, we selected all patients under the age of 20 who presented in 1994 with a bicycle-related injury. This provided a four-season sample of all injured child cyclists from a region that includes both urban and rural settings.

We then sent a questionnaire to all patients (or their parent/guardian if they were under the age of 16) who had consented to follow-up on the original CHIRPP form. A second mailing was sent one month later to those who did not reply to the first survey. The person who had been injured was asked to complete the questionnaire if possible.

We asked for information regarding four main areas that may have contributed to the injury: 1) the circumstances surrounding the injury event, 2) the bicycle on which the patient was injured, 3) the patient's riding experience, and 4) helmet use and safety precautions (see Table 1). Descriptive analyses related to helmet use were done using data from the mail survey, as CHIRPP data are likely associated with underreporting of the use of safety devices.

The KRISP data and the mail survey data were combined and analyzed using the Epi Info Database-Statistics System program, version 5.0 (public domain software, CDC, Atlanta).

Injury rates were calculated using the 1991 population of Kingston and Frontenac counties as the denominator. Only those cases with residency postal codes in these counties (for which the two study hospitals serve as the primary care institutions) were included in the numerator for the rate analysis.


In 1994, 348 patients under age 20 were recorded in the KRISP database as having presented with bicycling injuries. The mail survey was sent to the 280 patients (80%) who had given consent to follow-up on their CHIRPP questionnaire. After two mailings, 163 (58%) responses were received. The mail questionnaire was completed by the bicyclist in 45% of the cases; the mother, 44%; the father, 10%; and a grandparent, 1%.

Table 2 shows the demographics for the original 348 patients (group A) and the 163 patients (group B) who completed the mail survey. The age and sex distributions were similar between the responders and the original case group. The largest group of patients consisted of children aged 5-14 years (n = 269). Sixty-seven percent were males.

Topics questioned in a survey sent to 280 patients injured in bicycling incidents, children aged 0-19, Kingston and Region Injury Surveillance Program, 1994

Cause of the injury event
Location where the incident occurred
Conditio n of the riding surface
Grade of the riding surface
Use of helmet at time of injury event
Use of any other protective equipment

Safety features
Last maintenance period before the injury event
Frequency of maintenance
Frame size

III. PATIENT'S RIDING EXPERIENCE Overall riding experience (years)

Experience riding the bicycle being ridden (months)
Skill level (beginner, intermediate, advanced)
Riding frequency (times/week and miles/week)
Site(s) of most riding
Long-term medical problems acquired due to injuries
Previous bicycling incidents and the injuries acquired

IV. SAFETY PRECAUTIONS AND HELMET USE Helmet ownership and use before the injury event

Helmet ownership and use after the injury event
Source(s) of any previous safety advice about bicycling

V. DEMOGRAPHICS Age, sex, height, weight

Survey responder (patient, mother, father, other)

Demographics of 348 children who presented with bicycling-related injuries to the KRISP emergency department reporting centres in 1994
  Group Aa
(n = 348)
Group Bb
(n = 163)
Mean age (years) 11 11
Age groups:
number of cases (and %)
1-4 10 (3) 5 (3)
5-9 122 (35) 57 (35)
10-14 147 (42) 80 (49)
15-19 69 (20) 21 (13)
Male 67% 67%
a Group A = all 348 patients in the KRISP database b Group B = 163 patients who returned mail surveys

    We estimate the regional rate of bicycling-related injuries (brought to emergency medical attention) in children to be 9.4 events per 1000. Age-specific rates are provided in Table 3.

Bicycle-related injury rates (per 1000 population), children aged 0-19, Kingston and Region Injury Surveillance Program, 1994
  This study
(Kingston, 1994)
Thompson et al.8
(Seattle, 1990)
Age groups (years) Injuries
rate a
rate a
1-4b 8 1.2 12 0.8
5-9 107 12.7 102 6.7
10-14 120 15.4 116 8.1
15-19 57 6.8 43 2.6
ALL AGES(1-19) 292 9.4 364  
a Rates for females were approximately half those for males in all age groups
Data from Thompson et al. includes children aged 0-1 year

    CHIRPP Data Findings (n = 348): Circumstances of the Incidents and Breakdown of the Resulting Injuries

The 348 separate bicycling incidents resulted in a total of 432 injuries, with no deaths. We present data on only the primary injuries. All of the secondary injuries were minor.

Tables 4 and 5 provide breakdowns for the locations where the injury events occurred, the causes of these events, the injuries acquired and the body parts injured. One hundred and ninety-four of the incidents (56%) happened on a public road. Eighty-nine percent of the injuries required only minor treatment (e.g. dressings or sutures). Sixteen patients (5%) were admitted: ten with fractures, three for testicular hematomas, two for blunt intra-abdominal trauma and one for a concussion.

Location and cause of 348 bicycling incidents, children aged 0-19, Kingston and Region Injury Surveillance Program, 1994
Location of incident Number (%)
Public road 194 56
Yard/driveway 53 15
Public park/footpath 34 10
Parking lot/vehicle area 21 6
Playground/school/sporting area 11 3
Field/camping area 6 2
Other 29 8
TOTAL 348 100
Cause of incident Number (%)
Loss of control 266 76


with motor vehicle
24 7
with bicycle
9 3
with object/person
18 5
Fall 9 3
Stunt/horseplay 4 1
Equipment malfunction 4 1
Other 4 4
TOTAL 14 100

    There were 78 persons whose primary injury was to the head and neck (22%). Of these injuries, lacerations and abrasions made up 50%; minor head injuries and concussions, 25%; soft-tissue injuries and sprains, 22%; and there were two nasal fractures. Only one of the head injuries required hospital admission (for a concussion). Thirty-seven of the head injury events were caused by loss of control and ten were due to collisions with motor vehicles. The 16 other events involving motor vehicle collisions resulted in fractures and soft-tissue injuries, mainly to the extremities; 2 of those extremity fractures required admission.

Injuries acquired and body part injured in 348 bicycling incidents, children aged 0-19, Kingston and Region Injury Surveillance Program, 1994
Type of injury Number (%)
Soft-tissue injury:    
37 11
27 8
82 24
Fracture 85 24
Laceration 65 19
Concussion/unspecified closed head injury 23 7
Sprain 12 3
Dental injury 4 1
Dislocation 3 1
Other 7 2
LWBSa 3 1
TOTAL 348 101
Body part injured Number (%)
Upper extremity 154 44
Lower extremity 79 23
Head and neck 78 22
Trunk 31 9
Dental 4 1
Other 2 1
TOTAL 348 100


Mail Survey Findings (n = 163): Characteristics of the Bicycle, Cyclist's Riding Experience and Safety Behaviour

Mail surveys provided additional information regarding 163 of all injury events (47%). The breakdown of time, location and causes of these crashes, and the frequencies for the nature, body part and treatment of the injuries were similar to those reported above for the original 348 cases. Seventy-nine children (48%) indicated that they did most of their riding on roads and streets. In our survey, 74 (45%) reported wearing a helmet at the time of the injury event, while only 31 (19%) of these 163 patients reported helmet use on the CHIRPP form.

A comparison of the anatomic injuries sustained was made between the 74 helmeted patients and the 89 without helmets. Lacerations and abrasions were, again, the most common head and neck injuries in both groups. Of the nine minor head injuries and three injuries classified as concussions, only three minor head injuries occurred in helmeted patients.

Twenty-two children (13%) indicated having no experience riding the bike on which they were injured, forty-five (28%) indicated 6 months or less and another twenty-nine (18%) indicated 12 months or less. Therefore 96 of the injury events (59%) involved children who had been riding the bike for no more than one year (see Table 6). However, 42 of those 96 children had more than five years of riding experience, and an additional 35 had more than one year of experience. Only 22 of those children (23%) had their skill level rated as " beginner" on their questionnaire.

For the 96 cases with less than one year's experience on their bicycle, 55 of the bicycles were one year old or less. The bicycle age was unknown for 16 cases, many involving the use of a borrowed bike. Fifty-five of the bikes had received maintenance within the previous six months. The most common cause of the crash in these 96 children was loss of control, and this did not vary much between those who had more or less than one year's experience on the bike. They had comparable riding frequency habits. We could not determine whether or not these children were riding bicycles that were too large because accurate data on the bicycle frame size was not provided for many of the cases.

These results suggest that these 96 children, who most likely knew how to ride, were injured on bikes with which they were not familiar. If bike age and time of last maintenance are good predictors of bike condition, then poor condition of the bicycles seems an unlikely contributing factor in most of these incidents.

Even though only 74 of the 163 children (45%) were wearing a helmet at the time of the injury event, 119 (73%) did own a helmet. Of the 44 children who did not even own a helmet before, 26 (59%) had acquired one after the event; 5 of those 26 cases involved head injuries. Overall, 82 children (50%) reported wearing a helmet all or most of the time when they rode before the injury event, and 99 (61%) did so after the incident.

One hundred and fifty-three of the children or their parents (94%) in our survey group had at some time received advice about bicycling safety. The most common sources of information were at school (89% of those who reported receiving any information), from family or friends (52%) and on television (38%). Only 13 of these children (8%) had received safety advice from an emergency physician, and only 7 (4%) from a family physician; these were the two least common sources of advice.

Familiarity with new bicycles: breakdown on 163 bicycling incidents according to riding experience on the bicycle involved

  < or = 12 months
on the bike
(n = 96)
> or = 12 months
on the bike
(n = 67)
Age group (years)
1-4 3 3
5-9 45 20
10-14 45 55
15-19 7 22
Riding experience (overall)
None 2 -
<6 months 5 -
6 mo-1 yr 12 2
1-5 years 37 32
>5 years 44 66
Skill level (self-rated)
Beginner 23 11
Intermediate 51 44
Advanced 24 41
Unknown 2 4
Bicycle age (months)
0 (brand new) 4 -
1-6 29 -
7-12 23 5
>12 27 94
Unknown 17 1
Last maintenance (months before incident)
0-6 56 50
7-12 8 14
>12 36 36
Riding frequency (times/week)
0-3 18 23
4-7 65 52
>7 11 11
Unknown 7 14



We have described all 348 bicycle-related incidents and the resulting injuries that presented to the only two emergency departments serving the region of Kingston, Ontario. Our data provide a population-based sample of injured bicyclists who sought medical care in the emergency department setting in 1994. We believe the injury rates described here are conservative since persons living at the periphery of our regions are likely to have accessed other sites of medical care. These rates also underestimate the bicycle injury problem because they only include those persons seeking medical care in the emergency department. Yet the rates reported here are nearly twice those reported by Thompson et al.8 Due to the comprehensive injury reporting system in place in our emergency departments and the geographically distinct population they serve, we feel the bicycle injury rates cited here are the most accurate rates currently available.

This study is a descriptive series of cases identified by our CHIRPP site. Data on control children were not available to permit the quantification of risks for other than demographic factors. More detailed information on exposure variables for both cases and controls would be needed, and this was beyond the scope of this project.

In general, our findings support previous reports describing bicycling injuries. Most bicycling injury events are due to rider error; children aged 5-14 are injured the most frequently; and the most common injuries are soft-tissue damage or fractures to the upper extremity.

Even though our study found that most bicycle-related injuries were minor, head trauma and motor vehicle collisions were not associated with more severe injuries and they did not account for the majority of hospital admissions, as other studies have shown.4-7 However, there was not a large number of motor vehicle collisions to permit an adequate analysis of this subgroup. Most of the injuries in our study occurred on a public road, and the children reported that most of their previous riding was performed on public roads or streets. Only 45% were wearing helmets at the time of the incident, and only 50% wore helmets all or most of the time. We recorded 78 head injuries, only one of which required admission. Extremity fractures were by far the most frequent cause of admission. These results show that, even if helmet use were improved, bicycling injury events would continue to represent a significant problem in terms of childhood morbidity and health care costs.

Rider error is clearly responsible for the majority of bicycling injury events. If this—along with helmet compliance—could be improved, then we would likely see a significant concomitant reduction in the number of bicycling injuries. It is plausible that lack of experience on a specific bicycle is an important component of rider error. In 60% of bicycling injuries, we found that the children involved had less than one year's experience on the bicycle being ridden, even though most of those children were experienced riders. There are several possible explanations for this finding.

In most cases the bike was new and/or well maintained. Whether children were riding new or used bicycles, the bicycles may have been too big for them. Information on bicycle size was missing from too many of the cases to be properly assessed. Another explanation might be that more children are injured on new bicycles because they ride more often when they first get their bicycle and are therefore predisposed to a greater risk through increased exposure. If this were so, then we might have been misled to attribute their injuries to lack of experience on the bicycle rather than riding frequency. However, our study found that these children had riding habits that were no different than those of the children who had more experience on the bicycle. They were nearly all recreational riders who rode their bicycles on most days of the week. Although the great majority of them did not answer our question about how many miles they rode, it has been documented that child cyclists bike the fewest miles of any age group.8

The most plausible explanation for why many of the children in our study were injured is simply that they were not familiar with the bicycle they were riding. This suggests that riding a new (or different) bicycle might require more than just implementing old skills in a new situation.

Our study is one of the first to suggest that lack of familiarity with a bicycle, and not necessarily lack of bicycling experience, might contribute to bicycling " accidents" and injuries. Selbst et al. have reported findings similar to our own regarding experience on the bicycle.7 They found that 54% of the children in their study had been riding their bicycle for only six months or less before the injury event, 49% of all the bicycles were less than one year old and 24% were known to be in need of repair. However, Selbst et al. did not report on the overall previous riding experience of the children in their study. In a similar study, Cushman et al. reported that in 63% of their cases the bicycles were less than two years old, and in 74% the bicycle had been repaired the same year as the injury event. They did not report the experience of their riders.6

Both of these studies support our finding that children riding new bicycles are being injured frequently. We find it unlikely that poor bicycle condition would frequently contribute to a crash on a new bicycle. The question therefore remains whether or not children on new bicycles are at risk because of factors associated with the bicycle (e.g. improper size) or the cyclist's experience. It is likely a combination of both factors.

Perhaps children with bicycles new to them should be " refreshed" about safety procedures and given the opportunity to familiarize themselves with their new bicycles in a safe setting. Our study provides evidence that this group of riders should be targeted specifically for rodeos or similar courses that already exist. These measures could be implemented at the time of bicycle purchase, when both the parents and the child are available for bicycling safety awareness education. They could also be encouraged to buy helmets and get properly fitted to the bicycle.

Our study is limited in that it is descriptive in nature and does not address the issue of rider experience with a control group. Our conclusion that lack of familiarity with a bicycle might predispose to an injury event is therefore a hypothesis that should be assessed by a case-control study. The low response rate of 60% to the mail survey is a limitation. This survey was performed during the summer months when fewer people were home. It targeted young patients, often minors, which often required the survey to be completed by proxy (a parent). Our incidence estimates did not take into account ridership exposure. Hu et al. have recently reported incidence rates that incorporated into the denominator exposure by bike ownership, miles ridden and hours ridden.13 We did ask for this information for cases in our survey, but we found that parents had difficulty estimating hours and miles ridden by their children. Despite the noted limitations of our study, we believe that our conclusions point to potentially useful, new ideas in a field that has been limited to only a few preventive measures with poor compliance.

It is clear that helmet use alone will not succeed in preventing the morbidity associated with bicycling injury events, though this is an important component. Our study focused only on children who came to an emergency department with an injury. Of those who were surveyed, only 13% remembered being given bicycling safety advice by the emergency physician. This would seem an ideal opportunity to reinforce the importance of helmet use. With new helmet laws for children now being instituted across Ontario, physicians and retailers may be inspired to share some of the responsibility of educating children and parents, and of preventing unnecessary injuries. We especially hope that results from this study will further prompt preventive measures because it suggests that the rate of these injuries is nearly twice that of previous estimates.

We thank Lisa Hartling for her help in developing this project, Kathy Bowes for her advice and technical assistance, and Louise D'Aoust for her help with the mail surveys. This project was sponsored by the Kingston and Region Injury Surveillance Program (KRISP) and the Canadian Hospitals Injury Reporting and Prevention Program (CHIRPP).


    1. Rowe BH, Rowe AM, Bota GW. Bicyclist and environmental factors associated with fatal bicycle-related trauma in Ontario. Can Med Assoc J 1995;152(1):45-53.

    2. Sacks JJ, Holmgreen P, Smith SM, Sosin DM. Bicycle-associated head injuries and deaths in the United States from 1984 through 1988. JAMA 1991;266(21):3016-8.

    3. Weiss BD. Bicycle-related head injuries. Clin Sports Med 1994;13(1):99-112.

    4. Friede AM, Azzara CV, Gallagher SS, Guyer B. The epidemiology of injuries to bicycle riders. Pediatr Clin North Am 1985;32(1):141-51.

    5. McKenna PJ, Welsh DJ, Martin LW. Pediatric bicycle trauma. J Trauma 1991;31(3):392-4.

    6. Cushman R, Down J, MacMillan N, Waclawik H. Bicycle-related injuries: a survey in a pediatric emergency department. Can Med Assoc J 1990;143(2):108-12.

    7. Selbst SM, Alexander D, Ruddy R. Bicycle-related injuries. Am J Dis Child 1987;141:140-4.

    8. Thompson DC, Thompson RS, Rivara FP. Incidence of bicycle-related injuries in a defined population. Am J Public Health 1990;80(11):1388-90.

    9. Storo W. The role of bicycle helmets in bicycle-related injury prevention. Clin Pediatr 1992;31(7):421-7.

    10. Centers for Disease Control. Bicycle-related injuries: data from the National Electronic Injury Surveillance System. MMWR 1987;36:269-71.

    11. Lofthouse GA. Traumatic injuries to the extremities and thorax. Clin Sports Med 1994;13(1):113-35.

    12. Senzilet L. Children's Hospitals Injury Research and Prevention Program (CHIRPP). Technical report #2. Ottawa: Health and Welfare Canada, Laboratory Centre for Disease Control, Bureau of Chronic Disease Epidemiology; 1991.

    13. Hu X. Wesson DE, Chipman ML, Parkin PC. Bicycling exposure and severe injuries in school-age children. Arch Pediatr Adolesc Med 1995;149:437-41.

Author Reference

Hardy J Schwartz and Robert J Brison, Department of Emergency Medicine, Kingston General Hospital; and Department of Community Health and Epidemiology, Queen's University, Kingston, Ontario
Mailing address: Dr Robert Brison, Department of Emergency Medicine, Kingston General Hospital, 76 Stuart Street, Kingston, Ontario K7L 3V2

[Previous] [Table of Contents] [Next]

Last Updated: 2002-10-29 Top