Objectives. To delineate the mechanism of serious bicycle handlebar-related injuries in children and make recommendations for preventive strategies.
Methods. Prospective cross-sectional surveillance system of seriously injured child bicyclists supplemented by in-depth, on-site crash investigation to delineate specific injury mechanisms. Interdisciplinary analyses involved engineers, clinicians, epidemiologists, and biostatisticians.
Setting. The emergency department and in-patient trauma service of an urban level one pediatric trauma center between October 1995 and September 1997.
Participants. Patients under 18 years of age who were treated for serious bicycle-related injuries (Abbreviated Injury Scale scores of 2 or greater).
Results. The surveillance system identified two distinct circumstances for serious child bicyclist injury: 1) handlebar-related injuries associated with minor incidents (falls from bicycles) and 2) nonhandlebar-related injuries associated with severe incidents (bicycle-motor vehicle crashes). Crash investigations explored the minor incidents that resulted in serious handlebar-associated injuries. In the typical mechanism, as the child lost control of the bicycle and began to fall, the front wheel rotated into a plane perpendicular to the child's body. The child then landed on the end of the handlebar resulting in serious truncal injuries.
Conclusions. A discordancy exists between the apparently minor circumstances and serious injuries sustained by child bicyclists who impact bicycle handlebars. Recognition of the mechanism of handlebar-related injuries might aid the practitioner in early diagnosis of serious abdominal injuries in child bicyclists. This injury mechanism may be avoided through bicycle redesign that would involve both limiting rotation of the front wheel and modifying the ends of handlebars. An integrated approach involving a surveillance system to identify an injury hazard supplemented by in-depth, on-site crash investigations effectively provided the detailed mechanism of injury needed to develop interventions.
Aprimary goal of injury control research is to develop interventions that will reduce the incidence and severity of injuries. The process of translating research into interventions involves identifying injury hazards and elucidating the etiology of the hazard in sufficient detail for those who will develop the interventions. Application of this process has resulted in the identification of head injury as a significant source of morbidity and mortality in child bicyclists and resulted in the development of the bicycle helmet.1–6
Recent reports7–9 indicate handlebars as another source of injury among child bicyclists even among low-speed crashes.7 Effective countermeasures, however, have not yet been developed primarily because of the limited understanding of the injury mechanism. The majority of reports of handlebar-related injuries have been limited to descriptions of treatment strategies, operative techniques, and course of recovery. Impact with handlebars has been documented as producing traumatic abdominal wall hernia;10–13 renal, intestinal, liver, splenic, and pancreatic injuries;7–914–17 abdominal wall rupture;18 abdominal aorta rupture;19transection of the common bile duct;20 traumatic arterial occlusion;21 groin injuries;22,,23 and even death.24 Underlying organ injuries are often occult, as external bruising is infrequently present, and the signs and symptoms of organ injury do not present for hours.7 Although these reports have recognized the role of the handlebar in child bicyclist injury, there is insufficient information regarding the detailed mechanism of injury.
For effective interventions to be developed, the mechanism of handlebar-related injury must be understood. In the current study we proposed to 1) identify the basic circumstances surrounding child bicyclist injuries through use of a surveillance system and 2) elucidate the detailed mechanism of injury with on-site, in-depth crash investigations incorporating the expertise of engineers, epidemiologists, clinicians, and biostatisticians.
Injury Circumstance Evaluation (ICE) Study of Bicyclists, Pedestrians, and Motor Vehicle Occupants
The goal of the ICE Study is to identify significant injury hazards to children and to elucidate the mechanism of the injury hazard with sufficient precision to allow the development of interventions. To achieve this goal, the ICE Study combines a prospective, cross-sectional surveillance system of injured children and on-site, in-depth crash investigations to identify specific injury mechanisms. Previous results of the ICE Study revealed the mechanism of airbag-related deaths and serious injuries in children.25–27
The ICE Study recruits pediatric patients <18 years of age who were transported by Emergency Medical Services (EMS) personnel and treated for injuries sustained as bicyclists, pedestrians, or motor vehicle occupants at a level one urban pediatric trauma center servicing a five-county region. Excluded from the ICE Study are children for whom a history cannot be obtained. Consent for inclusion in the study is obtained according to a protocol approved by the Institutional Review Board of The Children's Hospital of Philadelphia.
The ICE surveillance system incorporates data collected from self-administered surveys completed by EMS personnel upon delivery of injured children to the emergency department and surveys administered by two trained parahealth professionals to children, witnesses, and relatives upon the child's admission to the inpatient trauma service for treatment of his or her injuries. For a given child, multiple respondents complete surveys to obtain the most complete history. Discordant responses are resolved by follow-up interviews. Surveys were developed by incorporating the expertise of engineers, clinicians, and epidemiologists through the approach of biomechanical epidemiology.28 Biomechanical epidemiology is a new concept that combines the strengths of engineering, medicine, and epidemiology through the design, execution, and analysis of injury research.
The ICE surveillance system is analyzed to identify circumstance-injury patterns. Cases representing these patterns are then subjected to on-site, in-depth crash investigation by Dynamic Science, Inc, (Washington, DC) a professional crash investigation team. Data collected by the investigators include scene and vehicle evidence and child bodily contact points that are used to determine how the injury event occurred and the kinematic movement of the child in response to the event. These data are analyzed independently by two professional investigators to elucidate detailed mechanisms of injury.
Reconstruction of the injury-producing events involves standardized procedures that relate scene and vehicle evidence to the injuries received by the child. Differences in interpretation are resolved by a third investigator. Protocols are available on request.
Descriptions of the injuries sustained are obtained from the medical record and coded according to the International Classification of Diseases, 9th Clinical Modification29 and Abbreviated Injury Scale (AIS) codes (see definition below) by two separate individuals (trained research assistants), one of whom is blinded to injury mechanism. Discrepancies are resolved by joint review of the medical records. Additional information obtained from the medical record includes gender and age.
The AIS30 and the Injury Severity Score (ISS)31were used to classify bicyclist injury severity. The AIS rates the severity of an anatomic injury from 1 (minor injury) to 6 (fatal injury) for each of six body regions: head/neck, face, chest, abdomen, extremities/pelvic girdle, and external. A score of 0 was assigned to patients who sustained no injuries. Because multiple injuries might occur within one or several body regions, Maximal AIS (MAIS) was defined for each body region and overall for any body region and was recorded for each subject. All subjects in this study had a MAIS of ≥2. The ISS was developed to account for overall injury severity. ISS is the sum of the squares of the three highest AIS scores, thereby accounting for multiple injuries sustained.
ICE Surveillance Data Collection for the Current Study
The subjects for the current study were the subset of the ICE Study population who were seriously injured bicyclists (AIS ≥2) treated at our regional trauma center. Only those more seriously injured children were included because they constitute a well-defined population who are triaged to the center for definitive trauma care.32,,33
As mentioned, the ICE Study uses a series of surveys to obtain detailed trauma histories that describe the injury circumstances. Answers used in the current study involve time of incident, descriptions of the sites of body impact, bicycle crash type (eg, struck and thrown on motor vehicle), direction of impact and fall, surfaces impacted, speed of vehicles, and other circumstance information. Specific questions addressing handlebar involvement in injury included anatomical descriptions of where the handlebars made impact with the child's body and what object the child's body first made impact. The questionnaire is available from the authors on request.
Child bicyclists were systematically classified into two impact groups based on survey responses. The handlebar group was composed of those whose survey responses indicated that they made impact with the handlebar. The remaining bicyclists were classified into the nonhandlebar impact group.
Survey responses were systematically classified into three event severity categories. Bicyclist collisions with a moving motor vehicle were classified as severe; bicyclist collisions with a stationary object (eg, pole or parked vehicle) or another bicyclist were classified as moderate; bicyclist collisions in which the child simply fell off the bicycle were classified as minor.
Gender was compared between the two impact groups (handlebar and nonhandlebar) by a Fisher's Exact test. Age was compared using a Student's t test for independent samples. Distribution of the most severely injured body regions was compared using an exact χ2 test. MAIS and event severity were compared between impact groups using a Kruskal–Wallis Exact test. ISS was compared between impact groups using a Mann–Whitney U test. Analyses were conducted using SPSS and StatXact. Statistical significance was set at P < .05.
On-site, In-depth Crash Investigations for the Current Study
Analysis of the surveillance data identified minor bicycle- related incidents resulting in serious injuries. Cases with this discordant circumstance-injury pattern for which a bicycle was available for inspection and consent was obtained were identified for targeted study using on-site, in-depth crash investigations for the determination of injury mechanisms. Specialized data collection forms were developed to record scene, vehicle, and injury data for injured child bicyclists and were used for the current study.
From October 1995 through September 1997, 107 bicyclists with MAIS 2 or greater injuries were identified from the ICE Study and formed the cohort of seriously injured child bicyclists for the current study. Over half (59%) of the 107 children in the study had multiple injuries. Overall, the 107 children in the study sustained a total of 190 injuries. These injuries were distributed by body region as follows: 46 head/neck injuries, 7 facial injuries, 5 chest injuries, 18 abdominal injuries, 61 extremity injuries, and 53 external/superficial injuries.
The distribution of the most severely injured body region for each child based on handlebar impact grouping is presented in Table 1. The handlebar impact group is comprised of children with injuries distributed among most body regions, but the majority of serious injuries occurred within the abdominal region. The nonhandlebar impact group had injuries distributed among all body regions, but the most severe injuries were to the extremities and head (P < .0001).
The abdominal injuries in the handlebar impact group were as follows: six splenic lacerations, two liver lacerations, three kidney injuries (two lacerations and one hematoma), and two pancreatic lacerations. Other handlebar impact-related injuries included one pneumothorax, one thigh impalement, one closed head injury with basilar skull fracture, and one radius fracture.
The distribution of demographic variables, MAIS, ISS, and event severity between the two impact groups is presented in Table 2. Overall, the majority of injured bicyclists were male (78%) and the average age was 10 years. There were no significant differences in gender and age for the handlebar impact and nonhandlebar impact groups. The handlebar and nonhandlebar impact groups had similar severities of injury as demonstrated by similar MAIS and ISS (P = .51 andP = .28, respectively). Median MAIS and ISS were 2 (range, 2–4) and 5 (range, 4–16), respectively, in the handlebar impact group and 2 (range, 2–5) and 5 (range, 4–34), respectively, in the nonhandlebar impact group.
There was a significant association between the event severity and impact group (P < .0001). No severe events (bicycle-motor vehicle crashes) resulted in handlebar impact; rather, the more minor events (falls from bicycles) were associated with handlebar impact.
Thirteen (76.5%) of the handlebar impact-related crashes occurred as a result of five minor crash types: 1) when the bicyclist lost balance after hitting a discontinuity in the riding surface (sidewalk pothole, grass-sidewalk interface, or curb), 2) when the child braked suddenly, 3) after the chain disengaged from the wheel (chain pop), 4) during a performance of a stunt, or 5) when the rider suddenly turned the front wheel. The four remaining handlebar impact-related crashes occurred under moderate crash circumstances in which the rider made impact with a stationary object or other bicyclist.
In both minor and moderate circumstances, upon falling from the bicycle, the child made impact with the handlebar end, the stemcrown (Fig 1), or crossbar (Fig 2). In 12 (70.6%) of the cases, the child made impact with the end of the handlebar. Two made impact with the stemcrown or crossbar. In the remaining three specific handlebar impact was unknown.
Surveillance data regarding handlebar-related injuries were confirmed and supplemented by in-depth, on-site crash investigations. Seven cases of handlebar-related injury were selected for investigation because the bicycle was available for inspection and consent was obtained. Events investigated included three bicycle mechanical failure, two loss of control, and two stunts. All of the bicycles were appropriately sized for the riders. Four subjects impacted the handlebar ends; 1 subject impacted the stemcrown; 1 subject impacted the crossbar; and 1 subject impacted either the handlebar end, the crossbar, or both. Five of the bicycles were stunt-type; two of which were self-assembled with parts from a variety of bicycles. The handlebars on the five stunt-type bicycles were regular handlebars with an exaggerated curvature toward the rider. The handlebars on the other two bicycles were the same regular handlebars with less curvature toward the rider. Protective rubber covering was present on six of the handlebar ends; none had shock-absorbing foam padding on the handlebar ends. One bicycle had exposed metal handlebar ends (Fig 3). The stemcrown (Fig 1) and crossbar (Fig 2) onto which two of the bicyclists fell were not protected or padded.
One mechanism of injury was common to four of the seven bicyclists: one stunt rider, two bicyclists who lost control, and one whose bicycle sustained mechanical failure. As the child began to fall, the front wheel rotated into a plane perpendicular to the child's body. The child then landed on the end of the handlebar and serious truncal injuries resulted. The injuries sustained included two liver lacerations, one renal laceration, and one splenic laceration. None of these involved penetration of the skin and five of the seven bicyclists had external bruising.
In two cases of mechanical failure, the child fell and landed on the crossbar or stemcrown. Injuries sustained included a flail chest with pneumothorax and a pancreatic fracture, respectively. In the case involving a rider performing a stunt, it could not be determined if the impact was with the crossbar or the handlebar end. In this case, a splenic laceration resulted. A typical case of impact with the handlebar end is presented below.
A 6-year-old boy arrived at our hospital via interhospital helicopter transport for treatment of a liver laceration. The child was riding his stunt bicycle at 4 miles per hour when his bicycle hit a discontinuity in the sidewalk. As the front tire began to cross-the concrete-grass interface, the child lost momentum, causing the front wheel to turn toward the right exposing the child's abdomen to the right handlebar, whose end was exposed rusted metal (Fig 3). He subsequently lost his balance and fell onto the right handlebar; together the bicycle and child fell to the ground. The child then lay at rest until a neighbor came over and helped him.
The child's mother called the pediatrician and described her son's abdominal pain. The pediatrician advised the mother to give her son acetaminophen and call back in a couple of hours. Disregarding the physician's advice, she brought her son to a local emergency department where an abdominal computed tomogram revealed a liver laceration that virtually transected the right and left hepatic lobes. He was immediately transferred by helicopter to The Children's Hospital of Philadelphia.
On arrival at our hospital, his heart rate was 124, respiratory rate was 32, and his blood pressure was 129/68. His Glascow Coma Score was 15 and his trauma score was 15 (−1 respiratory rate). On admission, he was febrile and the physical examination revealed a tender, distended abdomen. Laboratory studies were significant for hemoglobin of 10 g/dL (nL: 11.5–15.5 g/dL), sodium of 152 mmol/L (nL: 136–142 mmol/L), potassium of 3.32 (nL: 3.8–5.0 mmol/L), ionized calcium of 1.12 mmol/L (nL: 1.15–1.34 mmol/L), plasma glucose of 205 mg/dL (nL: 65–121 mg/dL), and albumin of 2.2 g/dL (nL: 3.5–5.2 g/dL). He was initially admitted to the pediatric intensive care unit, but because of tachycardia he was fluid-resuscitated and brought to the operating room where he received 3000 mL of packed red blood cells, 3000 mL of crystalloid solution, 1800 mL of fresh frozen plasma, and 650 mL of platelets. He underwent an immediate exploratory laparotomy during which a right hepatic lobecetomy and cholecystectomy were performed. His total blood loss was >3 liters. A T-tube was placed in the common bile duct, and he was returned to the pediatric intensive care unit. His postoperative course was initially unremarkable with discharge to home 11 days later after T-tube removal.
One week after discharge, the patient was readmitted complaining of abdominal pain and fever. An ultrasound revealed fluid collection in the right upper quadrant of his abdomen. Laboratory values were significant for a white blood cell count of 21.3 thou/μL (nL: 6.0–17.0 thou/μL), platelets of 659 thou/μL (nL: 150–400 thou/μL), segmented neutrophils of 82% (nL: 30–55%), and lymphocytes of 14% (nL: 30–55%). He was started on intravenous fluids and antibiotics. His fever continued. A chest radiography revealed a right pleural effusion whose drainage slowly tapered during his hospital stay. Fourteen days after admission he had minimal abdominal pain, and he was discharged to home.
This is the first study of pediatric handlebar- related injuries in which detailed circumstance and clinical data were collected prospectively to delineate the mechanism of this injury in sufficient detail for redesign of handlebars. This study used effectively a surveillance system supplemented by on-site crash investigations to delineate the typical injury mechanism in which a child, who lost control of the bicycle, began to fall. During the fall, the front wheel rotated into a plane perpendicular to the child's body. The child then landed on the end of the handlebar resulting in serious truncal injuries. Additional handlebar impacts resulted from bicycle mechanical failure. In these cases the child fell and landed on an unpadded stemcrown or crossbar.
Serious bicyclist injury from handlebar impact occurs with a history of an apparently minor incident, usually a fall. A previous case-control study of nonsevere child bicyclist injuries similarly found that minor circumstances, including low bicycle speed and riding on a sidewalk, were associated with emergency department visits. However, an explanation for these findings was not provided.34 In the present study's cohort of child bicyclists, 16% of serious injuries resulted from handlebar impact and none involved collision with a motor vehicle. This is in contrast to the remaining 84% of serious child bicyclist injuries that primarily involved bicycle-motor vehicle collisions.
The frequency of handlebar impact in producing serious injuries in child bicyclists was confirmed by a multi-institutional study of child bicyclist injuries reported to the National Pediatric Trauma Registry. Ten percent of the bicyclists enrolled in that study impacted the handlebars. In the National Pediatric Trauma Registry Study, none of the handlebar- related injured subjects sustained a head injury. Furthermore, among the nonhead-injured bicyclists, these handlebar impacts accounted for 22% of the reported injuries, thereby representing a significant source of injury. In agreement with the results of our study, none of these handlebar-related events involved a motor vehicle. (Baker SP, Fowler CJ, Winston FK, Li G, DiScala C.Sequelae of Head Injury in Child Bicyclists; Phase II.Submitted to: The George Snively Research Foundation. The Johns Hopkins University Center for Injury Research and Policy. October 1997).
Our results also appear relevant to the findings of Clarnette and Beasley7 who found that serious handlebar-related injury involves low-velocity crashes and often results in abdominal injury that may not be symptomatic until several hours after the injury. Because abdominal injuries can be occult, they may be missed by the diagnosing physician and the need for a thorough trauma history is essential to maximize efficient care.35 The case report presented in this study specifically highlights how handlebar injuries can become a missed or delayed diagnosis. Recognition of the mechanism of handlebar-related injuries might have aided the pediatrician in early diagnosis of a liver laceration. The evaluating physician should ask a series of questions regarding the bicycle crash in an effort to determine if the handlebar was involved in producing injury. If a handlebar were implicated in the injury causation, a follow-up history and physical eliciting signs and symptoms of shock should be pursued.
The case report of the 6-year-old child also brings attention to the use of the stunt bicycle in the home environment. Five of the bicycle crashes subjected to in-depth investigation involved stunt-like bicycles. The potential hazard of stunt bicycles in producing handlebar-related injuries is corroborated by the work of Sparnon et al in 198223 and Sparnon and Ford in 198616 in which stunt bicycles were implicated in scrotal injuries and intra-abdominal system injuries, respectively. These studies and the results of the current study elicit concerns about whether the home environment is appropriate for stunt bicycle use.
In addition to the type of bicycle, the size appropriateness of the bicycle for the child is important. Improper bicycle sizing may impede the child's handling of the bicycle, potentially predisposing the child to falling, and may expose more of the child's trunk to the handlebar. Although improper bicycle sizing was not apparent in this study among those cases subjected to in-depth crash investigation, the potential consequence of improperly sizing a bicycle for a young rider must be considered.
The importance of the trauma history has been demonstrated previously elsewhere. Trauma histories focusing on discordancies between injury mechanism and actual physical damage to the victim have been essential in the detection of child abuse36,,37 and domestic violence.38 Similarly, trauma histories have identified the role of air bags in child occupant fatalities in low speed crashes.25–27 The emergency department can provide essential information for injury prevention and treatment if a sufficiently detailed trauma history is obtained.
This study was limited to a single level one pediatric trauma center and included only children with serious injuries to be able to identify specific circumstances associated with these injuries. To obtain incidence and prevalence data regarding bicycle handlebar-related injuries, further study should include data from the community regarding children with no or minor injuries.33,,39Additionally, the sample size of the study does not allow for subgroup analysis of specific handlebar designs and their role in producing bicyclists' injuries. Previous studies performed on handlebar type were limited to describing the presence of protective padding on the ends of the handlebars.8 Further study incorporating a larger data set and expansion of parameters might permit the development of specific recommendations. Additionally, a larger study of the serious handlebar injuries should be conducted to determine whether manufacturers should be encouraged to produce bicycles with curved handlebars, especially for small and younger riders. Such a study could also be used to explore the benefit of design limitations, such as limitation of rotation of the front wheel.
Handlebar-related injuries could potentially be avoided by curving handlebar ends away from the rider and padding the handlebar ends. These handlebar redesigns are recommended for all bicycle types. Limitation of the front wheel rotation is also recommended, but it is recognized that restricting the free rotation of the front wheel might not be acceptable to some stunt bicycle competitors. Consequently, consideration should be given to limiting the use of stunt bicycles to controlled competition settings with experienced riders.
Clinicians can play an important role in the prevention and treatment of handlebar-related injuries. Through anticipatory guidance, clinicians can educate parents about choosing and maintaining their child's bicycle. This education might include: proper sizing of the bicycle to the child; appropriate type of bicycle for the child's age and skill level, maintenance of the bicycle to prevent mechanical failure; and maintenance of handlebar guards. If a child bicyclist is injured in a fall, clinicians should elicit a complete trauma history from EMS personnel, children, and witnesses. Identification of handlebar impact may be essential to identifying serious occult truncal injuries in child bicyclists.
This work was funded by Agency of Health Care Policy and Research Grant 5 U01 HD32828 as part of the Emergency Medical Services for Children Program.
We thank The Children's Hospital of Philadelphia Trauma Service, Philadelphia Fire Rescue, Frances Bents and Robert Schaar at Dynamic Science, Inc, Elisa Moll, Craig Rineer, Harry Jenkins, and Esha Bhatia for their contributions to this research.
- Received July 22, 1997.
- Accepted March 18, 1998.
Reprint requests to (F.K.W.) The Children's Hospital of Philadelphia, Abramson 706, 34th and Civic Center Blvd, Philadelphia, PA 19104.
- ICE =
- Injury Circumstance Evaluation •
- EMS =
- Emergency Medical Services •
- AIS =
- Abbreviated Injury Scale •
- ISS =
- Injury Severity Score •
- MAIS =
- Maximal AIS
- Li G,
- Baker SP,
- Fowler C,
- Discala C
- Sosin DM,
- Sacks JJ,
- Webb KW
- Sacks JJ,
- Kresnow M,
- Houston B,
- Russell J
- Seddon JA
- ↵Winston FK, Reed R. Airbags and children: results of a National Highway Traffic Safety Administration special investigation into actual crashes. In: 40th Staapp Car Crash Conference Proceedings. November 4–6, 1996; Albuquerque, NM. Society of Automotive Engineers, Inc; SAE 962438.27
- Hollands CM,
- Winston FK,
- Stafford PW,
- Lau HT
- ↵Puckett CD, ed. Educational Annotation of the International Classification of Diseases, 9th Clinical Modification. 4th ed. Reno, NV: Channel Publishing Ltd; 1995
- ↵Association for the Advancement of Automotive Medicine. The Abbreviated Injury Scale, 1990 Revision. Des Plaines, IL: Association for the Advancement of Automotive Medicine; 1990
- Baker SP,
- O'Neill B,
- Haddon W,
- Long WB
- Tingvall C
- Copyright © 1998 American Academy of Pediatrics