The aim of the study is to compare trends, circumstances and outcomes of single- versus multi-vehicle bicycle on-road crashes in Victoria, Australia, through the analysis of police records and hospital admissions between January 2004 and December 2008. The results show that over 80% of on-road single-vehicle bicycle crashes occurred as a result of the cyclist losing control of the bicycle with the remainder involving collisions with objects. Compared with multi-vehicle crashes, single-vehicle crashes were more likely to occur in the dark, in wet conditions and in rural areas. Over half of the cyclists hospitalised as result of on-road crashes were injured in single-vehicle crashes and this proportion seems to be increasing over time. Single-vehicle crashes were associated with hospitalised injuries as severe as those resulting from multivehicle crashes. The findings highlight the significant burden of serious injury associated with single-vehicle bicycle road crashes. Further research is needed to investigate in greater detail the risk factors of these crashes and the effectiveness of countermeasures to reduce their burden.
Statistics from Altmetric.com
The increase in the popularity of cycling as a healthy and sustainable mode of transport1 ,2 has resulted in a significant amount of research examining the risk factors for crashes between cyclists and motor vehicles.3–5 In contrast, very little is known about the characteristics of on-road single-vehicle bicycle crashes mainly due to their under-reporting in police crash data.6 ,7
Studies based on healthcare services data have found that single-vehicle bicycle crashes make up the majority of injurious cyclist crashes.8 ,9 The loss of control of the bicycle due to defective or wet road surfaces, poor lighting along with speeding and alcohol use among cyclists have been identified as the main risk factors.10 ,11 Previous studies also found that injuries associated with single-vehicle bicycle crashes are less severe than those involving motor vehicle crashes.8 ,12 However, most studies that examined single-cyclist crashes did not differentiate between on-road crashes and those that occur in off-road locations—two crash types that are likely to have different risk factors.
This study examines police crash records as well as hospital data in order to compare trends, circumstances and outcomes of on-road single-vehicle bicycle crashes to those involving collisions with other road users.
Data on cyclist road crashes in the state of Victoria, Australia, were extracted from police records and hospital admission data for the period January 2004–December 2008. Crashes reported to police included those that occurred on public roads where at least one person was killed or injured and the crash was attributable to vehicle movement. Police data did not include crashes that occurred off road including private property and parks. The data contained information on the demographic (eg age), behavioural (eg helmet use) and environmental (eg road type) crash characteristics.
Hospital cases were extracted based on the International Classification of Diseases (ICD-10) External cause codes V10–V19.13 Single-vehicle crashes were defined as those where the cyclist was injured in a non-collision crash (V18) or in a collision with fixed or stationary objects (V17). The other codes refer to crashes resulting from collisions with other road users. Injury outcomes of crashes were examined based on ICD-10 Chapter XIX (S00, T88). To avoid double counting, transfers to other hospitals, changes in care type within the same hospital and readmission within 48 h for the same injury were excluded.
The focus of the analysis was on-road crashes selected from hospital data based on ICD-10 fourth-character subdivisions, which indicate whether the person was ‘injured in a traffic accident’. A ‘traffic’ or ‘on-road’ crash is defined as any crash occurring on the public highway.13 Characteristics and injury outcomes of on-road single-vehicle cyclist crashes were compared with those of multi-vehicle crashes. In order to examine any significant differences between the two crash types, 95% CIs were computed for all proportions. Ethics approval for the study was obtained from The University of Sydney Human Research Ethics Committee.
There were 317 on-road single-vehicle bicycle crashes reported to police in Victoria between 2004 and 2008. This represents 5% of all 6432 police-reported cyclist crashes. Table 1 provides a comparison between characteristics of police-reported single-vehicle and multi-vehicle bicycle crashes. There were no notable differences in age distribution or helmet use between the two crash types. A higher proportion of single-vehicle crashes occurred on rural roads (13.2%) compared with multi-vehicle crashes (4.5%). Similarly the proportion of crashes that occurred on wet roads and during the dark was almost twice as high for single-vehicle crashes compared with multi-vehicle crashes. There were 40 fatalities (0.65%) associated with multiple-vehicle cyclist crashes compared with five fatalities (1.58%) resulting from single-vehicle cyclist crashes.
Between 2004 and 2008, 2159 cyclists were hospitalised following on-road single-vehicle bicycle crashes, over half (55%) of all hospitalisations for on-road bicycle crashes. The number of hospitalisations for single-vehicle crashes increased by 67% between 2004 and 2008 compared with a 37% increase in hospitalisations for multi-vehicle crashes (figure 1). The age distribution of cyclists hospitalised following single-vehicle road crashes was different to that reported in police data. This difference was particularly marked in the younger age group with the proportion of those aged 19 years and younger much higher in hospital data (31%) compared with police data (12.3%) (tables 2 and 3).
The proportion of head injuries (27%) was similar among cyclists hospitalised for both single- and multi-vehicle crashes. Intracranial injury, including concussion, was slightly more common among casualties of multi-vehicle crashes (10.5%) compared with those in single-vehicle crashes (8.2%). Similarly, the mean hospital length of stay in days was slightly higher for cyclists injured in multi-vehicle crashes (2.6) compared with single-vehicle crashes (2.3).
Most single-vehicle bicycle road crashes in both police records (82.6%) and hospital data (86.7%) resulted from non-collision crashes where the cyclist lost control of the bicycle with the remaining resulting from collisions with fixed or stationary objects (table 4).
While single-vehicle crashes represented only 5.2% of all cyclist road crashes in police data, they represented 55% of cyclist road crashes in hospital data, highlighting the under-reporting of these crashes to police. The difference in age distribution between the two datasets indicates this under-reporting is more common in crashes involving younger cyclists (19 years and younger).
On-road single-vehicle bicycle crash casualties, as a proportion of all on-road cyclist casualties in hospital data, increased from 54.3% in 2004 to 59.2% in 2008. Previous studies indicated similar trends, which have been linked to traffic calming measures reducing the risk of collisions with motorised traffic.14 Other potential explanations include the establishment of cycling infrastructure in Victoria that separates cyclists from other road users.
Both police and hospital datasets show that over 80% of single-vehicle bicycle road crashes occurred as a result of the cyclist losing control of the bicycle with the rest resulting from collisions with objects. Similar findings were reported in previous studies based on police data or surveys of injured riders, which were attributed to loss of control of the bicycle due to poor or defective road surfaces, wet slippery roads, poor lighting, speeding and alcohol use among cyclists, as well as bicycle mechanical failures.10 ,11 A detailed investigation of the crash scene of single-vehicle bicycle crashes found that at least 25% of crashes were related to the visual characteristics of bicycle facilities. The authors recommended the importance of edge markings, especially in curves of bicycle tracks, and improved conspicuity of road furniture.15 None of these studies differentiated between on-road and off road single-vehicle bicycle crashes.
The analysis of police data shows that, compared with multi-vehicle bicycle crashes, single-vehicle crashes were more likely to occur in the dark, in wet conditions and in rural areas. While it is difficult to speculate on these differences, it is possible that relatively lower traffic congestion in these conditions and potential differences in road types in rural areas contributed to a higher proportion of single-vehicle compared with multi-vehicle crashes.
Our study also found that hospitalisations associated with on-road single-vehicle bicycle crashes were as severe as multi-vehicle crashes as the proportion of head injuries and the average length of hospital stay did not vary significantly between these two crash types. Hospital length of stay was found to be a reasonable proxy for serious injury compared with more objective measures of injury severity.16 Previous studies based on healthcare data have indicated that injuries associated with single-vehicle bicycle crashes are less severe than those involving motor vehicles.8 ,12 However, unlike the current study, these studies also included off road crashes.
While the current study provides valuable insight into the characteristics of on-road single-vehicle bicycle crashes, it was based on hospital and police data that have a number of limitations. Police crash data, which contain valuable information on the circumstances of the crash, are known to under-report single-vehicle bicycle crashes and therefore have the potential of selection bias.6 ,7 In addition, contributory factors are based on the attending police officer's subjective opinion at the time of reporting and may not be informed by an extensive investigation of the crash. The hospital data, which better capture the prevalence of serious injury cases, lack information on the circumstances of the crash. In addition, some cases may have been included in both datasets but it was not possible to use record linkage of both datasets to identify them. Well designed population-based epidemiological studies can offer opportunity to examine indepth the risk factors of single-vehicle bicycle road crashes in order to identify appropriate countermeasures.
The study illustrates the significant burden of serious injury associated with on-road single-vehicle bicycle crashes. These crashes comprise over half of all hospitalisations due to on-road cyclist crashes and involve similar levels of injury severity compared with those associated with multi-vehicle crashes. Limited information on the circumstances of single-vehicle bicycle road crashes points to the need for further investigation of risk factors and the effectiveness of appropriate countermeasures. For instance, it would be useful to assess whether loss of control of the bicycle is related to poor riding skills or poor road conditions or both, and whether programmes targeting these risk factors are likely to reduce the burden of single-vehicle bicycle crashes.
What is already known on the subject
Cyclists have a high risk of serious injury and death compared with most other road users.
Single-vehicle bicycle crashes, including those that occur off road, are under-reported in police crash data but make up over half of hospitalisations as a result of cyclist crashes.
Loss of control of the bicycle appears to be the main cause of single-vehicle bicycle crashes, including those that occur off road.
What this study adds
Single-vehicle bicycle crashes that occur on public roads were the cause of more than half of hospitalisations due to on-road crashes involving cyclists.
The proportion of single-vehicle bicycle road crashes seems to be increasing over time.
Single-vehicle bicycle road crashes were as severe as those involving collisions with other road users, including motor vehicles.
Scotland's Cross-Party initiative to prevent injuries
The Royal Society for the Prevention of Accidents (RoSPA) in Scotland believes the new Cross Party Group on Accident Prevention and Safety Awareness could reduce the number of injuries of all kinds. Its first priority was to highlight potential poisoning when children access toxic cleaning products. There were 1657 deaths caused by accidents in Scotland in 2010/11 and ‘accidental injury’ remains the main cause of death for children after infancy across the UK. Editors comment: I agree with E Lumsden from RoSPA that it is impressive ‘to have accident prevention taken seriously at this level’ but I cannot agree with her view that they could easily ‘be prevented through simple measures like education and awareness campaigns.’
Bullying and suicide
The Journal of Adolescent Health released online a special issue of eight papers on the relationship between bullying and suicide. The findings reflected the views of a CDC panel. The main findings include: Bullying among youth is a significant public health problem; There is a strong association between bullying and suicide-related behaviours but causality cannot be established; Youth involved in bullying and suicide-related behaviours are also more likely to have other risk factors, such as depression, delinquency, physical and sexual abuse, and exposure to violence. Some evidence suggests ‘increased connectedness among youth and parents, other adults, and teachers may decrease bullying and suicide behaviours.’
This study was funded by VicRoads (Victoria's Road Authority), grant number Q23-00468. The authors wish to thank VicRoads staff, particularly Barry Scott, for their support and assistance with access to the data. Soufiane Boufous, Teresa Senserrick and Rebecca Ivers received salary funding from the National Health and Medical Research Council of Australia. Liz De Rome was supported by the NRMA-ACT Road Safety Trust.
Contributors All authors have substantially contributed to the conception and design of the study, interpretation of the analysed data, preparation of the paper and approved the final submitted version.
Funding VicRoads (Victoria's road authority).
Competing interests None.
Ethics approval University of Sydney Human Research Ethics.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.