Injury incidence rates of cyclists compared to pedestrians, car occupants and powered two-wheeler riders, using a medical registry and mobility data, Rhône County, France
Introduction
In some of France's major cities (Paris, Lyon, Lille, etc.), there has been an increase in cycling, mostly as a means of transport (Papon and De Solère, 2010). This is partly associated with local policies, such as the introduction of large self-service bicycle sharing schemes; also cycling is encouraged in the framework of sustainable development. We therefore need to know more about cyclist road risk, compared to other road user types.
In France and in most countries, crash data are provided by the police. However, police crash data suffer from under-reporting and selection bias; in developed countries it is estimated that only 39% of road casualties are reported and recorded by the police. This under-reporting varies mostly with injury severity, the presence/absence of a crash opponent and the road user type (Elvik and Mysen, 1999, Hauer and Hakker, 1988). Crashes involving cyclists are particularly less reported (Amoros et al., 2006, Langley et al., 2003); this is probably mostly due to bicycles not being considered as a proper means of transport. In France, cyclists have, when the crash involves a crash opponent, a 34% probability of being registered in police crash data vs. 44–55% for other road users (Amoros et al., 2006). This probability falls to 2% when the crash involves only the cyclist and no other crash opponent. In the Rhône County (population of 1.6 million inhabitants, with Lyon being the main city), a hospital-based road trauma registry provides a second source of data. The latter is much more complete than police data (Amoros et al., 2006). Indeed, over the period 2004–2007, police crash data recorded about 155 injured cyclists and the medical Registry recorded 1230 injured cyclists, per year (in the police data, the exclusion of off road crashes is not strict, and in the medical registry they probably account for less than 10% of the injured cyclists; Amoros et al., 2009a). Moreover, the medical registry takes into account both hospitalized individuals and individuals who were treated at the emergency department only.
To compare crash number between different types of road user, the exposure had to be taken into account: observed differences in crash number between different groups may be due to underlying differences in exposure to risk. Several studies have examined incidence rates of being injured taking the exposure into account (Beck et al., 2007, Bernagaud and Hurez, 2011, Christie et al., 2007, Gabet, 2005, Garrard et al., 2010, Jacobsen, 2003, Li and Baker, 1996, Licaj et al., 2011, Mercat, 2006, Pucher and Buehler, 2008, Pucher and Dijkstra, 2003, Sonkin et al., 2006, Tin Tin et al., 2010, Tin Tin et al., 2011). In France, three studies performed in the Lille conurbation (approx. 1.2 million of inhabitants) (Gabet, 2005), the Grenoble conurbation (approx. 712,000 inhabitants) (Mercat, 2006) and in the Lyon conurbation (approx. 1 million of inhabitants) (Bernagaud and Hurez, 2011) estimated injury rates using the police crash data and exposure measures from the corresponding regional household travel survey. Another study focusing on young people and on the characteristics of territories (deprived or non-deprived municipalities) was carried out using the medical registry accident data (Licaj et al., 2011). In other countries, accident data can also be obtained from police data, hospital-based data or from registries such as a death registry. Exposure data also generally come from household travel surveys.
The objective of the present study is to estimate injury incidence rates in the Rhône County taking the exposure into account. Rates are estimated for four main road user groups: car occupants, pedestrians, cyclists and powered two-wheeler (PTW) riders, separately for four injury categories (all-injury, hospitalization, serious-injury and fatal-injury) and by gender, age group and location. The exposure measures used here are number of trips, distance traveled or time spent traveling with a given means of transport. Measuring the exposure is based on a regional household travel survey, which includes the region covered by the hospital-based registry. This survey is restricted to the winter period of the years 2005 and 2006 and to typical weekdays. To correct this (to have estimates for full years), seasonality ratios were used, which were estimated from the 2007–2008 national transport survey. Given that the Rhône County is quite urbanized, a (population) ‘dense areas/non-dense areas’ variable was created using information on accident or trip location.
In the present study, trends of the injury incidence rates are also studied. For this purpose, the previous regional household travel survey, carried out in 1994–1995, is used. This survey was, however, restricted to Greater Lyon (Lyon and its suburbs). Thus, in order to make comparisons, the rates for the years 2005 and 2006 were also estimated from accident and exposure data restricted to the same area. The 1994–1995 survey was also carried out in the winter period and on typical weekdays; we used seasonality ratios estimated from the previous national survey (performed in 1993–1994).
Section snippets
Police-reported crash data
The French police are requested to write a report for every road crash causing at least one casualty. A road crash is officially defined as a crash involving at least one vehicle (motor vehicle or bicycle). The police report should describe everyone involved in the crash, and classify them as non-injured, slightly injured, seriously injured or killed. The slightly injured are those with out-patient status. Seriously injured are those who have been admitted to hospital. Killed are those who died
Seasonality ratios
For car occupants, the seasonality ratios were around 0.7–0.8; it means that exposure in the ‘outside RTS’ period was slightly lower than in the ‘RTS’ period (i.e. November to April on weekdays, outside school and public holidays). The lower car use during summer and/or week-ends, especially for local trips, can be explained by the decrease in commuting trips. Men and women had similar seasonality ratios. With regards to the number of trips, seasonality ratios are similar between age groups
Discussion and conclusions
As a summary, the present study provides estimates of traffic injury incidence rates by type of road users using three exposure measures (number of trips, distance traveled and time spent traveling) and based on a medical registry. Compared with car occupants and with regard to time spent traveling, cyclists were 8 times more likely to be injured, 12 to be hospitalized, 16 to be seriously injured, and 3 to be killed. Compared with car occupants, PTW riders were 42 times more likely to be
Acknowledgments
We wish to thank the following people for having participated in the data collection and data entry in the framework of the Association for the Registry of road traffic casualties in the Rhône (ARVAC: president Y.N. Martin) or in the framework of INRETS-UMRESTTE (B. Laumon, scientific consultant for the registry, and A. Ndiaye, physician, director of the registry): Ait Idir T., Ait Si Selmi T., Alloatti D., Amoros E., Andrillat M., Artru F., Asencio Y., Assossou I., Auzaneau F., Bagès-Limoges
References (39)
- et al.
Under-reporting of road crash casualties in France
Accident Analysis and Prevention
(2006) - et al.
Ottawa–Carleton commuter cyclist on- and off-road incident rates
Accident Analysis and Prevention
(1998) - et al.
A prospective cohort study on minor accidents involving commuter cyclists in Belgium
Accident Analysis and Prevention
(2012) - et al.
Incidence, severity and correlates of bicycling injuries in a sample of cyclists in Queensland, Australia
Accident Analysis and Prevention
(2011) - et al.
Exploring the male–female discrepancy in death rates from bicycling injury: the decomposition method
Accident Analysis and Prevention
(1996) - et al.
Contextual deprivation, daily travel and road traffic injuries among the young in the Rhone Département (France)
Accident Analysis and Prevention
(2011) - AAAM: Association for the Advancement of Automotive Medicine, The Abbreviated Injury Scale (1990 revision). Des...
- et al.
Actual incidences of road casualties, and their injury severity, modelled from police and hospital data, France
Eur J Public Health
(2008) - Amoros, E., Chiron, M., Ndiaye, A., Laumon, B., 2009a. Cyclistes Victimes d’Accident (CVA), partie 1: caractéristiques...
- Amoros, E., Thélot, B., Supernant, K., Guérin, A.C., Chiron, M., 2009b. CVA3 – Enquête auprès de 900 usagers de vélo;...
The injury epidemiology of cyclists based on a road trauma registry
BMC Public Health
All-cause mortality associated with physical activity during leisure time, work, sports, and cycling to work
Archives of Internal Medicine
Motor vehicle crash injury rates by mode of travel. United States: using exposure-based methods to quantify differences
American Journal of Epidemiology
Comparaison des risques d’accident entre les différents modes
How exposure information can enhance our understanding of child traffic “death leagues”
Injury Prevention
Incomplete accident reporting: meta-analysis of studies made in 13 countries
Journal of the Transportation Research Board
Cycling injuries in Australia: road safety's blind spot?
Journal of the Australasian College of Road Safety
Cited by (61)
Spatial dimension of accidents involving pedelecs 25 and conventional bicycles
2024, Transportation EngineeringInjury severity prediction of cyclist crashes using random forests and random parameters logit models
2023, Accident Analysis and PreventionComparison of injury severity between E-bikes-related and other two-wheelers-related accidents: Based on an accident dataset
2023, Accident Analysis and PreventionHow do road infrastructure investments affect Powered Two-Wheelers crash risk?
2023, Transport PolicySystematic literature review of 10 years of cyclist safety research
2023, Accident Analysis and Prevention
- 1
Present address: Hospices Civils de Lyon, Service de Biostatistiques, F-69000 Lyon, France.