Injuries to bicyclists in France: description of 1,541 casualties from the Rhône Road Trauma Registry

Abstract

As compared with 183 injured bicyclists recorded by the police during 1996, 1,541 were recorded in the Rhône Road Crash Trauma Registry, which includes, for each road crash casualty, the data provided by all the medical and surgical trauma emergency or care services concerned. The under-estimation by the police is especially large for bicyclists. Severity of injuries is increasing with age. Body regions most often injured are the upper (54% of casualties) and lower limbs (35%), the face (28%) and the head (20%). This body region distribution depends on age. 50% of AIS3 injuries are open or displaced fractures of the upper limb. Nearly all AIS4+ injuries are head injuries. Because of the exhaustiveness of the Registry, injuries to bicyclists show their real extent in comparison with other user types.

Objective

One of the most remarkable results of setting up a road accident trauma registry has been the appearance of the real importance of accidents among bicyclists who are the least well recorded road users in official road accident figures.

Thus the objective of this study is to evaluate the real significance of bicycle accidents and the characteristics of accidents in absolute terms and in comparison with other types of users, measure the seriousness of body injuries and analyze the influence of personal and circumstantial factors and, lastly, describe the injuries responsible for the seriousness of accidents to bicyclists.

Method

A trauma registry of road accident victims in the Rhône “département” (France, population of 1,600,000 habitants, chief city Lyon) has been in use since 1995. It is based on the collection of medical data from 78 public, private, civil and military centers dealing with crashes : emergencies, intensive care units, readaptation and surgery units, liable to receive the victims of crashes occurring in the “département du Rhône ”. The network also covers centers which, although outside the département, may admit eligible victim. As with any other vehicle, all bicycle and motorcycle crashes are included, whether another vehicle is involved or not. By contrast, pedestrians are only included if they are struck by a vehicle, motorized or not [Laumon, 1997].

In this article, we analyse the data obtained for cyclists involved in road crashes, with respect to the person (age and gender) and the accident (place of occurrence, antagonist). Characteristics of the victims and their injuries are compared with those observed in the Rhône Trauma Registry for the other three main types of road user (pedestrians, car users and moped-motorcycle users), and those taken from the official statistics established by the police [ONISR, 1997].

A description is then given of the severity of the injuries, measured by the ISS (the total of the squares of the highest AIS of the three body regions most injured, [aaam, 1990]), the body areas injured and the severity for each area, and the details of the serious injuries (AIS3+). We then analyse the links between the location and the severity of the injuries and the characteristics of the people injured and of the accidents, by cross-checking the variables two by two.

For the whole descriptive part, including the description of injuries, comparison results are only given if statistical tests are significant (χ² test, p<0.05%) and thus test figures are not indicated in order to simplify the text.

Finally we apply logistic regression, taking ISS9+ as the severity indicator (at least an AIS3 injury or several injuries leading to an equivalent severity). For a continuous observation study, the intercept of the model can be interpreted according to the basic risk R_0, and it is possible to estimate the really interesting parameter, the relative risk, from the Odds-Ratio (OR estimated by the exponential of the model parameters) by using the expression directly subtracted from their respective definitions, RR=OR/(1+R₀(OR−1)), with a specific calculation of the confidence interval of the relative risk [Martin JL, 1997].

Results

Of the 10,384 road accident victims of 1996 present in the Rhône Road Trauma Registry and presenting at least one injury coded by the AIS, 1,541 (14%) are cyclists. These figures should be set beside official statistics established on the basis of accident statements taken by the police and the gendarmerie during the same year and for the same geographic area: 4,537 victims, including 183 cyclists, i.e. 4%. Cyclists are the most under-evaluated type of user in official statistics.

Characteristics of bicyclists injured and their accidents

Age and gender

Of the bicyclists injured, 79% are male. Thus the bicyclists have a sex ratio of 3.8, between that of car users (1.0) and pedestrians (1.1) and that of moped-motorcycle users (4.9) in the Rhône Trauma Registry. This ratio does not differ from data belonging to the police. On the contrary, the share of children under 10 years old is much larger in our population (22% vs. 4%).

The age group most represented is that of the 5 to 9 years old for girls and from 10 to 14 years for boys. This distinguishes bicyclists from other traffic victims (the age group most prevalent for both sexes is from 20 to 24 years for car users and from 15 to 19 years for moped-motorcycle users). Figures for bicyclists are, however, closer to those for pedestrians, for whom the groups with the highest scores are those from 5 to 14 years for males and from 10 to 14 for females.

Of these 1,541 bicyclists injured in the Rhône, 1,481 live in this “département”, i.e. 96%. This type of user is therefore that for which the share of residents in the Rhône is the highest (reaching 87% for car users, 94% for pedestrians, and 93% for moped-motorcycle users).

The sex-ratio is 7.4 for bicyclists living outside the “département”, suggesting that bicycle use differs from one sex to the other (distances travelled).

For men, the bicycle comes after the car and motorcycles and mopeds but before walking, whereas for women, the car is first, then walking, then the bicycle and lastly mopeds-motorcycles.

If the incidence rates concerning only the residents having accidents in the Rhône are calculated, we obtain a raw incidence rate for the Rhône of 0.94/1000 (1.54/1000 for males, and 0.39/1000 for females), ignoring crashes outside the département and involving Rhône residents. According to usage in road accident studies, it is possible to match all the accident victims in the Rhône with the number of its inhabitants. This calculation gives a very similar raw incidence rate of 0.98/1,000 (1.60/1,000 for males, and 0.40/1,000 for females). Adjusted for age for the French population of 1996, this incidence rate is very similar.

Crash characteristics

As shown in the Table 1, nearly three quarters of cyclists were injured during a crash without an antagonist being identified, whereas 18.6% collided with a four-wheeled vehicle, 5.6% with a two-wheeled vehicle or pedestrian, and 1.6% with a fixed object.

Table 1.

Bicycle injuries in the Rhône (France) 1996: antagonists

Antagonist	n	%
None or not identified	1,143	74.2
Car	268	17.4
Bicycle	48	3.1
Fixed obstacle	24	1.6
Lorry, Coach, Van	19	1.2
Moped-Motorcycle	13	0.8
Pedestrian	10	0.6
Other	16	1.0
Total	1,541	100.0

The share of crashes without a known antagonist is characteristic of bicyclists, in comparison to that observed for car users (15%) or moped-motorcycle users (48%) in the Registry.

It is 1.7 times higher than that observed in the official statistics (44%), which are particularly inadequate for this type of crash [Laumon, 1997]. The difference in the total numbers cannot be explained however by this difference in recruitment: 348 injured during a collision with a vehicle feature on the register, as opposed to only 87 recorded by the police.

Information on crash location is often missing for cyclists recorded in the Register (whose objective is to collect data for medical purposes). As shown in the Table 2, this is the type of user for whom the share of “non-determined” crash locations (31%) is the highest (only 14% for pedestrians, 16% for cars, and 24% for moped-motorcycle users), which is a weakness of the study.

Table 2.

Bicycle injuries in the Rhône (France) 1996: crash location

Crash location	n	%	%(ignoring “not determined”)
Highway	1	0.1	0.1
Main roads	16	1.0	1.5
Secondary roads	65	4.2	6.1
Streets	671	43.5	63.4
Road network, without detail	80	5.2	7.6
Car park	24	1.6	2.3
Private road	64	4.1	6.1
Off the network	137	8.9	13.0
Not determined	483	31.3
total	1541	100.0	100.0

Bicyclists are characterised by the high number of accidents on private roads or off the road network (respectively 6% and 13% when ignoring the “non-determined” location), compared to car users (respectively 0.1 and 0.8%), moped-motorcycle users (0.8 and 3.5%) and pedestrians (1.4 and 2.3%).

They use secondary roads more frequently than pedestrians (6% vs. 1%), but less than car users (9%) and moped-motorcycle users (8%). They have fewer accidents on main roads than the other three types of user.

In comparison, the bicyclists having accidents recorded by the police are almost never injured in car parks or off the road network. This difference, inherent in the method of collecting the two bases, only partially explains the global difference between the numbers. Although the “officially” recorded bicyclists are, like those on the register, most often injured in streets, this sub-group only includes 136 individuals (as opposed to 671 for the register).

The place where the accident occurs is very strongly linked to the type of antagonist (Table 3): logically, it is more often motor vehicles for roads open to traffic.

Table 3.

Bicycle accidents: link between the accident location and the type of antagonist

	Highway, road or street		Other or not determined
No known antagonist	484	64.4%	658	83.5%
Antagonist without bodywork (pedestrian, two-wheeled vehicles, skater)	50	6.6%	37	4.7%
four-wheeled vehicle	206	27.4%	81	10.3%
Fixed obstacle	12	1.6%	12	1.5%
total	752	100.0%	788	100.0%

There is a strong correlation between the bicyclists age and gender and the type of accident. The share of bicyclists without a known antagonist regularly decreases with age (from 91% in 0 to 4 years old to 47% in people of 60 and over), and it is higher in males (20% vs. 14%).

Overall severity

If the ISS is used as indicator of overall severity (Figure 2), 63% of bicyclists present only minor injuries (ISS<=3, i.e. no AIS2 injury).

Figure 2.

Bicycle injuries in Rhône (France) 1996: severity according to ISS

Bicyclists are severely injured less often than moped-motorcycle users and pedestrians (Table 4). Compared to car users, they suffer minor injuries less often (ISS<=3). This order remains true when we only consider accidents occurring on public roads and streets.

Table 4.

Bicycle injuries in Rhône (France) 1996: severity (%), compared to other road users

Severity	Pedestrians n=1013	Car users n=5523	Moped-Motorcycle users n=1688	Bicyclists n=1541
ISS 1–3	58.7%	78.6%	59.5%	62.9%
ISS 4–8	25.1%	14.4%	28.0%	29.1%
ISS 9+	16.2%	7.0%	12.6%	8.0%
death	2.3%	1.3%	1.5%	0.4%

Six bicyclists died, i.e. 0.4% of the sample. They had also all been recorded by the police. This fatality is significantly less than that for other types of user in the Registry.

Severity generally increases with age (Figure 3). But the 10 to 14 years old are, on the contrary, injured more severely (ISS9+) than adults from 15 to 39 years old.

Figure 3.

Bicycle injuries in Rhône (France) 1996: overall severity and age

Although there is a strong correlation between gender and age (Table 1), the same number of severely injured (ISS 9+) was observed for both sexes.

The presence of an obstacle is associated with greater overall severity, especially if it is a fixed obstacle. In this case, 25% of the victims have an ISS of 9 or more, as opposed to 11% when the obstacle is a motor vehicle, and 7% when there is no identified obstacle, or when the obstacle is a pedestrian or a two-wheeled vehicle.

The accidents occurring on a private road or off the road network are associated with reduced severity. This is also true for those occurring in non-defined places, suggesting that the absence of this data results, at least in part, from the nature of these places which are unusual or difficult to classify. In this case, 6% of the victims have an ISS of 9 or more, as opposed to 10% for streets, main and secondary roads.

As personal factors (age) and those linked to the accident (place and antagonist) are related, logistic regression was performed to highlight the overall severity risk factors in injured riders.

The four main factors explored in the previous lines are used in a logistic regression to adjust the probability that a victim had an ISS equal to 9 or more. Table 5 shows Relative Risks associated with the different factors where value 1 indicates the reference level for each one.

Table 5.

Relative Risks and 95% confidence intervals from logistic regression fitted to 1,541 pedalcyclists

Factor		β	RR	95% C.I.
age	0–4	−0.15	0.87	0.27–2.75
	5–9	−0.13	0.88	0.47–1.61
	10–14	0.38	1.42	0.89–2.28
	15–39		1
	40–59	0.82	2.13	1.33–3.41
	60 & more	1.04	2.56	1.32–4.99
gender	male		1
	female	−0.04	0.97	0.62–1.51
crash location	other		1
	highway, road or street	0.44	1.51	0.93–2.46
collision with	nothing or pedestrian or two-wheeled vehicles		1
	four-wheeled vehicles	0.40	1.46	0.96–2.20
	fixed object	1.40	3.46	1.57–7.63

Once adjusted for other factors, the effect of age remains, i.e. the two age groups over 40 have an ISS of 9 twice as frequently as the 15–39 age group. There is no apparent difference in severity between the two sexes. The contribution of the factor characterising the possible crash with a vehicle or an obstacle is globally significant (χ²=8.8, p<2%), with an RR of 3.5 for crashes with a fixed object (significant in spite of the low number of observations). The RR associated with the location of the accident is not significant, the trend being towards greater severity on main and secondary roads nevertheless remaining on the same level of magnitude as with a univariate analysis.

Description of injuries

Localisation and severity of injuries

2,891 injuries were observed on the 1,451 cyclists. The injuries of the spine were distributed according to their level: cervical with neck, thoracic with thorax, and lumbar with abdomen.

Figure 4 shows the frequency and seriousness of injuries for each area. Minor injuries are in the majority for each region.

Figure 4.

Distribution of bicyclists for each region injured and each MAIS

The region most often affected is the upper extremity (54% of victims). Then come the lower extremity (35%), the face (28%), the cranium-brain (17%), the thorax (8%), the abdomen (5%), the neck (4%) and the skin with no specific area given (4%). 41% of the victims had head injuries (cranium, brain, face and/or neck).

The cranium-brain and the thorax are the only regions for which severe injuries were observed (AIS4+).

• For the upper limb (1,012 injuries described), 63% were skin or minor joint injuries AIS1, 26% were simple fractures or dislocations AIS2 essentially located at the level of the hand (10%), the forearm (9%) and the clavicle (8%), while 8% were the serious injuries described further in this paper.

• For the lower limb (659 injuries), 79% were minor injuries, 17% were major lacerations AIS2 or simple fractures (4% were fractures of the fibula, 4% were fractures or major lacerations of the knee, 3% were tibial fractures, 2% were fractures of foot bones, and 3% were the serious injuries described further on.

• For the face (558 injuries described), 76% of the injuries were minor skin injuries (abrasion, contusion, minor lacerations), 12% were minor injuries to the mouth AIS1 (lacerations) and teeth (fracture, loss). 3% of fractures observed were benign AIS1 and 3% were compound fractures AIS2, as well as 3% major lacerations AIS2.

• Concerning the cranium, (333 injuries described), 45% were minor injuries to the scalp, and 8% were minor symptoms (dizziness, cephalalgia). Mild problems of consciousness without neurological symptoms represented 31% of «injuries», and 12% were serious to maximum injuries (AIS3+, see further on).

• The thoracic region (118 injuries) comprises 67% minor lacerations or contusions, 21% simple costal (15%), spine or sternal fractures, and 12% serious or severe injuries (cf. further on).

• The abdomen (79 injuries) was affected in 76% of contusions and minor lacerations AIS1 to the trunk and in 9% of superficial lacerations or contusions AIS1 of the genital organs. The AIS2 injuries consisted in contusions to the internal organs (10%), or non-compound fractures (5%). A single AIS3 injury is described (retroperitoneal hematoma).

• In 92% of cases the neck (63 injuries) suffered minor AIS1 injuries (contusions, minor lacerations, sprains); 8% of injuries were AIS2 spinal apophysis fractures without contusion of the spinal cord.

• Lastly, injuries to the skin and sub-cutaneous tissue were all minor.

Comparison with other types of users shows that bicyclists are more often subject to head injuries than moped-motorcycle users(17.5% vs. 13.5), but less often than car users (23.8%) and above all less often and less severely than pedestrians (amongst whom 31% were subject to head injuries, and twice as many MAIS3+ as for bicyclists, i.e. 3.8%).

Bicyclists are more often subject to facial injuries than all the other types of users, but these facial injuries are never serious, as opposed to those of pedestrians and car users.

Likewise, bicyclists are more subject to injuries to the cervical region than all the other types of users, though the severity of these injuries is always minor or moderate (AIS 1 or 2), in contrast with what can be observed in car users and moped-motorcycle users in which these injuries can be fatal.

Injuries to the thorax are less frequent than for the three other types of users, and above all severe in only very exceptional cases (less than 0.1% of MAIS4+, i.e. one tenth of that for pedestrians, and one eighth of that for car users and moped-motorcycle users). Likewise, there are fewer injuries to the abdomen in bicyclists than in other road users and they are never severe. No AIS4+ injury was observed, whereas this is seen exceptionally among pedestrians (0.4%), car users (0.3%), and moped-motorcycle users (0.6%).

Bicyclists are the type of users most often subject to injuries of the upper limbs (53% as opposed to 50% for moped-motorcycle users, 37% for pedestrians, 23% for car users), and it is they for which these injuries are the most serious (more MAIS 2 and 3, meaning maximum severity for this region).

Lastly, bicyclists are subject to injuries to the lower limbs half as much as pedestrians and moped-motorcycle users (35% as against respectively 67 and 65%), but more often than car users (23%). These injuries are rarely serious (1.4% of MAIS3+, as against 7.3% for pedestrians, 5.6% for moped-motorcycle users, and 1.8% for car users).

To conclude, the severity for bicyclists, as for all road accident victims, mostly stems from cranium and brain injuries. In comparison to the three other types of road user, bicyclists are characterised by the frequency of minor and moderate facial injuries (AIS 1 or 2), and moderate AIS2 and severe AIS3 injuries to the upper limbs (simple and compound fractures).

Location of injuries according to the characteristics of the victim and the accident

The effect of age is evident regarding the location of injuries: Children under ten years old have a specific injury profile: the face is very frequently injured (43% vs. 23% in 10 years old and over), injuries to the cervical region are practically absent (only one case for 336 subjects, i.e. 0.3%, vs. 4.8% in 10 years old and over), injuries to limbs are less frequent (35% vs. 59% for over 10 years old for the upper limb, 25% vs. 62% for the lower limb).

Thoracic injuries are observed more frequently from 30 years onwards (17% as opposed to 5% in the youngest).

We did not notice a link between injury location and gender.

Regarding the circumstances of the accident, cranial and cerebral injuries are observed less frequently when no antagonist is identified (15% of victims with cranial and cerebral injuries vs. 24% with an obstacle, even when this is a pedestrian or a two-wheeled vehicle). However, no severe injury (AIS3+) in this region occurs with an obstacle with no bodywork such as a pedestrian or a two-wheeled vehicle, whereas 4% of victims who collide with a fixed obstacle or one with bodywork have this type of injury.

Injuries to the face are less frequent (18%) when the cyclist collides with an obstacle without bodywork (pedestrian or two-wheeled vehicle) than in other cases (28%).

Cervical injuries are particularly frequent when the obstacle is a four-wheeled vehicle (10.5% against 2.3% in the other cases).

Thoracic injuries are more frequent when the obstacle is identified (15% vs. 5%). They are often more serious when the obstacle is a four-wheeled vehicle (2.4% of AIS3+ as opposed to 0.6% in the other cases).

Likewise, abdominal injuries are more frequent when an obstacle is present (7% as opposed to 4% when no obstacle is present).

Whereas neither the presence nor the severity of injuries to upper limbs depend on the presence of an obstacle or its type, injuries to lower limbs are 1.5 times more common when the obstacle is a four-wheeled vehicle than in other cases (48% vs. 32%).

To sum up, it appears that serious injuries to the cranium and brain (the region where bicyclists’ injuries are the most severe) occur more often when the obstacle is a four-wheeled vehicle or fixed obstacle; this is also true for the thorax and the abdomen. Crashes with four-wheeled vehicles (car, HGV) also very often lead to injuries to the cervical region and to the lower limbs. Impact with an obstacle without bodywork (pedestrian, two-wheeled vehicle) rarely leads to facial injuries. The upper limbs are injured in a similar way whatever type of crash is involved.

Injuries to the head, abdomen and lower limbs are more frequent on main roads than elsewhere (respectively 37% vs. 17%, 19% vs. 4.5%, 62% vs. 34%), facial injuries on private roads (41% vs. 27%). It should be noted that these results are obtained with small numbers.

Nature of severe injuries to bicyclists (ISS>=9)

157 serious injuries (5.7% of all injuries) were recorded, i.e. 137 injuries AIS3, 13 injuries AIS4, 5 AIS 5 and 2 AIS 6.

All the AIS 5 and 6 injuries were to the brain, as were 11 AIS4 injuries out of 13 and 21 AIS3.

These 157 serious injuries concerned 108 victims (7% of the sample), 78 of which suffered from a single AIS3+ injury, 20 of two, 5 of three, 5 of four or more.

Of these,

• 55 people (3.6% of the population) had serious injuries (AIS3) only to the upper limbs, including 20 by two AIS3 injuries. These were serious isolated injuries consisting of open, displaced, or comminuted fractures: 26 of the radius, 8 of the humerus, and 1 of the ulna. When two serious injuries are associated, in 19 cases this was an association of two homolateral, open, displaced, or comminuted fractures of the radius and ulna, and once a biletaral compound fracture of the radius.

• 20 victims (six of whom died) had only serious injuries to the head, including 15 with a single injury described. In the case of an isolated severe injury, in 6 cases these entailed problems of consciousness on initial examination, including two with neurological signs and one with decerebration. The other isolated injuries were 3 extra or epidural hematoma, two fractures to the base of the cranium, two massive destructions of both cranium and brain, one cerebral contusion, and one cerebral edema. In the case of several serious injuries (five people), each had a bone injury (vault fracture in two cases AIS3 and 4, of the base in two cases AIS3, and major penetrating injury AIS5 in one case), associated with an edema AIS4 and 5 and/or a hematoma AIS4 and/or a hemorrhage AIS3 and/or problems of consciousness AIS3.

• 17 victims had one serious injury to the lower limb. These consisted of 8 fractures of the femur, 7 compound fractures of the tibia, 1 compound fracture of the pelvis, 1 fracture of the sacroilium.

• 10 people had serious injuries only to the thorax. These were compound rib fractures compounded with hemo or pneumothorax (3 cases), and multiple and bilateral fractures with the stable chest or with no details given.

• Only 6 victims were subject to serious injuries to several areas. For 15 victims, the ISS score reached or was higher than 9 in the absence of an AIS3 injury.

Fatalities

Six bicyclists died in 1996 in Rhône, leading to a mortality rate of 4/1,000,000. The six victims (4 males and 2 females), aged from 6 to 48 years old, had head injuries. These injuries were the cause of death for five of them. For the sixth, problems of consciousness on arrival were indicated (AIS3), associated with a poorly identified abdominal injury. All the cyclists were killed on a clearly identified road belonging to the public road network (4 streets, 2 main roads). Five of them crashed with a motor vehicle while the antagonist for the sixth was absent or not indicated.

Discussion and conclusions

This article describes a population of 1,541 cyclists with from slight to serious accidents, i.e. presenting at least one injury that can be given an AIS code, recorded in a trauma registry covering a French administrative area (département) with 1.6 million inhabitants. The incidence is 1.6/1,000 for males, and 0.4/1,000 for females. These values are observed without us having any knowledge about the exposure of different ages and genders to bicycle use in the Rhône. We know bicycle use varies greatly from one country to another, from one region to another and also between the town and countryside [roberts, 1997, welander, 1999]. The sex-ratio is comparable to those observed in the international literature [n.h.t.s.a., 1997, Kraus, 1996, Rivara, 1997, Rowe, 1997, Carré, 1995].

The number is 8.4 higher than that of official annual statistics for this French department, established on the basis of police data. As bicyclists are the type of user for whom this ratio is greatest, their relative share in all road accidents is thus 14% in the register as opposed to 4% for the police. Moreover, cyclists recorded in the registry differ from those in the official statistics: they are more often young children, injured on private roads or off the road network, without an antagonist and/or living in the administrative area. This different profile does not explain the enormous difference in numbers between the registry and police data, which have also been observed for adults injured on the public road network. Thus the official figures do not only considerably under-estimate the number of traditional accidents, as is already known [Harris 1990, IRTAD 1994, Aptel 1990, Laumon 1997, Koornstra 1998] though somewhat biased with respect to age and type of accident.

The study therefore sheds new light onto the real magnitude of the phenomenon in France, as well as giving precise knowledge of the injuries to bicyclists caused by road accidents.

The great majority of the victims (92%) has an ISS<=8, which is a figure very similar to that of Rivara [1997].

The region most often injured for the whole population is the upper limb (54% of victims). This is followed by the lower limb (35%), the face (28%), the cranium (17%), the thorax (8%), the abdomen (5%), the neck (4%) and the skin without details on the precise location (4%).

The severity and mortality of the accidents are mainly imputable to cranial and brain injuries, in agreement with the literature [rowe, 1995, zentner, 1996]. Of the victims (8% of the total) which were seriously injured (ISS9+), nearly half suffered from a serious injury (AIS3+) only to an upper limb, from one or more compound fractures.

Bicyclists can be distinguished from other road users by the frequency and seriousness of the injuries to the upper limb, the frequency of moderate injuries to the face and neck, and lesser overall seriousness.

The influence of the characteristics of the victim and the accident on the general seriousness was analyzed by logistic regression. It appears that seriousness (measured by the ISS9+) is more frequent over 40 years old, by taking into account the circumstances of the accident. An increase of seriousness with age has already been observed in the registry for pedestrians. This result agrees with the study carried out by Rivara et all. [1997] who noted in addition increased seriousness in under-twelve year olds, though this is not the case in our report.

Contrary to the injuries, the characteristics of the accidents are not described in detail in the registry, which is based on a medical collection of the data, mainly aimed at thoroughness in the medical area. The major part of the locations and obstacles not stated probably covers a location that is unusual or difficult to classify (therefore off the public road network), and therefore non-existent, which is consistent due to the fact that our bicyclists are younger and probably more concerned by their leisure than making a journey. In spite of the mediocre quality of the data on them, the influence of the location of the accident and the nature of the antagonist on its seriousness have been analyzed (table 5). The nature of the antagonist appears as a major factor of seriousness, in particular fixed obstacle, significant in spite of the low number of subjects concerned. Impact with a four-wheeled vehicle, which concerned only 18% of cases, is particularly harmful for the head, as emphasised by Borczuk [1995].

We have no information on wearing helmets and we hope to improve that point in the next future. However helmet wearing is optional and practiced little in France up to now. The seriousness of the cranial injuries observed, in agreement with the literature [Viano, 1997, Welander, 1999] could be an argument for its promotion [Thompson, 1996].

The quality of the medical data of the register and the network of medical corps, researchers, and victims that it implies through time, could permit in the future, besides the observation of the progression of the phenomenon of accidents involving bicycles, studying the risk factors and evaluating counter measures by specific studies.

Figure 1.

Bicycle injuries in the Rhône (France) 1996: age and gender