Measuring the incidence of injury among bicycle commuters on cycle networks and open roads: A pilot study

Background Government policy is to reduce motorised land travel. Nottingham has a system of segregated and unsegregated cycle pathways. Exposure to risk on these different types of cycle route and related injury rates are unknown.

Objective To develop simple, accurate methods of measuring distances travelled on different cycle pathways; to estimate the number of incidents (unexpected occurrences) per unit distance on different pathways, and the number yielding injuries.

Design Diary recording over two weeks by a random sample of cycle commuters.

Setting Nottingham city cycle routes used by hospital workers between home and work.

Method Four hundred bicycle storage area users at a National Health Service (NHS) hospital were asked to map their routes to and from work, to estimate the distance involved and to record the nature and location of any incidents occurring during cycle commuting over two weeks. Using a ruler, string and the map scale, the distances covered on each type of route were calculated. Calculated distances were compared with actual distances measured with a car mileometer. Incident rates were calculated for each type of pathway. The occurrence of any injuries sustained in these incidents was checked against Accident and Emergency (A&E) department records.

Results Only 267 proved eligible; 155 replied by the deadline; 87 agreed to participate, 68 declined. Eighty (44 males, 36 females) returned complete data. Cyclists' estimates of route distance, and the distances of the same routes calculated by measuring map lengths differed by 2.1 km on average (range -3.3 to +5.4 km). Distances calculated by measuring map lengths differed from actual distances measured by mileometer by only 0.5 km on average. We calculated that commuters covered 5368 km, averaging 4.6 km per journey (range 0.8 to 7.6 km), 57 per cent of it on roads. Twenty-eight cyclists reported 53 incidents (10 per 1000 km: 95%CI 7 to 13). Segregated cycle paths had the highest rates (43 per 1000 km: 95%CI 26 to 67). In 46 cases the incident involved taking action to avoid an obstacle, a pedestrian, another cyclist or a motor vehicle. No injuries were reported in these incidents and nobody involved in the incidents attended the hospital A&E department.

Conclusions It is feasible to record accurately over two weeks the exposure of commuting cyclists to travel on various types of urban cycle route, and the number and nature of potentially harmful incidents experienced. It is likely that the higher rate of injury on off-road cycle paths reported in other countries would be confirmed.