Another difficulty with studies that have compared bicycling and walking to other modes of transportation is that they may imply that the facilities and safety benefits of the different mode are comparable, when they are not. The idea of sharing road space is a difficult one for most motorists and administrators to accept, and the resulting discomfort and safety risks to motorists are a potential liability for cyclists. Research has suggested that the potential benefits of shared-space facilities are greater in countries where the modes are more dissimilar [ 85 ]. As most of the studies reviewed here were conducted in North America, the reader may question whether the conclusions reached are applicable to other countries. However, the only North American study reviewed here (that of Anderson and Collie [ 66 ]) found that shared space in urban areas supported greater levels of cycling by pedestrians than by motorists. This may have at least as much to do with shared-space architecture itself as the modal user. However, the study had no comparison with walking and therefore it is unclear how much of the safety benefit for pedestrians was due to modal substitution.
Assessing the effectiveness of interventions to reduce bicycle injury risk is challenging. For example, injury and crash data can only be collected for a limited period of time, and improvements are usually measured after time lags, which may exceed the typical duration of the intervention. Injuries can also be concentrated in localized areas, making it difficult to attribute the cause to any particular intervention. To date, most of the studies reviewed here reported on only one intervention with one cycle-related injury outcome or crash outcome. The extent to which the interventions are effective in reducing such outcomes (and injuries as well as crashes) is often difficult to determine: 3d9ccd7d82