The effect on cyclist safety of daytime running lights

On average, Danes cycle 1.3 kilometres per person per day. However, according to emergency department statistics, each year more than 16,000 cyclists are injured in Denmark. In an effort to improve cyclist safety we tested the effect of daytime running lights (DRL) on the number of cyclist accidents. Using a randomised controlled trial design, 5,380 voluntary cyclist participants were randomly assigned to a test group (using DRL continuously for one year) or a control group (using their usual light routine for one year). Participants were provided with accident surveys once each month. A total of 230 accidents were reported: 121 in the test group and 109 in the control group. Most accidents were multi-party accidents (MPAs), took place in winter and in daylight. Safety effect of DRL was estimated as incidence rate ratio (IRR) between the test and control group. The incidence rate is the number of accidents per person-month. Reports of single-bicycle accidents (SBAs) in daylight were used to correct for potential report bias following Madsen et al. (2013). For multiparty personal injury accidents (PIAs), the corrected overall effect of DRL was a 25% risk reduction. The effect was insignificant. For all multiparty accidents, the corrected overall effect of DRL was an insignificant 30% risk reduction. For multiparty accidents in the dark, the effect of DRL was a significant 71% risk reduction. This is surprising, since it would be fair to assume that DRL primarily would affects accidents in daylight, where cyclists typically do not have lights on. The finding is likely due to participants in the test group having much better lighting at night than participants in the control group—or simply because some cyclists in the control group non-compliantly with the Road Traffic Act did not use lights at night during the experiment.