Attention Blue Light! Evaluation of the training "Road safety during blue light driving" by the German Traffic Safety Council (DVR) and the German Social Accident Insurance (DGUV)

Status:

completed 11/2017

Aims:

The German Traffic Safety Council and the German Social Accident Insurance developed a one-day simulatorbased training especially for blue light driving. Goal of our project was to analyze how this training affects road safety during blue light driving in emergency medical services. Therefore, a multi-method evaluation of this training on the levels of reaction, learning, behavior, and results was conducted and the influence of further factors, like e.g. experience of the drivers or composition of the training groups, on the effectiveness of the training was tested. Options for potential improvements and necessary preconditions should also be deduced.

Activities/Methods:

To make sure that effects can be unambiguously attributed to the training an expanded intervention-control-group design was used. The control group was realized as waiting group and a further group was used to capture pure effects of measurement. Thus, an elaborate quasi-experimental design with up to four measurement times was used including a controlled pre-post-measurement (T1-T2) for the test of effectiveness, a waiting group measurement (T2-T3) for a test of replicability, and a fourth long-term measurement (T4) for a test of sustainability of effects. To ensure generalizability trainings were conducted in Saxony and southern Bavaria, Germany, in both urban and rural areas. All trainings were done by the same trainer. All employees of emergency medical services could participate as long as they drove at least 40 hours a month emergency ambulances or emergency doctor's vehicles and as long as their organization was willing to participate in the study. All participants were allocated to the experimental groups as evenly as possible. Methods used to measure changes on the different levels of evaluation were as follows. On the level of reaction participants had to fill in a survey of acceptance and satisfaction with the training directly afterwards and at later measurement times. For the level of learning questionnaires on attitudes regarding road safety and risk propensity were used as well as a newly developed knowledge test. Besides further questionnaires the level of behaviour was measured using aspects of driving behaviour in real traffic via telemetric data tracking (e.g., speed, longitudinal and lateral accelerations) and video recording. Finally, on the level of results data on stress and strain during real shifts and emergency operations were collected by physiological data (ECGs) and (shift-related) questionnaires on short- and long-term strain reactions. As an economic key performance indicator the length of emergency drives was measured.

Results:

On the level of reaction, participants evaluated the training very good directly after the training and quite good in the long run. A significant knowledge gain was achieved on the level of learning and risk perception especially for normal drives was increased by the training. This effect on attitudes was replicated in the waiting group; however, it was not sustainable. On the level of behaviour the training caused a decrease in mean and maximum speed during real emergency drives. These changes were replicable as well as sustainable in the long-term measurements. In the long run a decrease in longitudinal acceleration was also found. Sound proof for effects on the level of results could not be found. Strikingly, though, there was no difference in the mean duration of emergency drives despite the decreases in speed. For all other parameters that were measured on the levels of learning, behaviour and results no significant effects of the training were found.

As (excessive) speed is one of the biggest factors in traffic accidents, a reduction should increase road safety during emergency drives. Taken all results into account, the training shows (replicable and sustainable) positive but small effects and might be improved by some actions (like, e.g. heterogeneous composed training groups).

Last Update: