Citation

Strandroth J. Accid. Anal. Prev. 2015; 76: 133-140.

Affiliation

Swedish Transport Administration, Röda vägen 1, 781 89 Borlänge, Sweden; Chalmers University of Technology, c/o Lindholmen Science Park, Box 8077, 402 78 Gothenburg, Sweden. Electronic address: johan.strandroth@trafikverket.se.

Copyright

(Copyright © 2015, Elsevier Publishing)

DOI

10.1016/j.aap.2015.01.001

PMID

25645162

Abstract

When targeting a society free from serious and fatal road-traffic injuries, it has been a common practice in many countries and organizations to set up time-limited and quantified targets for the reduction of fatalities and injuries. In setting these targets EU and other organizations have recognized the importance to monitor and predict the development toward the target as well as the efficiency of road safety policies and interventions. This study aims to validate a method to forecast future road safety challenges by applying it to the fatal crashes in Sweden in 2000 and using the method to explain the change in fatalities based on the road safety interventions made until 2010. The estimation of the method is then compared to the true outcome in 2010. The aim of this study was to investigate if a residual of crashes produced by a partial analysis could constitute a sufficient base to describe the characteristics of future crashes.

RESULT: show that out of the 332 car occupants killed in 2000, 197 were estimated to constitute the residual in 2010. Consequently, 135 fatalities from 2000 were estimated by the model to be prevented by 2010. That is a predicted reduction of 41% compared to the reduction in the real outcome of 53%, from 332 in 2000 to 156 in 2010. The method was found able to generate a residual of crashes in 2010 from the crashes in 2000 that had a very similar nature, with regards to crash type, as the true outcome of 2010. It was also found suitable to handle double counting and system effects. However, future research is needed in order to investigate how external factors as well as random and systematic variation should be taken into account in a reliable manner.

Language: en