Knowledge and Information System SWOV (KISS)

Within the KISS project data is collected, analysed, enriched and made available to both SWOV  and non-SWOV stakeholders. This concerns data that is useful for road safety research; mainly data relating to casualties and crashes, exposure, risk and behaviour.

Research questions   

  1. Can the ambulance data be used for road safety policy?
  2. What is the actual number of serious road injuries in 2014?
  3. Is the number of cyclists who are seriously injured in non-traffic crashes indeed as high as the ICD10 coding suggests and what are the consequences for the number of serious road injuries and the policy of the Ministry of Infrastructure and the Environment (IenM)? ICD10 is the tenth edition of the International Statistical Classification of Diseases and Related Health Problems, a list of diseases maintained by the World Health Organisation that is applied internationally.
  4. Which data sources offer possibilities for future use?
  5. What is the best way for SWOV to maintain control of data and software, how do we validate the metadata?

Activities in 2015

Ambulance data holds information about the precise location of a crash.  This information is especially important for crashes in which no motor vehicle is involved as this crash type is not common  in the police register BRON. By linking the ambulance data to the Dutch hospital register LMR and by making frequency analyses and cross tables, the added value of ambulance data for road safety policy has been investigated.

The number of EVG 2014 was estimated by linking BRON data and LMR/LZB hospital data.. The transition to the ICD10 coding was corrected for by encrypting the data back to the ICD9 coding. In 2016 this conversion will be carried out in the opposite direction. This means that the ICD9 codes from previous years will be converted to ICD10. By calculating weights the actual number of serious road injuries was determined for the severities MAIS2 + and 3 +, crashes involving motor vehicles, as well as in non-motor vehicle crashes. The results have been published in SWOV-report R-2015-18 Serious road injuries 2014.

The conversion of ICD10 to ICD9 to determine the number of serious road injuries  2012/2013, showed that approximately one third of all hospitalized cyclists had sustained their injuries in a crash that was classified as a non-traffic crash. The reason may bet hat the crash did not occur on a public road or that there were other reasons to classify the accident as a non-traffic crash (no actual crash, patient blames him- or herself). Previously the estimated proportion of ‘not in public road crashes’ was considerably smaller (2.6%).

The research was part qualitative and part quantitative. The qualitative part consisted of interviews with coders of the LMR/LZB about their definitions and understanding of a (non-) traffic crash and of public road or pavement or verge. Other parties such as the Mountain bike association were also interviewed to see if points of reference could be identified that could explain the result. In the quantitative part of the research was a link with BRON was used to see if evidence could be found for the crash being a traffic crash and/or occurred on the public road after all.

Print this page


Project nr.



Start date: 01 Jan 2015
End date: 31 Dec 2015