Transport and Covid-19: responses and resources

Reporting on Serious Road Traffic Casualties

Combining and using different data sources to improve understanding of non-fatal road traffic crashes

To improve further road safety, it is necessary to have a better understanding of the real number of road traffic casualties, including serious injuries. This is made possible by linking different sources of accident data, including police and hospital data. This report, prepared by the IRTAD Group, reviews how serious injuries are defined in IRTAD countries and identifies and assesses methodologies for linking different sources of crash data.

Policy Insights

  • A complete picture of casualty totals from road crashes is needed to fully assess the consequences of road crashes and monitor progress.
  • Injury information should complement information on fatal crashes to give a fuller picture of road crashes. Information on injuries should become more important for international comparisons.
  • Police data should remain the main source for road crash statistics. However, because of underreporting problems and possible bias (for example with differing rates of reporting by vehicle type), police data should be complemented by hospital data, which are the next most useful source.
  • The data from hospital emergency departments, available in some countries, should be monitored regularly and researched to determine if they might shed more light on road casualties.
  • The assessment of the severity of injuries should preferably be done by medical professionals, and not by the police officer at the scene of the crash.
  • Medical staff should be trained in order to systematically classify (road traffic) injuries using ICD International Classification of Diseases and to assess severities with indices such as the Abbreviated Injury Scale (AIS) or the Maximum Abbreviated Injury (MAIS). This information -- without personal information -- should be made easily available for statistical purposes, policymaking and research.
  • Besides police data and hospital data, other data sources are available. These have a limited value on their own, and cannot replace police or hospital data, but can be used to build a more balanced and comprehensive picture, to enrich the main data sources, and as a quality check.
  • For linking data, the deterministic method is preferred if a unique personal identifier is available; otherwise the probabilistic method is a good alternative.
  • The six assumptions needed to use the capture-recapture method must be considered carefully. Using this method combined with linking police and hospital data may be appropriate to give a fuller picture of road casualties.
  • Having an internationally agreed definition of “serious” injuries will help the safety research community to better understand the consequences of road crashes and to monitor progress. Given the existing knowledge and practices, IRTAD proposes to define a ‘seriously injured road casualty’ as a person with injuries assessed at level 3 or more on the Maximum Abbreviated Injury Scale i.e. "MAIS3+".

Share this page