An Overview of Road Death Data Collection in the EU (PIN Flash 35)

The goal of this PIN flash report is to gather information on road death data collection in different PIN countries and to find out if and how countries cross-check or complement road death data recorded by the police with alternative sources. The aim is to provide information to exchange good practice on how to improve road death data collection and recording.

In all PIN countries the primary source of road death data collection consists of police records. Police data are, and should remain, the primary source of road death analysis. However, police resources are limited and increasingly under pressure from other tasks. There are also certain collision scenarios such as those involving non-motorised vehicles or single vehicle collisions where road deaths are more likely to go unrecorded by the police.

The actual number of road deaths can be determined even when not all road deaths are recorded by the police. This can be done by using complementary data sources that may be made available by national statistical offices.[1] Fourteen PIN countries rely on police data only for road death data registration.[2] Eighteen PIN countries link police data with other sources.[3] The most widely used additional data sources are hospital data, death certificates and records of unnatural death.

In general, police records provide an important overview of road death data. However, procedures complementing and linking police data with other sources should be applied in all PIN countries in order to detect changes in recording rates by the police. In addition, linking police and hospital records is essential for improving serious road traffic injury data collection. By improving the reporting rate of serious injuries, the reporting rate for the number of road deaths is likely to improve as well, as people who were first recorded as injured may, sadly, later die within 30 days.[4]

Experts from the great majority of the PIN countries consider that road death data collected in their countries by the police are accurate, even when procedures complementing police data are not in place. Yet some experts agree that certain road deaths might go unrecorded by the police in particular those following single bicycle collisions or bicycle collisions with pedestrians. As active travel is being encouraged, the number of collisions involving vulnerable road users might be increasing.

Among countries that assessed road death data recording rates by the police, Greece, Poland and the Netherlands were the countries that reported the biggest discrepancies between the number of road deaths recorded by the police and the actual number of road deaths. Comparison of police and hospital data in Greece over the period 1985-2015 revealed that the number of road deaths recorded by hospitals is 15% to 25% higher than the number of road deaths recorded by the police.  A pilot study conducted in Poland revealed that, based on the most moderate estimations, at least 5% of all road deaths in 2009 were not recorded by the police. The national statistical office in the Netherlands calculated that around 15% of all road deaths in 2016 were not recorded by the police and the number of road deaths not recorded by the police has been increasing over the years.

Some other countries might be affected by road death data underreporting by the police. The absence of procedures to link and complement road death data recorded by the police deprive these countries of a monitoring tool for the quality of official road death data reporting.

The absence of procedures to link and complement road death data reported by the police deprive some countries of a monitoring tool for the quality of official road death data reporting.

Countries that complement police data with alternative sources have different approaches. For example Sweden has a unique road death and injury information exchange system (STRADA) which facilitates information exchange between the police and hospitals under one platform, and a responsible authority – the Swedish Transport Agency – manages the database. In some countries, for instance Belgium, Finland, Germany and the Netherlands the National statistical offices and in Spain the Directorate General for Traffic, and not the police, are responsible for capturing the number of road traffic victims who died within 30 days by complementing police data with alternative sources.

This report also shows that, in the countries which could provide data, more than half of killed road users are recorded as dying at the site of a collision. But there are large differences between countries and there might be many possible reasons for this, including differences in the procedures or quality of the emergency services and hospital care. A high proportion of road deaths at the site might indicate lower reporting rates of deaths at a later date as underreporting can be related to deaths that occur later after a collision. These differences are difficult to interpret and should be further researched.

Official alcohol-related road death numbers based on police records are under-reported in a number of PIN countries as not all drivers, let alone all active road users, involved in fatal collisions are systematically tested for alcohol.

Drug-driving remains significantly less well understood than drink-driving. Some PIN countries already have legislation in place to allow the police to drug-test road users. Yet the scope of drug-driving deaths remains unknown as not all drivers or riders involved in fatal collisions are tested for drugs.

Another grey zone in road safety is road collisions related to (illegal) use of mobile devices. Data on the use of mobile devices at the time of a collision are not systematically collected or registered in the police database.

Press release.

Main recommendations to Member States

• Countries that have not yet estimated what proportion of road deaths are reported by the police should conduct a study to estimate it. This can be done by comparing individual police road death data records with other independent sources, such as death certificates, records of unnatural deaths and hospital data to identify the actual number of road deaths.
• If such a study shows that road deaths are missing from the police records, Member States should establish procedures to complement road death data recorded by the police with other sources, such as death certificates, records of unnatural deaths and hospital data. Use these procedures to detect changes in recording rates by the police over time, identify the missing cases and their characteristics and act upon improving the data completeness.
• Improve data linkages between police, health and justice institutions concerning collisions involving injured road users.[5]
• Consider establishing one platform for recording road traffic victims by the police, hospitals and other professionals dealing with them.
• Dedicate necessary financial and human resources to the police and require them to attend, register and follow-up all fatal and serious road traffic collisions.
• Introduce obligatory testing for alcohol of all active road users in all collisions resulting in road deaths or injuries and collect data systematically.
• Test all road users for drug use as a minimum when there is a reason to suspect drug consumption.
• Introduce procedures which allow police to verify whether a mobile phone was used at a time of a fatal collision by establishing information exchange between the police and mobile network providers.

Main recommendations to EU institutions

• Within the context of the revision of Regulation 2009/661 concerning Type-Approval Requirements for the General Safety of Motor Vehicles, mandate Event Data Recorders in all new vehicles and require the data to be made available for accident investigation.
• Adopt the SafetyNet recommended definition of a drink-driving death and update the CaDaS guidelines accordingly, whilst recognising that it differs among countries and does not include all alcohol-related road deaths.
• Standardise drug-driving monitoring methods by establishing a common framework for Member States to use.

[1] Derriks, H.M. and Mak, P.M, IRTAD (2007), Underreporting of Road Traffic Casualties, https://goo.gl/VbfTTo

[2] Bulgaria, Croatia, Czech Republic, Estonia, Greece, Hungary, Ireland, Latvia, Lithuania, Portugal, Poland, Slovenia, Slovakia and Serbia

[3] Austria, Belgium, Cyprus, Denmark, Germany, Finland, France, Italy, Luxembourg, Malta, the Netherlands, Romania, Spain, Sweden, Switzerland, UK and Israel.

[4] The EU definition of a road death: any person killed immediately or dying within 30 days as a result of a road collision.

[5] ITF-OECD IRTAD (2011), Reporting on Serious Road Traffic Casualties, https://goo.gl/aP99WK