|Investigation is focused on detection of offences and the prosecution of individual drivers. Few police are trained to look 'upstream' and identify opportunities to design out danger.
In Sweden, there is a multi disciplinary approach to collision investigation and the police consider what would have prevented the death, if not the fatal crash. Our police are not trained to differentiate between causation and prevention.
Road death investigation centre
Each year, DfT estimates the cost of crashes to the country, including the cost of police resources. In 2012, police costs were revised and dramatically increased after a 2011 DfT research project found much higher estimates than previously used (£1,848 for fatal collision and £245 for serious injury collision).
PACTS called for the Health and Safety Executive to establish a Road Death Investigation Unit in their report Beyond 2010–a holistic approach to road safety in Britain.
The need for a road crash investigation centre was a key question asked by PACTS’ Transport Safety Commission inquiry into Transport Safety – Who’s responsible.
In March 2017., PACTS held a conference on Collision Investigation and called for a National Collision Investigation Branch to be established.
Collision reporting and contributory factors
Collision investigation does contribute to injury prevention with data collection. Our understanding of collision circumstances is based on Stats 19 data, collected by the police in a voluntary agreement with the DfT.
Collisions are reported within days on the STATS19 database, with contributory factors being preliminary guesses, at best, and more likely biased guesses. DfT has acknowledged that as the police may have to testify in court on what they list as a contributory factor, that some, including speeding, will be unreported, until there is collaborating evidence.
RoadPeace has called for contributory factors to be recorded at the end of the investigation. We have also urged that fields be expanded to include 'Not available' so that police can differentiate between when a factor has not contributed to a collision and when they cannot tell if it has. ÊWe believe many factors would be listed as NA.