Guidance

Guide to severity adjustments for reported road casualties Great Britain

Updated 28 November 2024

Background on the introduction of injury-based reporting

The Department for Transport (DfT) publishes road casualty statistics based on the STATS19 data collection system. This well-established system sets out the variables and data standards to which each of the police forces in Great Britain submit data relating to vehicle accidents in which an injury has occurred.

Within STATS19, injured casualties are classified as seriously or slightly injured. Historically, this has been based on the judgement of recording police officers, using an agreed definition of serious injury (outlined in our notes and definitions).

However, over recent years, the coding of injury severity in many police forces has changed, as forces have adopted injury based reporting systems, where police officers record a casualty’s injuries from a defined list and the overall severity is determined from the most severe injury recorded. This removes an element of judgement from the coding of severity.

Eliminating the uncertainty in determining severity that arises from the officer having to make their own judgement means that the new severity level data observed from these systems using injury based methods is expected to be more accurate than the data from other systems.

Following the 2018 STATS19 review, it has been recommended that all forces move to this approach. However, the new approach has impacts on the level and trend in the number of serious casualties, so that a means of adjusting data for forces not using injury-based reporting was required to estimate figures comparable over areas and over time for use in the published statistics. This adjustment is particularly required when comparing or grouping data from systems using both injury-based and non injury-based systems.

This guidance provides brief details of the adjustment method and circumstances when we suggest that adjusted data should be used.

Definitions

IBRS: Injury Based Reporting System. A system used by police to record collision data, where casualty injuries are recorded and casualty severity then coded from the most severe injury. CRASH and COPA are examples.

CRASH: Collision Recording and Sharing system. This is a centralised system used by many police forces to record road traffic collisions.

COPA: Case Overview Preparation Application. This is a system used by the Metropolitan Police Service to record road traffic collisions.

Classification of injury severity within injury-based reporting systems

A list of the injuries from which police officers can select, and how these map to different overall severities, is shown in table 1.

Table 1: Classification of injury severity using injury-based reporting systems (such as CRASH or COPA)

Injury in CRASH	Detailed severity	Severity classification
Deceased	Killed	Killed
Broken neck or back	Very Serious	Serious
Severe head injury, unconscious	Very Serious	Serious
Severe chest injury, any difficulty breathing	Very Serious	Serious
Internal injuries	Very Serious	Serious
Multiple severe injuries, unconscious	Very Serious	Serious
Loss of arm or leg (or part)	Moderately Serious	Serious
Fractured pelvis or upper leg	Moderately Serious	Serious
Other chest injury (not bruising)	Moderately Serious	Serious
Deep penetrating wound	Moderately Serious	Serious
Multiple severe injuries, conscious	Moderately Serious	Serious
Fractured lower leg or ankle or foot	Less Serious	Serious
Fractured arm or collarbone or hand	Less Serious	Serious
Deep cuts or lacerations	Less Serious	Serious
Other head injury	Less Serious	Serious
Whiplash or neck pain	Slight	Slight
Shallow cuts or lacerations or abrasions	Slight	Slight
Sprains and strains	Slight	Slight
Bruising	Slight	Slight
Shock	Slight	Slight

Adoption of injury-based reporting by forces

Approximately half of English police forces adopted the CRASH system for recording reported road traffic collisions at the end of 2015 or the first part of 2016, although Surrey has been using the system since November 2012. In addition, the Metropolitan Police Service (MPS) switched to a new reporting system called COPA (Case Overview Preparation Application), which went live to police officers from November 2016.

In 2019, 10 police forces adopted CRASH. This includes all the Scottish police forces, Sussex and Lancashire. In 2021 Greater Manchester, West Yorkshire and Nottinghamshire joined CRASH, and in 2022 Dyfed-Powys became the first Welsh force to move to CRASH.

As at the end of 2023, the remaining forces use a wide variety of systems to report collisions, in which police officers use their own judgement and guidance to determine directly the severity of a casualty (‘slight’ or ‘serious’). It is hoped that the majority of forces will adopted injury-based recording during 2025.

In contrast CRASH and COPA are injury-based severity reporting systems where the officer records all injuries for the casualty (Table 1 above shows the link between injury and severity as used in the CRASH system). The injuries are then automatically converted to a severity level from ‘slight’ to ‘serious’ and the most serious severity is used to identify the casualty severity.

Adoption dates for injury based reporting by police force

Table 2 shows the police forces which use or have used either CRASH or COPA and the dates from which these systems have been used. Note that adoption dates are indicative as there can be phased introduction of new systems during transitions.

Table 2: Adoption dates for CRASH or COPA by police force

Police Force	System Used	Adoption Date
Surrey	CRASH	November 2012
South Yorkshire	CRASH	January 2013
Staffordshire	CRASH	May 2015
West Midlands	CRASH	November 2015
Warwickshire	CRASH	November 2015
Essex	CRASH	November 2015
City of London	CRASH	November 2015
Gloucestershire	CRASH	November 2015
Metropolitan Police Service	COPA	November 2015
West Mercia	CRASH	December 2015
Devon and Cornwall	CRASH	December 2015
Cumbria	CRASH	January 2016
Humberside	CRASH	January 2016
Kent	CRASH	January 2016
Norfolk	CRASH	February 2016
Suffolk	CRASH	February 2016
Durham	CRASH	March 2016
Northumbria	CRASH	April 2016
Bedfordshire	CRASH	April 2016
Hertfordshire	CRASH	April 2016
Cambridgeshire	CRASH	May 2016
Lancashire	CRASH	December 2018
Sussex	CRASH	April 2019
Police Scotland	CRASH	April 2019
Greater Manchester	CRASH	February 2021
Nottinghamshire	CRASH	February 2021
West Yorkshire	CRASH	April 2021
Dyfed-Powys	CRASH	June 2023

Overview of the severity adjustment method

The introduction of IBRS appears to have led to a change in the reported severity of road casualties. This can be explained by the change of reporting systems from Non-Injury-Based Reporting Systems (NIBRS), where judgement of the casualty severity is made by the reporting police officer, to IBRS, where the severity of the injury is determined automatically from the most severe type of injury suffered. It appears that some casualties that would have been categorised as ‘slight’ on NIBRS are recorded as ‘serious’ in IBRS. This became apparent from initial analysis of high level data suggesting that switching to CRASH and COPA added between 5 and 15% to the Great Britain total for ‘serious’ injuries (DfT, 2017).

The Office for National Statistics (ONS) Methodology Advisory Service have developed a methodology to quantify the effect of the introduction of injury based reporting systems (CRASH and COPA) on the number of slight and serious injuries reported to the police, and to estimate the level of slight and serious injuries as if all police forces were using injury-based reporting systems. This is described in detail in the final ONS methodology report. The final report was published alongside our 2018 statistical release to set out how this methodology was finalised.

This methodology has allowed us to produce adjusted figures for the changes in injury severity as experimental statistics. In 2018, we implemented the severity adjustment methodology for the first time and published adjusted figures in a limited number of our statistical tables. In 2019 we included adjusted figures in all of the statistical tables that were published alongside the annual report.

We also include probabilities of each casualty being serious under injury-based systems alongside the underlying dataset. Further guidance is given in the Annex to the adjustments method.

In 2024, the adjusted figures were reviewed by the Office for Statistics Regulation (OSR) and designated as accredited official statistics, confirming that they comply with the standards set out in the Statistics Code of Practice.

Impact on trends

Following the introduction of CRASH and COPA, the number of casualties recorded as serious has increased in Great Britain. Chart 1 shows the number of reported serious road casualties by police force over time, from two years before to two years after injury based reporting systems were introduced (year introduced, index = 100).

Chart 1: Reported serious road casualties by police force, two years before to two years after injury based reporting systems were introduced

Chart 1 shows that the size of the increase in serious casualties varies across police forces following the introduction of injury based reporting systems. A comparison line for non-IBRS forces has been included to show how reported serious casualties have changed in these forces. Some forces (such as the Metropolitan Police Service) show a more marked increase in serious casualties than others, while other forces (such as Durham) show a more stable trend in serious casualties over time. Devon and Cornwall and South Yorkshire have also been highlighted in Chart 1 to illustrate the variety of trends observed.

The differences in the impact of the introduction of injury based reporting systems is likely to depend on the practices within a police force that were in place before these new systems were introduced. For example, Durham have stated that having a relatively low number of casualties each year allows them to extensively validate how the severity of each casualty reflects the injury received, and that their previous system for severity recording was very similar to the CRASH approach. Whereas, larger forces might not be able to carry out extensive severity reviews and there might be more differences in practices between the large numbers of officers recording accidents.

Guidance on severity adjustment use

The flow diagram in Figure 1 shows an overview on when we consider that the adjusted and unadjusted figures are most appropriate to use, though all users should carefully considered their specific circumstances and we are happy to provide more detailed advice if helpful.

Figure 1: Guidance on when to use severity adjustments

Tables 3 and 4 below provide more in-depth guidance on when to use adjusted severities depending on the geographies and time series included in the analysis.

Comparing within the same geography or police force

Table 3: Use of adjustments when comparing within an area

Option selected	Result
Has used IBRS for the whole time period used in the analysis	No need to use adjusted severities
Has used IBRS for some of the time period used in the analysis	Severity adjustment needed
Uses IBRS but not for the time period I am examining	No need to use adjusted severities
Has not switched to using IBRS	No need to use adjusted severities

Comparing across geographies or police forces

Table 4: Use of adjustments when comparing across areas

Police force using IBRS	Time period	Result
All forces using injury based recording system	Whole time period is on injury based reporting system	No need to adjust, all the police forces analysed are using an injury based recording system over the time period we are analysing
All forces using injury based recording system	Some of the time period is on injury based recording system	Severity adjustment needed as some of the data analysed was recorded on IBRS and some was recorded on non-IBRS. There is a break in the series and comparisons are not reliable unless the adjustment is used to account for the change in the reporting system
All forces using injury based recording system	Whole time period in non_IBRS	No adjustment needed
Some forces using IBRS	Whole time period	Severity adjustment needed
Some forces using IBRS	Some of the time period	Severity adjustment needed
Some forces using IBRS	None of the time period	No severity adjustment needed
None of the police forces	Whole time period	No severity adjustment needed

Robustness of the adjustments - sensitivity analysis

In police force areas where injury-based reporting systems are not used, adjusted figures are estimates of the number of serious injures that would have been recorded were an IBRS in place. These estimates depend on the model used, and different approaches may lead to different estimates. To assess the robustness of the model to the inclusion of different variables, we have carried out some sensitivity analysis.

Forces joining Injury Based Reporting Systems

In 2021, we conducted sensitivity analysis to understand how adding an extra year of data affects the severity probabilities. When comparing adjusted figures between the 2019 and 2020 models, there was less than a 0.5% change on the annual trends for all seriously injured adjusted figures. When we looked at police forces breakdowns we noticed that the level of change is different depending on the police force. There is a higher percentage change for police forces that joined an injury based reporting system (IBRS) in 2019 (2 to 8%) with the highest percentage change for Police Scotland. Police forces that joined an IBRS in 2016 and the ones which are on a non-injury based reporting system (non-IBRS) had a percentage change of less than 2%.

Merging of forces

The model was updated in 2021 for use in producing 2020 adjustments. For this model run, all Scottish police forces have been grouped together into one police force called Police Scotland. The grouping of all Scottish police forces led to a slight increase in serious injuries, but had no impact on other police forces which have an IBRS. However, it affected police forces which have a non-IBRS due to the change in median police force.

Historical revisions to casualty severity

For the 2021 model, there was a revision to historic figures for slight and serious injuries impacting data as far back as 2012. Prior to 2020, whether a casualty was reported as ‘admitted to hospital’ factored into casualty severity calculations in the CRASH system. Since 2020 this is no longer the case. Historic cases since the introduction of CRASH have been updated and corrected to reflect this change. As a result some casualty severities have been downgraded between 2012 and 2019 compared to previously published numbers. Further information is available in the background guidance.

A sensitivity check was carried out to understand the effect of revising the severities on the model. This caused a decrease of 2 to 3% in the adjusted serious casualties compared to the model run using the pre-revision severities. These severity changes affected only casualties reported by police forces that have adopted CRASH. The correction to historic casualty severities affected police force severity adjustments differently depending on their size and the number of revisions for that police force. But there was no substantive changes to trends over time.

Nevertheless, the severity revision affected Gloucestershire police force slightly differently compared to other police forces. The revision caused the historic adjusted serious time series (prior to the adoption of CRASH) to be lower than unadjusted. While unusal, this is possible with the adjustment methodology used and reflects the fact that coding of injury severity is likely to have varied across forces prior to adoption of IBRS. It may be that Gloucestershire police were already recording a higher proportion of casualties as seriously injured, relative to other forces, so that the modelled impact of moving to IBRS is different.

Removal of mobile device indicator

For the 2022 model update, we removed the variable which was used to indicate whether a force collected the data via a mobile device, as this was found to have a minimal impact on the results and removing it was found to have a very small impact on the figures.

In the next few years, we also expect to update the model as further forces move to an injury based reporting system.

Future of severity adjustments

Severity adjustments are likely to be required while forces are using a mix of injury-based and non injury-based systems. However, when all forces have adopted injury-based reporting, adjustments will not be required except for historic data (before such systems were introduced in each police force area).

We will monitor the adoption of IBRSs as police forces implement the new STATS19 specification following the 2018 review, and provide an update in our future roadmap.

Your feedback

We welcome your feedback on this approach, specifically how you are using these statistics and whether this meets your needs. Please contact the road safety statistics team

Instructions for printing and saving

Depending on which browser you use and the type of device you use (such as a mobile or laptop) these instructions may vary.

You will find your print and save options in your browser’s menu. You may also have other options available on your device. Tablets and mobile device instructions will be specific to the make and model of the device.

How to search

Select Ctrl and F on a Windows laptop or Command and F on a Mac

This will open a search box in the top right-hand corner of the page. Type the word you are looking for in the search bar and press enter.

Your browser will highlight the word, usually in yellow, wherever it appears on the page. Press enter to move to the next place it appears.

Contact details

Road safety statistics

Email roadacc.stats@dft.gov.uk