Guidance

Road safety data and statistics: STATS19 review update and future plans

Published 26 September 2024

About this document

This document outlines the road safety statistics team’s future plans for development of the department’s road safety data and statistics, including an update on progress in implementing the latest STATS19 review.   

1. Background

The main source of data on road casualties in Great Britain is the dataset of personal injury collisions reported to police, known as STATS19.  The majority of the published road casualty statistics are based on STATS19 data.  The STATS19 collection is reviewed periodically.  The most recent review commenced in 2018, with final recommendations published in 2021 and rollout of the new specification from late 2023 in some police forces areas.  Section 2 gives an update on the review implementation.

As well as changes to the STATS19 collection, the review proposed a future data strategy for STATS19, to maximise the value of other data sources and linkages and reduce burdens on police.  Section 3 provides an update on progress since the review, and planned future priorities (subject to availability of resource).

In addition, the road safety statistics team regularly reviews the publication and dissemination of the road casualty data and statistics, and consults with users on planned changes.  Section 4 summarises recent and forthcoming changes and future plans.

2. STATS19 2018 review implementation

2.1 Main changes to STATS19 data and statistics

The changes to STATS19 arising from the review are summarised in the final recommendations document. In implementing the changes, a number of small amendments have been made to these recommendations.  Full details of the new specification, including variables and values are available in the updated STATS20 guidance.

As a summary, the main changes in the new specification which will relate to the published statistics include:

• all forces to adopt injury-based reporting, which will provide more granular data on collision and casualty severity, and in time, reduce the need for severity adjustments
• a change from the current contributory factors (CFs) to new road safety factors (RSFs)
• a new vehicle type to record powered personal transporters, including e-scooters, which should improve the quality and consistency of reporting
• a new variable to record results of roadside drug testing, which may facilitate the development of statistics on drug-driving in future
• coding of journey purpose using values consistent with the National Travel Survey
• extending the scope of seatbelt and helmet wearing variables

There are a number of other changes which will not impact greatly on the statistics published by the department, but which will affect users of the detailed dataset (for example, as supplied via data.gov.uk).  These include combining variables related to pedestrian crossing and carriageway conditions, and changes in variable values for some fields including road type, and vehicle manoeuvre, location and direction.  There will also be better tracking of collisions which are deleted from STATS19 where they are out of scope (for example, away from the public highway).

2.2 Adoption of new specification by police forces

The new specification (the ‘2024 specification’), implementing the above changes, was first adopted by 28 forces using the CRASH reporting system from early November 2023, though some collisions reported using certain mobile devices or online by the public are still based on the older specification (the ‘2011 specification’).  During 2024, a further 3 forces have adopted the 2024 specification.

It is currently hoped that the remaining 14 forces will adopt the new specification by early 2025, though this could be impacted by IT deployments in some cases.   

This phased rollout means that the majority of the data for 2023 was collected using the pre-2024 specification, 2024 data (scheduled for publication in 2025) will have a mixture of data for the two specifications and 2025 data (scheduled for publication in 2026) will be largely, though probably not entirely, based on the 2024 format. 

Timescales for inclusion of data from the new specification in published data and statistics are outlined below.

2.3 Progress in implementation of the 2018 review changes

As noted above, the vast majority of the data for 2023 was collected under the 2011 specification, so that the impact on published data is currently minimal.  However, in preparation for 2024, we have made a number of changes to the published data and statistics.  The main ones relate to:

Injury based reporting:  A majority of police forces are already using injury-based reporting.  We publish high level statistics on type of injury for these forces, and for 2023 have added the more granular breakdown of injury severity into the open datasets where it is available.  In future years, as the data becomes more complete, we will extend the reporting on detailed severity and injury type while retaining the legacy severity (killed, serious or slight) for consistency over time.

Road safety factors: In preparation for the change from contributory factors to road safety factors, we have presented an initial analysis by mapping from data collected as CFs to the new RSFs.  The data tables based on CF data are now also shown for RSF categories, and we have use the new RSFs to explore the ‘fatal 4’ factors in a new factsheet. 

In addition, subject to resources, we will present an initial look at other new variables (for example on powered personal transporters) for the first 6 months of 2024, alongside the provisional in-year statistics in November.

2.4 Future STATS19 reviews

STATS19 reviews are managed by the Standing Committee on Road Injury Collision Statistics (SCRICS), and typically take place every 5 to 10 years.  However, given the complexity of rolling out changes across different police forces, the timing of any future review will depend on the value of proposed changes set against the costs of implementation.  At this stage no date has been set for a future review.  This position will be reviewed annually by SCRICS.

The 2018 review adopted an approach of ‘digital first’, seeking to gather data automatically to reduce the burden of collection on police officers. As more forces adopt digital systems such as CRASH, it may be possible that changes to STATS19 can be made more easily on an annual basis, without formal reviews.  In the meantime, we will continue to record issues raised and proposed changes to STATS19, seeking to address these via other data where possible.

A summary of areas where we have received requests for changes to STATS19 are given in the table below.  We will add to this if any further suggestions are received.

Table 1: Issues raised for consideration in any future STATS19 review

Proposed change	Our response
Pedal confusion	This was considered during the 2018 review, but felt not to be something reasonable for a police officer to know at the time of collision or could record as a contributory factor.  Within STATS19, it may be possible to explore whether text captured in the accident circumstances field could help to identify collisions where pedal confusion was involved.

Trailers and towing

There is interest in capturing the weight of trailers involved in collisions, and where vehicle defects are recorded for vehicles towing, assigning them to trailers or towing vehicles.  This would support work for post-implementation review of recent legislation.

Automated vehicles (AVs)

With the passing of the recent Automated Vehicles (AV) Act, there is interest in how STATS19 might need to change to capture information needed to monitor the safety of AVs and their impact on overall road safety in Great Britain.

3. Future data strategy

Alongside the changes to the STATS19 collection, the review contained as ‘future data strategy’ report recommended exploration of additional data sources which could address some of the limitations of STATS19.  We have begun to assess the feasibility of accessing some of these, and aim to build on this further in the coming years.

3.1 Work since the STATS19 review   

The STATS19 review recommended the linkage of STATS19 to other data sources, both to reduce the burden on police officers, and to gather additional insights which may not be feasible for a police officer to collect as part of an initial assessment.  A number of suggested data sources were proposed for exploration.

Since the publication of the review report we have completed exploratory work on:

• Trauma data: we have completed a feasibility study linking STATS19 to data from the Trauma Audit Research Network (TARN), and used this to analyse vulnerable road user casualties.
• Fire services data: we have completed an initial feasibility study exploring linking STATS19 to Home Office data on road collisions attended by the fire service.
• Contributory factors data recording: we have worked with police to explore how the recording of STATS19 contributory factors may differ if based on information known following more in-depth investigation

In addition, we continue to work on developing statistics on drug driving collisions, with initial work focused on fatalities.

3.2 Future roadmap   

we hope to build on the initial work done in the next few years – ahead of any future STATS19 review. In addition to the above, there were a number of other datasets suggested for exploration in the review, for example data from the Driver and Vehicle Licensing Agency (DVLA).  There may be emerging needs arising from legislation on AVs and road safety policy and strategy.

Alongside the roll out of the new specification, we have set out our priorities for work over the next few years in a future data roadmap.  We have published this to provide users of STATS19 data with an overview of our current priorities, subject to resources and data access.  We intend to review this roadmap and feedback on the proposed priorities, or any suggestions for further work, is welcome.   

We have aligned potential work to the safe system pillars, with priorities including:

• Roads: seeking to add enhanced geospatial features through road network matching
• Vehicles: exploring DVLA and Driver and Vehicle Standards Agency (DVSA) vehicle data
• Road users: exploring DVLA driver data
• Post collision response: developing initial work to link STATS19 to trauma care and fire service data

We have also identified a number of areas where we hope to improve the understanding of STATS19 data quality and completeness, and the dissemination and sharing of STATS19 data, which are noted as part of the roadmap and also covered in the following section.

4. Dissemination of road casualty data and statistics

The STATS19 review and exploration of related datasets relate to data collection.  This section outlines recent changes and future plans related to publication and dissemination of the statistics produced from this data.

4.1 Overall approach

We anticipate that the future approach to dissemination of road casualty data and statistics will remain broadly as currently, including

• high-level provisional statistics published twice per year, 5 months after the end of the time period to which they relate (in November and May)
• final annual statistics and data, 9 months after the end of the year to which they relate (in September)
• annual statistics on drink-drive approximately 18 months after the end of the year to which they relate

We will continue to monitor user feedback on the frequency and timeliness of publications and seek opportunities to make improvements when possible, though there are no firm plans to make changes to the current arrangements at present.

We will also regularly review the content of the publications and datasets and seek user feedback on how the statistics could better meet their needs, as well as any proposed changes.

4.2 Recent changes to published data and statistics

Following a recent request for feedback on proposed changes in May 2024, we have made a number of changes to the published data and statistics which include:

• adding additional variables to our open datasets, including the more granular collision severity, and distance between collision location and driver home
• new analyses presented using the new road safety factors classification, in our fatal 4 factsheet
• new data tables on journey purpose and propulsion type (which includes electric vehicles)
• a change to the calculation of local authority casualty rates in table RAS0403
• no longer publishing e-scooter factsheets based on provisional data

We are still considering the best approach to revising STATS19 data, and have made no changes to the current approach for the time being.

4.3 Longer term plans

Rollout of new specification in published statistics and data

Given the adoption of the new specification outlined above, we anticipate that the published road casualty data and statistics will evolve over the next few years to reflect the changes in the data collected. The following is our current expectation, which may change depending on when the final forces adopt the new specification.

In September 2024:

• we have published data tables and analysis covering the new road safety factors, based largely on mapping from existing contributory factors data

In November 2024, alongside the provisional mid-year statistics:

• we intend to present a brief summary of other new variables based on data for the first half of 2024 (for example, on use of the new powered personal transporter (PPT) code)

In September 2025:

• we intend to publish statistics for 2024 aligned to the new specification, mapping from the previous one where necessary, and adding new tables on PPT and drug tests
• statistics on road safety factors will replace those on contributory factors (which will be archived), though the data will require merging data collected using both approaches
• some degree of severity adjustment will likely still be required as injury-based data for some police forces is likely to be limited
• time series in the published statistics may be aligned to new categorisations, but historic data will be left as originally collected meaning a few breaks in time series within the open dataset

In September 2026:

• we expect to be able to publish final statistics for 2025 based largely on data from the new specification
• we anticipate that all forces will be providing injury-based data so that the severity adjustments will be required only for historic data

Over the longer term:

• subject to data quality, we will explore the scope to produce statistics on drug-driving equivalent to what is currently done for drink-driving
• we expect to fix the severity adjustments for historic data so that they do not change as new years of data are added (subject to validation of this approach)

Severity adjustments

Following the introduction of injury-based reporting, figures for serious and slightly injured casualties are presented in published data tables both as reported by police, and adjusted to allow for changes in severity recorded.  Further details are in our guide to severity adjustments.

Currently, when new data are available, adjusted figures for forces not using injury-based reporting (including historic data for forces which subsequently move to an injury-based system) are updated which results in changes to previously published data.  This can impact on, for example, baseline periods used to monitor progress in casualty reduction.

As part of the 2018 STATS19 review, all police forces are required to adopt injury-based reporting, and we currently expect that this will be the case from the start of 2025.  This means that from the publication of the 2025 data (during 2026), it is possible that injury-based figures for all forces will be available.  At this point, it would be possible to fix the historic adjustments, so that historic data does not change each year, and it is our intention to do this subject to the data being sufficiently robust. This planned approach is subject to change, should the adoption of injury-based reporting in any police force be delayed past the start of 2025.

Data dissemination

In recent years, we have reduced the number of spreadsheet data tables included as part of the final statistics, and introduced a new data download tool and dashboard to allow users to access a wider range of more tailored extracts of the data. 

Alongside the 2023 final statistics, we have published new interactive tools which the road safety statistics should be able to maintain and develop.    

In future we expect to develop the new tool further, including:

• adding data on contributory and road safety factors
• adding exposure data to allow calculation of casualty rates where possible
• adding further maps and visualisations

In general, we anticipate making more use of the interactive tool with a smaller number of spreadsheet tables in future.

5. User feedback

We welcome feedback on any aspects of this roadmap, STATS19 data or the published road casualty statistics, which can be provided via the contact details below.

6. 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.

7. 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.

8. Contact details

Road safety statistics

Email roadacc.stats@dft.gov.uk