STATS19 and ONS death data linked to sharpen road fatality figures

The Department for Transport has taken initial steps to link its STATS19 road collision database with death registration data held by the Office for National Statistics, a move designed to give a more complete and reliable picture of how many people die as a result of road collisions in Great Britain.

STATS19 is the system police forces use to record details of road collisions, including casualties who die within 30 days of a crash. Because that 30-day rule is applied at the point of recording, some deaths that follow from a collision may not be captured, and others recorded as fatalities may later be reclassified. ONS death registration data, by contrast, captures every death certified in England and Wales, with cause of death determined by a clinician. Matching the two datasets allows researchers to check how well the two sources agree and to identify cases that appear in one but not the other.

The work published by the Department for Transport describes this as initial linkage work, meaning the methodology is still being assessed rather than used to produce revised headline casualty figures. The exercise looks at whether individuals recorded as killed in STATS19 can be traced to a corresponding ONS death record, and whether ONS records point to road collision deaths that STATS19 may have missed or recorded differently.

For ordinary road users, the practical significance lies in the longer term. If the two sources are successfully reconciled, the official statistics used to set road safety policy, target investment and measure progress against casualty reduction targets will rest on firmer foundations. Discrepancies between police-recorded and clinically certified death data have long been a known limitation in road safety research, and this linking work is a direct response to that gap.