PSNI Cannot Produce an Eyesight-Collision Total: The 13,847 Records Behind the Data Gap

A PSNI freedom-of-information response reveals that defective eyesight is not a standalone collision-causation category in its searchable records. Answering a three-year question would have meant reviewing 13,847 injury-collision files — an estimated 1,154 hours.

What the police can count — and what their system cannot

Road-safety debates often put precise figures beside speeding, distraction and drink-driving. But precise categories are not the same as a complete account of why collisions happen. A PSNI freedom-of-information response published on 20 January 2026 shows one consequential gap: the police could not readily supply the number of collisions in which defective or uncorrected eyesight contributed during 2023, 2024 and 2025.

In FOI/16677 the PSNI said defective or uncorrected eyesight was not recorded as a standalone causation factor in its central databases. To find relevant cases it would need to read individual incident material, searching for references to vision problems. For 1 January 2023 to 31 October 2025, the police identified 13,847 recorded injury collisions; it estimated a five-minute manual review per record, amounting to about 1,154 hours.

The police refused the request under the Freedom of Information cost provisions. That is a finding about data retrieval, not a finding that eyesight contributed to none of the collisions, nor that it caused a substantial number. Neither zero nor a guessed percentage can be derived from the disclosure.

The arithmetic explains why a simple public query failed

At five minutes a case, reviewing 13,847 files requires 69,235 minutes, or roughly 1,154 hours. PSNI explained the standard FOI appropriate cost limit for the force is £450, using a permitted calculation of £25 per hour, equivalent to 18 working hours of retrieval time. On its estimate, a comprehensive manual search would require more than 64 times that allowance.

The calculation is simple, but its significance is easy to miss. The police appear to possess the underlying incident records, yet the information cannot be extracted as a ready-made statistic. Police systems can therefore support detailed investigation of individual collisions without necessarily offering an accessible population-level answer to the question “how often was bad vision a factor?”

The 13,847 is not a denominator from which to calculate an eyesight-collision rate because the numerator is unknown. It covers injury-involved collisions, not all damage-only crashes, and the latest portion was provisional at the time of the response. Using it as a count of motorists failing vision standards would be entirely wrong.

What Northern Ireland drivers are actually required to see

The absence of an easily retrievable police count should not obscure a clear individual legal standard. Northern Ireland Highway Code rule 92 says motorists must be able to read a number plate in good daylight from 20 metres, or 20.5 metres for the older-style plate. Drivers who require glasses or contact lenses to meet that standard must wear them when driving. Police can require a driver to undertake an eyesight test.

The DVA practical-test procedure is more specific about character dimensions and the confirmation process if a candidate cannot read the first plate. It is not enough to assume a driving licence issued many years ago means eyesight remains unchanged. Vision can deteriorate between renewals, and a driver needs to meet the applicable standard whenever driving.

The number-plate test also does not capture every potential visual risk. The DVA’s visual-standard guidance considers field of vision and medical conditions, not only central distance acuity. Someone may read a plate but have a restricted visual field or another condition needing medical assessment. The proper response is to seek qualified advice and comply with any DVA reporting duty, rather than rely on an informal test as a certificate of fitness.

The overlooked difference between failing a sight test and contributing to a collision

Three different questions are commonly collapsed into one: does a driver meet the legal eyesight standard; was poor vision present at the time of a collision; and did vision problems contribute causally to that event? Those questions require different evidence. A licence declaration or number-plate check addresses compliance. A clinical examination can address a person’s visual function. Establishing collision causation may also require scene reconstruction, witness evidence, lighting conditions, vehicle position and distraction analysis.

Even if every incident form contained a “poor eyesight” box, the resulting number would need interpretation. A police officer may not know whether an involved driver had an uncorrected prescription or a previously undiagnosed visual-field problem. An optional or inconsistently completed field could understate the issue. Equally, coding a suspected vision issue would not necessarily prove that it caused the collision.

The point is to resist two bad headlines: that eyesight-related crashes are rare because the police do not publish a total, and that an unquantified national crisis can be inferred merely because the field is missing. Both would use the absence of data as if it were evidence of a particular outcome.

Practical implications for motorists, employers and commercial fleets

For an individual driver, the useful response is to maintain up-to-date prescription correction, seek examination when vision changes, and tell DVA when the applicable medical rules require notification. Night glare, difficulty distinguishing road signs, narrowing peripheral awareness or repeated problems judging distance are reasons to seek an eye-care assessment; they are not diagnoses and should not be ignored.

Employers whose staff drive for work should not treat a photocopied licence as evidence of current visual fitness. A proportionate driver-safety programme can remind staff of the standard, ask them to report changes and refer fitness concerns to appropriate medical and occupational-health professionals. Personal medical information should be handled confidentially and under relevant data-protection rules.

Older drivers deserve accurate guidance rather than a blanket assumption of risk. Medical suitability turns on a person’s visual function and conditions, not simply age. DVA guidance notes that some people with sight in only one eye may still meet driving standards, subject to adaptation and the applicable visual requirements. A headline equating one eye, glasses or age with unfitness to drive would misinform readers.

What reporting would close the gap

The obvious long-term question for NI road safety is whether collision recording could capture suspected vision-related factors without creating unreliable statistics. That would require definitions, officer guidance, appropriate medical safeguards and quality assurance. It is not simply a matter of adding a tick-box and publishing the total.

The next accountable question for the PSNI and DfI is whether a narrower, prospectively coded study could quantify how frequently vision concerns are reported, verified or ruled out. A research design would need to distinguish suspected impairment from confirmed failure of visual standards and avoid attributing blame without evidence. Until such a study exists, a reliable public answer may remain unavailable.

PureCar’s finding is deliberately narrower and stronger than the sensational alternative: PSNI could not extract a vision-collision count without reviewing 13,847 injury-collision files, and estimated the task at 1,154 hours. That reveals a measurable gap in the information available to policymakers and the public. Drivers, meanwhile, already have clear responsibilities that do not depend on a national percentage being published.

Sources and reporting method

FOI covers 1 January 2023 through 31 October 2025, with latter provisional. PSNI says eyesight not separately coded. No estimate of prevalence or causal responsibility made. This is an investigation of public statistics, not an account of a specific collision.

Primary documents consulted: PSNI FOI/16677 defective-eyesight collision records, 20 January 2026; Northern Ireland Highway Code rule 92, vision; DVA official guide to visual standards for drivers; nidirect driving practical-test eyesight requirements.