If you run a construction SME, RIDDOR can feel like two different jobs: knowing when something must be reported to the enforcing authority, and keeping a record that helps you learn, defend decisions and brief clients. Mix those up and teams either over-report noise — or under-record near misses until a serious event has no trail.
This note is a practical framing for small builders and contractors. It is not a substitute for the current Reporting of Injuries, Diseases and Dangerous Occurrences Regulations guidance on HSE’s site.
What RIDDOR is (and what it is not)
RIDDOR is a legal reporting and recording framework for certain work-related deaths, injuries, occupational diseases and dangerous occurrences. It is not a full incident-management system, and it is not the same as your internal “raise a near miss” culture.
HSE guidance emphasises that reporting duties sit with the responsible person for the work activity. On multi-contractor sites that can feel ambiguous — which is why clear internal ownership (“who dials / submits / confirms”) matters as much as the legal categories.
Reportable events construction SMEs most often meet
Categories and thresholds change only when the law or guidance changes, so treat the list below as a map of themes to check against current HSE material — not as a checklist carved in stone:
- Work-related deaths — rare, catastrophic, and always a peak-demand moment for evidence.
- Specified injuries — serious injury types defined in the regulations and explained in HSE guidance.
- Over-seven-day injuries — where a worker is unable to perform their normal work for more than seven consecutive days (again: confirm current wording).
- Dangerous occurrences — certain near-catastrophe events that must be reported even without injury.
- Occupational diseases — diagnosed conditions linked to work, where the regulations apply.
The operational point for SMEs: if supervisors cannot tell within hours whether something might be reportable, your process is too slow — not because everyone needs a law degree, but because delayed facts make both reporting and learning harder.
What to record even when you do not report
Most useful learning lives below the RIDDOR threshold. Near misses, first-aid cases, plant close-calls and “almost” scaffold issues should still be recorded with enough structure to answer later questions:
- What happened, where, and when (local time, not vague “morning”)?
- Who was involved or witnessed — including contractors?
- Immediate controls taken?
- Plant, task, weather or permit context?
- Photos or sketches while the scene still exists?
WhatsApp threads are a poor archive. They fragment timelines, lose attachments, and make client or insurer requests expensive.
Timing, ownership and “who reports”
HSE themes around RIDDOR stress prompt reporting for certain events and clear records. For a small contractor that usually means:
- Name a default owner (contracts manager, HSE lead, or nominated supervisor) who can escalate out of hours.
- Separate “first aid / make safe” from “decide reportability” so neither blocks the other.
- Keep a single evidence trail you can reconstruct later — not three notebooks and a phone gallery.
On CDM sites, principal contractors and contractors still need internal clarity: who gathers facts, who checks guidance, who submits, who tells the client.
Evidence that survives the first busy week
After a serious event, memories degrade fast. Teams that cope tend to standardise a short incident pack:
- Initial log with time stamps
- Injured person / witness statements when appropriate
- Plant and permit references
- Photos of conditions (not dramatised)
- Follow-up actions with owners and due dates
That pack supports RIDDOR where required — and supports learning where reporting is not required. It also reduces the scramble when a client asks for “everything you have” after a stoppage.
Common SME failure modes
- Recording only when someone insists — culture collapses under pressure.
- Treating RIDDOR as paperwork punishment — people hide near misses.
- No closed-loop actions — the same scaffold or plant issue repeats.
- Confusing insurance notification with RIDDOR — related, not identical processes.
Practical takeaway
Build a habit of structured incident and near-miss logging first. Layer RIDDOR decision support on top with current HSE guidance bookmarked and a named owner. You do not need enterprise theatre — you need a trail that works on a wet Tuesday when three trades are waiting and someone has just called from A&E.
Soft next step
If evidence trails and incident workflows are the pain, we can walk the RIDDOR-path module on a short demo.