How to Use: Incident & Near Miss Investigation Report Form (CDM 2015)

Purpose and outcome

This form helps you capture what happened, make the area safe, and investigate the real causes (not just “someone messed up”). Used properly, it will:

• ✅ Stop it happening again (through clear corrective actions with owners, dates, and proof).
• ✅ Provide a solid record for the Principal Contractor/Client (if required), insurers, and legal duties (including CDM 2015 where relevant).
• ✅ Show learning and improvement at site meetings until actions are closed out.

Good completion looks like: facts captured early, evidence referenced, witness notes taken calmly and separately, root cause explained clearly, and actions that are specific, owned, dated, and closed with proof.

Where this fits in the site workflow

Who uses it (roles and responsibilities)

When to use it (triggers and timing)

• Use it for: injuries/ill health, near misses, dangerous occurrences, property damage, environmental spills, or any “that was close” event.
• Start it: the same shift (once safe).
• Finish the main investigation: within 24–48 hours where possible.
• Review actions: at every site meeting / daily brief until closed.
• Update it: if new information comes in (CCTV, equipment checks, later medical outcome).

Before you start (what you need to gather)

Make safe first, then collect the basics while it’s fresh.

• People: names, company, role/trade, contact details, supervisor on duty, first aider details (if used).
• Time and place: exact location (plot/floor/zone), time of event, time reported.
• Evidence: photos (wide → close-up), a quick sketch/marked-up plan, witness names.
• Work info: task, plant/tools used, materials involved, weather/lighting/ground (if relevant).
• Controls and documents: RAMS, permits, brief/toolbox talk records, pre-use checks, inspection tags/logs, temporary works checks (if applicable).

⚠️ Keep it factual: record what’s relevant to the event. Don’t add unnecessary personal or medical detail.

How to complete it (step by step)

1. Immediate actions: make safe and look after people

   • Stop work if needed; isolate plant/tools; set an exclusion zone.
   • First aid and emergency services if required.
   • Inform Supervisor/PM, and the PC/Client where required by site rules.

2. Basic details: get the admin right

   • Project/site, exact area, date/time, who reported, who investigates, and the event type.
   • If you don’t know something, write “unknown” and come back—don’t guess.

3. What happened: facts first (simple timeline)

   • Task being done and step-by-step sequence.
   • What went wrong / nearly went wrong and the immediate cause.
   • Avoid blame words. Focus on conditions and controls.

4. People involved and injury details (if any)

   • List everyone involved and whether anyone was injured.
   • Keep injury notes brief and relevant (type, body part, treatment, time lost/restricted duties).

5. Evidence captured: reference it properly

   • Photos: record where stored (folder/app reference) and move them to your secure system.
   • Sketch/plan: even a quick hand sketch helps later.
   • Check inspection status and key documents (RAMS/permits/temporary works where relevant) and note document references.

6. Witness notes: separate, calm, in their words

   • Interview witnesses one at a time if possible.
   • Record where they were and what they saw/heard (their wording).
   • Signature is optional—don’t turn it into a confrontation.

7. Controls in place: what worked, what failed, what was missing

   • Use the controls section as a checklist (briefing, PPE, access, barriers, housekeeping, traffic, permits/isolations, inspections).
   • If a control existed but failed, write how (e.g., “barrier removed for access and not reinstated”).

8. Root cause: use 5 Whys to get past the obvious

   • Keep asking “Why?” until you reach something you can fix (planning, sequencing, supervision, competence, equipment, site set-up, procurement, time pressure).
   • Avoid stopping at “human error”. Ask what set people up to fail.

9. Corrective actions: make them strong and trackable

   • Write actions that change the job (method, protection, kit, layout, training, supervision).
   • Assign owner + due date for each action.
   • Define close-out proof (photo, updated RAMS, briefing record, inspection record, purchase/repair evidence).

10. Reporting checks: RIDDOR and notifications

    • If it might be RIDDOR, treat it as urgent and get competent advice (H&S adviser/PC guidance/HSE guidance).
    • Record reference numbers and who was told (PC/Client/insurer where needed).

11. Sign-off and action review: keep it alive until closed

    • Investigator and Supervisor/Manager sign to confirm the report and actions are agreed.
    • Use the action review section at site meetings to chase progress and remove blockers.

Review and approval (checks and escalation)

The reviewer should check the report answers these questions clearly:

• What happened? (clear timeline, location, task, immediate cause)
• What evidence supports it? (photos/sketch/doc references)
• What were the control gaps? (RAMS, supervision, equipment, protection, sequencing)
• What is the root cause? (beyond “human error”)
• Are actions strong enough? (specific, owned, dated, with proof)

Escalate for senior sign-off if any of these apply:

• Serious injury/ill health, hospital treatment, or likely time off.
• High potential event (could realistically have caused fatality or major injury).
• Dangerous occurrence indicators (collapse, major failure, uncontrolled release, etc.).
• Repeat events or a pattern (same type/location/task).
• Likely insurance claim, client concern, or regulator involvement.

Data protection (UK GDPR) – practical handling on site

• Need-to-know only: don’t share injury details in group chats or on notice boards.
• Keep it relevant: record the injury type/treatment outcome, not extra medical history.
• Secure storage: lock paper copies away; restrict access to digital folders.
• Photos: avoid unnecessary images of faces/injuries; store securely and label clearly.

Recordkeeping (storage, version control, retention)

• File it with the evidence: photos, sketches, RAMS/permit references, inspection checks, and close-out proof.
• Version control: if you update the report, date the update and note what changed.
• Retention: keep as long as needed for safety management and legal/insurance purposes, following your company retention rules (be consistent).
• Audit readiness: you should be able to show the report, action close-out, and how learning was shared.

How often to review or update

• Within 48 hours: confirm evidence/document checks are complete and actions are assigned.
• Every site meeting: review actions until all closed (record progress and blockers).
• Periodic trend check: monthly or per project stage, look for repeating themes and fix them proactively.
• After close-out: do a quick follow-up check that the new control is in place and being used.

Common mistakes (and how to avoid them)

Worked example (simple site scenario)

Scenario: Near miss. A labourer nearly falls stepping off a scaffold lift because a guardrail section was removed to move materials and not put back.

• Immediate actions: stop work on the lift, restrict access, reinstate guardrail, supervisor informs PC.
• Facts recorded: exact lift/level, task (moving boards), sequence, and the point the guardrail was missing.
• Evidence: photos showing missing rail position and access point; sketch of the lift; scaffold tag status; names of workers/witnesses.
• Root cause (example): materials movement not planned → RAMS didn’t cover loading method → no agreed rule for temporary removal/reinstatement checks → supervision didn’t verify after change.
• Corrective actions (stronger examples):
  • Update RAMS with loading/unloading method and “no rail removal without alternative protection and sign-off” (Owner + due date + re-brief proof).
  • Set up a defined loading point/sequence (layout change) to avoid ad-hoc rail removal (Owner + due date + photo/layout proof).
  • Add a start-of-shift interface check for scaffold edge protection where materials are handled (Owner + due date + record proof).
• Learning shared: 10-minute toolbox talk the same morning; record attendees and key message.

Quick site checklist (use alongside the form)

• ✅ Make safe, treat injuries, isolate hazards.
• ✅ Notify the right people early (including PC/Client if required).
• ✅ Capture evidence before it changes (photos, sketch, names, document refs).
• ✅ Facts first; avoid blame; use 5 Whys for root cause.
• ✅ Actions must be specific: owner + due date + proof.
• ✅ Review actions at every site meeting until closed.
• ✅ Store securely and keep it audit-ready (data protection).