Track root causes and corrective actions after a Canadian workplace incident

An incident log says that something happened. It does not show why controls failed, what was changed or whether the change worked. A corrective-action tracker should connect those stages without assuming the first explanation is the root cause. This is particularly important when several conditions combined: equipment design, supervision, work pressure, training or an unexpected change in the task.

This guide is a cross-Canada investigation workflow, not one national reporting rule. Occupational health and safety legislation differs across the federal, provincial and territorial jurisdictions. The Canadian Centre for Occupational Health and Safety (CCOHS) incident investigation guide recommends finding underlying causes and developing a corrective-action plan. Check the applicable regulator's notification, investigation, committee and record requirements for the worksite before applying a deadline.

Capture the event and preserve facts

Start with the incident location, date, people and activity, immediate harm, controls in place and evidence sources. Separate facts observed at the scene from later recollection. Record what had to be moved to rescue a person or make the area safe. Photograph relevant equipment only where it is appropriate and record its identity, condition and date. A stable event ID allows witness statements, maintenance records and action evidence to be linked without making one document carry everything.

Do not wait for a perfect root-cause analysis before taking immediate protective action. Stop or isolate an unsafe process where necessary, arrange care, notify the appropriate people and meet any legal reporting obligations. The tracker should distinguish interim containment from the permanent control chosen after investigation.

Ask why the control failed

Use an investigation team with knowledge of the work and, where required, committee or representative participation. Develop a timeline: normal procedure, actual conditions, deviation, failure and consequences. Test each explanation against evidence. “Worker error” may describe an act but often does not explain whether the worker had a workable procedure, suitable equipment, reasonable time and adequate supervision.

CCOHS advises looking for direct, indirect and basic causes and avoiding a blame-centred process. Record competing hypotheses and unresolved facts. If an alarm was reportedly ignored, establish whether it operated, was audible, was understood and triggered an achievable response. The tracker should not convert an untested suggestion into a final finding simply to move an item to “closed.”

Write corrective actions as controls

For each finding, record the hazard or system gap, proposed control, owner, due date, required resources, interim protection and approval. Prefer controls that remove or reduce exposure at the source where practicable. Training and reminders may be needed, but a new sign alone may not address an unsafe machine layout. State what is expected to change in the real workplace so the action can be verified.

Separate implementation from effectiveness. A new guard can be installed and still fail if it prevents normal work and is routinely bypassed. A revised procedure can be signed and still be unusable. Attach installation, inspection, briefing or procurement evidence to implementation; schedule a later observation, interview or audit to check effectiveness. Reopen an item if the control fails in operation.

Use clear closeout decisions

An action is complete when the agreed control has been implemented and verified by a competent reviewer; the investigation is closed when outstanding questions and required reporting are handled. These may be different dates. Preserve the decision maker and basis for accepting residual risk. If a contractor or landlord controls part of the remedy, record the interface and do not mark it complete on the strength of a request alone.

Review repeated incidents for patterns. Several “minor” manual-handling events may indicate one recurring design problem. The tracker should allow the organisation to see actions that are overdue or repeatedly reopened, without treating a count of closed items as proof that risk is low.

Example: repeated slips at a loading door

The first account says a worker slipped because they were rushing. The investigation finds water entering through a damaged seal, poor drainage, a temporary mat that moved under carts and no clear owner for building repairs. Interim barriers and cleaning reduce exposure; a repaired threshold and drainage change address the source. The file links the repair record, post-repair inspection and worker feedback. A reminder to “walk carefully” may support the response, but is not the main closeout evidence.

Assess a record workflow

The mapped Complys Canada OHS compliance page is the product page for assessing safety evidence records. Verify every feature claim from the current product implementation before use. The existing health and safety program guide owns broad programme design; this guide owns the narrower root-cause-to-control chain. A system can organise records, but cannot determine causation or certify a control's effectiveness automatically.

Choose measures that can be tested

Write an acceptance criterion for each corrective action before marking it complete. “Improve housekeeping” is difficult to verify; “install a fixed drainage channel, inspect it after rain and remove standing water from the loading route” gives a reviewer something observable. Name the person authorised to accept the work and the date of the follow-up check. If the control only works under ideal conditions, include a test under ordinary operational pressure.

Share the safety lesson with affected crews without exposing unnecessary personal details. Ask whether a similar condition exists at other sites, then record the answer. A closed local action should not hide an unresolved organisation-wide issue. Conversely, do not duplicate dozens of action tickets for one common change when a coordinated programme can show which sites have implemented and verified it.