From Incident to Prevention: How to Write Corrective Actions That Stick

Why Most Corrective Actions Don't Prevent the Next Incident

After a workplace incident, the pressure to act quickly is real. An investigation is completed, a report is filed, and corrective actions are assigned. But many workplaces find that the same or similar incidents happen again — sometimes in the same area, sometimes involving the same hazard. The problem is rarely that corrective actions were not assigned. It is that the actions assigned were the wrong kind. Telling a worker to 'be more careful' after an incident is not a corrective action. It is a note that closes an investigation without changing anything.

Why Most Corrective Actions Fail

• They address behaviour rather than the system or condition that enabled the behaviour
• They are written too vaguely to implement — 'improve housekeeping' instead of 'remove pallets obstructing aisle 4 by end of shift Friday'
• No individual is named as accountable, so no individual takes ownership
• No deadline is assigned, so completion is indefinitely deferred
• No one verifies that the action was completed — a form says it was done, but conditions were never actually changed
• They address the immediate cause without reaching the underlying root cause

Immediate Cause vs. Root Cause

Every incident has an immediate cause — the condition or action that directly resulted in the harm — and one or more root causes, which are the underlying system failures that allowed the immediate cause to exist. Corrective actions aimed only at the immediate cause are temporary: they fix the specific condition but leave the system that produced it unchanged. A worker slips on a wet floor. The immediate cause is the wet floor. The root cause might be a leaking overhead pipe that has been reported twice but not repaired, a cleaning procedure that does not require wet floor signs, or a drainage system inadequate for the volume of water produced by the adjacent process. Addressing root cause means changing the system so the wet floor cannot recur.

The Four Parts of an Effective Corrective Action

Every corrective action assigned after an investigation should include all four of these elements:

1. 1Specific action — describes exactly what will be done, not a general goal. 'Repair leaking pipe at column C-12' rather than 'fix plumbing issue'
2. 2Accountable person — names the individual responsible for completing it. Assigning an action to a department means no one is accountable
3. 3Completion date — sets a clear deadline. 'As soon as possible' is not a date; it is a way to defer indefinitely
4. 4Verification method — describes how and by whom completion will be confirmed after the action is taken

Questions to Ask When Drafting Corrective Actions

Use these questions to evaluate each action before it is assigned:

• Does this action prevent the same incident from happening again, or does it only respond to what already happened?
• Would this action have prevented the incident if it had been in place beforehand?
• Does this change a system, procedure, or physical condition — or does it instruct someone to behave differently?
• Who needs to act, and do they have the authority and resources to complete this within the assigned timeline?
• Is this action realistic given current workload, or will it slip without someone actively following up?

Tracking and Verification

An assigned corrective action is not a completed one. Every open action should appear on a tracking log that the JHSC reviews at its next meeting. The responsible person should confirm completion in writing, and a second person — ideally a committee member or supervisor not directly involved in completing the action — should verify that it was done as described and that it has actually changed the condition. If an action is marked complete but the underlying condition persists, the investigation is not finished.