You did the investigation.
You filled out the form.
You held the meeting.
And then the same type of incident happened again, which means you did not actually “solve” it, you documented it.
Recurring incidents are not usually a mystery.
They are a signal that the organization is treating investigation like a post-event ritual instead of a control-improvement process, so the paperwork closes while the risk stays open.
In practice, reoccurrence shows up when corrective actions are weak, when controls are not verified in the field, and when leadership accepts “trained the worker” as a finished outcome.
This blog is about getting brutally practical.
If incidents are reoccurring, you need to shift from “what happened” to “why the system keeps allowing it,” and then you need to prove the fix is working, not just hope it is.
If you investigate ten incidents and the same exposure repeats, then your investigation system is producing activity, not prevention.
That matters because incident investigation is supposed to generate corrective actions that reduce risk and prevent recurrence, not just explain the past.
OSHA is very direct that root cause analysis during incident investigation is about finding underlying causes so similar incidents do not happen again, which is the whole point of spending time on it.
If recurrence continues, it is telling you that either the root causes were missed, the corrective actions were not implemented, or the actions did not actually control the hazard in real work conditions.
https://www.osha.gov/sites/default/files/publications/OSHA3895.pdf
The good news is that recurrence gives you clean feedback.
It narrows the search to a few predictable failure points that show up again and again across industries.
A surface cause sounds tidy.
“Worker wasn’t paying attention.”
“Improper technique.”
“Didn’t follow procedure.”
Those statements might be partly true, but they are usually describing the last link in the chain, not the chain itself.
When you stop there, your “corrective action” becomes retraining or discipline, and the organization keeps the same conditions that produced the error in the first place.
The UK HSE pushes investigators to look at underlying and latent causes, including human factors and why human failures occurred, because that is where prevention actually lives.
https://www.hse.gov.uk/humanfactors/topics/investigation.htm
A corrective action can be “completed” and still be useless.
If the work environment is fast, messy, and high-pressure, then weak actions get crushed by production reality.
For example, “remind workers to wear face shields” will not beat fogging, poor fit, missing availability, or a supervisor who rewards speed over compliance.
If your actions do not change the conditions of work, then you are depending on perfect behavior to control risk, which means you are planning for failure.
This is the big one.
Teams implement actions, mark them complete, and move on, but nobody checks whether the change actually reduced exposure.
CCOHS frames investigation around finding facts that lead to corrective actions and looking for deeper causes rather than fault, but the hidden requirement is that corrective actions must actually prevent recurrence, which requires follow-up and verification.
https://www.ccohs.ca/oshanswers/hsprograms/investig.html
Verification is not “we sent an email” or “we updated the procedure.”
Verification is field reality: the control is present, used, and effective during normal operations, including night shift, peak workload, and awkward jobs.
This is a mindset shift that changes everything.
Instead of asking, “What did the worker do?” you ask, “Which control should have prevented this exposure, and why did that control fail or not exist?”
When you investigate controls, you get answers like: guarding design was bypassable, supervision tolerance was high, job planning was incomplete, preventive maintenance was overdue, or the risk assessment did not reflect actual work.
OSHA’s incident investigation guidance emphasizes a systems approach aimed at controlling underlying or root causes to prevent recurrence, which aligns with this control-focused framing.
https://www.osha.gov/incident-investigation
If a near miss had high potential, but you classify it as minor, you will assign a minor investigation and minor corrective actions, which means you just trained the organization to ignore a warning shot.
High-potential near misses deserve high-quality analysis, because their causes are often identical to serious injury and fatality precursors.
If corrective actions do not have a single accountable owner, a due date, and a close-out evidence standard, then they will get “sort of done” and then forgotten.
Even worse, people will assume someone else handled it, which means the hazard remains and everyone believes it is gone.
You can usually diagnose this without a deep audit, because the symptoms are consistent:
Start every investigation with a blunt question: what should have prevented this?
List the barriers that were supposed to be in place, then test each one against the evidence.
If the barrier existed, why did it fail?
If it did not exist, why was it missing?
This approach forces the team to stop blaming the last person in the chain and start fixing the system that set them up.
A simple test helps.
If the control relies on memory, perfect attention, or constant policing, it is weak.
If the control changes equipment, layout, sequencing, isolation, guarding, automation, or physical separation, it is typically stronger.
If the control builds in a check that is hard to skip, it is stronger.
You do not always need engineering controls, but you do need controls that match the risk and the work environment.
This is where most programs fall apart.
Your corrective action process should require an effectiveness check after implementation, not at the next annual review.
ISO-style management systems put heavy emphasis on corrective action and evaluating effectiveness as part of improvement, which is why verification and performance evaluation are baked into the structure.
https://www.iso.org/obp/ui/es/
Verification should answer two questions:
Did we implement what we said we would implement?
Did it actually reduce exposure and prevent recurrence under normal conditions?
Lagging indicators tell you what already hurt you.
Leading indicators tell you if controls are weakening before someone gets hit.
Examples include critical control inspection completion, corrective action effectiveness checks completed on time, preventative maintenance compliance, and supervisor field verification frequency tied to top risks.
If you do not measure control health, you cannot manage it, and recurrence will keep sneaking in through degraded barriers.
Training matters, but it is rarely sufficient as the primary corrective action for recurring incidents.
If retraining is required, it should be paired with a change in the job conditions, plus a competency validation that proves the skill in the field, not just a sign-in sheet.
If the same incident type happens again, the next investigation should be stronger, not just repeated.
That means bigger scope, better analysis, and higher-level accountability.
It is about matching response strength to the clear evidence that the system is not controlling the risk.
If you are stuck in the loop of “investigate, close, repeat,” you usually need an outside reset, because internal teams have already normalized the process and the language around it.
That is where we come in, and we do it in a way that is practical, not academic.
Calgary Safety Consultants can help you:
Recurring incidents are not a sign that your people are hopeless.
They are a sign that your controls are not holding, your corrective actions are not strong enough, or your verification loop is missing, which means the organization is learning the wrong lesson from its own data.
When you treat recurrence as a control problem and you prove effectiveness in the field, you stop chasing the same incident with a new form, and you finally start buying down risk in a way that lasts.
Occupational Safety and Health Administration (OSHA).
Incident Investigation.
https://www.osha.gov/incident-investigation
Occupational Safety and Health Administration (OSHA).
Incident Investigations: A Guide for Employers (OSHA 3895).
https://www.osha.gov/sites/default/files/publications/OSHA3895.pdf
Canadian Centre for Occupational Health and Safety (CCOHS).
Incident Investigation – Health and Safety Programs.
https://www.ccohs.ca/oshanswers/hsprograms/investig.html
Health and Safety Executive (HSE), United Kingdom.
Human Factors in Incident Investigation.
https://www.hse.gov.uk/humanfactors/topics/investigation.htm
International Organization for Standardization (ISO).
ISO Online Browsing Platform – Standards Database (including ISO 45001).
https://www.iso.org/obp/ui/es/
Recurring incidents usually mean the investigation process is producing documentation instead of prevention, because root causes were missed, corrective actions were too weak, or controls were never verified in the field.
Retraining depends on perfect behavior in imperfect conditions, which means it will lose against time pressure, poor equipment, missing supervision, and bad job design unless the work conditions and controls also change.
A surface cause describes the last visible mistake or event, while a root cause explains the underlying system condition that allowed the mistake to happen, such as inadequate guarding, poor planning, maintenance gaps, unclear supervision, or missing critical controls.
Calgary Safety Consultants is here to help you ensure compliance, enhance safety, and streamline your OH&S program. Don’t wait—fill out the form, and we’ll connect with you to discuss how we can support your business. Let’s get started!