ISO 45001 beyond the certificate

Worker participation is the clause that fails

ISO 45001 separated consultation and participation from general communication for a reason: it is the requirement that distinguishes a safety system from a safety folder. The standard is specific about what workers must be consulted on and what they must participate in, and it asks you to remove the barriers that stop it happening.

Most organisations can produce a health and safety committee and minutes. Fewer can show that workers were consulted before a control was chosen, that a hazard raised by a crew changed something, or that the reason a suggestion was declined got back to the person who made it.

• Consultation happens before the decision. Telling people afterwards is communication, and the standard treats them differently.
• Barriers are practical: meetings scheduled during shift, forms that require a computer, English-only documents, or a culture where raising something creates work for the person who raised it.
• Close the loop visibly. The strongest evidence of participation is a record showing what a worker raised, what was decided, and that they were told.
• Non-managerial workers must be involved in hazard identification and control selection specifically — not consulted in general terms.

Hazard identification that keeps up with the work

Hazard registers age badly. The work changes weekly — new plant, a different subcontractor, a modified sequence, a temporary arrangement that becomes permanent — while the register is reviewed annually. By month six it describes a site that no longer exists.

The fix is not more thorough annual reviews. It is making identification continuous and low-friction, so the register updates as a by-product of work rather than as a separate exercise.

1. 1.Make reporting take under a minute from a phone, in the field, with a photograph. Friction is the single largest determinant of how much gets reported.
2. 2.Trigger a register review on change: new task, new plant, new contractor, new substance, or an incident anywhere on site.
3. 3.Include the hazards the standard names but registers often miss — psychosocial hazards, fatigue, work organisation and how work is actually done rather than how it is planned.
4. 4.Feed incidents and near misses back into the register automatically, so the two are one system rather than two.
5. 5.Review high-risk task assessments with the crews who perform them, at the frequency the risk justifies.

The hierarchy of controls, applied honestly

Everyone can recite the hierarchy. Far fewer systems show evidence that elimination and substitution were genuinely considered before the control landed on training and PPE — which is exactly what an auditor is testing when they ask why a control was chosen.

The honest version records the reasoning. It shows that eliminating the task was assessed and rejected for a stated reason, that an engineering control was priced and deferred with a date, and that administrative controls are the interim measure rather than the answer. That record is also what protects you if the control fails and someone asks how the decision was made.

• Record the options considered and why each was rejected, not just the option chosen.
• Treat PPE as the last layer, never the only one. A risk assessment whose controls are all PPE and training is a finding waiting to happen.
• Where an engineering control is deferred on cost, put a date and an owner against it. An indefinite deferral is a decision to accept the risk permanently — make that explicit at the right level.
• Verify controls in the field. The gap between the assessed control and the installed control is where incidents live.

Contractors inside your system

Contractor management sits across procurement, operations and safety, which is why it falls between them. ISO 45001 asks you to coordinate with contractors on the hazards arising from their work and yours, and to apply your procurement processes to safety requirements.

The practical version is unglamorous: know who is on site, know their evidence is current on the day they are there, specify your requirements before they are engaged rather than at the gate, and monitor the work rather than trusting the paperwork that got them through the door.

This is also the area where systems most often diverge from reality, because the register lives with one team and site access with another. If those two lists disagree, the register is not a control.

Leading indicators worth measuring

Lagging indicators tell you what already went wrong, and in a small organisation they are statistically noisy — a good year and a bad year can differ by one event. Leading indicators tell you whether the system is being used, and they move before the harm does.

• Percentage of corrective actions closed within their due date, and the age of the oldest open one.
• Hazard reports per worker per month — falling numbers usually mean falling reporting, not falling risk.
• Proportion of planned inspections and audits completed on schedule.
• Time from a hazard being reported to a control being in place.
• Contractor evidence currency: how many active contractors have an expired document today.

Pick three, publish them where the crews can see them, and review them monthly. A small number of indicators that are genuinely watched will change behaviour more than a dashboard of twenty that nobody opens.

Questions we get asked