ISO 45001 Continual Improvement: Driving Long-Term OH&S Performance

Compliance stops at minimum requirements. Continual improvement never stops asking: How could we be better?

ISO 45001 Clause 10.3 is one sentence: “The organisation shall continually improve the suitability, adequacy and effectiveness of its OH&S management system.”

On the surface, this sounds aspirational but vague. In practice, it separates organisations that merely pass audits from organisations that build genuine safety excellence.

Continual improvement is not a special program you launch. It’s embedded in how you operate. It flows from your monitoring data. It’s triggered by your audit findings. It comes from worker suggestions. It emerges from management review discussion. And it’s powered by the PDCA cycle — a discipline that turns ideas into systematic change.

The organisations that excel at OH&S don’t do it through heroic effort. They do it through the accumulation of small, deliberate improvements, compounded over time. This article walks you through how to make continual improvement real and measurable.

Understanding Continual Improvement

Continual improvement is the disciplined commitment that your system is never good enough. Not because it’s failing, but because the potential for better always exists.

It’s important to distinguish:

Compliance: Meeting minimum legal and ISO 45001 requirements. Necessary but not sufficient. An organisation that merely complies will eventually fall behind.

Corrective action: Responding to nonconformities and incidents. Fixing what’s broken. Essential but reactive.

Continual improvement: Proactively making things better. Moving from responding to prevention. Building capability beyond minimum requirements. Strategic rather than tactical.

Many organisations confuse these. They invest heavily in corrective action (responding to incidents) and minimal effort in improvement (anticipating where incidents might occur). This is backwards. Improvement prevents incidents; correction merely responds to them.

The PDCA Cycle: The Engine of Improvement

The Plan-Do-Check-Act cycle is how you structure improvement. It’s simple in principle, profound in discipline.

The power is in the repetition and discipline. Most organisations attempt improvement once, it doesn’t work immediately, and they abandon it. PDCA acknowledges that improvement is iterative. You cycle until you find an approach that works.

Sources of Improvement Input

Where do improvement ideas come from? Everywhere. Your job is to make them visible and systematic.

1. Your Monitoring Data

If your LTIFR is declining, that’s positive, but why? What changed? Can you amplify it? If your near-miss reporting is flat or declining, that’s a red flag — either you’re missing opportunities or workers aren’t reporting. If your inspection completion rates are sliding, that’s a leading indicator of future problems.

Good monitoring asks: What trends do we see? What’s the trajectory? Are we moving in the direction we want? What adjustment could accelerate improvement?

2. Your Audit Findings

Internal audits surface gaps. Beyond correcting each nonconformity, look for patterns. If three audits found procedures not being followed, that’s not just correctable nonconformities — it’s a signal that your procedure design or training process needs improvement.

3. Incident Data and Investigations

Each incident teaches you something. Aggregate incidents by type or cause. What patterns emerge? A cluster of incidents involving the same hazard suggests your control needs redesign. Incidents involving new workers suggest your onboarding or training is inadequate. Incidents during high-pressure periods suggest your system isn’t resilient under stress.

4. Worker Suggestions and Feedback

Workers see hazards, inefficiencies, and opportunities you miss. But suggestions only surface if workers feel safe raising them. Create channels: suggestion schemes, safety committee, informal feedback through managers. Actively solicit: “What could we do better?”

Critically: respond visibly to suggestions. When a worker makes a suggestion and you act on it, word spreads. When you reject suggestions without explanation, people stop suggesting.

5. Management Review Discussion

This is where intelligence synthesises. Management review should explicitly ask: “Beyond fixing what’s broken, what could we improve?” Pick one or two areas per quarter and assign ownership. The improvement becomes concrete, not abstract.

6. Benchmarking Against Peers

How do you compare to similar organisations? What are they doing that you’re not? Industry associations, peer networks, published case studies can reveal improvement opportunities. The phrase “best practice” is overused, but learning from peers is legitimate.

7. Regulatory and Legal Changes

New standards, regulations, or court decisions can suggest improvements to stay ahead of legal change. If a regulation tightens, and you improve before the regulation takes effect, you’re improving proactively rather than reactively.

8. Technology and Innovation

New tools (digital incident reporting, predictive analytics, virtual reality training) can improve how you operate. The improvement isn’t the tool itself — it’s what the tool enables (faster analysis, better predictions, better training retention).

Leading vs. Lagging Improvement Strategies

There are two fundamentally different approaches to improvement.

Lagging Strategy: React to What Happened

You improve because you had an incident. You review how it occurred and fix the system to prevent recurrence. This is corrective action. It’s necessary, but it waits for something bad to happen before you act.

Example: A worker is injured because a hazard wasn’t controlled. You investigate, identify root cause, and implement control. Now that hazard is controlled.

Problem: You’ve only prevented that specific injury. Similar hazards elsewhere remain uncontrolled until you discover them through incident.

Leading Strategy: Anticipate What Could Happen

You improve because you see a pattern emerging. You notice that near-miss reports for a particular hazard are increasing. You notice that workers consistently find a procedure difficult to follow. You notice that a particular type of equipment has a design flaw that creates hazard exposure. You improve before an incident forces the issue.

Example: Your near-miss data shows that workers regularly encounter a hazard in a particular area. You redesign the control before anyone is injured. You prevent the incident entirely.

Advantage: You prevent incidents rather than responding to them. You’re making improvements constantly, based on patterns, not crises.

Mature organisations shift from lagging to leading. They build their improvement agenda around leading indicators and emerging patterns, not just incidents. They’re constantly asking: “What could go wrong?” rather than waiting to see what does go wrong.

Building a Continual Improvement Culture

Culture is everything. In a compliance culture, workers fear reporting problems (it makes their area look bad). In an improvement culture, workers eagerly surface issues (it helps the organisation get better).

How do you build improvement culture?

Model It from the Top

When the CEO publicly embraces an idea from a frontline worker, that’s cultural messaging. When management visibly acts on audit findings and incorporates them into strategy, that signals that improvement is valued. When leaders admit mistakes and discuss how to improve, they give permission for others to do the same.

Make Improvement Everyone’s Job

Don’t assign improvement to a committee. Make it part of every manager’s role. In your performance evaluation criteria, include: “What improvement initiative did you drive this year?” In safety committee meetings, dedicate 20 minutes to “Where can we improve?” In toolbox talks, ask workers: “What’s frustrating about our current process?”

Celebrate Small Wins

Don’t wait for big breakthroughs. Recognise incremental improvements. When a team redesigns a procedure to make it clearer, celebrate it. When a worker’s suggestion is implemented, publicly acknowledge the contribution. These signals compound.

Measure and Track Improvements

What gets measured gets managed. Maintain a log of improvements implemented. Track the impact of each improvement. At management review, report: “This quarter we implemented 4 improvements, resulting in X% reduction in near-miss incident latency, Y% reduction in procedure non-compliance, etc.” Quantify the value.

Allocate Resources to Improvement

If improvement is genuinely valued, allocate budget and time. Have people dedicated time (even 10% of their week) to improvement work. If every improvement project has to compete with operational demands, improvement will lose.

Moving from Compliance to Leading Practice

Leading practice means you’re intentionally creating an OH&S system that exceeds legal requirements. You’re not just checking boxes on ISO 45001; you’re building a system that prevents harm.

Practically, this means:

• Your monitoring includes leading indicators: You track near-miss rates, inspection completion, hazard close-out cycle times. You don’t just count incidents after they happen.
• Your incident investigations are rigorous: You invest in root cause analysis, using structured tools. You don’t settle for superficial explanations.
• Your worker engagement is genuine: You actively solicit input. You visibly act on suggestions. You share back what you learned and what changed.
• Your improvements are systematic: You use PDCA. You track what you tried, what worked, what didn’t. You learn from failures and iterate.
• Your culture values safety: Safety isn’t a compliance function separate from operations. It’s integrated. Managers’ bonuses reflect safety performance. Promotion decisions consider safety contribution.

This doesn’t happen overnight. But organisations that commit to leading practice see the returns: lower incident rates, higher worker engagement, better insurance outcomes, stronger community reputation.

Measuring Improvement Impact

You can’t improve what you don’t measure. For each improvement initiative, define:

Before state: What was the situation before? What metric captures it? What was the level?

Improvement action: What will you change?

After state: What metric indicates improvement? What’s your target?

Timeline: When will you measure improvement (2 weeks, 2 months)?

Verification: How will you gather evidence of improvement?

Example:

• Before: Incident investigations take an average of 21 days from report to final action. Target workers report near-misses; many don’t because they think the process is slow.
• Improvement: Establish dedicated investigator role; implement expedited review process for near-misses; daily briefing on incidents to keep momentum.
• After: Target is 7 days average.
• Timeline: Measure at 2 months.
• Verification: Pull incident register data. Calculate average cycle time. Conduct worker survey on confidence in reporting.

When you achieve the target, document it. Use it in communication to the organisation. At management review, celebrate it.

Benchmarking and Learning from Others

How do you know if your performance is good? Benchmarking against peers provides context.

Operational benchmarks: How do your LTIFR, TRIFR, near-miss rates compare to industry average? Where are you in the distribution? Top quartile, median, bottom quartile?

Procedural benchmarks: How does your incident investigation process compare to best practice? Are your leading indicators similar to what other organisations track? Is your continual improvement process typical or exemplary?

Industry associations, insurance consultants, and peer networks can provide benchmarking data. The goal isn’t to obsess over rankings, but to understand whether you’re performing well, adequately, or poorly, and where you should focus improvement.

Technology in OH&S Improvement

Technology enables improvement but doesn’t guarantee it. Digital tools help by:

• Capturing data faster: Mobile incident reporting means near-miss data is entered in real-time, not reconstructed days later.
• Enabling analysis: Software can automatically categorise incidents, identify trends, highlight emerging patterns.
• Improving training: Virtual reality, microlearning, and spaced repetition can improve retention and competence compared to traditional classroom training.
• Predictive capability: AI analysis of incidents and near-misses can identify likely future hazards before they manifest.
• Reducing friction: Digital inspection systems with offline capability mean inspections happen on schedule, not deferred due to inconvenience.

But remember: the technology is a tool. The improvement is in how you use the tool to change behaviour or capability.

Demonstrating Continual Improvement to Auditors

Certification auditors will ask: “Show me your continual improvement.” Here’s what they’re looking for:

Evidence of deliberate improvement efforts: Not just incident response, but proactive improvement. Can you show improvements you’ve implemented in the past two years? What was the reason for each? How did you measure impact?

Improvement strategy: Do you have a process for identifying, planning, implementing, and verifying improvements? Or do improvements happen ad hoc when someone thinks of something?

Resource allocation: Have you allocated budget, time, or people to improvement? Or is improvement expected to happen on top of existing workload (which means it doesn’t happen)?

Measurement: For each improvement, can you show before/after data? Can you demonstrate that the improvement actually occurred?

Iteration and learning: Have you tried improvements that didn’t work? How did you handle failure? Did you learn from it and try again? (This is actually a sign of maturity — failure is part of learning.)

Organisations that can walk an auditor through a portfolio of improvements (from initial idea through implementation through measurement) demonstrate genuine continual improvement. Those that have no systematic improvement process will struggle in audit.

Common Continual Improvement Failures

Failure 1: Improvement as a Special Program

You launch an improvement initiative once a year. It gets attention for a few months, then fades. Improvement becomes episodic rather than continuous.

Fix: Make improvement part of your ongoing rhythm. Have improvement discussion at every management review. Make improvement a normal part of how managers operate. It becomes continuous, not episodic.

Failure 2: Ideas Without Implementation

You identify lots of improvement opportunities but implement few. Ideas are brainstormed, discussed, and archived. Nothing changes.

Fix: Be selective. Choose one or two improvements per quarter. Give them real ownership, resources, and timeline. Execute them through to verification. Quality of execution matters more than quantity of ideas.

Failure 3: Measurement Without Action

You measure improvement but don’t communicate results or act on them. A dashboard sits in a folder. Data is collected but not discussed. Improvement remains invisible.

Fix: Share improvement results widely. At management review, discuss what worked and what didn’t. In toolbox talks, celebrate wins. When workers see improvement, they contribute more ideas.

Failure 4: Blaming Resources

You claim you can’t improve because you don’t have budget or staff. Improvement is deferred indefinitely.

Fix: Some improvements cost money; many don’t. Improving a procedure costs nothing but thinking. Better communication costs nothing. Empowering workers costs nothing. Start with low-cost improvements. As they show value, budget for bigger initiatives.

Practical Checklist: Building Strong Continual Improvement

• Establish a process for identifying improvement opportunities (monitoring, audits, incidents, worker feedback)
• At management review, explicitly discuss improvement priorities for the next period
• For each improvement initiative, define before state, action, after state, and measurement
• Assign clear ownership and allocate resources
• Implement through PDCA: plan carefully, do thoroughly, check for results, act on what you learned
• Measure improvement impact and document it
• Communicate results to the organisation
• If an improvement doesn’t work as planned, analyse why and iterate rather than abandoning it
• Maintain a portfolio of improvements implemented, showing what was changed and what resulted
• Celebrate and recognise improvements, especially those from worker suggestions
• Annually review improvement trends: Are we getting better? At what pace? In what areas?

FAQ: Continual Improvement in Practice