Leading vs Lagging Safety Indicators: What Each Measures and Why Both Matter

Most safety reporting is backwards-looking. Injury rates, lost-time frequency, total recordable injury frequency (TRIFR), incidents per million hours worked — all of these measure what has already happened. They are useful for trend analysis. They tell you nothing about the state of your defences right now.

A site can have a declining injury rate while its critical controls are degrading. Minor injuries drop because of housekeeping campaigns and PPE compliance. Meanwhile, the gas detection system is overdue for calibration, two critical control verifications have been missed for three consecutive months, and a TARP trigger has been acknowledged and not acted on. The lag indicators look good. The risk picture has not improved.

This is the problem that leading indicators are designed to address.

What Lagging Indicators Measure — and What They Miss

Lagging indicators measure outcomes: events that have already occurred. The standard set in mining and heavy industry includes:

• Total recordable injury frequency (TRIFR)
• Lost-time injury frequency (LTIFR)
• Medically treated injury frequency (MTIFR)
• Dangerous occurrence rate
• Near-miss frequency
• Environmental incident count

These indicators are necessary. They confirm whether the overall injury rate is moving in the right direction over time. They satisfy regulatory reporting requirements. They allow benchmarking against industry peers.

They miss two things. First, they are insensitive to low-frequency, high-consequence events. A site can have a zero injury rate for 18 months and then experience a fatality. The lag indicators provided no warning because fatalities do not accumulate gradually — they result from a specific combination of failures that lag indicators cannot detect in advance.

Second, they measure consequences, not causes. A drop in LTIFR could reflect genuine risk reduction, or it could reflect changed reporting thresholds, a workforce less willing to report, or a run of luck. The indicator does not distinguish between these. Acting on a drop in LTIFR as if it reflects improved controls, when it actually reflects changed reporting behaviour, leads to under-investment in the controls that actually prevent harm.

What Leading Indicators Track Before an Incident Occurs

Leading indicators measure inputs and process quality — the activities and conditions that determine whether a future incident is more or less likely. They focus on the state of the risk management system, not its outputs.

Effective leading indicators in mining and heavy industry include:

Critical control verification rates: What percentage of scheduled critical control verifications were completed on time this month? A rate below target means barriers are going unchecked. This is a direct measure of whether the critical control management system is functioning.

Overdue corrective actions: How many corrective actions from incident investigations or control failures are past their due date? A growing backlog means findings are not being closed. The conditions that produced the last incident remain unaddressed.

TARP activations and responses: How many TARP triggers were activated, and was the response executed correctly and on time? A pattern of missed or delayed responses indicates the TARP system is not operational in practice.

Control failure rate: Of the critical controls verified this month, how many failed their performance standard? A rising failure rate is a direct signal of barrier degradation.

Workforce participation in reporting: Near-miss and hazard report rates per hundred workers per month. A declining rate in a high-hazard environment is more likely to reflect reduced reporting confidence than a genuinely improving risk picture.

Scheduled risk assessment completion: Are risk assessments for high-risk tasks completed before work begins, or are JSEAs and SWMS being signed after the fact?

What a Balanced Indicator Set Looks Like in Practice

A balanced indicator set uses leading indicators to monitor the health of risk controls, and lagging indicators to confirm whether outcomes are improving over time. Neither type is sufficient alone.

A practical set for a mining operation might include:

Type	Indicator	Target	Frequency
Leading	Critical control verification rate	≥ 95%	Monthly
Leading	Overdue corrective actions (>30 days)	0	Monthly
Leading	Near-miss reports per 100 workers	≥ 2	Monthly
Leading	TARP response compliance rate	100%	Monthly
Lagging	TRIFR (rolling 12 months)	Site target	Monthly
Lagging	High-potential incident count	< prior period	Monthly
Lagging	Critical control failures resulting in incident	0	Monthly

The indicator set should be reviewed at least annually. Indicators that are consistently at target without any management action are not measuring anything that is actually at risk of failure. Replace them with indicators that track where the genuine exposure lies.

What Common Mistakes Look Like When Selecting Indicators

Too many lagging, too few leading. Most safety dashboards are 80% lagging. This reflects what is easy to measure, not what is most useful to track. If a site has 15 lagging indicators and two leading indicators, the reporting system is not oriented toward prevention.

Vanity leading indicators. Training completion rates and toolbox talk attendance are often used as leading indicators. They measure activity, not risk control effectiveness. A worker who attended a toolbox talk on isolation procedures and then conducted an isolation incorrectly is counted as a positive on the leading indicator. These are activity metrics, not risk metrics.

Leading indicators disconnected from risk. A leading indicator is only useful if it measures something that is causally connected to the outcomes you are trying to prevent. The question to ask of every proposed leading indicator: if this metric deteriorates, does that make a serious incident more likely? If the connection cannot be articulated, the indicator is probably not measuring the right thing.

Measuring the indicator instead of the risk. Once an indicator is established, there is pressure to maintain it. This can lead to gaming — completing verification paperwork without conducting the verification, or reporting near-misses selectively to manage the rate. An indicator that is being managed rather than monitored has lost its value.

What Leading Indicators Require From Your Data Systems

Leading indicators require data that is collected consistently, in real time, and connected to the risk management system. This is where spreadsheet-based systems fail.

In a spreadsheet environment, verification records sit in one file, corrective actions in another, near-miss reports in a third. Calculating a verification completion rate requires manual aggregation. By the time the monthly report is compiled, the data is 30 days old and the opportunity for early intervention has passed.

A connected risk management platform makes leading indicators practical: verification records are captured in the field on mobile devices and aggregated automatically; overdue actions surface on a dashboard without manual compilation; TARP activations are logged and their responses tracked against the required timeline. The indicators become real-time rather than retrospective.

Without that connection, leading indicators become a reporting exercise rather than a management tool. The data arrives after the window to act on it has closed.

---

For a breakdown of how RiskSight tracks critical control verification rates, overdue actions, and control health across sites in real time, see Risk Management Software for Mining, Construction & Heavy Industry.

Start a 30-day free trial with demo data included. No credit card required.