If Total Recordable Incident Rate (TRIR) or Days Away, Restricted, or Transferred (DART) rate has fallen for three consecutive quarters, leadership will reasonably want to see that as evidence the safety program is working. As EHS professionals, we should want the same thing.
The better question is what else changed. Are hazards being corrected faster? Are preventive maintenance tasks being completed? Are workers still reporting near misses? Are critical controls actually in place where serious exposures exist?
TRIR and DART cannot answer those questions because they measure outcomes that have already occurred. Leading indicators can tell us more about the prevention work happening before an injury.
What Lagging Indicators Actually Tell Us
TRIR, DART, injury severity, workers' compensation costs, and similar measures answer legitimate questions: how much harm occurred, how severe it was, and whether outcomes are changing over time.
The problem starts when an outcome measure becomes a proxy for the entire safety system. A low TRIR tells us that relatively few recordable cases occurred during the period being measured. It does not independently tell us whether machine guarding is effective, hazardous-energy controls are being followed, preventive maintenance is current, or known hazards are being corrected.
The boundary between leading and lagging measures is not always perfectly clean. What matters operationally is whether a measure gives the organization useful information early enough to influence the outcome it cares about.
Why a Falling Injury Rate Can Give Us False Confidence
Imagine a facility that worked 300,000 hours last year and experienced three recordable injuries. This year it has one. The rate improves substantially.
Now imagine that overdue corrective actions doubled, preventive maintenance completion deteriorated, near-miss reporting dropped sharply, and several critical inspections were missed. The injury outcome improved, but the condition of the system may not have.
Recordable injuries are also relatively infrequent events, particularly at smaller sites. A single case can move an annual rate considerably, so short-term changes should be interpreted carefully rather than treated automatically as evidence that risk has risen or fallen.
The 2005 BP Texas City refinery explosion illustrates the broader problem. The U.S. Chemical Safety and Hazard Investigation Board found that reliance on the refinery's low personal injury rate as a safety indicator failed to give a true picture of process-safety performance and the health of the safety culture.
The same problem can occur anywhere an organization relies on personal injury rates to represent risks those rates were never designed to measure. A low injury rate and a healthy safety system are related goals. They are not the same measurement.
Choose Indicators That Tell You Something Useful
Leading indicators look earlier in the chain. The useful question is not simply whether an activity can be counted, but what the count tells us about the condition of the safety system.
Consider corrective actions. Tracking the number closed may reward activity. Tracking the percentage closed by the required date, the age of overdue high-risk actions, or the time from hazard identification to verified control tells us considerably more about whether known hazards are being addressed.
The same test applies to inspections. “Safety observations completed: 412” sounds positive, but the number alone does not tell us whether the observations focused on meaningful exposures, identified significant hazards, or resulted in effective follow-through.
A stronger indicator helps the EHS team see something it can act on. Instead of only asking how many inspections were completed, ask whether the planned inspections occurred, what significant hazards they identified, and whether those hazards were controlled.
Build Indicators Around the Risk
It is tempting to start by asking what safety metrics other companies track. A better starting point is the operation itself: what risks, weaknesses, or control activities do we most need visibility into?
If mobile equipment is one of the site's highest risks, useful indicators might examine pre-use inspection completion, unresolved vehicle defects, pedestrian-control deficiencies, or corrective actions from traffic observations. If hazardous energy is a major exposure, the measures should reflect that system instead.
The objective is not to build a universal EHS dashboard. It is to measure whether the controls and activities that matter in your operation are working.
Connect Outcomes to Their Drivers
A practical scorecard connects the result you care about with the conditions or activities expected to influence it:
- Outcome
- Driver
- Measure
- Action
Outcome. What result are we trying to improve or protect? For example, reduce hand injuries associated with machine intervention.
Driver. What condition or activity should influence that outcome? For example, effective machine guarding, proper hazardous-energy control, and timely correction of identified deficiencies.
Measure. What can we track that tells us whether those drivers are functioning? For example, completion of guarding inspections, overdue high-risk guarding deficiencies, or verification of corrective actions.
Action. What will we do when the indicator moves in the wrong direction? For example, escalate overdue high-risk deficiencies, increase targeted field verification, investigate repeated findings, or remove equipment from service where required.
For each important measure, also define who owns it, when it is reviewed, what result requires attention, and what happens next. If a number turns red every month and nothing changes because of it, the organization is maintaining a dashboard rather than managing risk.
Read Near-Miss Reporting Carefully
Near-miss reporting is a good example of why leading indicators require interpretation. Imagine one department reports 40 near misses this quarter and another reports two. Which department concerns you more? There is not enough information to answer.
A higher reporting rate can reflect stronger worker participation and greater trust in the reporting process. A low number can mean fewer events, but it can also mean employees do not report what they see.
Read the count alongside report quality and follow-through. Are reports specific enough to act on? How quickly are they reviewed? Are corrective actions completed? Are similar events recurring?
The goal is not to maximize near-miss reports forever. It is to build a reporting and response system that identifies weak signals early enough to act on them.
Put a Clock on Known Hazards
Corrective-action aging is one of the clearest ways to see whether known risk is waiting for a decision.
A dashboard that reports 95 percent of actions closed can look excellent while the remaining 5 percent contains the highest-risk items on the list. Track risk and age together.
An overdue low-priority housekeeping item and an overdue critical machine-guarding deficiency should not receive the same management attention simply because both are seven days late.
- How many high-risk actions are overdue?
- What is the oldest unresolved high-risk action?
- How long does it typically take to move from hazard identification to verified control?
- Are temporary controls being monitored while permanent corrections remain open?
- Are actions verified before they are marked complete?
Corrective-action data becomes more valuable when it tells leadership where known risk is waiting for action.
Keep the Scorecard Small Enough to Manage
More metrics do not necessarily create better visibility. A smaller set tied to significant risks, known weaknesses, and important program activities is usually easier to manage and act on.
| What we need to know | Possible indicator |
|---|---|
| What harm has occurred? | TRIR, DART, severity, injury type |
| Are employees raising concerns? | Hazard and near-miss reporting trends |
| Are known hazards being controlled? | High-risk corrective-action aging |
| Are critical preventive activities happening? | On-time completion of selected inspections or preventive maintenance |
| Are required capabilities in place? | Training or competency status for safety-critical work |
| Are controls working in the field? | Verification findings for selected critical controls |
The exact measures should change with the operation. The principle should not: every metric should answer a question the organization actually needs answered.
Bringing Leading and Lagging Indicators Together
A useful safety scorecard does not simply describe last quarter. It tells us where the system is strong, where it is drifting, and where someone needs to act.
Lagging indicators still matter because they tell us whether harm occurred and whether outcomes are changing. Leading indicators give us more opportunities to intervene before those outcomes occur.
The test for any metric is simple: What will we do differently if this number changes?
If there is no answer, the metric probably does not belong on the dashboard.
Test Your Own Scorecard
Before treating the safety scorecard as complete, an EHS team should be able to answer these questions:
- Can we explain what decision each metric is intended to support?
- Does each major lagging outcome have at least one meaningful upstream measure where appropriate?
- Are our leading indicators connected to the site's actual risks rather than simply easy-to-count activities?
- Do we distinguish activity volume from the quality and effectiveness of the activity?
- Can we identify our oldest overdue high-risk corrective action right now?
- Do we interpret near-miss reporting alongside response and follow-through rather than treating the raw count as good or bad?
- Does every important indicator have an owner, review point, and defined response?
- Have we removed metrics that nobody uses to make a decision?
A “no” or “I’m not sure” identifies where the scorecard may be measuring safety activity without helping the organization manage safety performance.
