6 myths about TRIR that executives still believe

TRIR can fall while fatal exposure stays high. In 2024, the Bureau of Labor Statistics recorded 5,070 fatal work injuries in the United States, one every 104 minutes, which is enough to show why a recordable rate cannot be treated as a map of serious risk.

Across 25+ years in multinational EHS, Andreza Araujo has seen executives reward the number that is easiest to explain and miss the exposure that is hardest to see. In Safety Culture: From Theory to Practice, repeated decisions matter more than the dashboard headline because repeated decisions tell the truth about the site.

This article is for boards, plant directors and EHS leaders who need a safer way to read performance. The thesis is simple. TRIR remains useful, but only when it sits beside severity, precursor and control data, since none of those layers can be read from one rate alone.

Myth 1: lower TRIR means lower fatal risk

The myth sounds sensible because TRIR is a single number and single numbers feel decisive. When the rate drops, executives assume the site became safer, and that assumption is especially attractive when the monthly pack needs a clean headline.

The problem is that TRIR tracks recordable injuries, not the full chain of exposure that leads to a serious event. Bird and Heinrich's pyramid logic, together with James Reason's work on latent conditions, says the visible result is only the end of a longer path, which means a low rate can sit on top of the same weak barriers.

That is why the better question is not whether the rate improved, but which serious-potential events changed first. If you want the metric layer behind that question, SIF precursors give a better read on where fatal exposure is building.

What to do instead is simple. Keep TRIR, but stop using it as a proxy for fatality risk. Add serious-potential incidents, precursor counts and critical-control checks to the same review so the board sees the exposure path, not only the outcome.

Myth 2: one TRIR can compare every site fairly

The myth survives because one formula feels fair. If every site uses the same denominator, executives assume the same score means the same performance, even when the work mix, contractor load and exposure type are completely different.

That assumption breaks as soon as one site runs shutdowns, another site carries mostly steady-state work and a third site depends on heavy contractor activity. Patrick Hudson's maturity model is useful here because the same injury rate can sit on very different levels of control maturity and leadership discipline.

Across more than 250 cultural transformation projects supported by Andreza Araujo, one recurring pattern appears: the rate can look comparable while the operating reality is not. A single score hides whether the site is running high-energy work, whether contractors dominate the exposure or whether the reporting culture is thin.

What to do instead is to compare like with like. Segment the board view by work type, contractor share, high-energy tasks and exposure hours, because a plant, a project and a warehouse should not be judged against the same picture without context.

Myth 3: a clean TRIR proves controls are working

A clean TRIR often feels like proof because it is easy to celebrate and easy to explain upward. The trap is that control failure can grow quietly while the recordable count stays flat, especially when exposure is low or when people learn to report only after the situation has already hardened.

James Reason's model helps here because the event you can count is usually the last visible step in a longer alignment of weak conditions. A weak lockout, a rushed permit, a missed change in scope or a drifting supervision routine can all sit in the system before the first recordable appears.

That is why control quality has to be audited directly. The article on control effectiveness metrics is the right companion, because it moves the conversation from the result to the barrier that should have prevented the result.

What to do instead is to review critical-control verification on a fixed cadence. If the controls are healthy, the evidence should show it in field checks, permit quality, isolation discipline and response speed, not only in a clean injury count.

Myth 4: TRIR is objective enough for bonuses and board packs

The myth is tempting because TRIR is compact, familiar and easy to rank. Boards like it because the number fits on one slide, and managers like it because it looks measurable, but that convenience is exactly what creates distortion.

When bonuses depend on the rate, people start protecting the number before they protect the learning. Early reporting becomes a liability, first aid can be treated as noise and classification pressure creeps into the system, which is the wrong lesson to teach if the goal is honest control.

As Andreza Araujo writes in The Illusion of Compliance, a site can look orderly while the real exposure remains untouched. A TRIR-only bonus logic encourages that same illusion, because it rewards the appearance of control more than the control itself.

What to do instead is to separate incentive metrics from governance metrics. Let the board see TRIR, but pair it with severity-weighted events, precursor trends and critical-control performance so the conversation is about decisions, not about protecting a score.

Myth 5: zero recordables means low risk

Zero recordables looks like excellence because it is tidy and easy to publish. The danger is that a zero can come from low exposure, from luck, or from a reporting culture that has learned to keep small events invisible.

The issue is sharper in high-energy work, because a shutdown site, a lift-heavy project or a road-risk operation can sit on a quiet score while the next serious event is still possible. The board should not confuse a light month with a strong barrier system.

Andreza Araujo's experience across 30+ countries points to the same lesson. A site can look disciplined from the office window and still be one change in scope away from a serious event if the leaders never ask what work was actually done, what changed and what was deferred.

What to do instead is to ask about the work mix behind the zero. If the operation had heavy lifts, confined spaces, hot work, line breaks or road exposure, then the real question is not whether TRIR stayed flat. The real question is whether the controls held under pressure.

Myth 6: one metric can summarize serious risk by itself

The myth persists because leadership wants one headline. A dashboard with only TRIR feels decisive, but serious risk does not live in one dimension, and the board pays for that simplification when it misses the next weak signal.

A useful review needs at least three layers: frequency, severity potential and control status. Patrick Hudson's maturity lens and James Reason's latent-condition lens are both reminders that the system has to be read in parts before it can be understood as a whole.

That is also why the article on leading indicators matters. TRIR tells one part of the story, but it does not tell the board what changed in the work, which barrier slipped or where the next serious event is likely to come from.

What to do instead is to build a three-line board view. One line for TRIR, one line for severity and one line for critical-control or precursor status. If any of those lines move the wrong way, the executive team has a decision to make before the next meeting closes.

What to do now

Keep TRIR, but demote it from verdict to one input. The board should read it as a lagging signal that deserves context, not as a substitute for the risk review itself.

The practical upgrade is a short monthly pack that answers four questions: what changed in exposure, which barrier slipped, where did severity rise and what decision now belongs to the executive team. That pack works because it forces the conversation away from congratulating a number and toward controlling the work.

If the board can only answer the first question, it is managing a score. If it can answer all four, it is managing risk.

FAQ

Is TRIR useless? No. TRIR is useful as one lagging indicator, especially for reporting trends over time. The error is asking it to do the job of a full risk system.

What should replace TRIR? Nothing replaces it by itself. The better move is to pair it with severity-weighted events, SIF precursors and critical-control verification so the board can see the whole picture.

Which Andreza Araujo book fits this topic best? Safety Culture: From Theory to Practice fits best because it treats repeated decisions as the real signal. The Illusion of Compliance is the sharper companion when the number looks good but the work is still not under control.

For leaders who want a deeper review of the dashboard and the decisions behind it, Andreza Araujo's consulting work can help turn TRIR from a verdict into one line inside a more honest safety review.

Talk to Andreza Araujo