TRIR vs LTIFR vs DART: The Board-Level Choice Between Exposure, Rate, and Response

TRIR, LTIFR, and DART all describe injury performance, yet they answer different management questions. A board that treats them as interchangeable can approve a safety strategy that improves the report while leaving serious exposure untouched.

Picture a company whose total recordable injury rate falls while lost-time cases remain concentrated in a small number of high-energy tasks. The dashboard looks calmer, but the decision problem has not disappeared. Leaders need to know whether the organization is reducing exposure, preventing serious harm, or only changing how events are counted.

This comparison gives each metric a defined job. It also shows why no single rate deserves to become the board's safety score. The useful choice is not the most familiar acronym. It is the metric that fits the decision the board is about to make.

What should the board ask before choosing a safety metric?

The first question is not whether TRIR, LTIFR, or DART is better. It is what the board needs to decide. If directors are reviewing broad injury burden across a workforce, a total recordable rate may provide a useful signal. If they are testing whether injuries are removing people from work, a lost-time rate is closer to that question. If they are examining operational disruption, restricted work and job transfer deserve explicit attention.

The United States Occupational Safety and Health Administration describes recordkeeping categories and rate calculations through its recordkeeping guidance, while the U.S. Bureau of Labor Statistics publishes national injury and illness data through its Survey of Occupational Injuries and Illnesses. Those sources are useful reference points, although a company's internal definitions still need to be written and applied consistently.

A board should therefore specify the decision, the population, the exposure period, the denominator, and the action that follows a change in the number. Without that chain, the metric becomes a status symbol rather than a control on management attention.

Which criteria distinguish TRIR, LTIFR, and DART?

Use five criteria when comparing the metrics. The first is scope, meaning which event categories enter the calculation. The second is operational meaning, meaning what a change in the rate says about work. The third is sensitivity to reporting and classification choices. The fourth is usefulness for comparing sites or business units. The fifth is the management action that should follow a signal.

The comparison is not a ranking. It is a boundary map. Each metric becomes misleading when leaders ask it to answer a question that belongs to another measure or to field evidence.

What does TRIR reveal, and where does it mislead?

TRIR is useful when leadership wants a broad view of recordable injury and illness cases across a defined workforce and period. It can reveal whether the organization is experiencing a rising case burden, whether one site differs from comparable operations, and whether the case mix is changing after a management intervention.

Its weakness appears when the rate becomes the strategy. A lower TRIR can reflect fewer minor cases while high-energy exposures remain unchanged. It can also move because of workforce mix, hours worked, case classification, contractor boundaries, or reporting behavior. The number is real, but its meaning depends on the conditions around it.

The board should pair TRIR with a case-severity review and a serious-injury-potential review. The latter asks whether an event could have produced a fatal or permanently disabling outcome, even when the actual injury was less severe. That question connects the dashboard to the leading indicators that test control quality, rather than allowing a broad rate to absorb every concern.

When is LTIFR the more useful lens?

LTIFR is more useful when the decision concerns lost work time, return-to-work performance, or recurring conditions that remove people from their jobs. It can help leaders examine whether a site is preventing injuries that interrupt work and whether recovery arrangements are being managed consistently.

Lost time is not a complete severity measure. A serious exposure may produce no lost time because the task was interrupted quickly, the worker returned under modified duties, or the event was classified differently. A team that celebrates a low LTIFR without reviewing high-energy near misses can mistake an outcome measure for evidence that the hazard is controlled.

LTIFR deserves a second review of work design, task controls, and repeat mechanisms. The board can ask whether lost-time cases cluster around a process, contractor interface, shift pattern, or maintenance condition whose risk owner has not changed. That line of inquiry is stronger than asking managers to explain a rate in isolation.

What does DART add to the decision?

DART adds visibility into cases that involve days away, restricted work, or job transfer. It is valuable when leaders want to see how injuries affect the person's ability to perform normal work, because a case can create meaningful operational disruption without meeting a lost-time-only definition.

The metric still requires disciplined classification. A restricted-duty arrangement can look protective when it is genuinely suitable, yet it can also hide a weak control if the person is moved around the exposure without correcting the task. The board should ask what changed in the work, not only whether the person remained on the payroll.

DART works best beside evidence about control verification, action closure, and repeat exposure. The safety data-quality checks are relevant because a dashboard cannot distinguish a verified control change from an administrative closure unless the underlying record captures that difference.

How should a board compare the three metrics?

Use a decision matrix rather than a league table. Score each metric against the board's question, then record the evidence that must accompany the number.

This matrix prevents a common governance error. Leaders select a metric first and then reshape the question to fit it. The stronger sequence begins with the decision, selects the measure, and defines the evidence that can disprove a comforting interpretation.

Which metric should the board trust first?

Trust the metric that matches the decision, but do not trust any metric without its denominator, definitions, and case review. For broad injury burden, TRIR is usually the starting point. For lost work time, LTIFR is more direct. For restricted work and transfer, DART adds information that LTIFR leaves outside the frame.

None of the three should be the board's only safety measure. The UK Health and Safety Executive's annual statistics and performance information illustrate why injury outcomes need context about exposure, work activity, and the wider pattern of harm. Internal governance needs the same discipline, even when the dashboard is much smaller.

Andreza Araujo's *Safety Culture: From Theory to Practice* is useful here because it treats the gap between declared expectations and operating reality as a leadership problem. A board that rewards one rate without asking what work changed may be encouraging metric management rather than risk management.

How can leaders prevent a metric from becoming the target?

Set a review rhythm in which the number is only the first page. The next page should show case classification, serious-injury potential, control verification, unresolved high-hazard actions, and the owners who can change the exposure. If the rate improves while those fields deteriorate, leadership should treat the divergence as a warning.

Use the dashboard blind spots as a standing challenge in the monthly review. Ask whether the dashboard shows outcomes, whether it shows exposure, and whether it proves that controls work in the field. Then connect the answer to one decision with a named owner and a verification date.

A useful board conclusion is simple. TRIR describes broad recordable burden, LTIFR describes lost-time frequency, and DART describes cases that alter normal work. The safety decision is stronger when all three are interpreted against the exposure and control evidence that the rates cannot see.