DART Rate Explained: 4 Questions That Keep Absence Data From Misleading Leaders

A plant manager sees the DART rate fall for two quarters and removes a weekly review from the leadership agenda. The dashboard looks better, yet the maintenance backlog is growing, contractor hours are rising, and nobody has checked whether critical controls still work in the field. The number has improved. The decision may not have.

DART rate is a useful occupational safety measure when leaders read it as one piece of evidence. The trouble begins when the rate becomes a verdict on prevention. This explainer defines what DART counts, clarifies what it cannot show, and gives EHS leaders four questions for turning the number into a safer operational decision.

DART rate is the number of recordable cases involving days away from work, restricted work, or job transfer, multiplied by 200,000 and divided by hours worked. The rate helps compare certain injury outcomes across workforces, but it does not prove that exposure is controlled or that serious potential has declined.

What does DART rate mean?

DART stands for Days Away, Restricted Work, or Job Transfer. The United States Occupational Safety and Health Administration uses the measure within its recordkeeping system to identify cases that affected an employee beyond a basic medical-treatment record. The denominator of 200,000 hours represents the hours worked by 100 full-time employees over one year, which makes the result easier to compare across sites with different workforce sizes.

The calculation is straightforward, but its interpretation is not. A DART case describes an outcome that met a recordability and severity threshold. It does not describe how close another event came to causing permanent harm, whether a barrier failed during a high-energy task, or whether a team stopped work before exposure escalated.

Andreza Araujo's safety leadership work makes this distinction practical. In more than 250 companies and across 30+ countries, the relevant question has never been whether a dashboard can display a lower number. The question is whether the measure helps leaders see a decision that must be made before the next shift.

Why is DART rate not a complete prevention measure?

DART is retrospective. It records selected outcomes after a case has occurred, which means it cannot stand in for the condition of every control that should prevent a serious event. A site can report no DART cases while workers remain exposed to uncontrolled energy, suspended loads, vehicle interaction, or inadequate emergency readiness.

The measure also depends on classification, reporting, hours worked, and the type of work being performed. A contractor-heavy turnaround can change exposure sharply without producing a comparable shift in the rate. A smaller workforce can show a large numerical movement from one case. A strong result therefore needs context, not celebration.

James Reason's model of organizational accidents provides a useful lens because it separates visible outcomes from the conditions that allow them to develop. DART describes one visible outcome. Control verification, near-miss quality, maintenance status, and worker voice can reveal whether the layers behind that outcome are becoming stronger or thinner.

What should leaders ask before trusting a DART trend?

1. What exactly changed in the work?

Start with exposure, not the graph. Ask whether production volume, staffing, contractor hours, shutdown activity, overtime, materials, equipment, or task mix changed during the period. A lower rate after a quiet quarter does not carry the same meaning as a lower rate during a more demanding operating cycle.

The answer should be specific enough for an operations leader to verify. If the site cannot explain what work was performed, which high-energy tasks occurred, and how many people entered those exposures, the trend is not ready to guide a reduction in oversight.

2. What does the number leave out?

DART excludes many conditions that matter to fatality prevention. It does not measure a failed isolation that was caught before contact, a lifting barrier that held by chance, or a worker who stayed silent because the reporting route felt futile. It also cannot show whether a control was effective unless another source of evidence tests it.

This is why a dashboard should display exclusions beside results. Leaders should see the serious potential of near misses, overdue critical-control checks, open corrective actions, and unresolved worker concerns alongside DART. The purpose is not to make the dashboard crowded. It is to prevent a narrow number from becoming a broad claim.

3. Is the denominator still comparable?

Hours worked allow a rate to be normalized, but normalization does not remove operational context. Compare the current period with a period that had a similar workforce, task profile, and production rhythm. If the site moved from routine production to a major project, the same rate may represent a different exposure landscape.

Leaders should also examine whether the hours are complete and whether internal and contractor work are represented consistently. A clean calculation built on incomplete scope produces false confidence, even when the arithmetic is correct.

4. Which field evidence confirms the trend?

A DART improvement deserves confirmation from worksite evidence. Review critical-control verification, overdue maintenance, permit quality, supervision coverage, safety conversations, and the closure quality of corrective actions. The point is not to replace the rate with another favorite metric. The point is to connect the result to conditions that can still be changed.

In Safety Culture: From Theory to Practice, Andreza Araujo emphasizes the distance between documented compliance and lived practice. A DART trend becomes more credible when field evidence closes that distance by showing that people can identify, use, and challenge the controls that protect them.

How should DART rate appear on an executive dashboard?

Place DART where it can be read with its limits. Pair the current rate and trend with exposure context, critical-control verification, high-potential events, and the age of unresolved actions. Each measure should have a named owner and a decision threshold, because a dashboard that has no response rule is only a display.

The executive conversation should stay close to decisions. If the DART rate falls while control verification weakens, the dashboard should make that contradiction visible. If the rate rises after reporting improves, leaders should examine whether the organization is seeing more truth before treating the increase as failure.

When is DART rate useful?

DART rate is useful for monitoring selected injury outcomes, comparing similar operations over time, and asking whether a change in work has altered recorded severity. It can support resource decisions when leaders understand the data-generating process and keep the comparison fair.

It is not useful as a lone target, a ranking device, or proof that a site is safe. Those uses encourage attention to the number rather than to the conditions that produce it. Andreza Araujo's experience at PepsiCo South America, where the accident rate fell 50% in six months, reinforces the value of connecting performance measures with operational change rather than treating the metric as the change itself.

What is the practical definition to remember?

DART rate tells leaders about a defined group of recordable outcomes. It does not tell them whether the next high-consequence exposure is controlled. The practical standard is to read the rate, name its exclusions, check the denominator, and verify the field conditions that should make the result believable.

Frequently asked questions

A DART rate is most useful when it starts a disciplined conversation about exposure, control reliability, and decision ownership.