Key Takeaways

• TRIR, DART, and EMR remain valuable safety metrics, but they provide only a historical view of performance.
• Risk changes across projects, locations, and work phases, making a single incident rate an incomplete measure of safety.
• Factors such as company size, medical management, and regional reporting differences can significantly influence lagging indicators.
• TRIR does not reliably predict future performance or serious injuries and fatalities (SIFs).
• Improving operational readiness requires a broader approach that complements lagging indicators with additional safety measures.

Introduction

For decades, organizations have used Total Recordable Injury Rate (TRIR), Days Away, Restriction, and Transfer (DART), and Experience Modification Rate (EMR) to report safety outcomes, compare performance, and assess potential partners. They create a common language for understanding what has happened and how performance compares with an established benchmark.

These lagging indicators remain valuable, but the problem begins when historical safety metrics are treated as complete measures of organizational safety. A low TRIR can look reassuring, but it does not necessarily reveal whether risks are being controlled, whether work conditions are changing, or whether people and systems are prepared for what comes next.

In complex work environments, measuring operational readiness requires more context than a single rate can provide. Here are six reasons traditional safety metrics can leave important parts of the risk picture unseen:

1. Risk changes as the work changes

TRIR cannot fully reflect how risk shifts across projects, locations, and phases of work.

Consider a design-build contractor completing a multiyear project. Early phases may involve predominantly office-based design work with limited exposure to physical hazards. Later, teams may move to an active project site for construction, systems integrations, testing, and commissioning.

The organization may remain under the same industry classification, but its actual risk profile changes significantly. A higher TRIR during later stages could appear to indicate declining safety performance when it may also reflect a major change in the type and intensity of work.

Comparing one rate across different periods can therefore oversimplify performance. Safety leaders also need to consider the work as performed: where it takes place, which tasks are underway, who is exposed, and how conditions are changing.

2. One incident can distort the picture for a small organization

Traditional incident rates are less stable when they are calculated from a small workforce or limited exposure hours.

TRIR is scaled to represent recordable incidents per 100 full-time workers. For a small organization, a single incident can cause a sharp increase in the rate, making its performance appear significantly worse than that of a larger company.

Combining several years of data can reduce the effect of an isolated event, but the organization may have changed during that time. Its workforce, leadership, processes, scope of work, and safety practices may no longer reflect the conditions captured in the earlier data.

EMR can present similar limitations. Very small employers may not have enough qualifying exposure or premium history for the metric to apply reliably — or at all. Without sufficient context, comparisons between organizations of different sizes can create more certainty than the data supports.

3. Medical management can influence the reported outcome

TRIR, DART, and EMR can be affected by how an injury is treated and managed — not only by how or why it occurred.

Whether an injury becomes OSHA recordable may depend on the care an employee receives. Onsite first aid, telemedicine, clinics, and hospitals can lead to different treatment pathways and, in turn, different recordability decisions.

Workers’ compensation requirements add further variation. Rules governing provider choice and an employer’s influence over treatment differ across jurisdictions, so similar injuries may affect safety metrics differently.

EMR is also linked to claims costs, which can direct attention toward managing medical expenses and claim outcomes. Claims management has a legitimate role, but it is not the same as preventing exposure or strengthening controls.

When post-incident decisions can influence reported performance, the resulting rate should not be treated as a direct measure of the conditions that caused the event.

4. Global injury tracking is not always comparable

TRIR becomes harder to interpret consistently when organizations operate across different legal, medical, and compensation systems.

TRIR is based on U.S. OSHA recordkeeping criteria. Global organizations may apply those criteria across their operations, but concepts such as medical treatment, compensability, and work-relatedness are not defined or administered consistently worldwide.

An event considered occupational in one country may be classified differently elsewhere. For example, countries such as New Zealand have compensation systems that operate under a broader no-fault model, making it more difficult to distinguish work-related incidents in the same way other jurisdictions do. As such, similar events may be recorded differently across regions.

Therefore, creating one global rate requires significant administrative judgement. The final number may look standardized even though the incidents behind it were classified under different legal and reporting frameworks.

5. TRIR has limited predictive ability

TRIR describes historical outcomes, but it should not be treated as a reliable forecast of future safety performance.

Research examining 17 years of data from 10 construction companies — covering more than 3.2 trillion work hours — found that recordable incidents occur randomly across space and time. The analysis indicated that at least 100 months of data is required to reach a meaningful level of predictive power. Few organizations have the scale, consistency, or volume of data needed to meet that threshold.

Short-term changes in TRIR are therefore weak evidence that an intervention has succeeded or failed. Statistical variation and other operational changes may also affect the rate, while an effective improvement may take years to produce a visible shift.

Therefore, a precise-looking number can create a false sense of certainty. TRIR helps explain past experience, but it cannot show whether today’s workforce, suppliers, and worksites are ready for tomorrow’s risks.

6. TRIR does not distinguish incident severity

A lower recordable incident rate does not necessarily mean an organization has reduced its exposure to serious injuries and fatalities.

TRIR counts the volume of recordable incidents without clearly distinguishing between a minor case and an event with the potential to cause life-altering harm. Research has also found no discernible association between recordable incidents and serious injuries or fatalities.

Individual health factors can further influence recovery time, restrictions, and case outcomes. The same event may affect two workers differently based on their age, health, and other personal circumstances, even when the underlying cause and potential severity are similar.

A strong focus on reducing recordables can therefore shift attention away from the hazards most likely to result in serious injuries or fatalities. Effective safety measurement should help leaders prevent these hazards — not simply reduce the number of recordable incidents.

From reporting safety to understanding readiness

The central question now is not whether organizations should stop using lagging indicators. TRIR, DART, and EMR continue to support reporting, benchmarking, insurance decisions, and historical analysis. The more important question is what leaders expect those metrics to tell them.

Lagging indicators show what has already happened, but they do not fully explain how risk is changing, whether serious hazards are controlled, or whether an organization is prepared for future work.

A more complete approach combines historical outcomes with information about the conditions, capabilities, and actions that shape performance. That is where leading indicators become an essential part of the discussion, complementing lagging indicators to create a more comprehensive picture of safety performance. Together, they help organizations move beyond reporting past outcomes to understanding whether their people, processes, and operations are truly ready to work.

Explore the limits of traditional safety measurement

Download Avetta’s Understanding the role and limits of lagging indicators to examine how TRIR, DART, and EMR came to define safety performance, where their limitations emerge, and why a broader approach to measurement is needed. Be on the lookout for part two of this white paper, where we look at leading indicators and how to leverage them to build a proactive safety strategy.