RISKOPILOT Editorial Team, 16 September 2026, 4 min read
Why injury rates do not predict fatalities
TRIR and LTIFR measure how often people get hurt. Fatal risk follows a different mechanism, and boards need a different instrument.
The number boards trust most
Most boards follow safety through a total recordable injury rate (TRIR) or a lost-time injury frequency rate (LTIFR). Both are easy to compute and easy to compare, and both are usually shown as a single figure trending down.
That comfort is misplaced. The Construction Safety Research Alliance ran the largest statistical test of the recordable injury rate on 3.2 trillion worker-hours of member data. It concluded that there is no discernible association between the rate and fatalities, that recordable injuries occur almost entirely at random (96 to 98% of the variation), that the rate is not precise enough to report to several decimal places, and that in nearly every practical circumstance it is statistically invalid for comparing companies, business units, projects or teams.
An industry-scale warning
Mining shows the same pattern. Between 2017 and 2025, ICMM member companies reduced their recordable injury rate from 3.94 to 2.21 per million hours worked. Over the same years, fatalities stayed between 33 and 50 a year, with no steady decline.
ICMM draws the conclusion itself. In 2025, five of its fourteen members with no fatalities had injury rates above the membership average, and it warns that injury performance should not be used as a direct proxy for fatality risk. It also found ineffective control execution in 19 of the 26 fatal incidents it classified that year.
Two mechanisms, one workplace
Most recordable injuries come from low-energy events: a slip, a strain, a cut. Fatalities come from high energy, such as gravity, moving machinery, pressure or electricity, reaching a person because a barrier was missing or failed at the moment of demand. Reducing the first does not reduce the second, and a falling injury rate can coexist with growing exposure to fatal energy.
What to measure instead
The question that matters is whether the controls against fatal energy are present and working today. Critical control health (CCHI) answers it at three levels: a field gate before every high-energy task, indicators that show whether controls will still hold next month, and a board view in which the weakest indicator sets the readiness stage. The CCHI lab shows the method at work, including the moment an injury rate improves while fatal risk grows.
Sources
- Construction Safety Research Alliance. The tyranny of TRIR
- Hallowell, M. et al. (2021). The statistical invalidity of TRIR as a measure of safety performance. Professional Safety 66(4), 28 to 34
- ICMM (2026). Safety performance: benchmarking progress of ICMM company members in 2025