TRIR / LTIFR / DART Calculator

What the calculation does

The code takes four entries: hours worked, recordable cases, days-away cases and restricted-only cases. Each rate is a case count multiplied by 200,000 and divided by the hours. TRIR uses all recordable cases, LTIFR uses the days-away cases, and DART uses days-away plus restricted-only cases. A result appears only when hours are at least 1, the three counts are whole numbers, and days-away plus restricted-only cases do not exceed the recordable total, so an inconsistent entry is refused instead of corrected.

The rates are displayed to two decimals, but the benchmark comparison uses the unrounded values: the benchmark is subtracted, the difference is divided by the benchmark, and the percentage is shown to the nearest whole number. The benchmarks are fixed values for US private industry in 2024. There is no LTIFR benchmark, no non-US benchmark, no weighting for workforce size, no smoothing across years and no test of statistical significance.

Worked example

These figures are invented to show the arithmetic. A manufacturing site worked 520,000 hours in a year and logged 6 recordable cases: 2 with days away, 1 with restricted work only and 3 treated cases with neither outcome.

• TRIR: 6 × 200,000 ÷ 520,000 = 2.31
• LTIFR: 2 × 200,000 ÷ 520,000 = 0.77
• DART: (2 + 1) × 200,000 ÷ 520,000 = 1.15

With the manufacturing preset selected (2.7 for TRIR, 1.7 for DART), the unrounded TRIR is about 14.5% below the preset and is shown as -15%, and DART is shown as -32%. These are the values the calculator's own function returns for these entries. The comparison says only that this year's rates are numerically lower than the published ones.

Why the result can mislead

At 520,000 hours, one recordable case is worth 200,000 ÷ 520,000, about 0.38 of a rate point. A seventh case in the example would lift TRIR from 2.31 to 2.69, and an eighth to 3.08, while the same six cases over 10% more hours (572,000) would give 2.10. A gap to the benchmark of one or two cases is therefore within what a single event can produce.

The hours figure matters as much as the case count. Hours should come from payroll or time records and exclude paid time that was not worked, such as holidays and sick leave; an estimate of headcount times 2,000 hours can be off in either direction. The case count depends on classification: a site that records fewer minor cases will show a lower TRIR without being safer, and definitions of a lost-time injury differ between countries and companies. The benchmark covers a whole US industry for one calendar year, which may differ from your size, work mix and period.

Where the formula and definitions come from

• The US Bureau of Labor Statistics gives the formula (injuries and illnesses × 200,000 ÷ employee hours worked), explains that 200,000 stands for 100 employees working 40 hours a week for 50 weeks, and tells employers to leave paid non-work time out of hours. It also gives a DART variant using days-away and restricted or transferred cases: BLS: how to compute your firm's incidence rate.
• OSHA's general recording criteria list the outcomes that make a work-related case recordable: death, days away from work, restricted work or transfer, medical treatment beyond first aid, loss of consciousness and a significant injury or illness diagnosed by a health professional: 29 CFR 1904.7.
• The benchmark values come from the BLS table of 2024 industry rates, which reports total recordable cases, DART cases and cases with days away per 100 full-time workers: SOII Table 1.

The three pages were opened on 3 October 2026. They describe the US method; they do not validate this calculator or any particular result.