On a a construction site, a QHSE manager opens their spreadsheet in January to consolidate the data from the past year. Three accidents with stoppage, but how many hours worked exactly? Do temporary workers count?
The frequency rate they will calculate depends entirely on the reliability of these two figures, the numerator and the denominator. Even before discussing the formula, it is the quality of the input data that determines whether the indicator reflects reality or distorts it.
Recommended read : Discover how a sitemap optimizes the crawling and SEO of your website
Hours worked: the denominator that skews everything
Most calculation errors do not come from the formula itself, but from the volume of hours in the denominator. It is often believed that it is enough to multiply the workforce by the annual contractual time. In practice, the actual hours worked differ significantly from the theoretical hours.
Sick leave, vacations, partial unemployment, and training must be deducted. Conversely, the actual overtime worked must be added. Forgetting either one artificially inflates or deflates the rate.
See also : How to Choose the Best Fast Online Loan: Guide and Comparison 2025
The case of temporary workers poses a recurring problem. Their hours are sometimes accounted for by the temp agency, sometimes by the user company. Practices vary from site to site, even within the same group. For the calculation of the work accident frequency rate to remain consistent from year to year, a unique convention must be established and documented.
- Systematically deduct absence hours (sick leave, vacations, training) from the gross contractual volume
- Include declared overtime, including those from weekends and on-call hours converted into interventions
- Define in writing whether temporary workers fall within the site’s scope or remain accounted for by the agency
- Check the consistency between the payroll software and the QHSE spreadsheet, which often diverge by several thousand hours
Frequency rate formula and distinction from frequency index
The official formula used in France relates the number of work accidents with stoppage, multiplied by one million, to the number of hours worked over the period. This ratio, expressed per million hours, allows for comparison between sites or companies of different sizes.
Do not confuse frequency rate with frequency index: the frequency index relates the number of accidents with stoppage per thousand full-time equivalent employees. A frequency index of 30 means 30 accidents with stoppage per thousand employees over the year. Both indicators measure accident rates, but their denominators differ radically.
The frequency index is simpler to obtain (it only requires knowing the workforce), but less precise. It does not take into account the actual volume of exposure to risk. Two companies with the same workforce but very different working hours (massive part-time versus frequent overtime) will have the same frequency index for opposing exposure realities.
When to use one rather than the other
The frequency rate is suitable for internal management, site by site, when reliable data on hours worked is available. The frequency index is more suited for sectoral comparisons published by health insurance funds, as it loses all statistical meaning below a few thousand employees.
For a monthly dashboard on a site with fewer than 200 people, it is better to track the frequency rate over twelve rolling months rather than calculate an index on too small a workforce.
Base 200,000 or base 1,000,000: the trap of international comparison
French companies use a base of one million hours. In the United States, OSHA standardizes its incidence rates on a base of 200,000 hours, corresponding to 100 employees working 40 hours a week for 50 weeks. Comparing a French rate to an OSHA rate without conversion is like comparing kilometers and miles.
In international groups, the TRIFR (Total Recordable Injury Frequency Rate) is also encountered, calculated on a base of one million hours but including all recordable accidents, including those without stoppage. The numerator’s scope changes, not just the denominator.
For coherent group reporting, three points must be aligned between subsidiaries:
- The normalization base (200,000 or 1,000,000 hours)
- The scope of accidents accounted for (only those with stoppage, or all recordable ones)
- The treatment of commuting accidents, included in some countries but excluded from standard French calculations
Severity rate of work accidents: complementing the analysis of accident rates
A low frequency rate does not mean much if the few accidents that occurred resulted in long stoppages. The severity rate relates the number of lost days due to temporary incapacity, multiplied by a thousand, to the number of hours worked. It measures the intensity of the consequences, not their occurrence.
Cross-referencing frequency rate and severity rate reveals very different risk profiles. A logistics site may show a high frequency rate (many small musculoskeletal accidents) with a moderate severity rate. An industrial site with few accidents but one serious case will see the opposite.
Integrating indicators into the single document
These statistics directly feed into the update of the single document for assessing professional risks. When the frequency rate increases in a position or workshop, it is a concrete signal to reassess the associated risk rating and trigger an action plan.
The rolling monthly follow-up, rather than annual, allows for detecting a drift before it turns into a trend. On a medium-sized site, one additional accident can dramatically change the frequency rate, making the smoothing over twelve months essential to avoid disproportionate reactions to a one-off spike.
The frequency rate remains a tool, not an end. Its value depends on the rigor with which hours are collected and accidents are qualified. A stable indicator from one year to the next, calculated with the same conventions, is worth more than a figure flattered by a silent method change.