The frequency rate of workplace accidents is one of the most scrutinized indicators in annual safety reports. Its formula seems simple, but reliable calculation requires precise methodological choices regarding the scope of the accidents considered and the counting of hours worked. These choices, rarely detailed in dashboards, condition any sectoral or historical comparison.
Base 1,000,000 or base 200,000: two conventions that skew comparisons
The French formula is based on a million hours: (number of accidents with stoppage × 1,000,000) / hours worked. This multiplier, set by Social Security, produces a figure that is directly readable for workforces of several hundred employees.
Related reading : Everything You Need to Know About Home Renovation and Construction in Cher
Anglo-Saxon references (OSHA, TRIR, DART) use a base of 200,000 hours, which roughly corresponds to the annual work volume of 100 full-time employees. Therefore, a French frequency rate of 25 cannot be compared to a TRIR of 5 without explicit conversion.
Understanding the calculation formula for the workplace accident frequency rate starts with this methodological base distinction, as it explains why the figures published by French subsidiaries and their Anglo-Saxon parent companies sometimes seem contradictory.
Related reading : Practical Guide: Everything You Need to Know About H&M's Return Policy in 2024
To convert a French TF to an OSHA equivalent, simply divide by 5. The reverse operation (multiplying by 5) brings a TRIR back to the French scale. Without this step, any international comparison is misleading.
Hours worked: the denominator that changes everything
The frequency rate formula only works correctly if the denominator reflects the actual hours worked. Contractual hours are not suitable, as they include paid leave, public holidays, sick leave, and ongoing stoppages.
In practice, reliable calculation requires subtracting from the theoretical volume:
- Paid leave hours and compensatory rest days (RTT, recovery)
- Hours of absence due to illness, maternity, or ongoing workplace accidents
- Hours of external training if employees are not exposed to the usual risks of the position
This rigor regarding the denominator has a direct impact on the result. A company that uses gross contractual hours mechanically underestimates its frequency rate, as it artificially inflates the number of exposure hours. A discrepancy of several points can result from the choice of method for counting hours.
The best practice is to set the same reference period for the numerator (accidents) and the denominator (hours), and then document the chosen method to ensure reproducibility from year to year.
Commute accidents and accidents without stoppage: what the frequency rate does not count
The French frequency rate only considers workplace accidents that resulted in at least one day of stoppage. Minor accidents treated on-site, incidents without medical follow-up, and first aid without stoppage do not enter the numerator.
This minimum severity threshold enhances comparability over time and between sectors, but it creates a blind spot. A company that records numerous accidents without stoppage can display a reassuring frequency rate while presenting a high level of risk.
Commute accidents: a convention to establish
Commute accidents (home-work) are treated variably by companies. Social Security distinguishes them from actual workplace accidents. Some organizations include them in their overall frequency rate, while others systematically exclude them.
The problem arises when the convention changes from one year to the next, or when two sites of the same group apply different rules. Maintaining the same convention over time is the only way to preserve the consistency of a claims history.
Frequency rate and severity rate: why to cross-reference them systematically
An isolated frequency rate says nothing about the severity of accidents. A company can show a low TF while recording serious accidents with long stoppages. The severity rate, calculated by relating the number of stoppage days to hours worked (multiplied by 1,000), provides this complementary dimension.
The cross-referencing of the two indicators allows for distinguishing several profiles of claims:
- High TF and low severity rate: frequent but not severe accidents, often related to repetitive motions or slips and falls
- Low TF and high severity rate: rare but serious accidents, typical of sectors exposed to mechanical or chemical risks
- High TF and high severity rate: critical situation requiring a comprehensive review of the prevention system
Analyzing the trend over rolling twelve months rather than annual cumulative allows for earlier detection of deterioration. An isolated point in time has no statistical value, especially for small structures where a single accident can dramatically change the rate.
Statistical limits of the frequency rate for small workforces
The frequency index (number of accidents per thousand employees) is sometimes preferred to the frequency rate in small companies, as it does not require precise counting of hours worked. However, it loses nuance for structures that rely heavily on part-time or temporary work.
For workforces of fewer than a few hundred employees, the frequency rate loses its statistical significance. A single accident can cause it to jump from zero to an alarming level from one year to the next, without any change in the actual risk. Field feedback varies on this point: some safety professionals maintain monthly tracking for team awareness, while others favor proactive indicators (number of reported hazardous situations, rate of completion of corrective actions).
The frequency rate remains a management tool, not an end in itself. Its reliability entirely depends on the rigor applied to the numerator, the denominator, and the consistency of the conventions used. Any comparative analysis, whether between sites, years, or countries, assumes that the same calculation rules have been applied on both sides.