Standardizing HSE Reporting Across Countries: A Playbook

Why group safety numbers often don't add up

Ask five country HSE managers in the same group how many recordable injuries they had last quarter, and you may get five answers built on five different rules. One site counts a first-aid case, another does not. One country reports per 200,000 hours, another per million. One includes contractors, the next leaves them out. Hours worked are estimated in one place and taken from payroll in another.

Each local number may be correct by its own rules, but the group total is not comparable, and a falling injury rate can simply mean a site changed what it counts. For a multinational or industrial group that reports to a board, customers, insurers and sustainability frameworks, that is a real exposure. The fix is not more spreadsheets. It is a group reporting standard that every site applies the same way.

1. Separate legal reporting from group reporting

Every country has its own rules for notifying authorities about workplace accidents, and those rules do not change because your group wants consistent data. So design two layers:

Statutory layer: each site keeps meeting its own national notification and recordkeeping duties, owned by the local legal entity.
Group layer: one internal standard, applied everywhere, used for group KPIs, management review and external sustainability reporting.

A single record can serve both layers if it carries the extra fields each one needs. What you must avoid is letting a local legal definition silently become the group definition in one country but not in another. (This is general information, not legal advice. Confirm local duties with qualified advisers in each country.)

2. Write one group definitions dictionary

Most inconsistency starts with words. Agree a short dictionary, publish it in every working language, and point each definition to a recognized reference so it can be defended in an audit. You do not have to invent definitions; established ones exist:

Term	A recognized reference you can adopt
Occupational accident	ILO (16th ICLS, 1998): an unexpected and unplanned occurrence, including acts of violence, arising out of or in connection with work, resulting in personal injury, disease or death.
Recordable injury	US OSHA 29 CFR 1904.7: results in death, days away from work, restricted work or job transfer, medical treatment beyond first aid, or loss of consciousness, or is a significant injury or illness diagnosed by a licensed health care professional.
High-consequence injury	GRI 403: an injury that results in a fatality or from which the worker cannot, does not, or is not expected to recover fully to pre-injury health status within six months.
Serious injury or fatality (SIF)	ASTM E2920-26, a standard guide for recording priority occupational injuries and illnesses and SIF-related cases, written in part to allow consistent international comparison.

Pick one reference per term and write it down. Then add your own group rules for the grey areas that cause most arguments: what counts as first aid, when a restricted-work case starts, how to treat a case that becomes more serious later, and when an injury is work-related.

3. Fix the denominator: hours worked and the rate basis

An injury rate is only as good as the hours behind it. Two bases are common:

Basis	Formula	Used by
200,000 hours	cases × 200,000 ÷ hours worked	US BLS and OSHA (200,000 = 100 employees working 40 hours a week, 50 weeks a year)
1,000,000 hours	cases × 1,000,000 ÷ hours worked	ILO frequency rate

GRI 403 lets organizations calculate rates on either basis. The trap is mixing them. A site with 3 recordable injuries in 1,200,000 hours has a rate of 0.5 per 200,000 hours and 2.5 per million hours: the same performance, with numbers five times apart. Choose one basis for the group, convert everything into it, and print the basis next to every rate on every dashboard.

Hours need the same discipline. Agree whether hours come from payroll, time-and-attendance or an estimate, how overtime is treated, and how contractor hours are collected. ILO also expresses incidence per 100,000 workers, which uses headcount instead of hours; never compare a headcount-based rate with an hours-based one.

4. Decide who is in scope

Scope differences quietly distort group totals. Decide, and document:

Employees and contractors. GRI 403-9 asks for injury data both for employees and for workers who are not employees but whose work or workplace the organization controls. Record contractor cases in the same system, flagged as contractor, so you can report them together or separately. Our guide to safety data for ESG reports shows how the same records feed GRI 403 and the revised ESRS S1.
Commuting. ILO excludes commuting accidents from its core injury statistics and compiles them separately. Whatever you decide, apply it everywhere.
Visitors and the public. Track them, but keep them out of workforce injury rates.
New and divested sites. Set a rule for when an acquired site joins the group figures and how history is restated.

5. Use one event taxonomy and one severity scale

Beyond injuries, group leaders need comparable data on near misses, hazards, property damage and environmental events. Agree one list of event types, one set of injury categories, one root-cause list and one risk matrix. Keep the codes identical in every country and translate only the labels, so a report written in Bangla, Arabic or English rolls up to the same category.

Add a potential severity field next to actual severity. A dropped load that hurt nobody but could have killed someone deserves group attention; with only actual outcomes, it disappears from view. Our guide to SIF and potential-severity reporting shows how to classify these events consistently.

6. Standardize the workflow and the timeline

Definitions are not enough if sites report at different speeds and investigate at different depths. Set group rules for:

How soon an event must be entered after it happens, and who may enter it (ideally anyone, from a phone).
Which investigation level each severity triggers, and who must review and approve it.
How corrective actions are assigned, with an owner and due date, and how their effectiveness is checked before closure.

This also maps to ISO 45001:2018, where clause 10.2 covers incidents, nonconformities and corrective action, and clause 9.1 covers monitoring and measuring performance. A single group workflow makes it easier to show certification auditors the same process at every certified site.

7. Add leading indicators, defined just as tightly

Lagging rates move slowly and say little about tomorrow. (Our guide to leading vs lagging safety KPIs goes deeper.) Many groups add leading indicators such as near-miss reports, hazard reports, inspections completed against plan, overdue corrective actions and training hours. These need the same rigor: one definition, one counting rule and one normalization (per site, per hours worked, or per headcount). A rise in near-miss reports after a rollout often reflects better reporting rather than a worse site, so explain that to leadership before the first dashboard goes live.

8. Govern the data like financial data

Treat safety numbers the way finance treats revenue:

Owners: one accountable data owner per country or business unit.
Monthly close: a fixed date after which the month is locked, with a documented restatement rule for late reclassifications.
Quality checks: missing hours, open cases with no classification, contractor cases without a company, rates outside expected ranges.
Audit trail: every change to a classification keeps who, when and why, so a figure in the board pack can be traced back to the source record.

Checklist for group EHS leaders

#	Decision	Done?
1	Statutory and group reporting layers separated	☐
2	Definitions dictionary published, each term tied to one reference	☐
3	One rate basis (200,000 or 1,000,000 hours) shown on every report	☐
4	Hours-worked source agreed, including contractor hours	☐
5	Scope rules for contractors, commuting, visitors and acquisitions	☐
6	One event taxonomy, root-cause list and risk matrix; only labels translated	☐
7	Actual and potential severity on every event	☐
8	Reporting timeline, investigation levels and CAPA rules agreed	☐
9	Leading indicators defined and normalized	☐
10	Data owners, monthly close and audit trail in place	☐

Where software helps

You can write this standard in a policy document, but it only holds if the system people use enforces it. In AccelSafety, our enterprise HSE software, every site sits in one hierarchy from group to region, country, cluster and site; every country uses the same incident, hazard, near-miss, investigation and CAPA workflows; people report from a phone with photos; and group KPIs such as lost days, near misses, hazard reports, recordable injuries and training manhours drill back to the source records with their history. Access is scoped to each user's part of the hierarchy, and approvals and configuration stay with the roles you choose.

Planning a group-wide rollout? Read our six-step plan for moving from paper to digital HSE, or book a free demo set up with a structure like yours.

How to standardize HSE reporting across countries: a playbook for group EHS leaders