Research report · Critical controls
The injury rate tells leaders little about serious risk
Many operations have spent a decade driving recordable injury rates down. This report asks whether the controls that prevent a fatality have improved at the same pace, and how a leadership team would know.
Research report · PDF · Placeholder
Why this publication exists
Leaders need a measure of the controls that prevent serious harm
Before the Incident studies the time before a serious event: the precursors, the state of the controls that are meant to stop it, and the indicators leaders use to see both. It is written for EHS and safety leaders in heavy industry, and for the operations leaders and boards who share the risk.
From the current report
Three measures for one illustrative site
The measure a board usually sees can look settled while the measures that bear on serious harm move the other way.
Exhibit 3 Illustrative data The controls that prevent a fatality
In this illustrative site, a recordable rate of 0.8 sits beside 41 percent field verification
Three measures for one illustrative site, end of 2025
- 0.8 Recordable injury rate Per 200,000 hours worked, one site, 2025
- 41% Critical-control verifications done in the field Share of those planned, Q4 2025
- 14 Overdue energy-isolation verifications Count at the end of Q4 2025
SourceIllustrative data for design purposes only. Not a research finding and not to be quoted.
NoteValues are placeholders chosen to demonstrate the key-figure format. They describe no real site.
Appears inThe controls that prevent a fatality
Indicators and controls
A needle inside its band says the reading is normal. It does not say the control behind it was checked.
Latest briefings
-
5 min read
Briefing · Leading indicators
A green safety indicator should open a conversation
Lagging indicators tell a leader what has already happened. The useful question is what the same dashboard cannot see.
-
4 min read
Briefing · Critical controls
What a critical control verification should record
A tick in a box confirms that someone looked. A useful verification records what they saw, and whether the control would have worked.
-
4 min read
Briefing · Learning from events
Event reviews should describe conditions and controls
When a review ends at “human error,” the system that set up the error is left as it was.
From the briefing “A green safety indicator should open a conversation”
Exhibit 1 Illustrative data
In this illustrative review, most high-potential events had a precursor already on record
Where the earliest precursor was recorded, illustrative sample of high-potential events, count
SourceIllustrative data for design purposes only. Not a research finding and not to be quoted.
NoteValues are placeholders chosen to demonstrate the exhibit format. They describe no real events.
Appears inA green safety indicator should open a conversation
Data behind this exhibit
| Item | Value |
|---|---|
| Near-miss report | 11 |
| Maintenance backlog | 8 |
| Permit deviation | 6 |
| Audit finding | 4 |
| No prior record | 3 |
Cite this exhibit
Before the Incident, “In this illustrative review, most high-potential events had a precursor already on record,” Exhibit 1, in “A green safety indicator should open a conversation” (2026). Illustrative data; not a research finding and not to be quoted. Link
For the leadership team
Three questions for the next leadership review
- 1
Which of our critical controls were verified in the field this month, and what did the verification record?
- 2
Which high-potential events did we learn about because someone chose to report them?
- 3
Which indicator would change first if a critical control quietly degraded?
Controls that are holding
Most of the time nothing happens. The work is to know which controls are holding, before that changes.
Research
- Research report The controls that prevent a fatality 2026 · Placeholder
Research report · Critical controls
The controls that prevent a fatality
Many operations have spent a decade driving recordable injury rates down. This report asks whether the controls that prevent a fatality have improved at the same pace, and how a leadership team would know.
-
Research report · Verification
Verifying critical controls in the field
What a field verification should check, who should do it, and how to tell a verified control from a signed form.
-
Research report · Weak signals
Weak signals in permit-to-work
How the deviations, workarounds and late changes inside a permit system can point to the conditions for a serious event.
Subscribe
Get new research and briefings by email
- Reports on serious injuries, fatalities and the controls behind them, each with its PDF.
- Briefings you could take into Monday’s safety meeting.
- A notice whenever we correct a figure, so a number you have quoted stays right.