How to Build a Monthly Safety Decision Pack for the C-Suite in 30 Days

F2 practical guide for EHS directors, operations executives, plant managers, and safety data owners

A monthly safety decision pack is a short management document that connects safety evidence to decisions about exposure, controls, resources, and accountability. It is not a longer dashboard. Its purpose is to show what changed, what remains uncertain, who must decide, and how the organization will verify that the decision improved the work.

The usual failure appears when the pack becomes a monthly reporting ritual. The team copies last month’s charts, adds the latest injury rate, and explains every variance after the meeting has already moved on. The organization receives information, yet no leader is required to make a different decision.

Andreza Araujo’s work on safety culture emphasizes the distance between what an organization declares and what its routines make possible. That distinction matters at the executive level because a metric can look precise while the decision around it remains vague. A useful pack closes that gap by treating every important signal as a prompt for ownership and verification.

What you need before starting

Choose one executive sponsor who can change priorities, authorize resources, or pause work when evidence shows that a control is not reliable. Then appoint a data owner who can define the measures, document limitations, and challenge numbers that cannot be reproduced.

Collect the current definitions for injury rates, exposure measures, critical controls, overdue actions, serious near misses, and escalation records. Do not begin by choosing colors or chart types. Begin by asking which decisions the executive team needs to make before the next monthly review.

Use the existing safety dashboard blind-spot review to test whether the current reporting system confuses activity with control reliability. If the pack cannot answer that question, redesigning the layout will not solve the management problem.

Step 1: Define the decisions the pack must support

Write five to seven recurring executive decisions in plain language. Examples include whether a high-risk shutdown can proceed, whether a critical-control improvement needs capital, whether a contractor interface requires a different operating model, and whether an unresolved action should be escalated to the business president.

Each decision needs a threshold for attention, although the threshold should describe a management condition rather than pretend that one number can represent all risk. A failed verification of a critical control may require action even when the injury rate is stable.

Verify the step. Ask the sponsor to complete the sentence, “If this signal changes, I will decide…” Remove any measure that has no credible completion.

Common error. Starting with the data warehouse and discovering later that nobody knows what the meeting is supposed to decide.

Step 2: Separate harm measures from exposure measures

Place lagging measures such as recordable injuries, lost-time cases, and severity in one part of the pack. Place exposure measures in another part, including high-risk work volume, critical-control verification results, overdue safeguards, simultaneous operations, contractor interfaces, or known changes in staffing and process design.

TRIR and LTIFR remain useful when their definitions and denominators are stable, but neither rate can show every serious injury or fatality pathway. Andreza Araujo makes this warning central in *Far Beyond Zero*, the English gloss of her Portuguese title *Muito Além do Zero*. A target can become a distraction when it is treated as proof that the system is safe.

Verify the step. For each lagging measure, name the exposure evidence that can rise or fall without moving the rate.

Common error. Showing a green injury-rate tile beside no evidence about changing critical exposure.

Step 3: Choose a small set of decision-grade indicators

Select indicators because they change a decision, not because they are easy to export. A practical set may include critical-control verification quality, serious near-miss response time, overdue high-consequence actions, stop-work response quality, and the percentage of escalated concerns that received a documented decision.

Leading indicators are not automatically good indicators. The leading-indicator myths review explains why activity counts can create a false sense of prevention when the activity is detached from control performance.

Keep a data dictionary beside the pack. It should state the definition, inclusion rule, owner, time window, source system, and known limitation for every measure. This discipline protects the meeting from arguments about arithmetic when the real disagreement concerns meaning.

Verify the step. Ask two people who were not involved in building the measure to explain what action it should trigger.

Common error. Adding a new indicator because it looks sophisticated while removing none of the old measures.

Step 4: Add the context that changes interpretation

A number needs its operating context. Show the changes that can alter risk without appearing in a safety database, such as production ramp-up, maintenance backlog, overtime, new contractors, temporary equipment, leadership changes, or a compressed turnaround schedule.

Context does not mean filling the pack with narrative. Use a short panel with three questions. What changed in the work? Which control may be affected? What decision is needed before the next review? This format helps leaders distinguish a normal fluctuation from a change in the conditions that support safe work.

A monthly pack should also identify uncertainty. If a critical-control sample is incomplete, say so. A blank or qualified measure is more useful than a precise-looking percentage built from an unrepresentative sample.

Verify the step. Compare the pack with the operating plan and ask whether a major change in work would be visible to someone who did not attend the planning meeting.

Common error. Treating context as an explanation added after a poor result instead of as evidence that should shape the decision in advance.

Step 5: Create the decision register

Turn red signals into decisions rather than comments. For each item, record the issue, current exposure, interim protection, decision owner, required decision date, resources or authority needed, and the evidence that will confirm completion.

Operations should own changes to sequence, staffing, equipment, production priorities, and authorization. EHS can challenge the evidence and support the control review, but an advisory function should not become the permanent owner of a risk it cannot remove.

Use the 20-minute safety decision checkpoint as the operating rhythm between monthly reviews. The shorter checkpoint keeps urgent decisions from waiting for the next pack and prevents the monthly meeting from becoming an emergency queue.

Verify the step. Select one open item and confirm that the named owner has both authority and a realistic date.

Common error. Assigning an action to EHS because the concern was first detected by EHS.

Step 6: Design the pack around exceptions

Use a stable first page that answers four questions quickly. What changed? What is the most serious current exposure? Which decision is overdue? What evidence shows that previous decisions worked?

Put definitions, detailed trends, and supporting tables after the decision page. The executive meeting should not spend its first fifteen minutes searching for the point. A short pack is not automatically clear, so test it with a leader who did not build the document and observe where interpretation stalls.

Keep the visual language restrained. Red should indicate a defined management condition, not a team’s disappointment with a result. Green should mean that the stated control or decision was verified, not merely that a task was closed in a system.

Verify the step. Give the first page to the sponsor for three minutes and ask for the top two decisions without explaining the layout.

Common error. Designing the pack to reward reporting activity instead of helping leaders see exceptions.

Step 7: Run a dry review with the people closest to the work

Before the first executive meeting, review the draft with a supervisor, an operator or worker representative, an EHS specialist, and an operations manager. Ask whether the pack describes the work they recognize, whether any signal can be gamed, and whether the proposed response would change the task or only the paperwork.

This review is not a popularity test. Technical dissent is useful because it reveals where an executive measure loses meaning during translation. Andreza Araujo’s *Safety Culture: From Theory to Practice* supports the broader principle that culture is visible in routines, decisions, and relationships rather than in declared values alone.

Verify the step. Record at least three challenges and show how each one was accepted, rejected, or assigned for further evidence.

Common error. Asking only whether the page looks professional and never asking whether it changes the conditions for work.

Step 8: Establish the monthly review and verification loop

Set the same sequence every month. Review material changes, examine serious exposure, decide on unresolved items, confirm overdue commitments, and test whether completed actions changed the work. End by recording decisions in the same register that appears in the next pack.

Do not close an item because a procedure was revised or a training session was delivered. Close it when the changed control is visible under representative operating conditions and the people who use it can explain what is different. The verification method should fit the risk, which may require observation, document review, equipment testing, or a repeat assessment.

After three cycles, remove measures that never change a decision and strengthen measures that reveal meaningful uncertainty. The pack should become more useful through disciplined subtraction, not through constant expansion.

Verify the step. Compare three consecutive packs and check whether decisions became faster, ownership became clearer, or recurring exposure became less common.

Common error. Measuring success by punctual publication of the pack instead of by the quality and follow-through of the decisions it produces.

30-day implementation checklist

• Days 1 to 5, confirm the executive sponsor, data owner, decision list, and current definitions.
• Days 6 to 10, separate harm measures from exposure measures and document the data dictionary.
• Days 11 to 15, select decision-grade indicators and build the decision register.
• Days 16 to 20, add operating context, draft the first page, and identify data limitations.
• Days 21 to 25, run the dry review with frontline and operational participants.
• Days 26 to 30, hold the first executive review, record decisions, and schedule verification.

Do not wait for the first monthly meeting to discover that a critical control is weak. Use the draft pack on one current exposure, ask the decision owner to act, and test whether the evidence becomes clearer before the format is finalized.

A monthly safety decision pack earns its place when it makes risk harder to ignore and action easier to verify. The executive team does not need more safety decoration. It needs a reliable route from evidence to ownership, from ownership to changed work, and from changed work to proof.

For organizations building that route, Andreza Araujo combines engineering, creativity, and care in her safety-culture work. Explore Andreza Araujo’s resources on safety leadership and culture for a broader operating perspective.