Safety Decision Log: Build One in 14 Days

A maintenance team discovers that a temporary guard will remain in service for another week. The risk is known, the work is still scheduled, and everyone assumes that someone senior has accepted the exposure. That assumption is where control begins to weaken.

A safety decision log prevents unresolved risk from hiding inside email, meeting minutes, or a risk register that nobody reviews during live work. It records the decision that still has to be made, the evidence behind it, the person who owns the next move, and the authority that can accept or reject the remaining exposure.

This guide is for supervisors, EHS managers, maintenance leaders, and operations managers who need a practical record for decisions that sit between routine work and formal escalation. The method uses a 14-day setup period, although the review cadence should become part of normal management practice once the log is active.

What you need before starting

Start with one operational area where unresolved decisions already create delay, repeated discussion, or informal workarounds. A shutdown, maintenance backlog, temporary deviation list, or contractor mobilization can provide enough real material without turning the first version into a company-wide database.

Prepare four inputs. Bring the current risk register, the active work plan, the site’s escalation or risk-acceptance rules, and a leader who can remove barriers when the first entries reveal an authority gap. ISO 45001:2018 expects organizations to control operational risks and evaluate whether processes achieve their intended results, which means the log should lead to decisions and verification rather than become another storage location.

Keep the first version simple. A shared table with controlled fields is more useful than a sophisticated workflow that the night shift cannot access. Andreza Araujo’s *Make The Difference: Be a Leader in Health & Safety* treats leadership as an operating discipline, and that principle matters here because the record only works when leaders use it during real decisions.

Step 1: Define what belongs in the log

Write an inclusion rule before collecting entries. A decision belongs in the log when a known exposure requires a choice about stopping work, changing the design, adding resources, extending a temporary control, accepting residual risk, or escalating authority.

Exclude routine actions that already have a clear owner and approved method. If every minor task enters the log, urgent decisions disappear among administrative work. A useful test asks whether the work could continue safely without a management choice. If the answer is no, create an entry.

Use examples from the area you selected. An unavailable lifting device, an incomplete isolation boundary, a contractor scope that changed after mobilization, or a ventilation repair awaiting capital approval can all qualify when the decision affects exposure rather than convenience.

Step 2: Capture the exposure without solving it too early

Record what is known before writing a conclusion. Describe the task, location, people exposed, current control, duration, and credible consequence in plain language. State whether the information came from observation, a permit, an inspection, an incident review, or a technical assessment.

Separate facts from interpretations. James Reason’s work on latent failures shows why an immediate action can be only one part of a wider control problem. If the log says that an operator is careless when the evidence only shows a missing interlock, the wording has already narrowed the decision before the review begins.

Add a confidence field with values such as confirmed, partially confirmed, or assumed. The purpose is not to create mathematical precision. It is to show which uncertainty could change the decision and therefore deserves verification before the entry is closed.

Step 3: Separate the decision from the action

Write the unresolved question as a decision, then list the actions that support it. “Install a guard” is an action. “Can this equipment operate under the temporary barrier until the permanent guard is installed?” is a decision that requires evidence and authority.

Each entry should identify the options available, including the option to stop or defer the work. Avoid presenting one preferred action as if it were the only possible response, because that hides the tradeoff that leadership is being asked to accept.

Link the entry to the relevant risk-register item, permit, change record, or incident review. Risk Register Explained: 4 Fields That Keep It Live explains how the broader register stays useful when its controls remain connected to ownership and evidence.

Step 4: Assign one decision owner and one acceptance authority

Name one decision owner who must move the item. Then name the acceptance authority who can approve residual risk when the control cannot be restored immediately. These may be the same person, but the log should not assume that they are.

Do not assign ownership to a department alone. “Engineering” cannot attend a review, make a call, or explain why a deadline moved. Use a role and a named person, with a deputy when the exposure can remain active across shifts.

Set an escalation route beside the owner. If a supervisor can pause the task but cannot authorize a design change, the next authority should be visible before production pressure arrives. How to Set Safety Risk Acceptance Authority provides a related decision framework for this boundary.

Step 5: Set a deadline that matches the exposure

Choose a decision deadline based on exposure frequency, potential consequence, and the reliability of the temporary control. The deadline is not the date when the permanent project is expected to finish. It is the date by which the organization must decide whether the current arrangement may continue.

Use three dates when necessary. Record when the issue was identified, when the decision is due, and when effectiveness will be checked after the decision. This prevents a signed approval from being mistaken for a working control.

Mark overdue entries visibly and require a reason for any extension. An extension can be legitimate when new evidence changes the plan, yet an unexplained extension is a signal that the organization is carrying risk without an active decision.

Step 6: Review the log in the meeting where work changes

Put the log into an existing operational meeting rather than creating a separate ceremony that people can skip. Review new entries, decisions due before the next meeting, overdue items, and entries whose temporary controls are failing in the field.

Ask four questions for every priority item. What changed since the last review? What evidence is still missing? Who can decide now? What will workers do differently before the next review? The questions keep the conversation tied to control rather than allowing it to become a status report.

Use the same log across shifts when the exposure crosses a handover boundary, but do not treat access as proof of communication. The incoming supervisor should be able to explain the open decision, the temporary control, and the stop condition without searching through separate messages.

Step 7: Verify the decision in normal work

Close the decision only after checking that the chosen control survives ordinary production conditions. Observe the task when the crew is under the same time pressure, staffing pattern, and equipment constraints that created the original problem.

Collect evidence that matches the decision. A changed drawing can prove design approval, while a field observation is needed to show installation and use. A completed briefing can show communication, but it cannot prove that the barrier remains available during the task.

Record the verification result, the person who performed it, and any new exposure. If the result is not acceptable, reopen the entry or create a linked decision. Closure should describe what is different in the work, not only what document has been signed.

Step 8: Audit the log after 14 days

At the end of the setup period, review the log as a management-control sample. Count open, overdue, escalated, and verified entries, then read several records from different supervisors or work areas. The purpose is to test quality, not to reward a low number of open items.

Look for recurring weaknesses. If entries lack acceptance authority, the site has an accountability problem. If deadlines move repeatedly, the temporary-control process may be under-resourced. If decisions close without field verification, the site is measuring paperwork completion rather than control performance.

Use the findings to refine the inclusion rule, authority map, and meeting cadence. Decision Rights Matrix Explained: 4 Levels That Keep Safety Escalation Moving can help when the audit shows that people understand the risk but cannot identify who has the right to decide.

Final checklist for the first live review

• Confirm that every entry describes a decision, not only a hazard or action.
• Check that facts, assumptions, options, and evidence are visibly separated.
• Verify that one person owns the decision and one authority can accept residual risk.
• Review the deadline, escalation route, temporary control, and stop condition.
• Require field verification before closing the entry.

A safety decision log earns its place when it changes who acts, when they act, and what evidence proves that the exposure is controlled. It should make unresolved risk harder to ignore without pretending that a spreadsheet can replace technical judgment or leadership accountability.

For more practical guidance on risk management and safety leadership, explore the risk management articles and Andreza Araujo’s resources at andrezaaraujo.com.