After the event, establish what the controls actually did

An event puts pressure on everyone to explain it quickly. The first explanation is often a person’s last visible action. That may be part of the sequence, but it cannot tell us by itself why the work made sense at the time or whether the controls were capable of protecting anyone.

Build an account that can be checked

Begin with the available evidence: the scene, equipment, records, procedures, and accounts from people who were there. Record what each source can establish. Keep observations separate from interpretations, and mark gaps that still need to be resolved. A clear timeline helps, but a timeline is a starting point for questions, not a complete explanation.

Ask what people were trying to accomplish, what information they had, what changed during the task, and what tradeoffs the work required. Compare the planned process with the conditions people actually faced. A procedure may describe a safe path that was unavailable in practice; an informal adaptation may have helped work succeed many times before the event.

Test the controls

For each significant hazard, identify the control that was expected to prevent exposure or limit harm. Then ask whether it existed, was available, worked as intended, and remained effective under the conditions of the event. Do not treat a written control as proof that it functioned in the field.

• What was the control supposed to do?
• What evidence shows how it performed?
• What conditions weakened or bypassed it?
• Who will verify that a proposed change works during real work?

The response should fit the evidence. A reminder or retraining may have a place, but it cannot repair a poorly designed tool, missing capacity, or a control that fails under normal variation.

Know when to go deeper

An investigation develops a supported account of a particular event and tests the relevant controls. If that work exposes a broader question about how the task is routinely done, a Learning Team can explore it with the people who know the work. The two methods can inform each other while keeping their purposes clear.

Follow-through matters as much as the initial explanation. Assign owners, observe the change in use, invite feedback, and revisit the decision if the control still depends on people overcoming the same friction.