The problem

Corrective actions are where investigations, audits and risk assessments turn into change. They are also where many processes quietly lose their value.

A typical action list shows an owner, a due date and a status. When the owner marks the action as done, it is closed. Sometimes a second person verifies it. At that point, the action disappears from view, and everyone assumes the problem has been dealt with.

But "done" and "verified" do not tell you whether the action worked.

Why it matters

Actions exist to change outcomes: to remove a hazard, prevent a recurrence or strengthen a barrier. An action can be fully completed and properly verified and still fail to achieve that.

Consider a new procedure written after an incident. It can be completed (the procedure is written), verified (someone confirms it was issued and briefed) and still be ineffective, because the task is still done the old way in practice. Or a guard is fitted and inspected, but operators remove it because it makes the job harder.

When effectiveness is not reviewed separately, organisations report high closure rates while the underlying risk stays the same. Repeat incidents then appear to come from nowhere.

A practical approach

It helps to treat an action as passing through three distinct checks.

Closure. The action owner reports that the work has been done. This is a claim by the person responsible.

Verification. Someone other than the owner confirms that the action was carried out as described: the guard is fitted, the procedure is issued, the training took place. Verification answers the question "was it done?"

Effectiveness review. After enough time has passed for the action to have an effect, someone reviews whether it achieved its purpose. Effectiveness answers the question "did it work?"

Effectiveness reviews need evidence of outcome, not activity. Useful evidence might include:

  • Observations showing the new method is actually used.

  • Inspection results showing the control remains in place.

  • A reduction in related observations, near misses or defects.

  • Feedback from the people doing the work.

  • Follow-up checks at the location where the problem occurred.

Timing matters. An effectiveness review carried out on the day an action is closed usually adds little. Planning when and how effectiveness will be reviewed, at the moment the action is created, makes it far more likely to happen.

An effectiveness review can also conclude that an action did not work. That is a useful result. It should lead to a new action, not to quietly closing the old one.

What a good system should provide

A good action management system should:

  • Keep closure, verification and effectiveness as separate steps, each with its own record.

  • Record who verified the action and who reviewed its effectiveness.

  • Allow more than one effectiveness review over time.

  • Keep effectiveness reviews as history that is added to, not overwritten.

  • Link each action back to its source, such as an investigation finding, a risk assessment control or an audit result.

How Sospita addresses it

In the Sospita HSE Platform, actions are shared across modules. An action raised from an incident investigation or a risk assessment is a platform Action, with verification kept separate from completion.

Effectiveness is reviewed separately again. Actions support effectiveness reviews that are append-only, so each review is added to the action's history rather than replacing the previous one. In the Incident Investigation & RCA module, this makes "verified" and "effective" distinct states, and actions remain linked to the causes and barriers that produced them. How those causes are established in the first place is covered in Facts Before Causes: A Better Sequence for Incident Investigation.

The HSE Platform and the Incident Investigation & RCA module are currently coming soon.

Limitations and human responsibility

A system can make effectiveness reviews possible and visible. It cannot decide whether an action was effective.

Judging effectiveness requires people who understand the work, look at the right evidence and are willing to conclude that an action did not achieve its purpose. A review recorded without real evidence is no better than no review at all.

Some actions take a long time to show results, and some outcomes are hard to measure. In those cases, the review should state the uncertainty honestly rather than mark the action as effective by default.

Responsibility for deciding whether risk has been adequately reduced remains with competent people and the organisation.