Repeat Incidents: 6 Blind Spots That Keep RCA Shallow

Repeat incidents expose a failure that many RCA reports try to hide. The first investigation may have looked complete, the corrective actions may have closed on time, and the dashboard may have turned green, but the same exposure returned because the system learned how to finish the report rather than change the work.

This article is for EHS managers, plant managers, and investigation sponsors who see the same event family return under different names: hand injuries, dropped objects, vehicle-pedestrian conflicts, line opening leaks, LOTO deviations, falls from height, chemical splashes, or contractor shortcuts. The thesis is direct. A repeat incident is not only a second event. It is a management-system audit written in harm, because the previous investigation did not remove the conditions that made recurrence believable.

Why are repeat incidents more serious than fresh incidents?

A fresh incident can reveal a risk the organization had not understood well enough. A repeat incident reveals something harder to accept: the organization had a previous warning, processed it, and still allowed the exposure to return. That difference changes the leadership question from "what happened?" to "why did our learning system fail after it already had evidence?"

OSHA's incident investigation guidance warns that conclusions such as "worker was careless" or "employee did not follow procedure" stop short of root causes. That warning matters more in recurrence. If the second event receives the same explanation as the first, the report is no longer only shallow. It is recycling a failed theory of prevention.

Across 25+ years leading EHS in multinational environments, Andreza Araujo has identified a recurring pattern: companies often improve the investigation file before they improve the operating condition. The field then repeats what the document claimed had been solved.

Blind spot 1: treating recurrence as a new event

The first blind spot is opening a new investigation as if the event has no history. The team secures the scene, interviews witnesses, reviews photos, and writes a new timeline, although the most important evidence may sit in the previous report, previous near miss, previous audit finding, or previous action tracker.

Repeat incidents need family analysis. An event family groups incidents by the way harm could occur, not only by department, injury type, or immediate cause. A hand injury during cleaning, a near miss during jam clearing, and a maintenance cut during guard removal may belong to the same machine-access family even when the local labels differ.

The practical test is simple. Before the first cause meeting, ask whether the site has seen the same energy source, task, contractor group, shift condition, production pressure, equipment type, or supervisory decision before. If the answer is yes, the RCA should start with the previous learning package and ask why it did not travel into the work.

Blind spot 2: closing actions before testing the risk path

The second blind spot is treating a closed action as proof that recurrence risk fell. An action can close because a training record was signed, a procedure was revised, a toolbox talk was delivered, or a supervisor sent a photo. None of those artifacts proves that the risk path is now controlled.

For repeat incidents, closure should require an effectiveness test tied to the original risk path. If the event involved unexpected energy, the test should observe isolation, verification, re-energization control, and supervisor challenge in the field. If the event involved vehicle-pedestrian interaction, the test should observe traffic separation, blind corners, speed behavior, contractor compliance, and rule enforcement during a busy period.

The companion guide on testing corrective action effectiveness in 30 days gives the right discipline: verify changed work, not completed paperwork. A repeat incident after "all actions closed" is evidence that the closure standard was weaker than the hazard.

Blind spot 3: using a weaker lens after a low-severity repeat

The third blind spot appears when the repeat event has lower severity than the first one. Leaders may treat it as confirmation that the risk is manageable because "nothing serious happened this time." That is a dangerous reading. Lower severity can be luck, not control.

Frank Bird and Herbert Heinrich are useful here because their pyramid logic points EHS toward precursor events, although the exact ratios are not the point. The point is that repeated low-severity signals deserve attention when they share the same exposure path as a possible severe injury or fatality.

Andreza Araujo's Beyond Zero, glossed from Muito Alem do Zero, challenges the habit of using good lagging numbers as proof of safety. In recurrence analysis, that warning becomes practical. The question is not whether the repeat event was mild. The question is whether the same conditions could produce a worse outcome with a small change in timing, position, energy, or rescue response.

Blind spot 4: letting one site solve what the system created

The fourth blind spot is assigning recurrence to the local site when the condition is actually designed by the wider system. A plant may repeat contractor incidents because procurement rewards speed over competence. A warehouse may repeat pedestrian conflicts because the layout makes segregation impossible. A maintenance team may repeat LOTO deviations because shutdown planning compresses verification time.

James Reason's Swiss cheese model helps because it separates the visible active failure from latent conditions that leadership owns. In repeat incidents, those latent conditions are often not hidden. They were documented in the first investigation but converted into local reminders, retraining, or supervision notes because the real fix required money, design change, staffing, or production discipline.

This is where the sponsor matters. If the action owner has no authority over the condition that produces recurrence, the action tracker is pretending. Senior management should review repeat incidents for authority mismatch: who owns the exposure, who owns the action, and who controls the resource needed to remove the condition?

Blind spot 5: confusing a common cause with common wording

The fifth blind spot is semantic. Investigation databases often make recurrence hard to see because teams use different words for the same pattern. One report says "procedure not followed." Another says "inadequate risk perception." Another says "line of fire." Another says "poor housekeeping." The labels differ while the operational pattern stays the same.

EHS should build a recurrence map using event families and control failures rather than only cause codes. Compare task, energy, control, supervision routine, worker interface, contractor status, time pressure, and previous weak signals. This kind of map reveals recurrence that a dashboard may split across harmless-looking categories.

The related article on hindsight bias in incident investigation is useful because wording often carries judgment after the fact. If the organization keeps naming recurrence as individual failure, it will keep missing the repeated conditions that made the individual choice predictable.

Blind spot 6: protecting the previous investigation from scrutiny

The sixth blind spot is cultural. Teams do not like admitting that a previous RCA was incomplete, especially when managers approved it, auditors reviewed it, and corrective actions were celebrated as closed. A repeat incident forces the organization to audit its own learning, which can feel politically uncomfortable.

Andreza Araujo's The Illusion of Compliance, glossed from A Ilusao da Conformidade, is relevant because a system can look complete while remaining ineffective. The repeat event is the field's answer to the document. If the answer is recurrence, the organization should not defend the old report. It should re-open the learning path.

A mature review asks what the first investigation missed, which assumptions were wrong, which actions were too weak, which verification did not happen, and which leader accepted closure without proof. This is not an exercise in blame. It is a repair of the learning system that should have protected the next crew.

Comparison: ordinary RCA vs recurrence review

Repeat incidents need a different review frame because the organization is no longer starting from ignorance. It has a previous data point, and that previous data point must be tested as part of the evidence.

This comparison also changes meeting discipline. A recurrence review should not spend the first hour debating whether the worker made a mistake. It should ask why the previous learning failed to alter the conditions, controls, supervision, incentives, or design that shaped the new event.

What EHS should do in the first 72 hours after a repeat incident

In the first 72 hours, EHS should secure the current evidence and then pull every related record from the same event family. Include incident reports, near misses, observations, audit findings, maintenance records, permit deviations, stop-work records, and corrective-action effectiveness checks. The purpose is to see recurrence as a pattern before the new report narrows the story too early.

Second, compare the old corrective actions against current field reality. If the old action revised a procedure, test whether the task can be performed safely using that procedure under normal pressure. If the old action trained workers, verify whether supervisors can see and challenge the behavior. If the old action installed a control, inspect whether the control is present, used, maintained, and difficult to bypass.

Third, escalate authority mismatch. The article on barrier analysis before RCA helps here because recurrence often shows that the barrier was identified but never made reliable. If the fix requires engineering, staffing, procurement, layout, or production planning, the owner cannot be only the local supervisor.

FAQ

What is a repeat incident in safety? A repeat incident is a new event that shares a meaningful exposure path with a previous incident, near miss, audit finding, or corrective action. It may not have the same injury type or department, but the same risk condition has returned.

Does a repeat incident prove the first RCA was wrong? Not always, but it proves the first learning package was insufficient. The cause analysis may have identified part of the problem, although the actions, authority, verification, or leadership follow-through did not control the risk path.

How should EHS identify recurring incident patterns? EHS should map event families by task, energy, critical control, supervision routine, contractor status, time pressure, and previous weak signals. Cause-code dashboards alone often hide recurrence because teams use different words for the same operating pattern.

Why do corrective actions fail to prevent recurrence? Corrective actions fail when they repair memory instead of work. Retraining, procedure updates, and reminders are weak when the real drivers are design, workload, conflicting targets, missing authority, poor handover, or controls that nobody verifies.

When should repeat incidents be escalated to senior leaders? Repeat incidents should be escalated when the condition crosses sites, involves severe-risk exposure, repeats after actions closed, or requires resources the local manager does not control. Recurrence is often a leadership-system issue, not only a local investigation issue.

Recurrence is the audit the field writes.

Repeat incidents are uncomfortable because they expose the difference between documented learning and changed work. A company can finish the report, close the action, send the slide, and still leave the next crew facing the same weak condition.

The useful response is not a louder reminder. The useful response is a stricter learning system: family analysis, field verification, authority alignment, and leadership review of the previous investigation. For organizations that want RCA to change culture instead of decorate the archive, Andreza Araujo's safety culture books and ACS Global Ventures diagnostics give a practical starting point. Safety is about coming home.