In most safety management systems, the riskiest moment is not the hazard report. It is the signature that follows: the point where someone formally accepts a risk, approves a change to an operating procedure, or closes an investigation with a stated root cause. Those decisions shape operations for months or years, and they are surprisingly hard to reopen once filed. This article diagnoses why single-reviewer risk decisions go wrong and sets out a practical remedy: an independent second review of risk acceptance before the decision becomes final.
The problem: one signature, long consequences
Consider three common scenarios in a flight training or small commercial operation:
- A hazard assessment concludes that operating a particular aircraft type from a short, sloping strip carries an acceptable risk with existing mitigations, and the accountable manager signs it off.
- A change management review approves moving a maintenance task from a fixed interval to a condition-based approach.
- An occurrence investigation closes with “pilot handling” as the primary cause, and the corrective action is additional training for one individual.
Each decision might be entirely correct. But each has the same structure: one person or a small, tightly connected group assesses the evidence, reaches a conclusion and records it. After that, the decision becomes part of the system’s baseline. Future reviews tend to start from it rather than question it. Reopening a closed investigation or withdrawing an accepted risk assessment carries organisational cost and, frankly, some embarrassment. So it rarely happens unless something goes wrong.
Diagnosis: why single-reviewer decisions drift
When safety decisions turn out poorly, the cause is seldom incompetence. It is usually one of a small set of predictable pressures acting on a capable reviewer.
Anchoring on the initial framing
The person who receives a hazard report often frames the problem in their first reading, and that framing persists. If an occurrence is first described as a handling error, the investigation tends to look for evidence of handling error. Alternative explanations, such as a confusing checklist, an unclear procedure or a scheduling pressure, receive less attention.
Proximity to the operation
Safety managers in smaller organisations usually know the people, the aircraft and the routes personally. That knowledge is valuable, but it also creates blind spots. A procedure that everyone has “always done this way” can be hard to see as a hazard from inside.
Operational and commercial pressure
Risk acceptance decisions often arrive with a clock attached: a new route is due to launch, an aircraft needs to return to service, a contract depends on the change. Even a principled reviewer feels that pressure. The decision is not corrupted, but it is rushed.
Stake in the outcome
Sometimes the reviewer helped design the procedure under review, or will be responsible for implementing the corrective action. That stake is natural in a small organisation, but it makes genuinely neutral assessment harder.
The closure effect
Once a decision is recorded and the file closed, the system treats it as settled. Subsequent audits check that the paperwork exists, not whether the conclusion was sound. A weak decision can therefore persist unnoticed for a long time.
The solution: an independent second review before closure
The remedy is not more committees or longer forms. It is a narrowly scoped second look, by a person who gains nothing from either outcome, applied to the small number of decisions that are both high-consequence and hard to reverse. The design matters, so it is worth taking piece by piece.
1. Define which decisions qualify
Not every hazard report needs a second reviewer. That would overload the system and dilute attention. Set clear triggers, for example:
- Any formal acceptance of a risk assessed in the upper bands of your risk matrix.
- Any change to a normal or emergency procedure affecting flight operations.
- Any investigation closure where the primary finding attributes cause to an individual.
- Any decision that removes or relaxes an existing mitigation.
Writing these triggers into the SMS manual turns the second review from a favour into a routine step, which removes the awkwardness of asking.
2. Choose a reviewer who is genuinely independent
Independence means the reviewer did not participate in the original assessment, will not implement its outcome, and has no reporting line that makes disagreement costly. In larger organisations this may be a safety specialist from another department. In smaller ones, it is often an external party: a safety manager from a partner organisation, a retired investigator, or a peer from an industry safety network who agrees to reciprocal reviews.
3. Give the reviewer the evidence, not the conclusion
The most effective second reviews start blind. The reviewer receives the hazard report or occurrence data, the relevant procedures, any interviews or data downloads, and the operational context. They form a provisional view before reading the original assessment. Only then do they compare.
This sequencing directly counters anchoring. It also produces a more useful output: not “I agree” or “I disagree,” but a clear statement of where the two analyses converge and where they diverge.
4. Ask a narrow question
The reviewer’s brief should be specific to the one decision: “Is this risk acceptance justified by the evidence and mitigations described?” or “Does the identified root cause account for the evidence, and are there contributing factors the corrective action does not address?” A tight question keeps the second review proportionate and stops it turning into a general audit of the safety programme.
5. Structure the response
Ask the reviewer to return a short written note covering:
- Points of agreement with the original assessment.
- Points of disagreement, each with the specific evidence behind it.
- Any additional information that would change their view.
- Whether, in their judgement, the decision should proceed, proceed with amendment, or be reconsidered.
6. Resolve differences openly
When the two reviews differ, bring the original assessor and the second reviewer together with the evidence. The goal is not to decide who was right but to understand why their conclusions diverged. Frequently the gap reveals a missing piece of information, an assumption one party made and the other did not, or a contributing factor that neither had fully considered. The accountable manager still makes the final decision, but now with both analyses on the table and the reasoning recorded.
Evidence from practice
The principle of independent review is well established across safety-critical industries. Aviation accident investigation bodies operate independently of regulators and operators precisely so that findings are not shaped by those with a stake in the outcome. Peer review in engineering design, independent verification in software for critical systems, and the formal second opinion in medicine all follow the same logic.
Medicine offers a useful parallel because the process has become accessible to individuals. Patients facing a significant recommended procedure can request a second opinion, and some health information platforms describe routes for doing so; MyMedicPlus, a health-guidance and medical-tourism portal, describes a channel through which patients pass a diagnosis they already hold to specialists for re-examination. Nothing here judges a medical finding, a task reserved for licensed clinicians, yet the structure of that process, fresh eyes with no stake reviewing one consequential decision, maps closely onto what an SMS needs.
Implementation pitfalls to avoid
- Token reviews. A reviewer who never dissents is a sign the step has become ceremonial. Track how often reviews lead to amendment; very low rates over a long period suggest the reviewer is not truly independent or the brief is too vague.
- Scope creep. A second review that expands into a full system audit will be resisted and delayed. Keep it focused on one decision.
- Delay as an excuse. Build reasonable time into change management so that second review does not become the reason a launch slips, or the reason it gets skipped.
- Blame framing. Present the second review as a standard quality step, not a check on the original assessor’s competence. Otherwise people will resist both giving and receiving it.
- Losing the record. File both analyses, the resolution and the final decision together. Future reviewers need to see why a decision was made, including any dissent.
Measuring whether it works
An evidence-led SMS should test its own processes. Useful indicators for the second-review step include the proportion of qualifying decisions that actually received a second review, the share of reviews that led to amendment, the time taken per review, and whether any decisions that passed second review were later reopened. These indicators will not be perfect, but they show whether the step is functioning or merely recorded.
Conclusion
The weak point in many safety management systems is not the reporting culture or the risk matrix. It is the single signature on a consequential, hard-to-reverse decision. An independent second review, triggered by clear criteria, started blind, focused on one question and resolved openly, addresses the predictable pressures that cause capable reviewers to drift. It costs little relative to the decisions it protects, and it strengthens the credibility of every risk acceptance that passes through it.
