Fair review, real accountability and the conditions for improvement. The choice is not between finding a culprit and excusing everything. A useful review asks what happened, why it made sense at the time and what must change.
The false choice after something goes wrong
When a service fails, people understandably want an explanation. Leaders may feel pressure to show that someone has been held accountable. Staff may fear that any admission will be used against them. The result can be a narrow search for an individual mistake, or an equally unhelpful insistence that the event was simply a 'systems issue'. Neither is enough. A serious review must be fair to individuals while examining the conditions in which decisions were made. It must also be willing to identify poor conduct where the evidence supports it. Learning without blame is not learning without responsibility.
Consider an incident in which an important message did not reach the person who needed it. One account says an employee failed to pass it on. A wider review may find that the handover process relied on an informal conversation, that two teams used different contact lists and that the recipient was not on duty. Those findings do not erase individual decisions. They explain why the failure was possible and what could prevent a recurrence.
Start with the work as it happened
Policies describe work as imagined. Incidents occur during work as performed: under time pressure, with interruptions, incomplete information and competing demands. A review that judges every action against a perfect retrospective view will miss why a decision appeared reasonable in the moment. Reconstruct the timeline using documents, interviews and other available records. Mark uncertainties. Ask what information was available at each point and what alternatives were genuinely feasible. Avoid turning hindsight into a standard nobody could have met.
The purpose is not to construct a sympathetic narrative. It is to establish a credible one. Where accounts conflict, record the disagreement and the basis for any conclusion. Where evidence is absent, say so. Trust is damaged when a review sounds more certain than the underlying material allows.
Distinguish cause, contribution and coincidence
Not everything found during a review caused the outcome. A missing form may be important, irrelevant or evidence of a wider weakness. A training gap may have contributed, but so might an unclear instruction, unsuitable equipment or a decision made under pressure. Strong reviews resist the urge to collect every imperfection into one list of 'root causes'. They explain the plausible pathway between a condition and the event, and they distinguish immediate safeguards from longer-term improvements.
This distinction makes recommendations better. 'Retrain all staff' is easy to write and difficult to evaluate. 'Replace two conflicting instructions with one controlled version, assign an owner and test its use at the next handover' is more specific. It identifies a change, a responsibility and a way to check whether the change has worked.
Recommendations must survive the meeting
Many reviews end with a polished report and an action log. The harder question is whether the actions become part of normal work. Each recommendation should have an owner, a rationale, a realistic completion condition and a review point. The organisation should ask what new burden the change creates and whether that burden is proportionate. More forms are not automatically safer. A change that looks robust on paper but cannot be sustained may recreate the same vulnerability under a different name.
Independent consultancy can contribute by facilitating a fair review, clarifying evidence, improving recommendations and helping design a usable follow-up process. It should not pretend to replace formal investigations, legal duties or specialist regulatory roles where those are required.
What a good review leaves behind
The most important output is not a statement that lessons have been learned. It is a defensible account of what is known, an honest description of uncertainty, a small number of meaningful improvements and a way to tell whether those improvements have taken hold. People affected by an incident deserve more than a ritual assurance. Staff deserve a process that treats evidence seriously. Leaders need findings they can act on.
Fairness and accountability are not opposing values. Done properly, each makes the other possible.
From the essay to the work. These reflections inform how I approach defined consultancy projects: establish the problem, test assumptions, make responsibilities clearer and leave practical material that people can use.

