Incident review should change the system, not hide the error

The treatment has stopped, the client has entered the appropriate follow-up route and the affected equipment is no longer being used. The room is quiet again. This is often where an incident is most at risk of disappearing.

One impulse is to reduce it to a short note and move on before the event attracts attention. Another is to gather the team for an earnest conversation, agree that everyone will be more careful and call that a review. Neither response is enough. A real incident has created duties to the client, the record and the organisation. It needs a formal process with authority, evidence, owners and an end point that can be checked.

This is different from an educational case review. A case review can help practitioners examine judgement, discuss an anonymised scenario and learn how to make a better decision next time. Formal incident review has a narrower and harder job. It establishes what happened in an actual event, meets reporting and disclosure duties, determines responsibility, controls continuing risk and verifies that corrective action works. Learning may follow, but it cannot replace governance.

Start with a mandate, not a meeting

Before interviews and theories, define who is leading the review and what authority that person has. The reviewer needs access to the relevant clinical record, device information, training and authorisation records, service history, staff accounts and local procedures. They also need a clear route for issues that require notification to a clinical lead, laser-safety officer, manufacturer, insurer or regulator under local rules.

The scope should be written early. What event is being reviewed? Which client-care questions remain open? Is an item of equipment isolated? Could the same condition affect another appointment, device or site? Are there records at risk of being overwritten? These questions create immediate controls while the cause is still uncertain.

An interim control is not a final verdict. Pausing a handpiece, restricting a workflow or adding supervision can be necessary before the review is complete. The reason, owner and condition for lifting that control should be recorded. Otherwise a temporary restriction either vanishes quietly or remains forever without anybody knowing whether it is still useful.

The process also needs boundaries. Client care and follow-up continue through the qualified clinical route.

Employment questions follow the proper management process. Technical examination belongs to authorised personnel. The incident review connects these streams, but it should not let an unqualified committee diagnose the client, repair the laser or invent disciplinary rules after the event.

Preserve the record before the story settles

People form a coherent story quickly, especially after a stressful event. That is human, but it can make a later account feel more certain than the original observations. Obtain individual factual accounts while memories are fresh and before a group discussion aligns the wording. Ask what the person saw, heard, checked, understood and did. Do not begin with “Why were you careless?” or its polite equivalent.

Secure the original treatment record and its audit trail, device logs, error messages, photographs taken with consent, service records, relevant consumables, room schedule, training status and current versions of the protocol and instructions. Preserve access controls and confidentiality. An incident is not permission to copy a client's photographs into a general staff channel or to circulate medical details to everybody who is curious.

Keep observation separate from interpretation. “Cooling was no longer felt after the alert appeared” is an observation assembled from a report and a device event. “The practitioner broke the cooling system” is a causal conclusion. It may eventually be supported, contradicted or left unresolved. Labelling the two statements correctly prevents an early theory from hardening into the official record.

Unknowns are also findings. If the exact handpiece cannot be identified, if a parameter field is blank or if the current instruction was not accessible in the room, write that down. Do not fill a gap with the usual setting, copy a previous visit or edit a note so that it appears complete. Missing evidence may reveal a documentation weakness, and concealing it creates another one.

Build a timeline, then test the barriers

A useful timeline begins before the final action. It may include the booking information, health and medication update, recent UV review, consent discussion, room preparation, device check, model and handpiece confirmation, observed skin response, client feedback, interruption, stop, escalation and subsequent contact. The aim is not to produce a dramatic minute-by-minute narrative. It is to see when important information became available and where the process had a chance to catch the risk.

Then examine the barriers that were supposed to prevent or limit the event.

Was the relevant question present on the consultation form? Could the practitioner see a new answer before treatment? Did the interface clearly identify the device and mode? Were similar handpieces labelled so they could not be confused? Was cooling checked in a way that would reveal this failure? Was eye protection suitable, available and inspected? Could a practitioner stop without being pressured by the schedule?

Do not ask only whether a rule existed. Ask whether it was current, understandable, accessible and used. A protocol stored in a folder nobody can find during a problem is not a reliable barrier. Training completed once does not prove present competence. A double check performed by two people who both rely on the same unclear screen may produce two signatures and no additional safety.

This is where an individual action becomes a systems question. The action immediately before an incident still matters, but it sits inside equipment design, workload, handover, supervision, labelling, maintenance and local culture. Review each layer according to the evidence. Do not force every event into one favourite explanation such as “human error” or “poor training.”

Accountability needs precision

Systems thinking is sometimes misunderstood as a way to say that nobody is responsible. That is not the point. If a practitioner knowingly ignored a clear prohibition, used equipment without authorisation or concealed an event, that conduct must be addressed through the appropriate process. If a manager knew that cooling failures were recurring and kept the device in service, that decision also belongs in the review.

At the same time, naming the last person who touched the handpiece does not explain why the organisation failed to catch the risk. A practitioner may make an incorrect selection because two devices share an unclear interface, the asset label is missing and the handover is rushed. The selection remains their action. The conditions that made the error easy and detection unlikely remain the organisation's responsibility.

Fair review therefore asks two distinct questions. What was reasonably expected of each person given their role, training and the information available at the time? What did the system provide or fail to provide so that the safe action was visible and practical? Keeping both questions prevents blame from replacing analysis and prevents vague references to “the system” from dissolving individual duty.

The language of the finding should match the strength of the evidence.

Use confirmed, supported, possible and unknown deliberately. Give people an opportunity to correct factual errors in their account without bargaining over uncomfortable findings. Record disagreement rather than smoothing it away. A defensible conclusion is not the one everybody likes. It is the one another reviewer can trace back to evidence.

Corrective action is a testable change

“Staff reminded to be vigilant” is not a strong corrective action. It has no clear mechanism, no reliable completion point and no way to show whether the risk has changed. A useful action addresses a finding and names an owner, deadline and verification method. It might change device identification, restrict access, redesign a form, alter a maintenance trigger, improve escalation access or require observed reassessment before authorisation is restored.

Choose the level of control that fits the cause. More training may be appropriate when the expected skill was never taught or demonstrated. It is weak when the real problem is an ambiguous interface or a damaged component. A warning poster cannot repair a cooling fault. A second signature does not help if both staff members lack the same critical information.

Verification happens after implementation. Sample later records. Observe the revised workflow. Test whether staff can find the escalation route under time pressure. Confirm that an equipment change works in normal use and has not created another hazard. Review related incidents and near misses over an appropriate period. An action is complete when the defined control is operating, not when the email announcing it has been sent.

Client communication remains part of closure. Follow the clinic's legal and professional duties for honest information, continuing contact and access to questions. Do not turn a recognisable client story into training material or public proof of transparency. If the event later becomes an educational case, remove identifying details and teach the decision process, not the drama.

The final report should connect the event, evidence, findings, responsibilities, corrective actions and verification. It should also record what remains uncertain. Senior ownership matters here: someone with authority must accept residual risk, ensure required notifications are complete and decide when temporary restrictions can be lifted.

A healthy reporting culture is not one in which incidents carry no consequences. It is one in which consequences are accurate. People are accountable for their choices, leaders are accountable for the conditions they maintain and the organisation is accountable for proving that its correction works. That is how review protects the next client instead of merely producing a tidy explanation of the last one.