Preventing Laboratory Errors.
Errors in restorative work are usually described as somebody's mistake. They are more accurately described as a system that allowed a mistake to reach a patient.
Active failures and latent conditions
It is useful to distinguish between the immediate error — a margin misread, a bite record not seated, a component ordered incorrectly — and the conditions that made it likely.
Latent conditions are the ones worth attention, because they are the ones that can be changed. A prescription form that does not prompt for the shade guide. A packing routine that separates the paperwork from the case. A practice with no reliable way to reach a decision-maker. A laboratory with no defined point at which arriving records are inspected.
None of those is anyone's mistake. All of them make mistakes more likely, and all of them are fixable without asking anybody to try harder.
Information is lost at every handoff
A case passes through many hands: the clinician who forms the intent, the nurse who assembles the case, the courier, the person who receives it, the technician who makes it, whoever dispatches it, and the clinician who seats it.
At each transfer, only what is written down survives. Everything held in somebody's head is lost, and it is lost silently — nobody at the receiving end knows what they were not told.
This is why intent, constraints and priorities matter so much on the prescription. They are the information most likely to exist only in the clinician's mind and least likely to be reconstructible from the physical case.
Checklists work, and they work for a specific reason
The value of a checklist is not that it tells experienced people how to do their jobs. It is that it protects against the ordinary lapses that occur when competent people are busy, interrupted or working from routine.
The most useful ones are short and used at defined moments: before the record is taken, before the case is packed, and before the restoration is seated. Six items each, no more.
Before recording — undercuts, clearance in intercuspation and excursion, margin continuity, margin on sound tooth, adjacent teeth undamaged, retention features cut. Before packing — prescription complete, shade guide named, photographs attached where relevant, records included, case protected. Before seating — seats on the model, then fit, then contacts, then occlusion.
Standardization reduces variance
Every point at which a practice does something differently case to case is a point where something can be forgotten. One prescription format, one shade guide, one photographic protocol, one packing routine, one named contact.
This matters most in multi-clinician practices, where divergence between individuals produces variability that the laboratory experiences as inconsistency and attributes to the practice as a whole.
Standardize what does not need to vary, and record deliberately the things that do — individual material preferences, contact tightness, occlusal contact at delivery.
Verify where verification is cheap
Error correction cost rises steeply through a case. A clearance problem identified before the impression costs a minute of further reduction. The same problem identified at the seat appointment costs an appointment, a remake, and a patient's confidence.
So verification should be front-loaded deliberately: check the preparation before recording, check the record before dispatch, check the case on arrival at the laboratory, check the framework before veneering.
Each of those is inexpensive at the point it happens and expensive to skip.
Make it easy to raise a concern
A technician uncertain whether something is a problem will raise it if doing so is straightforward and will not if it is awkward. The same applies to a nurse who notices something on a case being packed, and to a clinician uncertain whether a record is good enough.
The conditions that make raising concerns easy are unremarkable: a named contact, a reachable number, a relationship where questions are treated as useful rather than as criticism, and a practice that does not react badly to being told something is wrong.
Laboratories that never query anything are not necessarily producing better work. They may have learned that querying is unwelcome.
Review near-misses, not only failures
The cases worth examining are not only the ones that came back. They include the ones that needed substantial adjustment, the ones where a query caught something at the last moment, and the ones that were nearly sent with something missing.
Near-misses are far more numerous than failures and they reveal the same latent conditions with none of the cost. A practice that notices it has twice nearly sent a case without a bite record has learned something useful.
Nobody needs a formal system for this. Noticing and mentioning is most of the value.
Both sides own it
The laboratory controls fabrication, inspection and dispatch. The practice controls the records, the prescription, the responsiveness and the delivery. Neither can fully compensate for weakness in the other.
The restorations that fail are disproportionately the ones where a weakness at one end met a weakness at the other — an ambiguous margin met by an assumption rather than a phone call, or a query raised and not answered.
Which is why the most effective single intervention available to either side is a working relationship in which questions get asked and answered quickly.
In short
- Fix latent conditions rather than asking people to be more careful
- Only what is written down survives a handoff
- Use short checklists at defined moments: before recording, packing and seating
- Front-load verification — correction cost rises steeply through a case
- Review near-misses; they reveal the same causes without the cost
More on clinical
Discuss a case
Where a case sits between options, a short conversation before preparation is usually quicker than resolving it afterwards.