The appointment is at eleven. The patient took half a day off, crossed the city, and is sitting in your waiting room on time. The crown is not. Somebody calls the lab, the lab says it will come by evening, and the receptionist walks out to explain to a person who did everything right that they will have to come again. The temporary goes back on. The chair slot is gone. And in the patient's memory, this is not the lab's failure. It is yours.
Any clinic doing crowns, bridges and dentures is running a small manufacturing supply chain with an outside vendor. There is a specification, a production lead time, a delivery date, a quality check on receipt, and a monthly bill to reconcile. In most practices this entire operation is managed with a whiteboard, a WhatsApp group and the senior assistant's memory.
Lab work is the only part of dentistry where you routinely promise a patient a date you do not control. That is worth saying plainly, because it explains why the fix is not about working harder. It is about tracking.
Every case out at the lab is an appointment you have already promised
The moment the impression is taken, the front desk books the fitting. That booking is a commitment made on another business's production schedule, usually on the strength of a verbal turnaround the lab quoted once, months ago, for a different kind of case.
When it slips, the damage is never just the delay. A patient who took leave has to take leave again. A temporary sits in the mouth longer than it was designed to, and a proportion of those will debond or fracture before the real unit arrives, which is a fresh emergency appointment you will not bill for. A prime slot in a full book goes empty on the day. And the case slides out of the treatment plan's rhythm, which for some patients is the exact moment the plan stops progressing altogether.
None of that is exotic risk. It is the ordinary cost of promising dates on faith.
What a lab order has to carry: work type, shade, due date
A lab order is a specification, and specifications are written down. The minimum is small enough that there is no excuse for not having it, and complete enough that the lab never has to phone the clinic to ask what you meant.
- The patient, and the exact tooth or teeth the work is for
- The work type — porcelain-fused-to-metal crown, zirconia, a removable partial, a complete denture, a night guard — stated the same way every time
- The shade, recorded at the chair while the patient is still in it
- The date you need it back, recorded on the order itself — the system flags the case the moment that date slips
- The dentist who will fit it, so the fitting is booked against someone real
- Anything the lab cannot infer from the impression: pontic design, occlusal clearance, a note about the opposing arch
Shade deserves its own paragraph because it causes more remakes than any other single field, and because it is the one nobody wants to own afterwards. It is decided in a moment of daylight with a shade tab against a wet tooth, and then it has to survive the journey to a technician who has never met the patient. A photo in a WhatsApp thread is useful evidence and terrible record-keeping — three weeks later nobody can find it, and the thread has two hundred messages about other patients in between. The shade belongs on the order, next to the work type and the date, where the same person who took it wrote it down.
The whiteboard only the senior assistant can read
Here is the real system in most clinics: a whiteboard with initials and half-dates, a diary at the front desk, a WhatsApp group with the lab, and one senior assistant who genuinely knows where every case is. It works beautifully, right up until she takes two weeks off for a family wedding, or leaves for the clinic that offered her three thousand more.
There is a quick test. Walk to the front desk right now and ask which cases are at the lab and which of them are late. If the honest answer involves calling somebody, or scrolling a chat thread, the clinic does not have a system — it has a person. People are wonderful and they take leave.
The WhatsApp thread is the sneakiest part of this, because it feels like a record. It is not searchable in any way that answers a useful question. You cannot ask a chat thread which four cases are due back this week, or how many jobs sent to that lab last month came back after the promised date. Every message in it was written to solve a problem that afternoon.
Ordered, received, fitted: the three dates that matter
Track every case through three states, each with its own date: ordered, with the expected return date attached; received, meaning the work is physically in the clinic and has been looked at; and fitted, meaning it is in the patient's mouth and the case is closed.
The states matter because the gaps between them each tell you something different, and they point at different people.
- Ordered to expected return — the date the lab committed to on the order — and expected to received, the lateness that accumulates lab by lab
- Expected return to received — lateness, which accumulates into that lab's real performance record rather than your impression of it
- Received to fitted — your own delay, a case sitting in a drawer at reception while the patient waits for a call nobody made
That third gap is the uncomfortable one. Clinics that track only whether a case is back will blame the lab for delays they created themselves — work that arrived on Tuesday and was fitted the following Monday because nobody was watching the received list. You cannot fix what you have quietly misattributed.
Remakes and adjustments, and who pays for them
Sooner or later a unit comes back with the wrong shade, an open margin or a bite that is high in a way no adjustment will save. Then comes the conversation nobody enjoys: was it the impression, or was it the lab. If the order was a phone call, that conversation is one memory against another, and the clinic usually absorbs the cost because a patient is waiting and arguing is slower than paying.
If the order was a written spec with the shade and work type on it, the same conversation takes two minutes and stays professional. Labs are far more willing to own a remake when the record is unambiguous, because they are not being accused, they are being shown.
Track the remake as its own order, linked to the original, rather than editing the first one and pushing its date out. This is a small discipline with a large consequence: a clinic that overwrites the original order every time something is remade will end each year with a lab performance history that has quietly cleaned itself. Minor occlusal adjustments are normal dentistry and not worth logging as failures. A pattern of them coming from one lab, on one type of work, is data you will only ever see if you kept the cases separate.
Costing lab work into the treatment price
A lab bill is a direct cost attached to one specific case for one specific patient. Most clinics treat it as a monthly overhead — one payment to the lab at the end of the month, filed under expenses — and in doing so lose the ability to say what any individual crown actually earned.
Do the arithmetic on one case and it becomes obvious why this matters. A zirconia crown quoted to the patient at a price, minus what that lab charged for that unit, minus the chair time across two or three sittings, is the real margin. Run the same subtraction across a month of crown-and-bridge work and you will usually discover that the case mix you thought was profitable and the case mix that actually is are not the same list.
This is also the only way to check the lab's monthly statement. A bill listing thirty jobs is uncheckable if your own record of what you sent lives on a wiped whiteboard; you sign it because disputing it would take an afternoon you do not have. It is entirely checkable if every order is a dated record with a work type against a named patient. And when the invoice goes to the patient, lab charges should sit as their own line alongside the procedure and any prescription, so the bill reads as an honest breakdown rather than one lump sum a family has to take on trust.
The daily list of what is due back
Five minutes, every morning, before the first patient. Three questions, in this order.
- For every fitting appointment on today's list, is the work physically in the clinic — not promised, in the clinic
- Which cases were due back today or earlier and have not been received
- Which cases fall due in the next two working days, so a slip can still be absorbed
The value is entirely in the timing. Calling the lab the day before turns a delay into a reschedule with notice, which patients forgive easily. Calling at eleven fifteen while the patient sits in your waiting room turns the same delay into an insult. Same lab, same slip, completely different reputation.
One rule enforces the whole routine: never confirm a fitting appointment for a case that has not been received. The confirmation call the front desk makes anyway becomes conditional on the work being in the building, and the clinic stops promising things it cannot see.
The patient never meets your lab. Every promise the lab breaks is recorded in their memory against your name.
Running two labs, and tracking both
Almost every established clinic uses more than one lab. The nearby one for removable work and repairs, because it is quick and cheap. The good one across town for anterior aesthetics, because the technician there can actually match a central incisor. Perhaps a milling centre for zirconia. Splitting work by strength is sound practice; it is also the point at which one whiteboard stops being able to hold the truth.
Once each order carries the lab it went to, three questions become answerable per lab: what turnaround they actually deliver against what they promise, how often their work comes back for a remake, and what each work type really costs you there. That is the difference between a feeling and a record.
It changes the annual pricing conversation too. Do not tell a lab they are slow — they will disagree, politely, and nothing will change. Show them their own dates from your own orders, and ask what they need from you to hit the promised ones. Labs that are good at their job respond well to that, and the ones that do not respond have told you something useful.
The mechanics required here are unglamorous. A lab order recorded against the patient with work type, shade and an expected return date. Three states with dates on them, so ordered, received and fitted are distinguishable. Lab charges that reach the invoice as their own line. And a pending-lab-work list that flags overdue cases on its own, so the morning check takes five minutes instead of five phone calls.
Patients judge a dental clinic on very few things, and one of the biggest is whether it does what it said it would do on the day it said it. That promise is only as good as your worst-tracked case.