Explainer

How the estimate is built

Four data sources, the order they are tried in, and why Mesio would rather leave the number blank than quote one it cannot stand behind.

Every procedure Mesio proposes carries what the patient is likely to owe. That number is stitched together from four different sources, each of which answers a different part of the question. This is what they are, the order they are tried in, and where Mesio stops rather than guess.

The arithmetic, and the two halves nobody has

What a patient owes is what you charge, minus what the carrier pays. Both halves have to come from somewhere.

The first half is yours. You upload your fee schedule once, as a PDF, a photo of a printout, or a spreadsheet, and it becomes an editable table. A code your schedule does not list stays blank rather than being filled from a neighboring code, because a fee invented by analogy is exactly the kind of number a patient repeats back to you.

The second half depends entirely on whether you are in network with that carrier.

In network, the answer exists. Out of network, nobody publishes it.

In network the carrier applies its percentage to a contracted rate, and that rate is in your contract. Mesio reads what the plan covers for that specific patient today, the percentages, the deductible, what is left of the annual maximum, the frequency limits, and prices against the contract. Add a procedure halfway through the appointment and it prices immediately.

Out of network there is no contract. The carrier still pays a percentage, but of an amount it decides for itself, called the allowed amount, and it will not tell you in advance what that will be. This is the hard part of the problem, and it is where the other data sources come in.

What carriers actually do out of network

Plans define the allowed amount one of two ways. Some pay against a fixed schedule of their own. Most define it as a point on a curve of what dentists in your area charge or are allowed for that procedure, which is why plan documents say things like the ninetieth percentile of usual and customary.

So Mesio holds payments benchmarking data. It is keyed to your area and to the individual procedure code, and each one is a distribution rather than a single figure, running from the low end of the market to the high end, in two separate families: what dentists bill, and what plans allow. That distinction matters, because a plan that pays a percentile of billed charges and one that pays the same percentile of allowed amounts produce very different numbers.

Finding which point on the curve your carrier actually pays

Knowing the curve is half of it. The other half is knowing where on it a particular carrier sits, and this is the part that gets better the longer you use Mesio.

For the larger carriers, the plan documents say. That is the starting assumption, and it is a real one. But documented language and actual behavior drift apart, so Mesio does not stop there.

If you let Mesio read your ledger, it can see what that carrier has actually paid your office, and it prices that carrier the way it has genuinely been behaving rather than the way its documents describe. The two are often not the same, and your remittances are the only place that difference shows up.

The bar for adopting a conclusion is deliberately high. Where the evidence does not clearly support one, Mesio keeps the more conservative answer rather than committing to a poorly supported one.

Because a carrier tends to pay the same way across its whole book, payments on any procedure teach Mesio where that carrier sits for every procedure. A handful of crowns can sharpen the estimate on work you have never billed that carrier before.

The order it tries, best evidence first

Every figure on the screen carries which rung it came from, in plain words. A number the carrier itemized and a number inferred from a curve look different and read differently, because a front desk that cannot tell them apart will eventually quote one as the other.

Where it deliberately errs, and where it refuses

When Mesio has to estimate, it aims slightly low on what the carrier will allow, which means aiming slightly high on what the patient owes. A patient quoted a little too much is a conversation. A patient quoted too little is a surprise bill, and the practice wears it.

The refusals are as deliberate:

• Your fee is never used as the out-of-network base. It is what you bill, not what the carrier allows, and treating the two as the same is the single most common way these estimates go wrong.
• It will not read past the ends of a curve. If a plan pays at a point the benchmarking data does not cover, that is a nil answer, not an extrapolation.
• It will not price a code your fee schedule does not list.
• It will not invent a carrier or a patient. An unresolvable carrier, a plan that could not be found with the identity on file, or a practice with no zip saved all produce a blank.

Those blanks are the correct answer, and they tell you to send a pre-treatment estimate, which is what you would do anyway when the payer has not spoken.

Making yours better

• Your fee schedule. The single biggest improvement, and a ten-minute job.
• Your NPI and zip. Without the zip there is no local curve, and every out-of-network estimate falls back.
• Which carriers you are in network with. A checklist in settings.
• Your ledger. This is what moves you from what plans say they do to what your carriers actually do with you. What leaves the practice is a carrier, a code, amounts and a month. Never a patient.

A dial in Settings → Insurance shows what share of what you bill can be priced from your own data. It is not a task that completes. It is an accuracy reading.