Nobody has measured dental case acceptance

If you own a dental practice you have been told your case acceptance is somewhere between a third and a half, and that this is bad. The number arrives in a webinar, a dashboard, a consultant's opening slide, or a benchmark report you had to give an email address to read. It is always low enough to be alarming and always attached to something you can buy.

We went looking for where the number comes from. Not the blog post that repeated it, but the organisation that first produced it, the year, the sample, and the definition of what was being counted. We could not find a single independent measurement of general-dentistry case acceptance in the United States. Not one. Every figure in general circulation originates with a practice-management software company, an analytics vendor, a consultant who sells case-acceptance training, or a trade column written by one of those.

That is not a complaint about bias. It is a specific, checkable problem: none of them publish a denominator.

Fourteen averages that cannot all be right

Here is what the category currently calls the average case acceptance rate, side by side, with who published each one. The bars are drawn against a stated axis of 100%, so nothing here is scaled to make a point.

Thirty-four to seventy-eight. The range only holds together because almost none of these sources says whether it is counting procedure line items, treatment plans, patients, or dollars, and those four questions have four different answers in the same practice on the same day.

The company that supplies both the indictment and the confession

Five of the figures above come from one corporate parent. Henry Schein One is a joint venture between Henry Schein, Inc. and Internet Brands, and it sells Dentrix and Dentrix Ascend, Jarvis Analytics, Lighthouse360, and Software of Excellence. It is the largest single supplier of case acceptance statistics in the industry.

In 2019, its Software of Excellence brand published research in the British Dental Journal, hosted on nature.com and carrying a disclaimer that it was submitted by a commercial organisation and had not been peer reviewed. The finding was that only 16% of practices formally track treatment plan acceptance at all, while 95% of respondents claimed they convert at least one plan in two.

Their own consultant explains why the number is wrong

The most useful thing we found was not a contradiction between companies. It was an explanation, published by the vendor itself, of why its own figure understates reality.

When a clinician offers a patient a choice of an implant, a bridge, or a partial, many practices enter all three into the software as separate planned procedures. The patient accepts one. The other two are never marked as declined, because from the practice's point of view nothing was declined: the patient chose. But the software has three procedures presented and one accepted, and it reports 33%.

The denominator is permanently inflated with treatment no patient was ever meant to accept. Every alternative ever offered sits in it forever. A practice that presents more options carefully looks worse than one that presents a single take-it-or-leave-it plan.

What a stated denominator looks like

There is one measurement we found that does all the things the vendor reports do not. Greenlee and colleagues, published in the American Journal of Orthodontics and Dentofacial Orthopedics in 2024, ran a prospective observational cohort through the National Dental Practice-Based Research Network: 91 participating practitioners, 345 adult patients, six geographic regions, three and a half years, funded by two National Institute of Dental and Craniofacial Research grants. No vendor, no consultancy, no software company.

It reports that 78% of patients accepted the most highly recommended plan, 19% accepted the second, and 3% accepted the third. Effectively nobody walked away. It also states its own limitations plainly: practitioners and patients self-selected, and patients were enrolled after treatment had begun, which biases toward acceptors.

We are not offering 78% as the new benchmark, and it would be dishonest to. That study measured adult anterior open bite in orthodontic practices. It is not a general-dentistry national figure and nobody should quote it as one. What it demonstrates is the shape of the gap: when you count patients deciding rather than procedure lines closing, the number roughly doubles. The vendors are not measuring your patients' willingness. They are measuring your data entry.

How a number with no source becomes an ADA benchmark

You have probably heard that the American Dental Association benchmarks case acceptance at 75-80%. It does not, and never has.

That figure traces to a single page in the ADA's Guidelines for Practice Success series, dated 2016, listing ten metrics a practice might track. The case acceptance entry reads, in full: “Are at least 75-80% of case presentations accepted?” It is a question. It sits alongside equally round rules of thumb: collect 98%, keep overhead under 63%, grow new patients 10-15% a year. There is no data behind it, no sample, and no source. The list is even mis-numbered, skipping 08.

The ADA Health Policy Institute does publish rigorous, nationally representative dental research with stated denominators, such as that 15.2% of the US population needed dental care and did not obtain it between 2013 and 2016, for instance, or that 13% reported cost barriers to dental care in 2023 against 4-5% for other health services. It has never published a treatment-plan case acceptance benchmark, in any year, in any form. Every citation of “ADA HPI case acceptance data” we checked resolved either to that 2016 tip sheet or to nothing at all.

A citation that does not contain the number citing it

One more, because it is documentable rather than rhetorical. Jarvis Analytics, a Henry Schein One company, states on its blog that “actual data proves that the reality lies between 25-50%”, and hyperlinks that claim directly to a July 2015 Dental Economics column by Roger Levin.

We read the column. It contains no case acceptance percentage of any kind. Its only figure is that 67.3% of surveyed dentists fall short of a 90% target. The 25-50% range is not in it, in any form, anywhere.

The 61% figure has a related problem. It comes from a 2016 Levin Group survey in which dentists reported their own acceptance rates. The same consultancy later published that dentists overestimate their own case acceptance by as much as 30 percentage points, which, if true, invalidates the survey that produced its own most-quoted number.

The finding the category does not repeat

Planet DDS, working with DentalXChange, did something the others have not: it tested whether case acceptance predicts anything. Across roughly 8,500 practices it measured the correlation between case acceptance rate and revenue growth.

The correlation was −0.09. Essentially zero. A practice's case acceptance percentage told you approximately nothing about whether it was growing. That finding is published, and it is not repeated anywhere, because a metric that predicts nothing is difficult to sell a dashboard for.

What we do with this, and what we would suggest you do

We build recare and reactivation systems for dental practices, so the honest disclosure is that we are in the same category as everybody quoted above. Which is exactly why there is no case acceptance percentage anywhere on this site, and why the dental pages carry a method for measuring rather than a figure to beat.

If you want a number that means something, the benchmark is not the industry. It is your own practice, last quarter, counted the same way twice:

• Count patients, not procedure lines. One patient offered three alternatives is one decision, not three.
• Mark the alternatives you did not expect to be taken. If your software cannot flag an option as an alternative rather than a rejection, that is the single change worth making before you measure anything.
• Set an observation window and keep it. Accepted at 7 days, 30 days and 90 days are three different numbers, and a report that does not say which one it used is not a report.
• Separate diagnosed-and-unscheduled from presented-and-declined. They are not the same failure and they do not have the same fix.
• Compare only to yourself. Nobody else's denominator is knowable, so nobody else's number is comparable.

That gives you something the entire benchmark industry cannot: a figure whose definition you know, measured the same way twice, on a sample you can inspect. It will not match anybody's report. That is the point, and the reasoning behind the system we build around it is published in full on the dental practices page.