00K Average Loss: Embezzlement in Dental Practices

Somebody inside a dental practice typically steals about $109,000 before anyone catches on, and the scheme runs for close to two years before it surfaces, according to Prosperident's forensic casework in dentistry. That is not a rounding error on a $1.2 million practice. At a realistic 35% flow-through on incremental production, replacing $109,000 of stolen net cash means producing roughly $311,000 of additional dentistry, for free, just to get back to even.

Our view, and it is the uncomfortable one: nearly every dental embezzlement case is the owner's doing. Not the hiring. The design. When one person posts adjustments, runs the day sheet, prepares the deposit, and gets the bank statement, you haven't hired a thief, you've built a job where theft is undetectable and then filled it with a human being. Plenty of owners will push back on this, and the pushback is reasonable: you cannot run a six-person front office like a bank's treasury department. We think that objection is wrong, and the rest of this piece is why.

Half the Profession, a Billion Dollars, and Nobody Talks About It

The scale is not anecdotal. A 2007 ADA study found that 17.5% of surveyed dentists reported their primary practice had been embezzled. More recent figures are worse: DOCS Education, citing Prosperident founder David Harris and 2020 ADA survey data, puts the share of affected US dental offices at roughly half, with average cases exceeding $100,000 before detection and real files running to $160,000 and beyond. Industry-wide, the Michigan Dental Association Journal pegs annual embezzlement losses in dentistry at over $1 billion.

Cross-industry data says the same thing in a different dialect. The ACFE's Occupational Fraud 2024: A Report to the Nations found a median loss of $100,000 in health care occupational fraud cases, with an average of $721,000. The gap between median and average is the part owners should read twice. Most cases are survivable. A minority are practice-ending, and nothing about the first six months of a case tells you which one you're in.

What Most Owners Get Wrong: Guarding the Cash Drawer While the Adjustments Walk Out

The specific mistake, made most often by owners of solo and two-doctor practices who grew from four staff to twelve without ever changing the money workflow, is treating embezzlement as a cash problem. So they count the petty cash. They install a camera over the front desk. They make a rule that two people have to be present when the deposit bag is sealed.

Cash theft is the amateur version, and it is capped by how little cash walks through a modern dental practice. The expensive version lives in the practice management software, and it looks like this:

• A patient pays $1,400 in cash or by check. The payment never gets posted. The balance gets written off as a courtesy adjustment, a "bad debt," or a senior discount.
• An insurance check arrives, gets posted to the wrong patient's ledger, and the float covers last month's theft. That's lapping, and it's why these schemes run for years instead of weeks.
• Refunds get issued to real patients at addresses the embezzler controls.
• Credit balances get quietly cleared, because nobody audits credit balances.
• Payroll gets padded: unapproved overtime, a bonus line, a reimbursed "supply run."

Here is why this beats your instincts. Every one of those schemes leaves your collection percentage looking normal, because the receivable disappeared along with the money. Your production looks normal. Your day sheet balances, because the person who created the imbalance is the person who reconciles it. The only trace is a slow, boring drift in your adjustment line and your overhead percentage, and most owners read a rising overhead number as a supply-cost problem or a staffing problem. We'd call that the single most expensive misdiagnosis in private dental practice.

The second thing owners get wrong is the profile. They screen for the person who seems shady. The files say the opposite: long-tenured, deeply trusted, the one who never takes a full week off, who insists on handling the bank run personally because it's easier, who gets defensive about the one report you asked for. Trust is not the opposite of risk. Trust is the delivery mechanism.

The $1.2M Practice That Really Lost $185,000

Run the numbers on a practice collecting $1.2 million a year with a 62% overhead, so about $456,000 of owner profit before debt service. Say a trusted business coordinator takes the average: $109,000 over 22 months, which is about $4,955 a month. Against monthly collections of roughly $100,000, that's just under 5% of the top line, which is small enough to hide inside normal seasonality and a bad August.

Now the real damage. The $109,000 came off the bottom line, not the top, so replacing it with dentistry at a 35% marginal flow-through takes about $311,000 in additional production. That's roughly four extra months of a mid-sized practice's output, delivered for nothing.

Then there's enterprise value, which almost nobody computes. The theft suppressed reported collections by about $54,500 a year. If you sell on anything resembling a multiple of collections, call it 0.75x, those two depressed years cost you about $40,875 of practice value at the closing table, assuming a buyer will even normalize for a fraud you can't document. Add a forensic investigation, your attorney, your CPA's remediation time, and the hiring and retraining cycle, and $35,000 is conservative. The total event is about $185,000.

Recovery? Partial at best. Restitution orders get paid at pennies on the dollar because the money is gone, spent on ordinary life. Fidelity bonds and employee dishonesty riders pay out, but only to the extent you can prove the loss, and proving the loss is exactly what two years of unreconciled adjustments make hard.

Against $185,000, price the control. An outside bookkeeper who reconciles your bank statement to your practice management deposit report every month, and who reports only to you, runs maybe $650 a month in most markets. That's $7,800 a year. You are buying down a $185,000 tail risk for 1.7% of the loss, annually. We don't know many purchases in dentistry with that shape.

Three Controls That Actually Break the Scheme

Forget the twenty-point checklist. Three controls do most of the work, and they work because they break the loop between creating a discrepancy and hiding it.

One: the bank statement never touches your team. It goes to your home address or to an email only you open. Every month you compare total deposits on the statement to total payments posted in the software for the same period. Two numbers. If they don't match within a day or two of timing, you ask why, in writing. This one habit kills most lapping schemes, because lapping requires that nobody ever compares the ledger to the bank.

Two: adjustment authority is separated from payment posting, and adjustments get reviewed by name. Pull an adjustment report monthly, sorted by user and by type. You're looking for concentration: one operator generating most of the write-offs, courtesy adjustments outside your stated policy, round numbers, adjustments posted after hours. Set a hard rule that any single adjustment over $250 needs your initials before it posts. Our call: if your software lets staff delete or backdate transactions without an audit trail, turn that permission off this week, and if it can't be turned off, that's a reason to change software.

Three: mandatory vacation, and you cover the desk. Make it a benefit, not a test. Two consecutive weeks a year where the business coordinator is fully out and somebody else runs the money. Schemes requiring daily maintenance fail in week one. It's also the cheapest audit you'll ever run, because you'll see with your own eyes how the deposit actually gets made.

Why Your CPA Isn't a Safety Net

Owners assume the accountant would catch it. The accountant is reconciling what hit the bank against what you told them hit the bank. If a payment never entered the system, there's nothing to reconcile and no variance to flag. A compiled year-end statement is not an audit, and your CPA isn't claiming it is.

Same with the camera over the front desk. It records the cash drawer, which is the smallest pile of money in the building. Nobody watches footage looking for a write-off.

And the one we'd push back on hardest: the idea that going fully cashless or outsourcing billing solves it. Outsourced billing moves exposure, it doesn't delete it. Someone still posts payments, someone still has adjustment rights, and now that someone doesn't work in your building where you'd notice the new car.

If You Think It's Happening Right Now

Do not confront. Do not ask a leading question at the morning huddle. The first thing a suspected embezzler does is destroy the audit trail, and the second is tell the rest of the team a story that costs you three good employees. Do not terminate first either; a quiet firing without documentation forfeits your insurance claim and leaves you nothing to bargain with on restitution.

Call a forensic specialist who works in dentistry, and loop in your attorney before anyone else knows. Preserve data access and backups quietly. Say nothing to your spouse's hygienist friend. Then let the professionals build a case you can actually collect on.

Four Moves Before Friday

Redirect your bank statements to an address your team can't reach, today. Pull a twelve-month adjustment report by user and read it line by line, flagging anything round, late-night, or concentrated in one person's login. Open your software's permission settings and remove delete and backdate rights from every account except yours. Then schedule the two-week mandatory vacation for whoever handles money, and put it on the calendar before the end of the month.

One more thing worth doing: embezzlement shows up in your books as overhead you can't explain, so know what your overhead should be before you go hunting for what's wrong with it. If your total overhead has drifted two or three points over the last couple of years and supplies and labor don't explain it, that drift is a lead, not noise. Run your own numbers on our overhead calculator, compare your adjustment and collection lines against the benchmark, and find out whether you're looking at a cost problem or a control problem.

What we're tracking next: whether practices that moved to centralized or outsourced billing show lower loss rates or just longer detection times. If the data shows locked-down audit trails alone cut average losses materially, then this is a software problem and we'll say so. Until then, we hold the position: the owner designs the opportunity, and your most loyal employee and your largest fraud risk are often the same person, because you gave only one of them both sets of keys.