What a Denial Costs Your Practice Once You Count the Rework, Not the Write-Off

The number on the dashboard is the wrong number

Your billing report shows two things about denials: the rate, and what got written off. Both are real. Neither is what denials cost you.

A denial costs a person's time. Somebody opens the remittance, reads the reason code, pulls the chart or the registration record, fixes something, resubmits or writes an appeal, and then checks back in three weeks to see whether it worked. That hour is paid for. It just is not paid for on a line called denials. It sits inside the salaries of your billers, your front desk supervisor, and sometimes your coder and a nurse, which means it never shows up where anyone is looking for it.

So most practices run with a cost that is large, recurring, and invisible at the same time. And because nobody sees it, nobody asks the question that follows from it: which of these denials are worth working at all?

What the rework actually costs

The figures that circulate for rework cost come in a wide band. The most commonly cited range is roughly 25 dollars for a straightforward correction up to about 118 dollars for a complex denial, against a few dollars to send a claim out clean the first time. We would treat that range as directional. It is repeated widely, and it traces mostly to vendor and MGMA-derived material rather than to a single primary study.

There is one firmer data point, and it comes from the hospital side. Premier's survey of member hospitals and health systems, fielded at the end of 2023, found that providers who fought denials spent an average of 43.84 dollars in administrative cost per claim across private payers, and 47.77 dollars on Medicare Advantage claims. That figure excludes clinical labour. Premier put the total cost of claims adjudication for providers at 25.7 billion dollars, and judged about 18 billion of it potentially unnecessary.

Two things about that number matter more than the number itself.

First, it is an average. A demographic correction that takes a biller four minutes and a medical necessity appeal that needs a physician letter are both in it. Your cost per denial is not one figure. It is a distribution, and the shape of that distribution is what decides your decisions.

Second, it is priced at wage. The honest price of a rework hour is what that person would otherwise have been doing. A biller working a 60 dollar denial is not working the 1,400 dollar claim sitting past 90 days. A front desk lead fixing registration errors from last month is not preventing this month's. The opportunity cost is the part that never appears anywhere, and in a lean billing office it is usually the larger part.

Why nobody computes it

It is not negligence. The data structure hides it.

Your practice management system knows the denial, the reason code, the payer, and the dollar amount. It does not know how long anyone spent on it, because nobody logs time against claims. Your payroll knows what the billing team costs. It does not know how that cost splits between clean work and rework. The two records never meet, so the cost of a denial exists in neither.

Rate reporting makes it worse. A denial rate of 8 percent sounds like a single problem with a single owner. In practice it is a pile of different failures, with different fix costs, created by different people, most of whom are not in the billing office. We wrote about one of those upstream sources in prior authorization, where the failure is procedural before it is financial.

Some denials are not worth working

This is the part nobody selling denial management will tell you, because it shrinks the thing they are selling.

If a denial is worth 38 dollars and the rework costs 45 in time, working it loses money even when you win. If the appeal succeeds half the time, the expected recovery is 19 dollars against the same 45 dollars of effort. A practice that insists every denial gets worked is, on a meaningful slice of its volume, paying staff to lose money slowly.

The industry numbers suggest practices already sense this. A commonly repeated figure holds that roughly 65 percent of denied claims are never resubmitted. On the other side, Premier's analysis found that roughly 70 percent of denied claims were ultimately overturned and paid, which tells you most of those denials were wrong and recoverable. Both can be true at once. Practices are abandoning denials, and they are abandoning them by accident rather than by decision: whichever ones fell to the bottom of the queue, not whichever ones were cheapest to lose.

That is the real failure. Not that low-value denials get dropped, but that the drop is random. A deliberate threshold, set by denial type and payer, puts the hours where the recovery is. An accidental one puts them wherever the queue happened to be sorted that morning.

We are not going to publish a threshold here, and not out of coyness. The break-even depends on your staff cost, your payer mix, your overturn rate by reason code, and your timely filing windows, and a number borrowed from someone else's practice will be wrong for yours in a direction you cannot predict.

Stop working them, and fix the cause

Writing off a class of denials without fixing why they happen is just a slower write-off. The decision to stop working a denial type only makes sense paired with the decision to stop creating it.

And this is where the pattern gets encouraging. Front-end and preventable errors account for the majority of denials in most analyses. Figures commonly cited from HFMA and AMA-derived benchmark data put demographic and eligibility errors alone at somewhere between 30 and 60 percent of the total, depending on the source and how a denial is counted. The range is wide enough that we would not quote a single figure as settled. It is narrow enough to say the biggest pile of rework usually starts at the front desk, not in coding.

That changes the economics completely. A registration error costs almost nothing to prevent and a full rework cycle to fix. A denial type you have decided not to work is a denial type whose cause you can now attack with the hours you just freed up. The same hour, moved upstream, is worth several hours downstream.

It also changes who owns the problem. The person who creates most of your denials is usually measured on how fast patients get checked in, and almost never sees what happened to the claim afterwards. Rework cost lands in billing. The cause lives somewhere else. Until those two are connected, training the front desk again is a reasonable thing to do and it will not hold.

What this means before you buy anything

Denial management software is a large, crowded category, and most of it is built to make working denials faster. That is useful. But faster rework of a denial you should not be creating, or should not be working, is efficiency applied to the wrong problem. You get quicker at losing the same money. If you are sizing a purchase, the approach in our piece on automation ROI applies directly: price the baseline honestly before you price the tool.

The questions that decide whether a tool pays back are ones your own data can answer:

• What does a denial cost you by reason code, not on average?
• Which denial types are you currently abandoning, and did anyone choose that?
• What share of your rework traces to registration and eligibility, and does the person who registered the patient ever find out?
• For the denials you do work, what is your overturn rate by payer?

If the answers show that a handful of causes create most of the rework, the cheapest fix is probably upstream and probably not software. If the answers show genuinely complex, high-value denials that need skilled appeals, then a tool or a specialist may earn its keep. Either way, you want to know which one you are before the demo, not after the contract.

The honest edge

An article can tell you that the write-off is the wrong number and that some of your denial work is costing more than it recovers. It cannot tell you which denial types those are in your practice, what your break-even is by payer, or where in your intake the most expensive errors start. That takes sitting with your remittance data and your workflow for a while.

If your billing team is busy all month and the denial rate does not move, the problem is probably not effort. Talk to us about where your rework is actually going, and whether the fix is a tool, a threshold, or a front desk that finally sees its own claims.