If denials are consuming your team’s capacity while margins compress, I work through this with clients at HARRIS CareTracker every day.
Sixty five percent of denied claims are never resubmitted. That is the most revealing number in revenue cycle management, and almost no one treats it as the strategic warning it is.
It tells you something the industry doesn’t like saying out loud: the denial system doesn’t need to defeat you on the merits. It just needs to outlast you. Make the pursuit expensive enough, and a predictable share of providers will walk away on their own.
Most RCM companies measure their denial operation by overturn rate, and they’re proud when that number hits 60 or 70 percent. It sounds like strength. But overturn rate hides the one figure that actually determines whether your appeals operation makes money or quietly loses it: what each recovered dollar costs you to win.
Denials Are Rising, and the Reason Isn't Clinical
Start with what the data shows, because the numbers tell a story the merits can’t explain. Initial denial rates have climbed to roughly 15% of all claims, up from the 10 to 12% range of a few years ago. More than 90% of medical groups report denials rising year over year. And the growth is concentrated in prior authorization and eligibility friction, not medical necessity. These aren’t clinical judgments. They’re administrative obstacles.
When denied claims are appealed, roughly 85% are ultimately recoverable. If the overwhelming majority of appealed claims are winnable, then most denials were never about whether the claim was correct. They were about whether you’d do the work to prove it. A denial you’ll win 85% of the time on appeal isn’t a decision. It’s a dare.
The 65% Is the Business Model
If 85% of appealed claims are recoverable but 65% of denied claims never get resubmitted at all, the system isn’t built to win disputes. It’s built to discourage them. Industry estimates put the revenue providers abandon to denial write offs and underpayments at 20 to 40 billion dollars or more each year. That’s not money lost to invalid claims. Its money lost to claims nobody pursued.
A payer doesn’t need to defeat every provider. It needs a predictable share of providers to decide that chasing a denial cost more than the denial is worth. The friction does the work the merits couldn’t. High touch prior auth workflows, manual status inquiries, multi cycle documentation requests: each one raises the cost of pursuit until, for a meaningful slice of claims, the rational move is to write it off. Proving you’re wrong was never the point. Making being right too expensive to bother with is the entire play.
Whether any single payer engineers this deliberately is harder to prove and beside the point. The mechanism operates regardless of intent. Administrative friction produces predictable provider drop off, and that drop off is worth tens of billions a year. The behavior is the strategy, whether or not anyone drew it on a whiteboard.
Overturn Rate Measures Effort, Not Profit
A 70% overturn rate tells you how hard your team worked. It tells you nothing about whether that work paid off.
Consider what it actually measures and what it conceals. You win 70% of your appeals but only appeal 30% of your denials. Your overturn rate looks excellent while your true recovery rate is a fraction of what it implies. The metric counts your wins and silently ignores the denials you never touched, the labor each win consumed, the time to cash, and every claim you wrote off because it sat below your effort threshold.
Overturn rate is a vanity metric. It feels like a scoreboard. The financially serious numbers are the ones top operations already track net collection rate, cost to collect (which typically runs 3 to 8% of revenue), denial write off rate, and denials sorted by root cause into preventable versus non preventable. Cost per appealed dollar is the sharpest extension of cost to collect, applied at the level where the actual decisions get made.
The Math Makes This Concrete
Appeals aren’t free. The cost to rework a denied claim runs from roughly 25 to 118 dollars depending on complexity, and the cost to appeal specifically often lands between 40 and 100 dollars or more per claim, frequently across multiple touches. Hold those costs against the recovery and a threshold appears immediately.
Start with a strong, high value segment. Average denied claim value of 120 dollars, a 70 percent appeal success rate, and a 60 dollar cost to work each appeal. Across 10 appeals, you spend 600 dollars and recover seven wins at 120 dollars, or 840 dollars.
| High Value Segment | Results | ||
|---|---|---|---|
Claim Value | $120 | ||
Cost to work each appeal | $60 | ||
| High Value Segment | Results | ||
|---|---|---|---|
Recovered (7 wins of 10) | $840 | ||
Total appeal cost (10 claims) | $600 | ||
Cost per recovered dollar | $0.71 | ||
Seventy one cents spent to recover a dollar. Thin, but positive. Now change one variable. Same 60 dollar cost, same 70 percent win rate, but a 60 dollar claim instead of 120.
| Lower Value Segment | Results | ||
|---|---|---|---|
Claim Value | $60 | ||
Cost to work each appeal | $60 | ||
Recovered (7 wins of 10) | $420 | ||
Total appeal cost (10 claims) | $600 | ||
Cost per recovered dollar | $1.43 | ||
You spent a dollar forty three to recover a single dollar. You are losing money on every appeal in this segment, and your overturn rate is still 70 percent. The metric looks strong. The economics don’t. That crossover point, where pursuing a valid, winnable claim becomes economically irrational, is the most important number in your building. And almost nobody calculates it.
Yes, You Should Sometimes Write Off Valid Claims
This is the conclusion that makes people uncomfortable, so let me be precise.
A claim can be valid, winnable, and still irrational to pursue when the labor required to win it exceeds the recovery. Pursuing it isn’t diligence. It’s a voluntary transfer of your margin, recorded in your statistics as a win.
Disciplined RCM companies should consciously release the category of appeals where the labor math favors the payer. Not because those claims are invalid, but because pursuing them is precisely the behavior the friction was designed to extract. The skill isn’t surrender. It’s knowing exactly where your threshold sits, defending every claim above it with full force, and releasing the claims below it on purpose so your team’s capacity flows to the recoveries that actually move margin.
Appealing everything that looks winnable is how administrative friction converts your work ethic into someone else’s profit.
The Threshold Isn't Fixed
The payer’s entire advantage is the labor cost of an appeal. Collapse that cost and the threshold moves in your favor.
This is where the situation stops being bleak. The threshold is a direct function of what an appeal costs you, which means anything that lowers that cost pulls more claims back into rational, recoverable territory. Automated transactions run roughly 80 to 90 percent cheaper than their manual equivalents. A manual prior authorization costs 11 to 15 dollars per transaction and a manual status inquiry 7 to 10 dollars, and automation strips most of that away. Lower the cost of pursuit and the threshold moves downward, reclaiming claims that manual operations are forced to abandon.
Fragmented technology is where this war gets lost quietly. When denial identification lives in one system, documentation in another, payer rules in a third, and analytics nowhere at all, every appeal carries maximum labor cost. That pins your threshold exactly where the friction wants it. The levers that move it are concrete.
Front end prevention through eligibility verification and claims scrubbing. The cheapest appeal is the denial that never happens, and prior auth and eligibility are exactly where denials now concentrate.
Root cause analytics that separate systematic friction patterns from genuine errors, so you attack whole categories of denials at the source instead of working them one claim at a time.
Automated, structured denial management that compresses the time to identify, document, and submit an appeal, driving the per appeal cost down toward the automated benchmark.
Cost per appealed dollar reporting segmented by value band, payer, and denial reason, so the threshold stops being invisible and becomes a number you manage deliberately.
Together these levers reshape the math. Prevention shrinks the pool of denials. Analytics shows you which survivors are friction patterns worth attacking at the root. Automation cuts the labor cost on the appeals you do pursue. And the reporting makes the threshold visible, so you defend above it and release below it on purpose rather than by instinct.
You Can't Manage a Threshold You Can't See
Most RCM companies cannot even calculate their cost per appealed dollar. The data needed to find the threshold lives in disconnected systems that were never built to talk to each other. Denial data sits in one place, labor and cost data in another, recovery data in a third. You can’t opt out of a dynamic you can’t measure, so you appeal by habit, which is the precise behavior the friction rewards at your expense.
At HARRIS CareTracker, we work with RCM companies to build this kind of denial intelligence infrastructure. The platform consolidates eligibility verification, claims scrubbing, structured denial management, payment posting, and analytics into one cloud based environment, which means denial patterns, appeal labor, and recovery data live together instead of scattered across systems that hide the threshold. That consolidation is what makes cost per appealed dollar measurable. And a number you can’t measure is a margin you can’t defend.
Stop Measuring Wins. Start Measuring What Wins Cost You.
The denial economy depends on a single assumption: that providers will keep spending labor to chase claims right up to the point where chasing them costs more than the recovery, and then quietly write off the rest. Sixty five percent never resubmit. Tens of billions go uncollected. And the overturn rate on the wall measures, with uncomfortable accuracy, how faithfully an operation has played the role the friction assigned to it.
This isn’t cynicism. It’s arithmetic. A valid claim is not always a claim worth fighting. The most sophisticated RCM companies will be the ones that know exactly where their threshold sits, have the technology to push it lower, and have the discipline to manage to it.
Frequently Asked Questions
What is cost per appealed dollar?
Should an RCM company ever write off a valid claim on purpose?
How do you lower your cost per appealed dollar?
About the Author
Thomas Koehl is a 30 year health technology veteran and currently Director of Marketing at Harris CareTracker. Prior leadership roles at QRS Healthcare Solutions focused on supporting revenue cycle management partners. Following Hurricane Katrina, he served as Director of a large New Orleans medical clinic that delivered care to over 32,000 patients. Koehl has testified before the U.S. House Committee on Energy and Commerce as an expert witness on disaster healthcare delivery. He also volunteers as COO of International Medical Alliance, a nonprofit providing free medical care to impoverished communities in developing countries. He writes about the business, strategy, and human side of health technology for the practitioners and leaders living it day to day.
Follow me here for more breakdowns, and follow HARRIS CareTracker for product updates and resources.
