Why 60% of denials are never appealed (even though 70-80% of appeals win)
The largest source of revenue leakage in outpatient billing isn't denied claims. It's denied claims your team never had time to fight. Here's the math and the playbook.
If you run a medical billing operation, you already know your denial rate. It’s somewhere between 9% and 15%. You probably can’t recite it to two decimals, but you’d guess close.
What you might not know, because the industry doesn’t talk about it loudly, is that 60% of those denials never get appealed. The biller looks at the worklist, sees a CARC code, and the claim sits.
Now stack the next number on top: when appeals are actually filed, 70 to 80% of them win.
That gap is where your revenue lives. Not in the denials. In the denials that became silence.
The math nobody wants to do
A small example. One clinic, 50 claims a day, 250 working days. 11.8% denied, the 2024 industry average, up from 10.2%. That’s about 1,475 denied claims a year.
At a $150 average claim value, that’s $221,000 a year in denied revenue per clinic.
Now apply the gap.
- Appeals filed today: 40% of denials × 70% win rate = $62,000 recovered
- Appeals filed in a world where you actually fight all of them: 93% × 75% = $154,000 recovered
The difference, about $92,000 per clinic per year, is sitting in a queue because nobody had the bandwidth to write the letter.
For a billing company with 25 clinics on its book, that’s $2.3M in recoverable revenue per year. None of it requires a new contract, a new EHR, or a new client. It’s already yours.
Why your team isn’t filing them
The honest answer is capacity. But that’s not the only answer. Three other things are happening:
1. Worklist sort order. Denials come back in your EHR sorted by date. That’s the worst possible sort. The denial worth $1,200 is buried next to the $42 one with the same CARC code. By Friday afternoon, your appeals specialist is making decisions based on which claim is on the top of the screen, not which is worth fighting.
2. Appeal letter ergonomics. The biller has to open the payer’s policy PDF (if they can find it), scan to the relevant section, draft an argument that cites it, copy that into a Word doc, and upload it through a portal that doesn’t autosave. Average time per appeal: 22 minutes. Most billers can do six of these in a workday, and only if nothing else interrupts.
3. Tribal knowledge. The senior biller knew which Aetna denials were soft (file once, win), which BCBS denials needed a peer-to-peer, which Medicaid denials were a waste of time. She left in March. The patterns went with her.
What “closing the gap” actually looks like
The instinct is to hire. Another biller, another appeals specialist. The math doesn’t work, at $25/hour fully loaded, you’d need to recover ~$50K/year per new hire just to break even. And the throughput cap is still 30 claims per person per day.
The real lever is change the ratio of work-per-appeal. Reduce the per-appeal cost from 22 minutes to 4. Suddenly the same team is filing four times as many appeals, and the recovery curve shifts.
That’s where AI agents earn their keep, not by replacing a biller, but by collapsing the prep work that makes a biller useful only six times a day. A denial classifier reads the CARC and the EOB, looks at the historical win rate for that code at that payer, ranks the claim by expected recovery. An appeal drafter pulls the right paragraph from the payer’s policy text, drafts the letter with the citation already attached, hands it to the biller for review. The biller becomes a reviewer instead of a writer.
That’s a 5x throughput change, not a 1.2x change. It’s what closes the 60% gap.
What to measure
If you want to know whether this leak exists in your operation today, three numbers tell the story:
- Appeal-to-denial ratio. How many appeals filed last month divided by denials received last month. National average is around 40%. Anything below 50% is leaving money on the table.
- Average days from denial to appeal. If it’s over 14, you’re risking the appeal window on the denials you do file. CARC 197 windows can be as short as 30 days.
- Win rate by CARC code. If you have this in a spreadsheet you’re ahead of 80% of your peers. If you don’t, the first month of work is just building it.
Those three numbers are enough to know whether you have a $50K leak or a $500K leak. Either way, it’s worth a Tuesday afternoon to find out.
Stats in this post: 11.8% industry denial rate from 2024 industry reports (up from 10.2%). 60% of denials never appealed and 70-80% appeal win rate are widely cited in MGMA and AHIMA literature.