Here is a number that deserves more attention than it gets: U.S. providers spent more than $25 billion fighting claim denials in 2023, up 23 percent year over year, according to figures cited in HFMA's January analysis of payer AI and denials. That is not the value of denied claims. It is the administrative cost of arguing about them - the letters, the records pulls, the re-submissions, the hold time.
HFMA has since launched an initiative to build a universally accepted definition of a clean claim, with the explicit goal of eliminating tens of billions in related administrative waste. The premise is simple: if payers and providers agreed on what a complete, accurate claim looks like, most of the back-and-forth would never start. That standardization will take years. Your organization's clean claim rate is something you can measure this month.
What a clean claim audit actually looks like
Most billing teams know their denial rate. Far fewer know their first-pass yield - the share of claims that go out complete, correct, and payable on first submission. The audit is straightforward:
- Sample a month of submissions - 200 to 500 claims across payers and service lines.
- Grade each claim against the payer's own requirements: eligibility verified, authorization attached, codes supported by documentation, demographics and payer IDs correct.
- Separate errors from rejections. A claim bounced by a clearinghouse edit never became a denial - but it cost a day. Count both.
- Trace each defect to its origin. Registration, eligibility, authorization tracking, coding, or charge entry. The origin, not the denial code, is where the fix goes.
The output should be a short list, not a long one. If the audit produces twenty findings, the sample was probably graded inconsistently, or the organization has deeper process problems than one quarter can absorb. Most teams find three to five root causes explaining the large majority of defects. That concentration is the good news: three process fixes, each with a named owner, usually move first-pass yield within a single quarter. Start with eligibility - it is the most common root cause in most audits and the cheapest to fix.
Teams that run this for the first time typically find a small number of root causes producing most of the rework - and that most of them live upstream of the billing office.
Why the $25 billion keeps growing
The cost of fighting denials grows faster than claim volume for two reasons. First, payers adjudicate faster and deny more precisely: the same HFMA analysis reports initial denial rates around 11.65 percent for 2025, per Kodiak Solutions data, with some rejections arriving within seconds of submission. Every automated denial triggers a manual response somewhere.
Second, denial work is labor that scales with complexity, not volume. A new payer policy, a new prior auth list, a new documentation requirement - each one multiplies the effort per claim. Staffing cannot keep up with a cost curve like that, which is why the number rose 23 percent in a single year.
There is also a selection effect at work. Claims that would once have been denied and abandoned now generate appeals, because organizations increasingly track recovery rates and realize the money is winnable. Fighting denials is becoming more common precisely because it pays - which means the cost curve keeps rising until prevention shrinks the pool being fought over. Every claim cleaned upstream is a claim that never joins that contest, at any price. The $25 billion figure will keep climbing until the industry takes that trade seriously; your first-pass yield is the one variable fully under your control.
The takeaway
The $25 billion is an industry-level number, but it is paid practice by practice, in biller hours and delayed cash. The clean claim audit is the cheapest intervention in revenue cycle: it costs a week of analysis and tells you exactly where your share of that waste originates. Fix the top three causes and the rework queue shrinks on its own. That upstream discipline - verifying eligibility, confirming authorization, grading claims before submission - is precisely the layer QHB's agents automate, so the $25 billion problem stops being your problem one claim at a time.