Understanding Revenue Cycle: Beyond Denial Rates Analysis

Every revenue cycle meeting starts with the same chart. Denial rate, month over month, usually trending down if the practice has invested in front-end verification and prior authorization staff. Everyone nods. The number looks good.

And it’s telling you almost nothing about whether the practice is actually being paid what it’s owed.

A denial is a specific, visible event: a payer rejects a claim outright and sends back a reason code. But not all revenue leakage appears as a denial. Some losses show up as paid claims reimbursed below the contracted amount. Others appear later as recoupments. Others never appear at all, because the service was never billed.

That’s the uncomfortable part: a clean denial rate does not mean a healthy revenue cycle. It tells you how often claims are being rejected outright. It says much less about claims that were quietly underpaid, reversed months later, or never submitted in the first place.

Three kinds of invisible money

Paid, but underpaid. A claim adjudicates. The payer issues a remittance. No denial appears anywhere on it. And the amount is still wrong, because the allowed amount does not match the contracted reimbursement rate for that code, that payer, that plan. Nothing about the claim looks abnormal in the practice management system — it shows as paid. The only way to catch the gap is to compare what was actually paid against what the contract says should have been paid, line by line, which most billing workflows are not built to do automatically.

Paid, then taken back. A recoupment reverses money that already sat in the practice’s revenue for a period of time — sometimes through an offset against a future payment that has no obvious connection to the original claim. A biller reconciling this month’s deposits against this month’s claims can easily miss the link back to a service rendered months earlier, because the reversal doesn’t read as a denial. It reads as a routine payment adjustment.

Never billed at all. A service may never become a claim because it wasn’t billed, because it was held during an unresolved authorization or documentation question, or because it was submitted after the payer’s timely filing deadline. Because there’s no accepted claim to reconcile against, this is the hardest category to detect retrospectively — there’s simply no record for anyone to audit.

Why the reporting architecture can’t see it

None of this is necessarily a diligence failure. Denial reporting commonly relies on CARC and RARC codes — the standardized reason codes that explain claim adjustments and denials. That architecture is useful for answering “why was this claim rejected.” It is not, by itself, a complete answer to “was this claim paid the right amount?” Contract information may sit in PDFs, amendments, and fee schedules rather than in a system that can automatically compare it against remittance data. So the comparison simply doesn’t happen unless someone builds and maintains a process for it.

The blind spot isn’t theoretical. Medicare itself reimbursed hospitals at roughly 83 cents for every dollar of cost in 2024, resulting in more than $100 billion in hospital underpayments, according to the American Hospital Association. That’s a cost-to-reimbursement gap, not the same thing as a commercial contract underpayment — a different mechanism entirely.

The clearest illustration of where that money hides comes from Ensemble Health Partners, a revenue cycle firm that works with hospitals and health systems on underpayment recovery. In a Becker’s Hospital Review discussion of its 2022 client results, Ensemble reported recovering more than $200 million in underpayments for clients, and said 70 percent of that amount was identified outside the standard payment-variance report. In other words, most of the money wasn’t sitting where the standard report was looking.

Meanwhile, the industry’s attention stays concentrated on the visible half of the problem. Experian Health’s 2025 State of Claims survey found that 41% of providers reported denial rates of 10% or higher, with missing or inaccurate data, authorization issues, and incomplete patient registration information among the leading causes. That’s a real and measurable problem. But it’s also the part of the revenue cycle that the standard denial dashboard is specifically designed to capture — which means attention, staffing, and software keep flowing toward denial management while underpayment, recoupment, and unbilled-service leakage compound quietly underneath it.

What actually surfaces this money

For a medical practice, none of this requires a new department. It requires redirecting existing reconciliation work toward comparisons that currently aren’t being made.

Build an expected-versus-actual comparison at the line level, not the claim level. A claim can show as “paid” while individual CPT lines within it were reimbursed below contract. Modeling each payer’s fee schedule and running remittances against it automatically turns underpayment from something a biller might notice by instinct into something a report flags by exception.

Track recoupments separately from denials, tied to the original date of service, not the date the money was taken back. That’s the only way to connect a reversal months later to the encounter that generated it, and to check whether the underlying audit finding was even correct.

Watch adjustment codes, not just denial codes. CARC 97, for example, indicates that a procedure or service is not payable separately — a legitimate outcome much of the time, not automatically an error. What’s worth reviewing is a repeated pattern on the same code combination, since that’s where a genuine misapplication of a bundling edit would show up.

Review downcoding as a pattern, not an assumption. If a practice’s 99214 claims are consistently reimbursed at 99213-level amounts for one payer and not others, that pattern may warrant review of the remittance, the coding, and the documentation before it’s escalated to the payer — not a conclusion that the payer got it wrong.

And periodically sample authorization-adjacent decisions that never became claims. This is the hardest category to instrument, because there’s no claim to pull. It usually means going back through scheduling and documentation for a defined period and checking, service by service, whether something that should have been billed wasn’t.

For practices that do not have the staff or systems to perform this analysis routinely, an external revenue-cycle review can help identify recurring payment variances and organize the documentation needed for payer follow-up. WCH supports practices with this type of review.

The number that actually matters

None of this argues for ignoring denials — Experian’s data makes clear that’s real, worsening work. But a practice that has driven its denial rate down while never modeling contracted reimbursement, never separating recoupments from new denials, and never checking the gap between what was scheduled and what was billed has optimized the metric it can see and left the larger one unmeasured.

The better question for a year-end revenue cycle review isn’t “what’s our denial rate.” It’s “what share of contracted revenue did we actually collect — across everything we were owed, paid claims included.” That number is harder to produce. It’s the number that tells a practice whether the green dashboard is describing reality, or simply describing the part of reality the dashboard was built to track.

Sources