Underpayments and Overpayments
A denial announces itself. A variance does not. An underpayment and an overpayment point in opposite directions and share the property that makes both hard: each one arrives as a payment, posts without incident, and looks exactly like the process working.
Updated 8 min read
Reviewed by Anwaar Tayyab
Director of Billing Operations ·
On this page
Key takeaways
- Neither produces an error. The claim pays, the arithmetic balances, the account closes — the only signal is a comparison nobody is obliged to make.
- Finding either requires an external reference: the contracted rate. Without it, an underpaid claim is indistinguishable from a correctly paid one.
- An underpayment is not a contractual adjustment. Adjusting to the contracted rate is the agreement working; paying below it is the agreement not being honored.
- An overpayment is not your money, and identifying one carries obligations that vary by payer and program.
Why both are invisible, for the same reason
Most revenue-cycle failures come with a signal attached. A denial arrives with a reason code on a claim that did not pay. A rejection arrives in an acknowledgment report. An aging claim shows up in a bucket that grows. Each one gives somebody something to notice.
Variances give nobody anything. A claim underpaid by a third still pays, still posts, still closes its balance, and still appears in reporting as collected. An overpayment is even more thoroughly disguised: it arrives as extra cash. Nothing in the ordinary operation of a billing office is designed to be suspicious of money that showed up.
The signal has to come from outside the transaction
Underpayments: paid, and short
An underpayment is a claim paid at less than the contracted rate. The critical distinction is from the thing it most resembles — and it resembles it closely, because both are a gap between what was billed and what arrived. Finding one is where this article ends; contract-based underpayment recovery covers what it takes to get the money back, and why that is markedly harder in this direction than in the other.
| Contractual adjustment | Underpayment | |
|---|---|---|
| What it is | The billed charge reduced to the contracted allowed amount. | The allowed amount itself falling below the contracted rate. |
| Is it expected? | Yes. It is the agreement working exactly as written. | No. It is the agreement not being honored. |
| Is it a loss? | No. Nobody was ever going to pay it. | Yes — and it is revenue the contract entitled you to. |
| How you can tell | The gap is between the charge and the allowed amount. | The gap is between the allowed amount and the contract. You need the contract to see it at all. |
| What it looks like on the remittance | Normal. | Also normal. That is the entire problem. |
The last row is the point. There is no visual difference on a remittance between a claim adjusted correctly to the contracted rate and one allowed below it. The document reports what the payer decided; it does not report whether the decision matched the agreement.
Which means detection is a data problem before it is a billing one. Finding underpayments requires holding the contracted rates in a form that can be compared against each payment, systematically. A practice without that reference is not failing to catch underpayments — it has no mechanism by which catching one would be possible, except by accident. Run at the moment of posting, that comparison becomes a check on the size of each contractual adjustment before the account closes — validating the contractual adjustment against the loaded contract — so a payment allowed below the contract is held for review while the account is still open. And even with the reference, an underpayment that has already closed to a zero balance — often because the shortfall itself was cleared with an adjustment — falls off every worklist, since nothing is owed; catching that one is its own discipline, a zero-balance review of the accounts that look done.
Overpayments: money that is not yours
An overpayment is an amount received beyond what was owed. It can come from a duplicate payment, a payer's own processing error, two plans both paying as primary, a patient paying more than their final share, or a claim later adjusted downward.
It surfaces as a credit balance — an account showing more received than owed. And a credit balance is genuinely difficult to prioritize, because every incentive runs the wrong way: the cash has already landed, nothing is outstanding, no patient is calling, and no payer is chasing. It is the only item in the revenue cycle where doing nothing feels like a positive balance.
The obligation is real and it is specific
What is safe to say operationally is that credit balances need a review process rather than a queue that gets worked when someone notices — because nothing about a credit balance will ever make someone notice. Once one is confirmed, resolving it is its own procedure: refunding a credit balance to the payer or patient whose money it is, never writing it off.
Recoupment: when the payer takes it back itself
A payer that identifies its own overpayment may not ask for a check. It may recoup — recovering the amount by withholding it from a later payment. The money comes back out of the next remittance, against a completely different set of claims.
This is worth expecting, because it produces a deposit that does not match the claims on the remittance and looks exactly like an error. The recovery is not attached to any of those claims; it is reported separately, at the provider level, and the payment total the remittance states is the claim payments less that recovery. So the claims add up to one figure, the bank shows a smaller one, and both are correct — the difference was taken back against an older claim. Posting has to reflect the payment and the offset — recording the recovery against the older claim it belongs to rather than the current ones it arrived with, which is what Posting a Payer Offset is about — and a team that does not know recoupment is a normal payer behavior will spend real time investigating a discrepancy that is working as designed.
Which is why this lands on reconciliation
What the invisibility costs
Both variances distort the numbers the practice uses to judge itself, and they distort them in the direction of reassurance.
An underpaid claim counts as collected. It contributes to the net collection rate as a success and closes out of days in A/R on schedule. A practice with a systematic underpayment problem does not see a metric decline — it sees perfectly respectable metrics, computed from payments that were each quietly short. An overpayment does the reverse, inflating collections with money that will have to go back.
This is the honest limit of collection metrics
Common questions
How is an underpayment different from a contractual adjustment?
A contractual adjustment reduces the billed charge to the contracted allowed amount — that is the agreement working as written, and it is not a loss because nobody was ever going to pay it. An underpayment is the allowed amount itself falling below the contracted rate: the agreement not being honored, and revenue you were entitled to. Both appear as a gap between what you billed and what arrived, and there is no visual difference between them on a remittance. Only the contract distinguishes them.
Why don't our collection metrics show underpayments?
Because they cannot, by construction. The collection rates measure how much of what the payer decided you were owed was actually collected — they are computed from the payer's own numbers, so a claim underpaid by a third counts as fully collected and closes out of A/R on schedule. A systematic underpayment problem does not produce a declining metric; it produces respectable metrics built from payments that were each quietly short. Only comparison against the contract reveals it.
The deposit is less than the claims on the remittance add up to. What happened?
Check for a recoupment before treating it as an error. A payer that has identified its own overpayment may recover it by withholding the amount from a later payment rather than requesting a check. The recovery is reported separately from any claim on that remittance, at the provider level, and the payment total the remittance states is already the claim payments less that recovery — so the deposit is smaller than the claims alone while both the remittance and the deposit are correct. The difference was taken back against an older claim. Posting has to reflect both the payment and the offset.
We have a credit balance. Can we just apply it to the patient's next visit?
That is not a question this site can answer for you, and the reason is that it is a legal one rather than an operational one. An identified overpayment is not the provider's money, and the rules governing how it must be handled — and how quickly — vary by payer, by program, and by state, with federal programs imposing requirements a commercial contract does not. What applies to a specific credit balance belongs with your counsel, your contract, and the rules of the program involved. What is safe to say is that credit balances need a deliberate review process, because nothing about one will ever prompt you to look at it.
Key terms in this article
Defined once, on their own pages.
Continue learning
Where to go next.
Refunding a Credit Balance
What to do with the credit balance an overpayment leaves: confirm it, find whose money it is, and return it.
Payment Reconciliation
Where a recoupment is usually first noticed, and why.
From Billed Charge to Collected Dollar
The arithmetic a variance quietly breaks without producing an error.
Net collection rate
The metric that counts an underpaid claim as a success.
What Is a Claim Denial?
The failure that does announce itself — and why that makes it easier.
Authoritative sources
- 42 CFR 401.305 — Requirements for reporting and returning of overpayments (opens in a new tab)
eCFR. The Medicare Parts A and B rule: a person who has received an overpayment must report and return it; a person has identified an overpayment when the person knowingly receives or retains one; and the deadline that follows identification is suspended while an acknowledged OIG or CMS self-disclosure is pending.
- HHS-OIG — Health Care Fraud Self-Disclosure (opens in a new tab)
Office of Inspector General, U.S. Department of Health & Human Services. The protocol the Medicare report-and-return rule points to, by which an individual or entity subject to civil monetary penalties voluntarily discloses self-discovered evidence of potential fraud — and, in OIG's own terms, avoids the costs and disruptions of a government-directed investigation and civil or administrative litigation.
- Remittance Advice Resources and FAQs (MLN Booklet) (opens in a new tab)
Centers for Medicare & Medicaid Services. States that provider-level adjustments usually do not relate to any specific claim or service line, gives a deduction from payment for a prior overpayment as an example of one, and gives the check amount as the sum of all claim-level payments less the provider-level adjustments.
- HFMA MAP Keys — industry-standard revenue cycle KPIs (opens in a new tab)
Healthcare Financial Management Association. HFMA's own standardized key performance indicators for the revenue cycle, published with definitions, inclusions and exclusions, and the data source each one draws on — the set that includes net days in accounts receivable and net days in credit balance.

