Repeat Procedure Modifiers: Which Edit Are You Answering?
When a second line carrying the same code as the first is rejected, the reflex is to add a modifier and resubmit. That works only if the modifier speaks to the edit that actually stopped the claim, and three different systems produce a rejection that looks the same from the outside. A repeat modifier answers exactly one of them — and for the bundling edits, it does nothing at all.
Updated 9 min read
On this page
Key takeaways
- A repeat modifier asserts that the same service was genuinely performed again. It is an answer to duplicate logic, which matches on code, date, and provider.
- It does not bypass an NCCI procedure-to-procedure edit. CMS states this in terms: the repeat modifiers are not PTP-associated modifiers, and using them does not bypass a PTP edit.
- The repeat modifier for a laboratory test is on the PTP-associated list even though the two procedure repeat modifiers are not — three modifiers that all mean “again”, split across that boundary.
- For a unit-limit edit, the route exists only where the edit is adjudicated per claim line. Where the edit is a date-of-service edit, reporting the code on separate lines does not help.
- Which repeat modifier applies turns on who performed the second service and whether it was a procedure or a laboratory test — not on why the repeat happened.
- None of this is a way to get a real duplicate paid. The record has to show the second service happened.
Three edits that look identical from the outside
A claim line can be stopped because the payer thinks it already has that service, because policy says the code should not be reported alongside another code on the claim, or because the number of units exceeds what the code is expected to carry. All three arrive as a line that did not pay. They are different mechanisms with different remedies, and a modifier chosen for the wrong one is simply ignored.
| The edit | What it tests | Do repeat modifiers reach it? |
|---|---|---|
| Duplicate logic | Whether this line matches an earlier claim or line on the elements the payer matches on — typically the same code, the same date, the same provider. | Yes. This is the question a repeat modifier exists to answer. |
| NCCI procedure-to-procedure edits | Whether two different codes on the claim are a pair that policy says is not reported together. | No. CMS states that the repeat modifiers are not PTP-associated modifiers and that using them does not bypass a PTP edit. |
| Medically Unlikely Edits | Whether the units of service reported for one code exceed the value that code carries. | Sometimes. Only where the edit is adjudicated as a claim line edit rather than across the whole date of service. |
The middle row is the one that costs time. A repeated service and a bundled pair feel like the same complaint — “the payer thinks this is already covered” — and they are not the same edit, so the same modifier cannot serve both.
Read the remittance before choosing the modifier
Which repeat modifier, and what decides it
There are three, and the choice between them turns on facts about the second service rather than on the reason it was needed. That is worth stating plainly, because the instinct is to select a modifier that explains why — and none of these do.
- Modifier 76 — the same practitioner performed it again
- The identical procedure or service, repeated by the practitioner who performed it the first time, on the same date. The assertion is that this is a second performance, not a second submission of the first.
- Modifier 77 — a different practitioner performed it again
- The same procedure, repeated on the same date by someone other than the practitioner who did it first. The distinction from 76 matters because the payer's duplicate matching includes the provider, so the fact pattern the claim is asserting is different.
- Modifier 91 — a repeated clinical laboratory test
- The laboratory case, and the one governed differently from the other two. Where a laboratory repeats a component test as a medically reasonable and necessary service on the same date, the repeat is reported with this modifier appended. Its scope is the repeated test, not a repeated procedure.
The asymmetry worth memorizing
The unit-limit case, and the lookup it depends on
Where a code's reported units exceed the value assigned to it, whether the claim can say “these were separate performances” depends on how that edit is adjudicated — and that is a per-code field, not a judgment. The files carry an adjudication indicator for every code.
Look up how the edit adjudicates
One value means the edit is applied per claim line. The others mean it is applied across the whole date of service. The route described below exists only for the first.Where it is a claim line edit, separate lines are available
CMS states that appropriate use of the distinct-service, repeat, and anatomic modifiers may be used to report the same code on separate lines of a claim, and that each line is then separately adjudicated against the code's value. The mechanism is not that the modifier raises the limit; it is that it produces more than one line for the limit to be applied to.Expect contractor-level restrictions on top
The same passage adds that claims processing contractors have rules limiting the use of these modifiers with some codes. So a route that exists in national policy can still be closed for a particular code by the contractor adjudicating the claim, which is a lookup against that contractor rather than an inference.Where it is a date-of-service edit, this does not apply
The absolute date-of-service edits are set where units above the value would be contrary to statute, regulation, or subregulatory guidance. Splitting the same units across lines does not change the total for the date, and the edit is applied to the date.
Two indicators, two different questions
What a repeat modifier does not do
The modifier is an assertion about what happened, submitted on a claim, and nothing at the moment of payment verifies it. That makes it exactly as strong as the record behind it and no stronger — the general point How a Modifier Changes Adjudication makes about the whole family, and it applies here with a specific edge: a repeat is one of the easiest assertions to make and one of the hardest to reconstruct afterward if the documentation does not show the second service separately.
- It does not fix a resubmission. Where the first claim was wrong and a corrected version is being sent, the mechanism is a replacement claim, not a second claim with a repeat modifier on it. Sending the correction as a new claim is what creates the duplicate in the first place.
- It does not describe why the repeat was needed. Medical necessity for the second performance is a separate question, answered by the record and by the payer's coverage policy, not by the modifier.
- It does not turn a re-do into a billable repeat. Repeating work because the first attempt did not produce a usable result is a different fact pattern from performing a service again because the patient needed it again, and the two are not interchangeable on a claim.
- It does not survive a record that shows one service. The assertion is that something happened twice. If the documentation describes it once, the modifier is unsupported — worse than absent, because an unsupported modifier is a claim the practice cannot stand behind on review.
Common questions
Our claim hit a bundling edit. Will a repeat modifier get it through?
No. CMS states the point directly: the repeat modifiers are not procedure-to-procedure-associated modifiers, and using them does not bypass a procedure-to-procedure edit. Those edits are about two different codes reported together, which is a different question from whether one code was performed twice. The modifiers that can participate in a bundling exception are a defined list, and whether an exception is available at all for a given pair is a lookup against that pair's correct coding modifier indicator before any modifier is chosen.
What is the difference between modifier 76 and modifier 77?
Who performed the second service. Modifier 76 asserts the same practitioner repeated it; modifier 77 asserts a different one did. The distinction is not cosmetic, because duplicate matching typically includes the rendering provider — so the two modifiers describe genuinely different fact patterns and the claim should carry whichever is true. Neither says anything about why the repeat was clinically necessary.
Can we split units across lines to get past a unit-limit edit?
Only where that code's edit is adjudicated as a claim line edit, which is recorded per code in the published files. Where it is, CMS states that appropriate use of the distinct-service, repeat and anatomic modifiers may report the same code on separate lines and each line is adjudicated separately against the value. Two cautions come with it: contractors have rules limiting the use of these modifiers with some codes, and where the edit is an absolute date-of-service edit, splitting lines changes nothing because the limit applies to the date.
Is modifier 91 just the laboratory version of modifier 76?
Functionally similar, procedurally not, and the difference catches people. Modifier 91 is on CMS's list of procedure-to-procedure-associated modifiers; modifiers 76 and 77 are named as not on it. So the laboratory repeat modifier can participate in a bundling exception where the pair's indicator allows one, and the two procedure repeat modifiers cannot participate at all. Anyone reasoning from the lab case to the procedure case will reach the wrong conclusion.
The payer denied our second line as a duplicate. Do we appeal or resubmit?
It depends which of the three situations produced it, and that decision belongs to the duplicate-denial article rather than to modifier selection. In outline: a true duplicate is correctly denied and there is nothing to appeal; a correction sent as a fresh claim is fixed by resubmitting it as a replacement; and only where two genuinely distinct services were read as one repeat is there something to argue — where the stronger route is usually a corrected claim carrying the modifier and documentation that distinguishes them, rather than an appeal.
Key terms in this article
Defined once, on their own pages.
Continue learning
The neighboring modifier questions, and the denial this one answers.
The X Modifiers: Saying Which Kind of Distinct
The bundling case a repeat modifier cannot reach, and the indicator lookup that comes before choosing one.
Modifiers That Break Out of a Global Period
The staged, returned-to-theater and unrelated cases — a different family for a different question.
How a Modifier Changes Adjudication
Why an unsupported modifier is worse than a missing one, and what a payment modifier can and cannot invoke.
Duplicate Claim Denials
The denial this modifier answers — and the two situations where it is not the remedy.
Coding, Modifiers & Edits
The cluster: what modifiers assert, which edits read them, and what the record has to show.
Denial Code Decoder
Look up what a remittance code is telling you before deciding which edit fired.
Authoritative sources
- CMS — Medicare NCCI Policy Manual, Chapter I: General Correct Coding Policies (opens in a new tab)
Lists the procedure-to-procedure-associated modifiers (the anatomic set; global surgery modifiers 24, 25, 57, 58, 78 and 79; and 27, 59, 91, XE, XS, XP and XU), and states that modifiers 22, 76 and 77 are not among them and do not bypass a procedure-to-procedure edit. Defines the correct coding modifier indicator values, and the unit-limit adjudication indicator values — where an edit adjudicates per claim line, appropriate use of the distinct-service, repeat and anatomic modifiers may report the same code on separate lines, each separately adjudicated, subject to contractor rules limiting those modifiers for some codes. Also states the laboratory case: a component test repeated as a medically reasonable and necessary service on the same date is reported with the repeat laboratory modifier appended.