CO-97: bundling denials and when a modifier is honest

What does CO-97 bundled into another service mean?

CO-97, payment adjusted because this service/procedure is included in the payment/allowance for another service/procedure that has already been adjudicated, is a Claim Adjustment Reason Code that says the payer is not refusing to pay for the service — it is refusing to pay for it a second time 1Ref 1X12 (2026).Claim Adjustment Reason Codes.That CARCs, including CO-97, are the standard X12 code list explaining why a claim or service line was paid differently than billed, used here as the definition of what CO-97 signals: the service is already paid for inside another adjudicated line.. Somewhere else on the same claim, or on a claim already processed for that date, a different code already covers it.

That framing matters because it changes what kind of problem you're looking at. A CO-97 is not a documentation gap and it is not a coverage dispute; it is the payer's bundling logic doing exactly what it was built to do. Most CO-97s are correct, and the work is confirming that, not fighting it.

How NCCI edits decide what's bundled

For Medicare and most payers that follow its lead, the National Correct Coding Initiative publishes procedure-to-procedure edits that name which code pairs cannot both be paid for the same patient on the same date by the same provider 2Ref 2Centers for Medicare & Medicaid Services (2026).NCCI for Medicare.That NCCI procedure-to-procedure edits define which code pairs Medicare will not pay together, each carrying a modifier indicator of 0, 1, or 9 that governs whether any modifier can separate the pair, used here to explain both the bundling mechanism and the modifier-59/XE/XS/XP/XU distinction.. Each edit pairs a column-one code with a column-two code the column-one code is presumed to already include.

Every PTP edit also carries a modifier indicator — 0, 1, or 9 — that decides whether a modifier can ever separate the pair. A 0 means never: no modifier unbundles that combination, full stop. A 1 means a modifier can apply if the clinical facts genuinely support it. A 9 means the edit doesn't apply to that pair at all. Looking up the specific pair in the public NCCI edit files, before assuming a modifier will fix anything, tells you which of the three situations you're actually in.

Reading the RARC before you decide anything

The Remittance Advice Remark Code riding alongside a CO-97 often names the specific code the service was bundled into, supplying the one detail the CARC alone doesn't carry 3Ref 3X12 (2026).Remittance Advice Remark Codes.That RARCs supply the supplemental explanation on an 835 remittance beyond the CARC, maintained by X12 with a public list, used here as the read that names the specific code a CO-97 was bundled into.. Pulling that code and checking its NCCI modifier indicator is the fastest way to know whether you have any option here at all.

A remark code that names a bundling partner with a modifier indicator of 0 closes the question immediately — there is nothing to appeal and nothing to resubmit, because no combination of documentation changes an edit that never allows separation. A remark code pointing at a modifier-indicator-1 pair is the one worth a second look, since that is the only category where the two services might genuinely have been distinct.

When a modifier honestly separates two services

A modifier is the correct tool only when the two bundled services were actually distinct in fact — a different anatomic site, a separate patient encounter, or a procedure unrelated to the one it's paired with — not when the goal is simply getting a denied line paid. Modifier 59 and its more specific successors, XE (separate encounter), XS (separate structure), XP (separate practitioner), and XU (unusual, non-overlapping service), each assert a specific kind of distinctness, and the claim has to actually support whichever one is used 2Ref 2Centers for Medicare & Medicaid Services (2026).NCCI for Medicare.That NCCI procedure-to-procedure edits define which code pairs Medicare will not pay together, each carrying a modifier indicator of 0, 1, or 9 that governs whether any modifier can separate the pair, used here to explain both the bundling mechanism and the modifier-59/XE/XS/XP/XU distinction..

The test worth applying honestly: would the documentation stand on its own if a payer's reviewer pulled the note and asked why these weren't one service? If the answer requires stretching the facts, the modifier is unbundling, not honest coding, and the safer move is accepting the CO-97 as correct rather than manufacturing a distinction the record doesn't support.

MUE denials are a related but different animal

A Medically Unlikely Edit caps the number of units of one code that can be billed for one patient on one date — a unit-count ceiling, not a code-pair rule — and CMS publishes those values separately from the PTP edit files 4Ref 4Centers for Medicare & Medicaid Services (2026).Medically Unlikely Edits.That a Medically Unlikely Edit caps the units of service one provider reports for one patient on one date, published separately from the PTP edit files, used here to distinguish a unit-cap denial from a CO-97 code-pair bundling denial.. A claim that exceeds an MUE denies for a different reason than a CO-97, even though both trace back to the same NCCI program.

Confusing the two wastes time: appealing a unit-cap denial as though it were a bundling dispute, or vice versa, means arguing the wrong point to the payer. Check which edit type actually produced the denial before drafting anything, since the fix for one — reducing units billed to the allowed cap — does nothing for the other.

The CARC on the remittance usually settles which one you're looking at: a CO-97 names another service as the reason, while an MUE-driven denial reads more like a units-exceeded rejection on the line itself. When both a bundling edit and a unit cap could plausibly explain a denied line, the RARC is the tiebreaker, since it's built to state which specific rule fired.

Commercial payers bundle by their own rules too

Medicare's NCCI edits are public and specific, but a commercial payer's own bundling logic is set by that payer's own published policy and is not guaranteed to match NCCI pair for pair. UnitedHealthcare, for one example, publishes its reimbursement policies on its own provider portal, and that published policy — not a Medicare edit file — is what actually governs a UnitedHealthcare claim 5Ref 5UnitedHealthcare (2026).UnitedHealthcare Policies and Protocols.That UnitedHealthcare publishes its own reimbursement policies on its provider portal — cited only as UnitedHealthcare's own named example of a commercial payer setting bundling rules that are not guaranteed to match NCCI edit for edit, never as what all payers do..

That means a pair that unbundles cleanly under Medicare's modifier indicator can still bundle under a different commercial payer's own edit, and the reverse is just as true. Checking the specific payer's own policy before assuming NCCI logic transfers directly is worth the few minutes it costs, because your contract with that payer, not a Medicare rule, is what actually controls the claim.

Building the habit that prevents most CO-97s

Most CO-97 denials are predictable before submission, because the code pair that triggered them was already sitting in a public edit file. Running a claim through a scrubber that checks NCCI pairs before it goes out catches the combination in seconds, at a point where fixing it costs nothing.

When a CO-97 does land despite that check, treat it first as a confirmation the bundling logic worked as designed, and reserve the modifier-and-appeal path for the specific, documented cases where the two services really were separate. That discipline keeps the rare honest unbundle from getting lost in a pile of appeals that were never going anywhere.

A short reference sheet of the code pairs your own practice bills most often, with each pair's modifier indicator noted next to it, turns this from a lookup you repeat every time into a one-glance check at the point of coding — before the claim goes out, not after the remittance comes back.