Modifier 22 and Increased Procedural Services

Modifier 22 sits in a specific regulatory list. Under 42 CFR 414.40 (opens in a new tab), CMS establishes uniform national definitions of services, the codes that represent them, and the payment modifiers to those codes — and paragraph (b)(3) enumerates the payment-modifier policies it must maintain: assistant-at-surgery, multiple surgery, bilateral surgery, split surgical global services, team surgery, and unusual services. Unusual services is the last item, and it belongs there. It is a payment modifier with a national policy behind it, exactly like the others.

What makes it unlike the others is what the national policy turns out to say. Each of the other categories on that list resolves, eventually, to an instruction a computer can carry out. Bilateral surgery resolves to an indicator on the code and an arithmetic rule. The professional and technical split resolves to a division of the relative value units that were already computed. Assistant-at-surgery resolves to a per-code indicator and a ceiling expressed as a fraction. In each case the modifier changes a number, and the system knows which number and by how much before the claim arrives.

That is worth sitting with, because it inverts the usual question. Practices ask what modifier 22 pays, and the honest answer is that the question has no form. There is no fraction, no multiplier, no indicator value, and no national amount — not because CMS declined to publish one, but because publishing one would contradict the instruction that no relative values may be established for these services at all. A number would defeat the mechanism.

Every other property of this modifier follows from that. Individual pricing needs something to price against, so documentation has to arrive with the claim rather than be available on request. Individual pricing is done by a human, so the line does not finalize on the same schedule as its neighbors on the same claim. And individual pricing has one predictable failure — nothing to read — which is why CMS was able to specify in advance what the remittance says when that happens.

The manual's requirement is short and it is a list of two, not one. The biller must provide a concise statement about how the service differed from the usual, and an operative report with the claim. Practices reliably supply the second and skip the first, which is the wrong way round: the operative report describes the operation, and the reviewer's question is not what happened but what about it exceeded the code.

The concise statement

A short account of how this service differed from the usual service the code describes. It is the only document in the pair that answers the question actually being asked, and it is the one most often missing. It is not a restatement of the procedure; it is the comparison — what was present here that is not contemplated by the code, and what that required.

The operative report

The contemporaneous record that substantiates the statement. It carries no weight on its own for this purpose, because a reviewer reading an operative note without being told what to look for is being asked to construct the practice's argument for it.

The NCCI Policy Manual sets the substantive bar for when any of this is appropriate, and the bar is higher than “this took longer”. The modifier is not to be reported unless the services performed are substantially more extensive than the usual services included in the procedure the reported code describes. Extra time on its own is evidence of something; it is not the assertion. The assertion is about the content of the work relative to what the code already contemplates.

Before any of the substantive questions arise, there is a mechanical one that is easy to miss and settles the matter. CMS instructs that modifier 22 should only be reported with procedure codes that have a global period of 0, 10 or 90 days.

The global period is a per-code field in the fee schedule, and its other values are not surgical periods at all — they are markers for codes the concept does not apply to, including add-on codes and codes with no global period of their own. So a diagnostic study that was genuinely harder than usual, or an add-on code reported alongside a difficult primary procedure, is outside the modifier's stated scope regardless of the merits. The field decides availability, and it was set for reasons that have nothing to do with this case.

This is also the cleanest example of a rule worth generalizing beyond this modifier: the fee schedule records decisions about a code long before anyone treats a patient under it, and several of those decisions are dispositive of what a claim may assert. How a modifier changes adjudication sets out why the indicator rather than the modifier is what decides, and global period modifiers covers what the period itself governs.

This is the most common misuse, and the correct-coding manual forecloses it in a single sentence. Modifier 22 is not an NCCI procedure-to-procedure-associated modifier, and its use does not bypass a PTP edit. Neither do the repeat-procedure modifiers. The set that can move an edit is enumerated in the manual, and this modifier is expressly excluded from it.

The reason is structural rather than punitive. A bypass modifier makes a claim about separateness — a different encounter, a different anatomic site, a different specimen — which is the only thing that can answer an edit asserting two codes describe one episode of work. Modifier 22 makes a claim about magnitude, and magnitude is not responsive to that assertion. More work on one procedure is still one procedure.

What the manual does describe is a different route, and it is genuinely useful and almost never reported accurately. Each edit carries a correct coding modifier indicator — where its value forbids any bypass, no modifier will separate the pair, and that is the end of the question about the second code. But where the work genuinely qualifies as unusual, the manual says the physician may report the pair's surviving Column One code with modifier 22, and the contractor may then evaluate whether the unusual service justifies additional payment.

The manual's own worked example is the operating-microscope code, 69990, whose payment CMS limits to a list of procedures. Where it is reported alongside two other codes and one of them is off that list, an edit prevents payment for it — and, as the manual observes, claims systems cannot tell which of the two procedures the microscope was actually used for. That is a case where the edit is correct on the information available and wrong on the facts, and the route the manual offers is a resubmission to the contractor with the modifier attached, so a person can resolve what the edit could not see. NCCI procedure-to-procedure edits covers the edits themselves and how the indicator is read.

Buried in the chapter's adjudication instructions is something unusual: CMS specifies, for contractors, the exact remittance advice messaging to return when a claim carries modifier 22 and no additional documentation. It names the group code, the claim adjustment reason code, the remark code, and the beneficiary-notice message, and notes that the combination conforms to an industry operating-rules scenario.

The codes themselves are 252 with remark N706, returned under the contractual-obligation group code. Their published wording belongs to the code-set maintainers and is not reproduced here, but their effect is plain enough: the adjustment says an attachment was required to adjudicate the line and was not there. It is an information failure, not a judgment about the case.

The corollary is that a practice cannot tell, from the fact of a modifier 22 line being unpaid, whether its argument failed. Two entirely different things produce an unpaid line here — the documentation never arrived, or it arrived and did not persuade — and only one of them is worth an appeal. The group and reason codes distinguish them, which is the practical reason to read the line rather than the claim.

It does not say what modifier 22 pays, because nothing does. The instruction that contractors may not establish relative values for these services is precisely a statement that no rate exists to publish, and any figure offered elsewhere is either one contractor's practice or somebody's recollection. The honest planning assumption is that the amount is unknown until it is adjudicated.

It also does not say how long the review takes, how often additional payment is granted, or what any particular contractor asks for beyond the two documents the manual names. Those vary, they are not published as national policy, and a number here would be read as a standard.

Everything above is Medicare's national policy for the physician fee schedule. A commercial plan may recognize the modifier, may price it under its own rules, may require the documentation through a different channel, or may not accept it at all — and where the agreement incorporates a payer's own payment policies by reference, the answer is in that document rather than in this one. Reading a payer contract covers finding the term that governs.

Finally, it does not resolve the question of whether the work should have been reported with a different code entirely. Where the service performed is not the service the code describes — rather than a more extensive version of it — the question is a coding one, and unlisted procedure codes covers what happens when no code fits.