How we apply CMS's rules

Nothing on this site is typed in, scraped from another tool, or estimated. Each quarter we load CMS's own published files — the relative value file that carries every code's RVUs and payment indicators, the geographic practice cost indices, the ZIP-to-locality crosswalk, the NCCI add-on code edits, and the national payment amount files — normalise them, and commit the result. Pages are built from that committed copy, so the figure you see was computed from a file version we can name.

The two conversion factors are not a choice we made: since CY2026 CMS publishes one for clinicians who qualify as Advanced APM participants and one for everyone else. Every surface on this site lets you switch between them, and every figure states which one it used. Both are taken from the final rule linked in Check us against CMS.

A fee schedule amount is three relative value units, each scaled by its own geographic index, summed, and multiplied by the conversion factor. Work is the physician's effort, practice expense is the room, staff and supplies, malpractice is the liability premium share.

Worked on 64483 Njx aa&/strd tfrm epi l/s 1 — a single-level lumbar transforaminal epidural injection — in an office, at national average indices (every GPCI 1.000) and the standard conversion factor:

That figure is not written into this page — it is computed here by the same engine call /code/64483 makes, so the two always agree, and a data refresh updates both. Open it beside this page and check.

One exception to the formula. For 515 codes in this year's file, federal law caps the fee schedule amount at the hospital outpatient rate for that locality. We apply the cap from CMS's own published payment amount files rather than recomputing it, and the figure says so when it bites.

The geographic indices are why a fee schedule amount is not one number. Each index applies to its own component before the conversion factor, so a locality with expensive rent moves the practice-expense share without touching the work share. The same code, in an office, at the standard conversion factor:

A $122.00 spread on one injection, from the indices alone. Which locality applies to you is set by your practice's ZIP code, not your state — CMS's crosswalk splits several states into multiple localities and puts some counties on their own. Every locality control on this site takes a ZIP and resolves it against that crosswalk.

CMS attaches a set of single-character indicators to every row. They are the payment policy: whether the code is paid at all, what its global period covers, whether it splits into professional and technical components, and which modifiers may change its price. The definitions below are rendered from the same module that generates the indicator chips on every code page, so the two cannot drift apart.

A modifier only changes a payment when the code's own indicator says it can. That is the rule this site is built on: we never offer you a modifier the indicator refuses, and when a modifier is valid but changes nothing we say so rather than hiding it. Each card below shows the CMS field that decided the answer, the rule the engine cited when it applied it, and what it works out to.

−26 and −TC: not arithmetic

For a diagnostic test with professional/technical indicator 1, CMS publishes three separately valued rows: the whole test, the physician's interpretation, and the equipment-and-staff portion. So we do not split a percentage off the global amount — we price the row CMS published for the component you picked.

−50: bilateral

−51 and the multiple-procedure rule

When several procedures subject to the standard ranking are performed in one session, the highest-valued one is paid in full and the rest at half. The reduction is a property of the session, not of a single code, so it is the one rule that cannot be shown on a code in isolation.

The two page systems apply this differently, on purpose. A procedure page prices a fixed combination of codes, so it ranks them for you automatically. On the model page you are assembling lines that may or may not share a session, so the reduction is an explicit −51 chip per line — we do not guess at your operative sessions.

Second surgeons: −80/−81/−82, −62 and −66

−52 and −53: reduced and discontinued

CMS publishes a separately valued row for a discontinued procedure on just 4 codes, all colonoscopies. Where that row exists we price it, so 44388 Colonoscopy thru stoma spx is $353.72 completed and $177.02 with −53. Everywhere else, a reduced or discontinued service is priced by your Medicare contractor from your documentation. There is no published percentage, so we do not invent one — those modifiers are not offered rather than shown with a made-up figure.

An add-on code is one CMS says is never billed alone. We determine that from a single field: the global period. CMS assigns the value ZZZ — “add-on code; related care is included in the primary's global period” — to 981 codes in this year's file, and that is the site's only classifier. Every surface reads it, so a code's role never depends on which page you arrived from.

A second CMS file, the NCCI add-on code edits, answers a different and narrower question: which primaries a given add-on may accompany. It classifies nothing here. We use it only to populate the list of acceptable primaries — and for 367 of those add-on codes it names none, either because it publishes the pairing as contractor-defined or because there is no edit row for the code at all. Either way the acceptable primary is whatever your contractor accepts, and an add-on with no usable list is still an add-on. 61781 Scan proc cranial intra is one of them; its page says “contractor-defined” on the face of it rather than leaving you to wonder why the list is empty.

Pricing is then ordinary: the add-on is a line of its own, priced by the same formula, at the units you billed. What is not ordinary is that it escapes the multiple-procedure reduction — CMS gives add-ons multiple-procedure indicator 0, so an add-on is never the “second procedure” that gets halved.

One honest wrinkle: CMS's edit file lists billable pairs, and an add-on's acceptable primary is occasionally itself an add-on — the second level of a procedure that already has a first. A named primary is therefore not a promise that the primary is separately billable on its own.

Every estimate on this site is a fee schedule allowed amount: what Medicare recognises for a code, in a place, under a set of modifiers. Several things stand between that and money in an account, and we would rather you learn them here than discover them later.

Sequestration

The Medicare payment sequester reduces what is actually paid, not the fee schedule amount. Every figure here is the allowed amount before it.

NCCI edits and bundling

We price the codes you enter. We do not tell you whether they may be billed together — procedure-to-procedure edits, modifier-51 exemptions and medically-unlikely unit limits are coding questions, not pricing ones.

Mid-year budget neutrality

We ship the conversion factor as published in the final rule, and pick up corrections on the next data refresh. The retrieval date above tells you which version you are reading.

OPPS and ASC payment

The facility's side of the claim is a different fee schedule. We do apply the OPPS dollar cap where CMS caps a PFS code, but we do not price the facility's own payment.

Anesthesia

Anesthesia codes (status J) are paid on base units plus time against a separate anesthesia conversion factor. They carry no RVUs, so the formula on this page does not apply to them.

Beneficiary cost sharing

The allowed amount is the total Medicare recognises, before the Part B deductible and the patient's 20% coinsurance are split out of it.

Non-Medicare payers

Commercial and Medicaid rates are often quoted as a percentage of Medicare, but they are set by contract. Nothing here is a commercial rate.

The point of a page like this is that you do not have to take our word for it. These are the rules and the datasets themselves, at the exact file versions this copy of the site was built from.

Found a figure that disagrees with CMS's look-up tool? That is a bug and we want it — the feedback widget on any page reaches us directly.