Modifier 26 and TC: Billing One Service in Two Halves

Some services described by a single code contain two economically distinct things. One is a physician's cognitive work: reading the study and producing the report that a treating clinician will act on. The other is everything required to produce the study in the first place — the machine, its maintenance, the consumables, the technologist's time, and the room it all sits in.

Those two can be furnished by one organization or by two. An imaging center that owns the scanner and employs the radiologist furnishes both. A hospital that owns the scanner while an independent radiology group reads the images furnishes one each. The component modifiers exist so that a single code can describe the same service in either arrangement, with each party claiming what it did.

The professional component

The physician work — the interpretation and the written report. Claimed with 26. The report is not incidental to this component; it is most of what the component is.

The technical component

The resources that produced the study — equipment, supplies, staff time and the space. Claimed with TC. The party that bore those costs is the party that bills it.

The global service

Both halves, furnished by the same entity, billed on one line with no component modifier. This is an assertion, not a default.

This is the point at which most component-billing errors are already determined, before anyone chooses a modifier. Whether a code has a professional and a technical component is a property recorded against that code in the fee schedule, alongside the other payment policy indicators that decide how a published amount is adjusted for a particular line.

The indicator distinguishes several situations that a biller has to treat differently: codes that genuinely split into two payable halves; codes that exist only as a global service and cannot be divided; codes that are professional work only, with no technical half to claim; codes that are technical only; and codes for which the concept does not apply at all. How a modifier changes adjudication owns that mechanism in general, and explains why this site does not publish the indicator values themselves — they are per code, revised annually, and reading them from the current file is the only correct way to know.

The same reasoning repeats for every payment policy that has an indicator, which is why learning it once is worth more than memorizing modifier pairings. The bilateral procedure modifier works the same way — four situations rather than five, and a documented disagreement between two authorities about the resulting claim form on top.

One consequence is worth stating separately, because it is the error that survives longest without being noticed: the place of service reported on a component line describes where that component happened. The technical half happened where the equipment is. An interpretation produced somewhere else did not happen there. Place of service on professional claims owns the field itself, and the place of service code lookup lists the codes; the component-specific point is that splitting a service can split its setting too.

The arrangement that causes the most trouble is the one where a practice orders a study, has it performed or read by someone outside the practice, pays that party, and then bills the payer for the component itself. This is a real and often legitimate arrangement. It is also specifically regulated, and the regulation removes the thing that usually motivates it.

42 CFR 414.50 (opens in a new tab) addresses a physician or other supplier billing for a diagnostic test performed or interpreted by a physician who does not share a practice with them. Where it applies, payment to the billing supplier is the lowest of three amounts.

1. The performing supplier's net charge to the billing supplier.
2. The billing supplier's own actual charge.
3. The fee schedule amount that would have been allowed if the performing supplier had billed the payer directly.

Two details in the regulation do most of the work. The net charge must exclude “any charge that is intended to reflect the cost of equipment or space leased” between the parties — which forecloses recovering the margin through a lease running the other way. And “share a practice” is a defined term with two limbs: one turning on the share of a physician's professional services furnished through the billing entity, the other on being an owner, employee or contractor performing the service in the billing entity's own office location.

It is easy to think of the technical component as a machine and a technologist, which makes billing it feel like a question about equipment ownership. Two conditions in 42 CFR 410.32 (opens in a new tab) make it a question about the service instead.

1. 01

   It has to have been ordered by the treating physician

   Diagnostic tests “must be ordered by the physician who is treating the beneficiary … and who uses the results in the management of the beneficiary's specific medical problem.” A test ordered by someone not managing the problem, or whose result nobody uses, fails at the order rather than at the modifier — and no component modifier repairs that.
2. 02

   It has to have been furnished under the right supervision

   The regulation sets three levels. General supervision means the procedure is furnished under the physician's overall direction and control, but the physician's presence is not required. Direct supervision means the physician is present in the office suite and immediately available to furnish assistance and direction. Personal supervision means a physician is in attendance in the room while the procedure is performed.
3. 03

   Which level applies is a fact about the code

   The required level is carried per code in the fee schedule, next to the component indicator. So two studies performed on the same afternoon in the same room can carry different supervision requirements, and whether the practice met each one is part of whether it can bill the technical half.

• Both parties bill globally. Each furnished one half and each omitted its modifier. The first line paid stands, the second denies as a duplicate, and the party that lost the race has to correct rather than appeal.
• A component modifier on a code that does not split. The fee schedule says the code has no such component; the modifier does not create one. The fix is upstream — check the indicator before building the charge, not after the denial.
• The technical half billed where supervision was not met. Nothing on the claim reveals this, which is precisely why it survives until an audit reads the schedule against the requirement.
• A purchased component billed at the practice's own rate. The payment limitation applies whether or not the practice applied it, so the difference is not revenue — it is an overpayment waiting to be identified.
• The place of service left at the practice's default. The technical component happened where the equipment is. A default that says otherwise is a false statement on a claim about a fact the payer can check.