Modifier 91: The Narrowest Repeat Rule in Lab Billing

Modifier 91 has the narrowest defensible use of any modifier in routine laboratory billing, and CMS says so in unusually plain language. The Medicare Claims Processing Manual instructs contractors, when telling laboratories that these modifiers exist, to emphasise that they have very narrow application — and adds that any evidence of excessive use will be referred to the Program Integrity Unit for further review. The same passage opens by stating that improper use of these modifiers is likely to indicate a fraudulent or abusive circumstance.

That is not boilerplate. It is a description of how modifier 91 volumes are actually monitored, and it should shape how the decision is made.

The Decision Rule

The governing text is Chapter 16, Section 100.5.1 of the Claims Processing Manual, and it is worth reading as a set of gates. Modifier 91 is appropriate when:

1. The test was performed more than once on the same day for the same patient.
2. It was necessary to obtain multiple results in the course of treatment. This is the substantive test, and it is narrower than “the test was repeated for a good reason.” The clinical purpose must be the series of values itself — serial cardiac markers, serial potassium during correction, glucose monitoring across a treatment course — not a single reliable answer.
3. The test is paid under the clinical laboratory fee schedule. The manual states twice that these modifiers may be used only for laboratory tests paid under that schedule.
4. No standard code already describes the series. Where a HCPCS code exists for the sequence of results — the manual gives glucose tolerance tests and evocative/suppression testing as examples — that code is the answer, and modifier 91 is not.

The three exclusions, stated as exclusions

Section 100.5.1 lists what modifier 91 may not be used for, and these three account for most misuse:

• Tests rerun to confirm initial results. A confirmatory repeat produces one reportable answer, not multiple results in the course of treatment.
• Repeats due to testing problems with specimens or equipment. A haemolysed sample, a recollection, an instrument fault, a quality-control failure — all of these are the cost of producing the first result.
• Any other reason where a normal, one-time, reportable result is all that is required. This is the catch-all, and it is the sentence to apply when the first two do not quite fit. If one good answer was the clinical goal, modifier 91 does not apply however many times the analyser ran.

Read together, these mean the question is never “was the repeat justified?” It is “was the second value itself clinically needed?” A laboratory that repeats a test because the first run was untrustworthy has done the right thing scientifically and has still not generated a billable repeat.

Panels, Components and the NCCI Edits

The most common legitimate use of modifier 91 is not a standalone repeat at all — it is a repeated component of a panel, and understanding it explains a large fraction of laboratory bundling denials.

The NCCI Policy Manual (Chapter 1, Section N, and Chapter 10, Section C) sets out the structure. Where a laboratory performs all the tests in a CPT-defined organ or disease-oriented panel, it reports the panel code rather than the components — and the panel codes are ordered as panels. The NCCI program then contains edits pairing each panel code, as the Column One code, with each individual component code as a Column Two code, so a component reported alongside its own panel normally denies.

Those edits allow NCCI PTP-associated modifiers to bypass them where one or more of the individual tests is repeated on the same date of service. The repeat testing must be medically reasonable and necessary, and modifiers 59 or 91 may be used to report it. The manual then points straight back at Chapter 16, Section 100.5.1 for the necessity standard — so the panel-component bypass does not relax the rule, it inherits it.

Concretely: a comprehensive metabolic panel drawn in the morning and a repeat potassium at midday during active correction is the pattern the bypass exists for. The panel is reported as a panel; the repeated component is reported separately with modifier 91.

Note also that modifier 91 is an NCCI PTP-associated modifier — the Policy Manual lists it alongside 27, 59, XE, XS, XP and XU — so unlike modifier 76, it genuinely can clear a bundling edit with a Correct Coding Modifier Indicator of 1. That asymmetry between two superficially parallel “repeat” modifiers is one of the more useful facts in this area.

What the Record Must Show

Modifier 91 assertions are tested against the ordering record and the result record together, because the laboratory’s own worksheet cannot establish clinical necessity on its own.

• A separate order for each reported instance, with a time. A single order with a repeated run behind it is one test with a rerun; two timed orders reflecting a clinical decision to obtain a further value are two tests.
• Times on the results. The whole justification rests on the sequence, so the sequence has to be visible. Two identically-timestamped results supporting a modifier 91 claim will not survive review.
• Clinical documentation of why the further value was needed. Titration of a therapy, monitoring during correction, a serial protocol, a change in the patient’s condition. The strongest records show the second result being acted on.
• Nothing indicating a technical rerun. This is the practical risk in a laboratory information system, because instrument and QC records are exactly where reruns get logged. A modifier 91 claim whose underlying LIS record reads “repeat — sample integrity” or “rerun after QC failure” contradicts itself. Sites with mature CLIA quality control processes generate excellent evidence of when a repeat was technical — which is useful for quality and equally useful for showing which repeats are not billable.

The Misuse That Triggers Denials and Program Integrity Referral

1. Confirmatory repeats. Excluded expressly.
2. Repeats after specimen or instrument problems. Excluded expressly. This is the highest-volume error in real laboratories because it is the highest-volume kind of repeat.
3. Modifier 91 where a series code exists. Glucose tolerance and evocative/suppression testing are the named examples; billing components with modifier 91 instead of the series code misstates what was done.
4. Modifier 91 on tests not paid under the clinical laboratory fee schedule. Outside its scope by the manual’s own terms.
5. Modifier 91 in place of laterality or component modifiers. Contractor guidance notes it does not replace RT, LT, 50 or the digit modifiers, and it has nothing to say about the professional/technical split — see modifiers 26 and TC.
6. High-volume routine use. Even where individual claims are arguable, the manual’s referral language means aggregate volume is itself a signal. A laboratory whose modifier 91 rate is materially above peers should expect the question.

One adjacent mechanic worth knowing for dialysis billing: for ESRD automated multi-channel chemistry tests, the Claims Processing Manual directs that line items with modifier 91 used in conjunction with the CD, CE or CF modifiers be included in the 50/50 rule calculation. Modifier 91 does not remove a test from that arithmetic.

Modifier 91 vs. Its Nearest-Confused Siblings

Modifier 91 vs. modifier 76 — the one that matters

Both mean “this was done more than once today,” and they are strictly domain-separated.

	Modifier 91	Modifier 76
Applies to	Clinical laboratory tests paid under the clinical laboratory fee schedule	Procedures and services generally — surgeries, imaging, injections
Necessity standard	Necessary to obtain multiple results in the course of treatment	The repeat was clinically necessary
Concerned with who repeated it?	No	Yes — modifier 77 is used where a different practitioner repeated it
Bypasses an NCCI PTP edit?	Yes — it is an NCCI PTP-associated modifier	No — the manual states 22, 76 and 77 are not
Technical rerun	Expressly excluded	Expressly excluded (equipment failure)

Contractor guidance closes the loop from both directions: the modifier 76 page states that it is not appropriate with laboratory codes and that modifier 91 should be appended instead. If you are choosing between them, the fee schedule the test is paid under decides it, not the clinical narrative.

Modifier 91 vs. modifier 59

Section 100.5.1 names both as available for same-day repeat laboratory testing, which is why they are so often conflated — but modifier 91 is the specific instrument for clinical laboratory fee schedule tests, and general NCCI guidance directs that the most specific applicable modifier be used rather than modifier 59. Where a repeat lab test is what happened, modifier 91 says so precisely; modifier 59 says only “distinct,” and invites the reviewer to ask distinct in what way. See modifier 59 and the X{EPSU} subset modifiers for the full hierarchy.

Modifier 91 vs. modifier 90

Different question entirely. Modifier 90 (reference or outside laboratory) identifies who performed a test the billing entity did not perform itself. Modifier 91 says how many times. They can coexist on a claim and neither substitutes for the other.

Modifier 91 vs. modifier QW

QW identifies a test as CLIA-waived, performed by an entity holding a valid certificate of waiver. It speaks to regulatory authorisation to perform the test, not to repetition or payment adjustment. A waived test repeated for genuine serial monitoring could carry both. CASRAI covers the underlying framework in CLIA certification, CLIA-waived point-of-care testing and high-complexity testing requirements.

Modifier 91 vs. modifiers 26 and TC

Clinical laboratory fee schedule tests generally have no professional/technical split to make — that is what the PC/TC indicators for laboratory physician interpretation codes encode. So a modifier 91 test is almost never also a 26 or TC question. Anatomic and molecular pathology, which do split, are paid differently and sit outside modifier 91’s scope.

Where Jurisdiction Matters

Section 100.5.1 is national, and it is unusually prescriptive, so there is less room for jurisdictional divergence on the core rule than elsewhere in the modifier set. What varies is enforcement posture and local coverage: contractors differ in their frequency edits for specific analytes, in the Local Coverage Determinations governing how often a given test is covered — and where a repeat is expected to breach a frequency limit, the question stops being modifier 91 and becomes an ABN and modifier GA question, in what documentation they request on review, and in how aggressively they act on utilisation outliers. Novitas, Noridian and First Coast Service Options each publish their own modifier 91 material, and their laboratory LCDs are not identical.

Non-Medicare payers commonly accept modifier 91 but apply their own frequency limits and medical-policy criteria, and some require the repeat justification in the claim note rather than on request. Verify against the adjudicating payer. Where no guidance is available, the manual’s own framing — multiple results genuinely needed in the course of treatment, never a rerun, never where a series code exists — is the durable structure.

Frequently Asked Questions

When can modifier 91 be used?

When a clinical laboratory fee schedule test is performed more than once on the same day for the same patient and it was necessary to obtain multiple results in the course of treatment — and no standard code describes the series.

Can modifier 91 be used when a test is repeated to confirm a result?

No. Confirmatory reruns are one of the three exclusions stated in Chapter 16, Section 100.5.1, along with repeats due to specimen or equipment problems and any situation where a single reportable result is all that is required.

What about a repeat because the specimen was haemolysed?

Not billable with modifier 91. Testing problems with specimens are expressly excluded.

Can modifier 91 be used on a panel component?

Yes, and this is its most common legitimate use. Where a panel was performed and one component is medically necessary to repeat on the same date, the component may be reported with modifier 91 appended, bypassing the NCCI edit that pairs the panel with its components.

What is the difference between modifier 91 and modifier 76?

Domain. Modifier 91 is for clinical laboratory fee schedule tests; modifier 76 is for repeated procedures and services generally, and contractor guidance for modifier 76 explicitly redirects laboratory repeats to modifier 91. They also behave differently against NCCI edits: modifier 91 is an NCCI PTP-associated modifier and modifier 76 is not.

Does modifier 91 apply to tests paid under the Physician Fee Schedule?

No. The manual restricts these modifiers to laboratory tests paid under the clinical laboratory fee schedule.

How much modifier 91 use is too much?

There is no published threshold, but the manual directs contractors to emphasise the modifier’s very narrow application and to refer evidence of excessive use to the Program Integrity Unit. Treat a modifier 91 rate materially above peer laboratories as a question to answer internally before someone else asks it.

Related CASRAI Resources

• Modifier 76: repeat procedure by the same physician — the non-laboratory counterpart
• Modifiers 26 and TC — and why most clinical lab tests have no component split
• Modifier 59 and the X{EPSU} subset modifiers
• CLIA quality control requirements — where technical reruns are properly recorded
• CLIA certification and CLIA-waived point-of-care testing
• Point-of-care testing
• Laboratory developed tests and FDA regulation
• Modifier GA: waiver of liability and the ABN — the instrument for a lab test expected to exceed an LCD frequency limit
• Modifier KX: what you are attesting to
• The False Claims Act in billing

Sources: CMS Medicare Claims Processing Manual, Publication 100-04, Chapter 16, Sections 100.5.1 and 100.6; CMS National Correct Coding Initiative Policy Manual, Chapter 1, Sections E and N, and Chapter 10, Section C (revision date 1/1/2026); Noridian Healthcare Solutions modifier 91 and modifier 76 guidance. CPT is a registered trademark of the American Medical Association; descriptors are summarised here rather than reproduced. General reference material — verify against your MAC’s current guidance and the payer’s own policy.