What CPT Category II codes are

CPT Category II codes are four digits followed by the letter F. The AMA calls them "a set of supplemental tracking codes that can be used for performance measurement." They report that something was done or that a test came back at a particular value. They do not report a billable service.

Because they describe work already bundled into an evaluation and management service, they carry no relative value units. A Category II code on a claim is worth exactly $0.00. The AMA is explicit that using them is optional, that they are not required for correct coding, and that they may never substitute for a Category I code.

That is the mechanism, and it is also why they are useful. A Category II code is a free field on a claim you were already submitting, and it carries a clinical result that the claim otherwise leaves invisible to the plan.

$0.00
What a Category II code bills for. No RVUs, no payment, no effect on the rest of the claim.
2 codes
What a blood pressure encounter needs: one for the lowest systolic reading, one for the lowest diastolic.
1P, 2P, 3P, 8P
The only four modifiers the AMA permits on a Category II code.

Why they close HEDIS gaps without a chart chase

A HEDIS numerator can be met two ways. Administrative data, meaning claims and encounters, is scored automatically. Everything else has to arrive as supplemental data or come out of a chart, and both cost money.

Chart retrieval is the expensive end: a vendor requests the record, an abstractor reads it, and for a hybrid measure NCQA can ask for primary source verification of the whole process. A supplemental data feed is cheaper per record but has to be built, mapped, validated against NCQA's standard data requirements, and maintained by someone.

A Category II code skips both. It rides the claim the practice was already sending, so the result lands in the plan's administrative data with no new feed and no retrieval request. That is the difference between an A1c of 6.8 that closes a gap in February and an A1c of 6.8 that gets discovered by an abstractor in October, or never.

People often describe Category II codes as "the cheapest supplemental data." Strictly speaking they are not supplemental data at all. They are administrative data, which is what makes them cheap.

The Category II codes by HEDIS measure

These are the codes quality teams actually use. Each one reports a result or a completed action, and each one maps to a measure a Medicare Advantage plan is graded on.

MeasureCodeWhat it reports
Controlling High Blood Pressure (CBP) and Blood Pressure Control for Patients with Diabetes (BPD)3074FSystolic less than 130 mm Hg
3075FSystolic 130 to 139 mm Hg
3077FSystolic 140 mm Hg or higher
3078FDiastolic less than 80 mm Hg
3079FDiastolic 80 to 89 mm Hg
3080FDiastolic 90 mm Hg or higher
Glycemic Status Assessment for Patients with Diabetes (GSD)3044FHbA1c below 7.0%
3051FHbA1c at least 7.0% and below 8.0%
3052FHbA1c at least 8.0% and no more than 9.0%
3046FHbA1c above 9.0%
Eye Exam for Patients with Diabetes (EED)2022FDilated retinal exam read by an ophthalmologist or optometrist, with retinopathy
2023FDilated retinal exam read by an ophthalmologist or optometrist, without retinopathy
2024F7 standard field stereoscopic photos, with retinopathy
2025F7 standard field stereoscopic photos, without retinopathy
2026FValidated eye imaging, with retinopathy
2033FValidated eye imaging, without retinopathy
3072FNo retinopathy on the prior year's exam
Transitions of Care (TRC), medication reconciliation post-discharge1111FDischarge medications reconciled with the current outpatient medication list
Care for Older Adults (COA)1159FMedication list documented in the medical record
1160FReview of all medications by a prescriber or clinical pharmacist
1170FFunctional status assessed
Advance care planning1123F, 1124F, 1157F, 1158FAdvance care plan or surrogate documented, or the discussion documented

The nine Category II measure ranges run 0001F through 7025F: composite measures, patient management, patient history, physical examination, diagnostic and screening results, therapeutic and preventive interventions, follow-up outcomes, patient safety, and structural measures, plus a separate non-measure reporting range at 9001F to 9007F. The codes above sit mostly in the history and results ranges, which is where result values live.

How to report them so they actually count

The reporting rules are small and easy to get wrong, and getting one wrong quietly voids the code.

  • Report the date of the test, not the date you read it. For A1c and eye exam codes, the date of service is the date of the result. If a lab drawn in March is reviewed at a June visit and coded June, the measure sees June. The exception is 3072F, which uses the current-year office visit date.
  • Send two codes for blood pressure. One for the lowest systolic reading of the encounter and one for the lowest diastolic. A systolic code alone does not close CBP.
  • Bill at $0.00. Or $0.01 if the billing system refuses to accept a zero-dollar line. Either way the code does not change what the claim pays.
  • Report medication reconciliation where it happened. 1111F is reportable whether the reconciliation was done by phone or during the transitional care management office visit.
  • Document both COA medication codes. 1159F and 1160F cover different things, and the review under 1160F has to be performed by a prescribing practitioner or a clinical pharmacist.

The four modifiers, and why they do not close a gap

The AMA allows exactly four modifiers on a Category II code, and no others: 1P for a medical reason the service was not performed, 2P for a patient reason, 3P for a system reason, and 8P for an action not performed with no reason specified. They exist so a physician quality program can distinguish a legitimate exclusion from a miss.

They also do the opposite of what a gap-closure workflow wants. A code carrying 1P, 2P, 3P or 8P reports that the thing did not happen. It will not put the member in a HEDIS numerator. Staff who append a modifier out of habit are documenting a failure in a field the plan reads.

The AMA further restricts them: the exclusion modifiers are only intended to be used when the parenthetical notes or reporting language for that specific measure allow it.

Common mistakes

What goes wrongWhat it costs
Only the systolic code is sentCBP and BPD both need a systolic and a diastolic code. The gap stays open
A1c coded to the visit date instead of the result dateThe result can land outside the measurement window
Habitual 8P on codes where nothing was excludedReports the action as not performed. The member is scored as a miss
A Category II code used in place of a Category I codeThe AMA prohibits it and the encounter goes unbilled
Codes suppressed by a clearinghouse scrubber as "non-payable"The line never reaches the plan. Worth checking before blaming the practice
Eye exam coded by the primary care office as the retina specialist's examAllowed. Any provider may report the EED code, using the date of the test
Waiting for a supplemental data feed to carry a result the claim could have carriedTurns a free administrative hit into a build-and-validate project

How Pelica handles Category II coding

The blocker is rarely willingness. A practice will happily add 3074F if someone tells it, at the moment of the encounter, that the member's CBP gap is open and that the reading in front of them closes it.

Pelica ingests claims, EHR, pharmacy, lab and ADT feeds into one live member record, scores the open gaps continuously, and pushes the specific code that closes each one into the practice's workflow before the claim goes out. The Quality & Stars copilot runs the same logic across every payer at once, so a practice sees one list rather than six portals. Customers have seen a 41% improvement in quality gap closure.

Related terms

HEDIS measures covers the measure set these codes feed. ECDS explains the reporting method that is replacing hybrid chart abstraction, and why claims data matters more every year. CBP and GSD are the two measures where Category II coding moves the rate the most.

Sources