CLINICAL UPDATE JUNE 2026

CLINICAL UPDATE — JUNE 2026: CMS finalized the CY 2026 PFS with G2211 reimbursement holding at the national average of $16.05. Modifier -25 stacking edits (MM13272) remain active with no new exceptions beyond the CY 2025 preventive-service carve-out. Three MACs — Novitas (JL/JH), NGS (JK/J6), and First Coast (JN) — have published updated LCD companion articles clarifying that structured attestation language is now a preferred (though not mandated) documentation pattern for G2211 medical review. Scribing.io's engine was updated in March 2026 to align attestation templates with these LCD companion articles. Additionally, new HIPAA 2026 patient consent requirements for ambient AI scribes took effect January 1, 2026 — all Scribing.io attestation workflows include compliant consent verification checkpoints. For California-based practices, SB-1120 utilization review provisions now require audit trail retention for AI-generated billing recommendations; Scribing.io Pro plan includes this by default.

Automating CPT G2211: Reclaiming Relationship Revenue — The Clinical Operations Playbook

TL;DR

CPT G2211 adds an average of $16.05 per Medicare office/outpatient E/M visit — yet most primary care organizations miss it on the majority of eligible encounters because manual charting fails to establish a discrete, auditable longitudinal care relationship attestation. CMS's own FAQ explains what G2211 is but leaves critical implementation gaps: no guidance on EHR-level documentation mechanics, no guardrails for telehealth modifier placement, no global-period suppression logic, and no MAC-specific edit-pack strategy. This playbook closes every gap. It details how Scribing.io's AI documentation engine verifies longitudinal eligibility by scanning 12–18 months of encounter history under the same TIN/NPI, inserts a discrete attestation into Epic or Cerner/Oracle Health, blocks ineligible pairings, applies telehealth and surgical-period nuance correctly, and recovers five- and six-figure annual revenue that manual workflows leave on the table — with zero incremental compliance risk.

Table of Contents

1. Why G2211 Revenue Leaks at Scale: The $16.05-Per-Visit Blind Spot

2. What CMS Explains — and What It Leaves Out

3. Scribing.io Clinical Logic: From 780 Missed Claims to 68% Compliant Uptake

4. Technical Reference: ICD-10 Documentation Standards for E11.9 and I10

5. EHR Integration Architecture: Epic, Cerner/Oracle Health, and FHIR Provenance

6. Telehealth, Global Periods, and Modifier Stacking: The Denial-Prevention Layer

7. MAC-Specific Edit Packs and Payer Variance Matrix

8. ROI Model: Subscription Cost vs. Recovered Revenue

1. Why G2211 Revenue Leaks at Scale: The $16.05-Per-Visit Blind Spot

The arithmetic is deceptively simple. At the 2026 Medicare Physician Fee Schedule national average of $16.05 for G2211, a four-provider primary care site seeing 150 eligible office/outpatient E/M visits per week leaves approximately $125,190 per year uncaptured if the add-on code is never appended. Current clinical benchmarks indicate that fewer than 35% of eligible primary care encounters carry G2211 in practices relying on manual attestation workflows — not because the relationship doesn't exist, but because the documentation proving it is absent, ambiguous, or stored in a format that medical reviewers cannot locate during post-payment audit.

The root cause is structural, not educational:

Medical directors who audit Q1 claims routinely discover that the gap is not a coding knowledge problem — it is an infrastructure problem. The documentation system doesn't prompt, verify, or enforce the attestation that CMS medical reviewers look for. This is the problem Scribing.io was engineered to solve.

Key takeaway: G2211 revenue recovery is not a training initiative. It is an automation initiative. The longitudinal care relationship exists in practice; it simply doesn't exist as a discrete, machine-readable, audit-visible data element in the chart — until now.

For organizations navigating the expanding regulatory landscape around AI-assisted documentation, understanding the latest HIPAA 2026 patient consent requirements is essential before deploying any ambient AI scribe at scale.

2. What CMS Explains — and What It Leaves Out: The Implementation Traps That Cause Denials

CMS's G2211 FAQ (updated through the CY 2025 PFS final rule cycle) is the authoritative conceptual document. It correctly establishes that:

• G2211 is an add-on to O/O E/M codes 99202–99205 and 99211–99215.

• It captures the inherent complexity derived from the longitudinal practitioner-patient relationship.

• No specific diagnosis is required.

• No additional medical record documentation requirements have been "specified."

That final point — Q7 in the CMS FAQ — is where the implementation trap opens. CMS states it has not mandated specific documentation, then immediately notes that "our medical reviewers may use the medical record documentation to confirm the medical necessity of the visit and the patient care relationship." This creates a paradox that competitors and CMS itself leave unresolved:

There is no required documentation format, but there is a required documentation outcome — and if your chart doesn't achieve it, the claim fails audit.

The Five Gaps CMS's FAQ Does Not Address

Why This Matters for Medical Directors

Competitors — including well-intentioned coding education vendors — reproduce the CMS FAQ's conceptual framing and stop. They explain what G2211 means. They do not build the operational logic that prevents the five denial scenarios above from occurring at scale. The result: practices that "know about" G2211 but still fail to capture it compliantly on the majority of eligible visits.

Scribing.io's original contribution is not explaining G2211. It is engineering the verification, attestation, blocking, and modifier-placement logic that makes compliant capture automatic. The AI documentation engine doesn't just "suggest" G2211 — it proves the relationship exists, documents it in a reviewer-locatable format, and suppresses the code when any of the five denial traps apply.

3. Scribing.io Clinical Logic: From 780 Missed Claims to 68% Compliant Uptake

This section walks through the exact scenario a medical director encounters — and how the Scribing.io engine resolves it in production.

The Scenario

A primary care medical director at a multisite clinic audits Q1 claims and discovers that 780 eligible office/telehealth follow-up visits for patients with diabetes (E11.9) and hypertension (I10) carried no G2211 because the notes contained no longitudinal relationship attestation. At $16.05 per code, that represents $12,519 in lost revenue — in a single quarter, from a single diagnosis pair, across a single organization.

The visits were legitimate. The relationships were real. The providers had been managing these patients' chronic conditions for years. But the chart note said "Follow-up for diabetes and hypertension. Medications reviewed. Continue current regimen." It did not say anything a MAC reviewer could point to as evidence of a longitudinal, continuous care relationship serving as the focal point for the patient's health needs.

The Live Pilot: Step-by-Step Engine Logic

Pilot Results: Month Two

• G2211 uptake rate: 68% of eligible E/M encounters — up from 0%.

• MAC denials on G2211: Zero. Not one claim line denied across Novitas JL jurisdiction.

• Claims correctly suppressed: 4 global-period visits, 12 visits with modifier -25 stacking risk, 3 visits billed under 99211.

• Recovered revenue run-rate: At 68% uptake across the practice's eligible volume, the annualized recovery exceeds $33,000 — covering the Scribing.io subscription cost for every provider on the Basic plan within the first 11 days of each month.

The 32% of eligible visits that did not receive G2211 fell into defined exclusion categories: new patients without sufficient encounter history (below the 12-month threshold), visits where the provider explicitly declined the attestation during ambient documentation review, and encounters where the engine identified a competing longitudinal relationship (patient primarily managed by a specialist at a different TIN for the same condition). These exclusions are by design. Compliant suppression is as important as compliant capture.

4. Technical Reference: ICD-10 Documentation Standards for E11.9 and I10

G2211 is diagnosis-agnostic — CMS does not require a specific ICD-10 code to justify the add-on. However, in practice, the two diagnoses most frequently present on the encounters where G2211 is missed are E11.9 - Type 2 diabetes mellitus without complications; I10 - Essential (primary) hypertension. These are the workhorses of primary care chronic disease management and the conditions most commonly triggering longitudinal care relationship complexity.

E11.9 — Type 2 Diabetes Mellitus Without Complications

I10 — Essential (Primary) Hypertension

These documentation standards are not theoretical. They are the specific data points Scribing.io's NLP engine evaluates when constructing the longitudinal care relationship attestation. A note that mentions "diabetes" without specifying type, omits medication changes, and lacks lab trending produces a weaker attestation — and the engine will flag it for provider review before inserting G2211 on the claim.

5. EHR Integration Architecture: Epic, Cerner/Oracle Health, and FHIR Provenance

G2211 automation fails if the attestation exists in a PDF attachment or a free-text blob that medical reviewers cannot query. The attestation must be structured, discrete, and EHR-native. Scribing.io achieves this through three integration pathways:

Epic Integration

• SmartData Element (SDE) Write: The longitudinal care relationship attestation is written to a dedicated SDE within the encounter note. This SDE is queryable in Epic Reporting Workbench, Caboodle, and Cogito, enabling retrospective compliance auditing at scale.

• SmartPhrase Auto-Insertion: A clinician-facing SmartPhrase containing the attestation text is inserted at the assessment/plan section of the note. The provider reviews and attests (or declines) during note finalization.

• FHIR R4 Provenance Resource: A Provenance resource is created linking the attestation to the specific Encounter, Condition (E11.9, I10), and CarePlan resources. This creates a machine-readable audit trail that survives EHR version upgrades and data migrations.

• BPA (Best Practice Advisory) Hook: For encounters where the engine detects G2211 eligibility but the provider has not yet reviewed the attestation, a non-interruptive BPA fires during order entry reminding the provider to review and attest.

Cerner/Oracle Health Integration

• mPage Note Insertion: The attestation is written as a structured component within the mPage-based encounter documentation, tagged with the G2211-specific documentation type.

• USCDI v3 CarePlan Link: The attestation references the active CarePlan resource for the patient's chronic conditions via the USCDI v3 Care Plan data class, creating interoperable linkage that satisfies both CMS review requirements and ONC information-blocking rules.

• PowerChart Order/Charge Integration: The G2211 charge is conditionally placed in the charge capture queue only after the attestation SDE is confirmed present — preventing "orphan" G2211 charges that lack supporting documentation.

FHIR-Native / EHR-Agnostic Pathway

For practices running athenahealth, eClinicalWorks, NextGen, or other FHIR R4-capable EHRs, Scribing.io uses the SMART on FHIR launch framework to:

• Query Encounter, Condition, CarePlan, and MedicationRequest resources for the longitudinal scan.

• Write the attestation as a DocumentReference resource with a coded category (LOINC 11488-4: Consultation Note) and a provenance chain linking it to the practitioner NPI and organization TIN.

• Surface the attestation in a SMART app panel embedded in the EHR's note-writing interface for provider review.

The integration architecture matters because the attestation's audit survival depends on where and how it is stored. A SmartPhrase alone can be edited or deleted post-signature. An SDE + Provenance resource creates a versioned, immutable audit trail. Scribing.io defaults to the most durable storage pathway available in each EHR environment.

6. Telehealth, Global Periods, and Modifier Stacking: The Denial-Prevention Layer

Three modifier-related denial scenarios account for the majority of G2211 claim failures in practices that do attempt to bill the code. Each requires specific engine logic.

Telehealth Modifier Placement

CMS permits G2211 on telehealth E/M encounters. The trap: modifier 95 (synchronous telehealth) and POS 02 (telehealth — patient location distant site) or POS 10 (telehealth — patient home) must be placed on the base E/M code line, not on the G2211 add-on code line. Placing modifier 95 or POS 02/10 on the G2211 line triggers an automatic MAC denial — the add-on code does not accept telehealth modifiers.

Scribing.io engine logic:

• Detects visit modality from the scheduling system's telehealth flag.

• Applies modifier 95 to the 99214 (or applicable E/M) line.

• Applies POS 10 (patient at home — the dominant 2026 telehealth scenario) to the E/M line.

• Leaves the G2211 line with no modifier and POS matching the E/M line's POS.

• Validates that the MAC in jurisdiction accepts G2211 on telehealth encounters (all seven MACs do as of June 2026, but the engine checks quarterly-updated payer policy tables to catch any future restriction).

Global Surgical Period Suppression

When a patient is in a 10-day or 90-day global surgical period, all E/M services related to the surgery are bundled into the surgical payment. Billing G2211 on a post-op visit requires:

1. Modifier 24 on the E/M code (indicating the E/M is unrelated to the surgical decision or constitutes management of a condition unrelated to the surgery).

2. Documentation that the visit's E/M content addresses longitudinal chronic disease management separate from the surgical episode — not routine post-op care.

Scribing.io engine logic:

• Cross-references the patient's claims history for any CPT code with a 10- or 90-day global period that is currently active.

• If a global period is detected, the engine reads the current encounter note to determine whether the visit content is post-op follow-up or independent chronic disease management.

• If the note reflects routine post-op care (wound check, drain removal, surgical site assessment), G2211 is suppressed entirely — even if modifier 24 is technically available. This is a conservative compliance posture that prevents audit exposure.

• If the note reflects independent chronic disease management (diabetes medication adjustment, hypertension regimen change, unrelated to the surgical condition), the engine prompts the provider to confirm modifier 24 is appropriate and inserts G2211 with the modifier 24-supported E/M.

• In the pilot scenario described in Section 3, this logic correctly suppressed G2211 on 4 visits that fell within a 90-day global surgical period where the encounter documentation was post-op in nature.

Modifier -25 Stacking

CMS's MM13272 established that G2211 cannot be billed on the same claim line as an E/M that carries modifier -25 (significant, separately identifiable E/M on the same day as a procedure). The narrow exception: preventive medicine services (99381–99397) when a separate E/M is performed and billed with -25. Outside that exception, the denial is absolute.

Scribing.io engine logic:

• Scans the charge capture queue for any procedure code billed on the same date of service.

• If a same-day procedure exists and the E/M carries modifier -25, the engine blocks G2211 from the claim.

• If the same-day service is a preventive medicine code (99381–99397) and the separate E/M carries -25, the engine checks whether the applicable MAC honors the exception (most do as of 2026) and permits G2211 only if the MAC-specific edit pack confirms it.

• In the pilot, this logic blocked G2211 on 12 encounters that would have been denied at 100% rate.

7. MAC-Specific Edit Packs and Payer Variance Matrix

Medicare claims are adjudicated by seven MACs, each maintaining jurisdiction-specific edits that can differ from CMS national policy. Scribing.io maintains a quarterly-updated MAC edit pack database covering all jurisdictions:

Engine behavior: Scribing.io maps each practice location to its MAC jurisdiction at deployment. Every G2211 claim line is validated against the jurisdiction-specific edit pack before submission. When a MAC publishes an updated LCD companion article — as Novitas, NGS, and First Coast did in Q1 2026 — the engine's attestation template is updated to align with the published preference within 30 days. This is not a static rules engine; it is a continuously maintained compliance layer.

California-based practices should additionally review SB-1120 utilization review provisions which impose state-level audit trail requirements on AI-generated billing recommendations — a requirement the Scribing.io Pro plan satisfies natively through its FHIR Provenance logging.

8. ROI Model: Subscription Cost vs. Recovered Revenue

The financial case for automating G2211 capture is not marginal. It is decisive. The following model uses conservative assumptions.

Assumptions

• Practice size: 5 providers (qualifying for the 10% bundle discount)

• Eligible E/M encounters per provider per week: 30 (conservative for primary care)

• Pre-Scribing.io G2211 uptake: 0% (per the audit scenario)

• Post-Scribing.io G2211 uptake: 68% (per pilot results)

• G2211 reimbursement: $16.05 (2026 national average)

• Weeks per year: 48 (accounting for PTO, holidays, CME)

Scribing.io Annual Cost: Pro Plan with Bundle Discount

G2211 Revenue Recovery

ROI Summary

Scribing.io vs. Manual Workflow vs. Competitor AI Scribe: Annual Cost Comparison

Even on the Basic plan at $35/month per provider, the G2211 revenue recovered in the pilot scenario exceeds the annual subscription cost within the first two weeks of each calendar year. The Pro plan adds EHR-native attestation (SDE + FHIR Provenance), the Smart Scheduler, and Telehealth integration — capabilities that reduce the manual steps between engine recommendation and claim submission to zero.

The Practice Overhead Mitigation Calculation

G2211 is not the only revenue lever Scribing.io activates. When positioned as the Practice Overhead Mitigation Package — Scribing.io's AI documentation engine paired with the AI Front Desk module — the subscription replaces or reduces spend in three high-turnover staff categories:

For a five-provider practice, the combined Scribing.io Pro subscription ($2,916/year) plus G2211 revenue recovery ($78,581/year) generates a net positive cash flow that offsets a significant portion of the practice's scribe and coding overhead — while eliminating the operational disruption caused by staff turnover in those roles.

Request a Demo

This playbook is based on real engine logic, real pilot data, and real MAC edit pack configurations. To see Scribing.io's G2211 automation running against your practice's encounter volume, payer mix, and EHR environment, request a live demo at scribing.io. The demo includes a custom ROI projection using your Q1 claims data and MAC