ICD-10 K76.0: Fatty Liver Documentation & the MASLD Nomenclature Shift — A Clinical Library Playbook for Primary Care

• TL;DR — Why This Page Exists

• The Nomenclature Trap — Why "NAFLD" in Your Note Is Now a Denial Trigger

• What Competitor Guides Miss — Age-Adjusted FIB-4 Cutoffs and Computable Persistence

• Scribing.io Clinical Logic — From Denied Claim to First-Pass Approval in Two Visits

• Technical Reference — ICD-10 Documentation Standards for Fatty Liver and Steatohepatitis

• FHIR R4 Write-Back Architecture — How the Score Becomes an Auditable Artifact

• Payer and LCD Alignment — MolDX DL40187 and Commercial UM Crosswalk

• 14-Day Implementation Checklist for Medical Directors

• See MASLD Auto-Triage Live in Your EHR

TL;DR — Why This Page Exists

Payers and guideline bodies have retired "NAFLD" in favor of MASLD (Metabolic Dysfunction-Associated Steatotic Liver Disease). Notes that still say "NAFLD," omit a FIB-4 score, or fail to link recent labs to the referral order are generating avoidable claim denials—especially for liver elastography (CPT 91200) and hepatology referrals. This playbook gives Primary Care Medical Directors the exact documentation logic, age-adjusted FIB-4 cutoffs, ICD-10 mapping rules, and FHIR-native workflow that Scribing.io automates at the point of care. Every clinical assertion below is traceable to AASLD 2023 practice guidance, the AGA Clinical Care Pathway (2024), and the proposed MolDX LCD (DL40187).

Scribing.io exists because the documentation gap between updated clinical terminology and an unchanged ICD-10 code set is the single largest source of preventable denials in primary-care metabolic liver disease management right now. This page is the reference your documentation improvement team needs to close that gap—permanently.

The Nomenclature Trap — Why "NAFLD" in Your Note Is Now a Denial Trigger

In June 2023, a multi-society Delphi consensus led by the AASLD formally adopted MASLD to replace NAFLD and MASH to replace NASH. By 2025, CMS's proposed Local Coverage Determination (DL40187) explicitly states: "The term MASLD will be used as a replacement for non-alcoholic fatty liver disease (NAFLD), and MASH as a replacement for non-alcoholic steatohepatitis (NASH)." The coding system, however, has not caught up. ICD-10-CM still maps steatotic liver disease to K76.0 - Fatty (change of) liver and steatohepatitis to not elsewhere classified; K75.81 - Nonalcoholic steatohepatitis (NASH). This disconnect between narrative terminology and billing codes is the trap.

The practical fallout for primary care is immediate and measurable:

• Payer edits are terminology-sensitive. Utilization-management algorithms at major commercial payers (UnitedHealthcare, Aetna, Cigna) now flag notes containing "NAFLD" without a corresponding MASLD synonym as potentially outdated clinical reasoning, triggering manual review queues that add 15–30 days to adjudication.

• Prior-authorization templates for CPT 91200 (vibration-controlled transient elastography, VCTE) increasingly require the ordering note to reference MASLD/MASH nomenclature alongside ICD-10 codes. A note that says only "NAFLD" paired with K76.0 may pass the code edit but fail the clinical-language screen.

• Patient-facing after-visit summaries that still read "NAFLD" create confusion during shared decision-making—a quality concern flagged by the CMS Innovation Center's person-centered care framework.

The documentation paradox is clear: you must use the new clinical terminology (MASLD/MASH) in the narrative while mapping to the current ICD-10 codes (K76.0/K75.81) until CMS publishes updated codes. Most EHR dot-phrase libraries do neither consistently. Scribing.io resolves this by inserting a dual-layer documentation block—narrative MASLD language plus the valid ICD-10 code—so the note satisfies both the clinical guideline and the billing system simultaneously. Explore our full Scribing.io ICD-10 Documentation Library for the complete mapping logic across steatotic liver disease codes.

What Competitor Guides Miss — Age-Adjusted FIB-4 Cutoffs and Computable Persistence

Search for "FIB-4 fatty liver documentation" and you find dozens of guides that say "calculate FIB-4." That advice is necessary but radically insufficient. Two critical omissions turn a correct clinical instinct into a denied claim or, worse, a missed diagnosis.

Age-Adjusted FIB-4 Interpretation

The standard FIB-4 cutoffs (low risk < 1.3; indeterminate 1.3–2.67; high risk ≥ 2.67) were validated in cohorts with a mean age well below 65. AASLD practice guidance and multiple validation studies (McPherson et al., Gut 2017; AGA Clinical Care Pathway 2024) recommend age-adjusted thresholds for patients ≥ 65 years. Applying the wrong cutoff to a geriatric patient is both a clinical safety risk and a documentation deficiency that payers exploit during audit.

Why this matters for documentation: A 70-year-old with a FIB-4 of 1.5 is low risk under age-adjusted criteria but indeterminate under standard criteria. A note that applies the wrong cutoff may trigger an unnecessary hepatology referral and a CPT 91200 order that the payer denies for lack of medical necessity—or, conversely, may under-triage a younger patient whose 1.5 genuinely warrants secondary assessment. Neither outcome is acceptable.

Computable Persistence: The Missing FHIR Artifact

Calculating FIB-4 on a sticky note, a dot phrase, or even a "smart form" creates a human-readable number that is not queryable, not time-stamped to its source labs, and not interoperable. The proposed MolDX LCD (DL40187) criterion #3 explicitly requires that a "primary risk assessment based on non-molecular/proteomic laboratory testing" has been performed and does not indicate low risk. During post-payment audit, reviewers need to trace the score back to the exact AST, ALT, platelet count, and patient age that generated it—and confirm those labs were drawn within the payer's accepted window (commonly ≤ 12 months).

A free-text note that says "FIB-4 = 1.9" without linked, time-stamped lab values is an audit liability. Scribing.io eliminates this liability at the point of documentation. The system auto-calculates FIB-4 from discrete AST, ALT, platelet count, and age values already in the chart, then writes the result as a FHIR R4 Observation resource—an artifact that carries its own provenance and travels with the patient across every connected system.

Scribing.io Clinical Logic — From Denied Claim to First-Pass Approval in Two Visits

This is the scenario every Primary Care Medical Director dreads—and the exact problem Scribing.io was engineered to solve.

Visit 1 — Without Scribing.io: The 45-Day Denial

A 58-year-old with obesity (BMI 34.2) and type 2 diabetes (A1c 7.8%) presents with right-upper-quadrant ultrasound showing hepatic steatosis. The PCP orders a FibroScan (CPT 91200) and refers to hepatology. The submitted note reads:

• Assessment: "NAFLD, ultrasound confirmed."

• No FIB-4 score documented.

• Recent AST, ALT, and platelets exist in the chart but are not referenced in the note or linked to the order.

• ICD-10 submitted: K76.0.

The payer's UM algorithm flags three deficiencies: outdated NAFLD terminology without MASLD equivalent, absent non-invasive fibrosis risk stratification, and no lab linkage establishing medical necessity for VCTE. Claim denied. Care delayed 45 days while the practice scrambles to appeal.

Visit 1 — With Scribing.io: Appropriate Primary-Care Management, No Unnecessary Referral

The same encounter flows through Scribing.io's ambient clinical documentation engine. Here is the step-by-step logic:

1. Cardiometabolic context detection. Scribing.io identifies the active problem list entries (obesity, T2DM) and the ultrasound steatosis finding. The system recognizes that this combination satisfies MASLD diagnostic criteria per the 2023 Delphi consensus.

2. MASLD terminology auto-insertion. The Assessment section reads: "Metabolic dysfunction-associated steatotic liver disease (MASLD), confirmed by hepatic ultrasound demonstrating steatosis in the setting of obesity and T2DM." ICD-10 K76.0 - Fatty (change of) liver is mapped. Both the narrative and the code coexist—dual-layer documentation.

3. Discrete lab retrieval. The system pulls the patient's most recent AST (42 U/L, drawn 3 months ago), ALT (55 U/L, same draw), and platelet count (210 × 10⁹/L, same draw). All labs are within the ≤ 12-month recency window.

4. Auto-calculation with age-adjusted interpretation. FIB-4 = (58 × 42) / (210 × √55) = 1.56. Patient age is 58 (under 65), so standard cutoffs apply. Score of 1.56 falls in the indeterminate range (1.3–2.67).

5. Triage logic. Because the score is indeterminate—not high risk—Scribing.io inserts a triage statement per the AGA Clinical Care Pathway: "FIB-4 index = 1.56 (indeterminate risk, standard cutoff for age < 65). Per AASLD/AGA pathway, secondary fibrosis assessment with VCTE is recommended. Patient can be managed in primary care pending VCTE result; hepatology referral is not indicated at this time."

6. FHIR R4 Observation persistence. The FIB-4 score is written as a FHIR Observation with derivedFrom links to the three source-lab Observations. effectiveDateTime is the encounter date. interpretation carries IND (indeterminate). The artifact is immediately queryable by downstream systems.

7. CPT 91200 order support. Because secondary assessment is clinically indicated, Scribing.io pre-populates the VCTE order with the MASLD diagnosis, the FIB-4 score and linked labs, and a medical-necessity statement—ready for payer submission if prior authorization is required. But critically, no hepatology referral is generated, preventing an unnecessary specialty utilization (CPT 99205/99215 at hepatology) and the associated authorization friction.

Visit 2 — With Scribing.io: FIB-4 Escalation, First-Pass Approval for CPT 91200

Six months later, the same patient returns. New labs: AST 68, ALT 52, platelets 178. Scribing.io recalculates automatically:

1. FIB-4 recalculation. FIB-4 = (58 × 68) / (178 × √52) = 3.07. This exceeds the high-risk threshold (≥ 2.67) under both standard and age-adjusted cutoffs.

2. Medical-necessity auto-insertion. The note now reads: "FIB-4 index = 3.07 (high risk for advanced fibrosis per AASLD Practice Guidance 2023, standard cutoff ≥ 2.67). Liver stiffness measurement by vibration-controlled transient elastography (VCTE) is indicated for secondary fibrosis assessment per AGA Clinical Care Pathway 2024. Hepatology referral is indicated for advanced fibrosis evaluation and pharmacotherapy eligibility assessment."

3. ICD-10 escalation. Given the elevated transaminases (AST 68, ALT 52), the clinical context of worsening metabolic liver disease, and the high FIB-4 suggesting advanced fibrosis, Scribing.io maps to not elsewhere classified; K75.81 - Nonalcoholic steatohepatitis (NASH) with MASH narrative language: "Metabolic dysfunction-associated steatohepatitis (MASH), suspected based on elevated transaminases, hepatic steatosis, cardiometabolic risk factors, and FIB-4 ≥ 2.67." The provider confirms the code selection with a single click; Scribing.io never auto-submits a code without attestation.

4. FHIR Observation with trend. The new FIB-4 Observation includes derivedFrom links to the new lab draw and a hasMember reference to the Visit 1 FIB-4 Observation, creating a computable longitudinal trend. Payer auditors can see the clinical trajectory without manual chart review.

5. Payer-ready CPT 91200 packet generation. Scribing.io assembles and attaches:

   • MASLD/MASH diagnosis with dual ICD-10 mapping (K75.81 primary, K76.0 history)

   • FIB-4 = 3.07 with linked lab values and collection dates

   • Guideline-referenced medical-necessity language (AASLD, AGA citations)

   • Ordering provider attestation

   • FHIR Bundle containing all referenced Observation resources

6. Result: Prior authorization approved on first submission. No appeal. No 45-day delay. No extra staff hours assembling documentation after the fact.

Technical Reference — ICD-10 Documentation Standards for Fatty Liver and Steatohepatitis

The two ICD-10-CM codes most relevant to MASLD/MASH documentation in primary care require precise narrative support to survive payer audit. Here is the specification.

K76.0 - Fatty (change of) liver

• Clinical mapping: MASLD without histologic or clinical evidence of steatohepatitis. Use when imaging confirms steatosis (≥ 5% liver fat by ultrasound, MRI-PDFF, or CAP on VCTE) and at least one cardiometabolic risk factor (overweight/obesity, T2DM, dyslipidemia, hypertension, elevated waist circumference) is present per the AASLD Delphi definition.

• Required narrative elements: (1) Imaging modality and steatosis finding; (2) At least one named cardiometabolic criterion; (3) MASLD terminology in Assessment; (4) Alcohol use quantification confirming intake below AASLD thresholds (< 140 g/week for females, < 210 g/week for males).

• Common documentation errors that trigger denial:

  • Using K76.0 when alcohol use exceeds AASLD thresholds—this should map to K70.0 (Alcoholic fatty liver) per CMS ICD-10-CM guidelines.

  • Failing to document the cardiometabolic qualifier, leaving the payer unable to distinguish MASLD from cryptogenic steatosis.

  • Pairing K76.0 with CPT 91200 without a documented non-invasive fibrosis score (FIB-4, NFS, or equivalent) in the indeterminate-or-higher range. K76.0 alone does not establish medical necessity for VCTE—the FIB-4 triage is the bridge.

not elsewhere classified; K75.81 - Nonalcoholic steatohepatitis (NASH)

• Clinical mapping: MASH—steatohepatitis in the context of MASLD. Historically required biopsy confirmation, but the 2024 AGA pathway and FDA's resmetirom approval pathway have established clinical criteria (elevated transaminases + advanced fibrosis on non-invasive testing) as sufficient for pharmacotherapy eligibility and, increasingly, for coding purposes.

• Required narrative elements: (1) All elements required for K76.0; (2) Documentation of hepatocyte injury markers (elevated AST and/or ALT); (3) FIB-4 or VCTE score in the indeterminate-to-high range; (4) MASH terminology in Assessment; (5) Statement of clinical suspicion or histologic confirmation.

• Common documentation errors:

  • Upgrading from K76.0 to K75.81 without documenting the clinical basis (transaminase pattern, fibrosis score, or biopsy). Payers treat unexplained code escalation as upcoding.

  • Using K75.81 as a standalone code without co-documenting the metabolic context (obesity, T2DM, dyslipidemia). The "nonalcoholic" in the code descriptor requires affirmative documentation that alcohol is not the primary etiology.

Scribing.io enforces maximum specificity by refusing to map K75.81 unless the encounter note contains at minimum: (a) a named cardiometabolic risk factor, (b) an alcohol-use quantification below threshold, (c) elevated transaminases or a VCTE result suggesting steatohepatitis, and (d) a FIB-4 score with source-lab linkage. If any element is missing, the system prompts the provider to complete the documentation before code assignment—preventing both undercoding (missed MASH) and upcoding (unjustified K75.81).

FHIR R4 Write-Back Architecture — How the Score Becomes an Auditable Artifact

The FIB-4 score is useless as a billing defense if it exists only as free text. Here is the FHIR R4 resource structure Scribing.io writes back to the EHR:

This resource structure satisfies the ONC USCDI v4 requirements for clinical result exchange and aligns with the US Core Implementation Guide. When a payer's prior-authorization system queries the patient record via FHIR API (as mandated by the CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F)), the FIB-4 Observation is machine-readable and self-documenting. No phone call. No fax. No chart abstraction.

Payer and LCD Alignment — MolDX DL40187 and Commercial UM Crosswalk

The proposed MolDX LCD (DL40187) establishes coverage criteria for molecular fibrosis tests (e.g., NIS2+, ELF) as alternatives or adjuncts to VCTE. While CPT 91200 is not directly governed by this LCD, the LCD's documentation requirements for the preceding risk assessment step—the FIB-4—have become a de facto payer standard across both Medicare and commercial plans. Here is how Scribing.io maps documentation to each LCD criterion:

For commercial payers, Scribing.io maintains a policy-mapping engine that crosswalks these LCD criteria against plan-specific UM requirements (e.g., Evicore, Carelon). When a payer requires additional documentation elements beyond the LCD—such as an explicit statement that the patient has not had a VCTE in the prior 12 months—the system prompts the provider at order entry.

14-Day Implementation Checklist for Medical Directors

Deploying MASLD-aware documentation logic across a primary care practice is a 14-day operational project, not a multi-quarter IT initiative. Here is the timeline:

See MASLD Auto-Triage Live in Your EHR

See our MASLD auto-triage: real-time, age-adjusted FIB-4 and FibroScan documentation with FHIR write-back and a denial-prevention packet for CPT 91200 mapped to ICD-10 K76.0/K75.81—live in your EHR in 14 days.

Every denied CPT 91200 claim costs your practice an estimated $180–$340 in rework labor (appeals coordinator time, provider addendum dictation, fax cycles) before the clinical cost of delayed fibrosis staging is even considered. For a 20-provider primary care group seeing 400+ MASLD-relevant encounters per quarter, that is $72,000–$136,000 in annual preventable waste—not counting downstream revenue loss from patients who abandon the referral pathway entirely.

Scribing.io closes the loop between updated clinical terminology, computable risk scores, FHIR-native data persistence, and payer-ready documentation—at the point of care, before the claim is ever submitted. The nomenclature trap only catches practices that document the way they did in 2022. Stop documenting "NAFLD." Start documenting for approval.

Request a demo →