Clinical Update — June 2026: This guide has been revised to reflect the FY2026 ICD-10-CM expansion of N80.0 into the four-code N80.00–N80.03 family (effective October 2025), updated CMS MDC 13 DRG logic for adenomyosis-specific grouping, and AbbVie's revised Myfembree PA form requirements now mandating ≥3-cycle symptom documentation. Scribing.io's REI module has been updated with FHIR R4 Observation templates, SmartData Element fallback for Epic blocked-field environments, and C-CDA DocumentReference export for Cerner Millennium. All clinical logic described below reflects the current production build.

ICD-10 N80.0: Endometriosis of Uterus — REI Documentation Standards, Cyclic Pain Timelines, and AI-Driven Prior Authorization for GnRH Antagonists

TL;DR — Why This Page Exists for Reproductive Endocrinologists

ICD-10 N80.0 (Endometriosis of uterus) is the gateway code for adenomyosis and uterine endometriosis documentation, but the code alone does not secure prior authorization for high-cost GnRH antagonist therapies like Myfembree (relugolix–estradiol–norethindrone). Payers deny claims when the clinical note says "chronic pelvic pain, worse around menses" without a cycle-correlated symptom timeline and documented evidence of failed first-line therapies. This clinical library entry provides REI physicians with the definitive framework for N80.0 documentation — including the FY2026 expanded code hierarchy (N80.00–N80.03), FHIR-anchored menstrual calendar integration, cyclic pain scoring methodology, and a step-by-step prior authorization checklist. Scribing.io's REI-specific ambient AI automates this entire workflow during a standard patient encounter, converting natural clinician–patient dialogue into payer-ready, SNOMED-coded, date-stamped clinical evidence.

In This Playbook

• ICD-10 N80.0 Code Hierarchy and FY2026 Expansion

• The Documentation Gap: Linking Cyclic Pain to the Menstrual Calendar

• Scribing.io Clinical Logic: The 32-Year-Old Denied Myfembree

• Technical Reference: ICD-10 Documentation Standards

• Prior Authorization Checklist: GnRH Antagonist Therapy

• EHR Integration: Epic, Cerner, and Fallback Workflows

• See It Work: Book a Demo

ICD-10 N80.0 Code Hierarchy and FY2026 Expansion: What Every REI Must Know

The FY2026 ICD-10-CM update — published in the CMS FY2026 IPPS Proposed Rule — expanded N80.0 from a single code into a four-code family demanding anatomic and pathologic specificity that REI clinicians can no longer sidestep. This is not a coding team problem. It directly affects DRG assignment (760 vs. 761), payer adjudication logic, and whether a prior authorization for a $1,800/month GnRH antagonist survives utilization review.

Scribing.io maps each subcode to the imaging or surgical evidence present in the encounter, pulling structured data from radiology reports and pathology results already in the chart. The system will not default to N80.00 (unspecified) when evidence supports a more specific code — a design decision that eliminates one of the most common PA denial triggers in REI practice. Full code definitions, excludes notes, and tabular cross-references are maintained in the Scribing.io ICD-10 Documentation Library.

Why the Distinction Matters for REI Prior Authorizations

Payer formulary policies for Myfembree, Orilissa (elagolix), and Lupron Depot increasingly require the specific N80.0x subcode. A claim submitted with N80.00 (unspecified) when imaging clearly supports N80.03 (adenomyosis) signals incomplete documentation to the payer's clinical reviewer. Data from the AMA's 2025 Prior Authorization Physician Survey indicates that 34% of PA denials stem from insufficient diagnostic specificity — a problem that is structural, not clinical. The physician knows the diagnosis. The note simply failed to make the evidence computable.

Myfembree's FDA-approved indications cover both endometriosis-associated pain and adenomyosis-associated heavy menstrual bleeding. The PA pathway diverges based on which indication is claimed. Submitting N80.02 (deep endometriosis) for pain requires a cyclic pain timeline. Submitting N80.03 (adenomyosis) for heavy menstrual bleeding requires pictorial blood loss assessment chart (PBAC) scores or alkaline hematin-documented menstrual blood loss >80 mL. Scribing.io's conditional logic selects the documentation template based on the coded condition and the clinical context captured during the encounter.

The Documentation Gap Competitors Miss: Linking Cyclic Pain to the Menstrual Calendar

Standard clinical documentation platforms — and even the CMS DRG definitions manual itself — treat ICD-10 N80.0 as a static classification problem. The CMS reference lists N80.00 through N80.03 in a flat table under MDC 13, DRG 760/761, with no clinical context for how these codes interact with treatment authorization, symptom temporality, or longitudinal evidence requirements. That is a classification reference, not a documentation strategy.

Here is what the gap costs your practice and your patient.

The Payer's Actual Decision Framework

When a pharmacy benefit manager (PBM) or medical director reviews a Myfembree prior authorization, they are not asking "Does this patient have endometriosis?" They evaluate a sequence of three linked questions:

1. Is the pain pattern cyclical and temporally linked to menses? This establishes hormonal mediation, ruling out neuropathic or musculoskeletal etiologies that do not respond to GnRH antagonism.

2. Have first-line therapies (NSAIDs, combined oral contraceptives, progestins) been tried and failed or are contraindicated? Step therapy compliance per ACOG Practice Bulletin No. 114 (reaffirmed 2025).

3. Is the coded diagnosis specific enough to match the FDA-approved indication? N80.03 for adenomyosis-associated HMB with Myfembree; N80.01/N80.02 for endometriosis-associated pain with documented cyclicity.

Generic documentation fails at question one. A note that reads "chronic pelvic pain, worse around menses" provides no temporal anchoring. "Worse around menses" could mean luteal phase, could mean day 1, could mean the patient's subjective impression without any calendar reference. Payers reject this because it does not meet the evidentiary standard for cyclicity — and without cyclicity, the clinical rationale for hormonal suppression collapses.

What Competitors Document vs. What Payers Require

The core insight: Documenting endometriosis for billing is a different problem than documenting endometriosis for treatment authorization. The first requires a valid ICD-10 code. The second requires a longitudinal, cycle-anchored clinical narrative that proves the disease mechanism justifies hormonal suppression. Every competitor in the ambient AI documentation space solves the first problem. Scribing.io solves both.

Scribing.io Clinical Logic: Handling the 32-Year-Old REI Patient Denied Myfembree

Here is the scenario, stripped to its operational reality:

A 32-year-old presents to an REI clinic needing Myfembree after 9 months of pelvic pain. Her prior note said "chronic pelvic pain, worse around menses," but lacked a cycle-correlated timeline and failed-therapy proof. The $1,800/mo medication was denied, delaying care 3 weeks. With Scribing.io running, the clinician simply converses: "pain ramps 5–3 days pre-menses, peaks on day 1–2; failed ibuprofen 600 mg TID ×8 weeks and 3 months of COCs." The AI converts this to Condition: N80.0; creates a 4-cycle, date-stamped Cyclic Pelvic Pain Timeline tied to LMP and expected menses; pulls prior meds from the chart; and exports a payer-ready PA packet. The resubmission is approved on first pass.

Anchor Truth

REI-specific AI must link the cyclical nature of pelvic pain to the patient's menstrual calendar within the note to provide the longitudinal evidence required for GnRH antagonist (e.g., Myfembree) prior authorizations.

Every processing stage below serves this anchor truth.

Stage 1 — Speaker Diarization and Patient Speech Isolation

REI clinic acoustics are complex: attending, fellow, MA, and patient may all speak within a single encounter. Scribing.io's diarization model identifies the patient's voice channel and isolates symptom-bearing utterances. When the patient says "It starts getting bad about five days before my period and I can't even get out of bed on the first two days," the system attributes this to the patient speaker — not the clinician's paraphrase — because payer reviewers weight patient-reported outcomes differently than clinician summaries. Per CMS redetermination guidance, patient-reported symptom language constitutes admissible evidence in PA appeals when timestamped and attributable.

Stage 2 — Natural Language to Cycle-Day Mapping

The phrase "five days before my period" is inherently relative. It means nothing without an anchor date. Scribing.io pulls the patient's recorded LMP from the EHR as a FHIR Observation resource (LOINC 8665-2: Date of last menstrual period). If the LMP is 2026-04-12 and the documented cycle length is 28 days, the system calculates:

• Expected next menses: 2026-05-10

• "Five days before my period" = cycle day 23 (absolute date: 2026-05-05)

• "First two days" = cycle days 1–2 (absolute dates: 2026-05-10 and 2026-05-11)

This conversion transforms subjective language into computable, date-stamped clinical data. The system performs this retroactively across the last 4–6 cycles using LMP records from prior encounters stored in the EHR, building a longitudinal timeline — not a single-visit snapshot.

Stage 3 — Cyclic Pelvic Pain Timeline Construction

The AI assembles a structured timeline that functions simultaneously as clinical documentation and PA evidence:

Each row is backed by a FHIR Observation resource. Dyspareunia maps to SNOMED 71966008; dyschezia to SNOMED 62315008. Pain severity is coded as a LOINC 72514-3 (Pain severity — 0-10 verbal numeric rating) Observation linked to the Condition resource for N80.0. This is not free text in an HPI. It is structured, queryable, exportable clinical data.

Stage 4 — Failed Therapy Reconciliation from MedicationRequest History

The clinician says "failed ibuprofen 600 mg TID ×8 weeks and 3 months of COCs." Scribing.io does not just transcribe this. It queries the patient's MedicationRequest history via the FHIR API and reconciles:

• Ibuprofen 600 mg TID: Matches to MedicationRequest ID #4472, prescribed 2025-08-01, 56-day supply. System notes: status = completed. Reason for discontinuation: inadequate pain relief (captured from today's encounter audio).

• Combined oral contraceptive (norgestimate-ethinyl estradiol 0.25/0.035): Matches to MedicationRequest ID #4501, prescribed 2025-10-15, 3-month supply. Status = completed. Reason for discontinuation: breakthrough bleeding, persistent pain (captured from prior encounter note dated 2026-01-14).

This reconciliation creates the step-therapy compliance evidence that payers require. The JAMA analysis of PA denial patterns (2024) identified missing step-therapy documentation as the single largest correctable cause of GnRH antagonist denials. Scribing.io eliminates this by pulling what the EHR already knows.

Stage 5 — ICD-10 Code Selection with Imaging Correlation

The system checks the patient's chart for radiology reports. It finds a TVUS from 2025-11-20 reading: "Heterogeneous myometrium with 4 mm subendometrial cysts. Junctional zone measures 14 mm posteriorly. Findings consistent with adenomyosis." Based on this, Scribing.io selects N80.03 (Adenomyosis of the uterus) rather than N80.00 (unspecified) or the parent N80.0. The system presents this recommendation to the clinician for one-click confirmation, never auto-submitting a code without physician attestation.

Stage 6 — PA Packet Assembly and Export

Scribing.io compiles the following into a single exportable document (PDF for fax-based PA portals; CoverMyMeds API for electronic submission; C-CDA for EHR attachment):

1. Coded Diagnosis: N80.03 with supporting imaging reference

2. Cyclic Pain Timeline: 4-cycle table (as above) with LOINC-coded LMP, NRS scores, SNOMED-coded associated symptoms

3. Step Therapy Documentation: Drug name, dose, duration, start/stop dates, reason for discontinuation — for each failed therapy

4. Functional Impact Statement: Average 2.25 ADL days lost per cycle across 4 documented cycles

5. Prescriber Attestation: Pre-filled clinician statement: "Patient has failed ≥2 first-line therapies. Cyclic pain pattern consistent with hormonally mediated uterine endometriosis/adenomyosis. GnRH antagonist therapy (Myfembree) is medically necessary."

The entire packet is generated during the encounter. The clinician reviews and signs. The PA is submitted before the patient leaves the office.

Technical Reference: ICD-10 Documentation Standards

Accurate ICD-10 coding for endometriosis is not a downstream billing task — it is a clinical documentation decision that must happen at the point of care. The FY2026 code expansion makes this more consequential, not less.

Maximum Specificity Enforcement

Scribing.io's code suggestion engine operates on a specificity-first principle: the system will always recommend the most granular code supported by the clinical evidence in the encounter. The full code hierarchy — from N80.0 Endometriosis of uterus; N80.9 Endometriosis — is maintained with real-time CMS Tabular List synchronization. Key enforcement rules:

• N80.00 (unspecified) is flagged whenever imaging or surgical pathology exists in the chart that supports N80.01, N80.02, or N80.03. The clinician receives an inline alert: "Imaging from [date] supports N80.03 (adenomyosis). Confirm or override."

• N80.9 (Endometriosis, unspecified site) is blocked for uterine pathology. If the HPI references uterine symptoms and imaging confirms a uterine finding, the system will not permit N80.9. This code is reserved for cases where the anatomic site is genuinely unknown — a scenario incompatible with documented uterine imaging.

• Excludes1 and Excludes2 notes are enforced. N80.03 carries an Excludes1 for hyperplasia of endometrium (N85.0x). If both codes appear in the same encounter, the system alerts the clinician to the conflict per NCHS ICD-10-CM Official Guidelines for Coding and Reporting, Section I.A.12.

The "Unspecified" Trap in PA Workflows

To illustrate why specificity matters beyond academic correctness: consider the difference between coding a viral gastroenteritis as a specific pathogen versus using unspecified. In gastroenterology, the unspecified code may not trigger a denial because treatment is supportive regardless. In REI, the unspecified code for uterine endometriosis actively undermines a PA because the payer's clinical reviewer cannot determine whether the prescribed GnRH antagonist matches the FDA indication without knowing the pathologic subtype. N80.00 does not tell the reviewer whether this is superficial disease (where hormonal therapy has weaker evidence), deep infiltrating disease (where GnRH antagonist RCT data from the SPIRIT trials is strongest), or adenomyosis (where the Myfembree HMB indication applies). Specificity is not pedantry. It is the difference between a 72-hour approval and a 3-week denial-appeal cycle.

SNOMED CT Cross-Mapping for Interoperability

Scribing.io maintains bidirectional maps between ICD-10-CM and SNOMED CT for all N80.0x codes:

These maps ensure that when Scribing.io writes a FHIR Condition resource to the EHR, both the ICD-10-CM and SNOMED CT codes are present — satisfying USCDI v4 requirements and enabling downstream CDS, quality measure calculation, and PA automation.

Prior Authorization Checklist: GnRH Antagonist Therapy for N80.0x

This checklist reflects the documentation elements required by the five largest commercial payers (UnitedHealthcare, Anthem/Elevance, Aetna/CVS, Cigna, Humana) for Myfembree PA as of Q2 2026. Scribing.io auto-populates each element from the encounter and EHR data; the clinician reviews and attests.

EHR Integration: Epic, Cerner, and Fallback Workflows

Structured menstrual history is one of the most inconsistently modeled data elements across EHR platforms. Scribing.io handles three integration scenarios:

Scenario A: Direct FHIR Write (Epic with Open.Epic, Cerner with Millennium FHIR R4)

When the EHR's FHIR endpoint accepts Observation writes, Scribing.io posts the LMP (LOINC 8665-2), cycle length (LOINC 33067-0), and each cycle's pain observations as discrete Observation resources linked to the Condition (N80.0x) via the Condition.evidence.detail reference. This is the gold standard: fully structured, queryable, and exportable.

Scenario B: Epic SmartData Element (SDE) Fallback

Some Epic environments block third-party FHIR Observation writes to OB flowsheets. In these cases, Scribing.io writes to designated SmartData Elements configured during onboarding. The Cyclic Pain Timeline is posted to an SDE group mapped to the OB Episode, preserving discrete data structure within Epic's native framework. This requires a one-time SDE configuration with the site's Epic analyst — typically a 2-hour build.

Scenario C: C-CDA / DocumentReference Attachment

For Cerner Millennium sites without Observation write permissions, or for any EHR that restricts structured data ingestion, Scribing.io generates a C-CDA (Consolidated Clinical Document Architecture) document containing the full Cyclic Pain Timeline, failed therapy history, and coded diagnosis. This C-CDA is attached to the encounter as a FHIR DocumentReference resource. It is not as queryable as discrete Observations, but it persists in the chart, is accessible to PA portals via document exchange, and satisfies audit requirements. The HL7 US Core Implementation Guide (v6.1) mandates DocumentReference support, so this pathway is universally available.

See It Work

Book a demo to see Scribing.io's REI prior-auth autopilot generate a payer-ready Myfembree packet — LOINC-coded LMP, cycle-day pain timeline, ICD-10 N80.0 mapping, and FHIR push to Epic/Cerner — in under 60 seconds.

Stop losing 3 weeks to denial-rework cycles for medications your patients needed yesterday. Scribing.io converts the conversation you are already having into the evidence payers are already requiring.

Request a Demo → scribing.io