ICD-10 F50.81 Binge Eating Disorder: The PMHNP Documentation & Coding Authority Guide

TL;DR — What This Guide Covers and Why It Matters

ICD-10 code F50.81 (Binge Eating Disorder) requires far more than a diagnostic label in the clinical note. Payers routinely deny reimbursement and prior authorizations—especially for lisdexamfetamine (Vyvanse®)—when documentation lacks quantified binge frequency, a specific onset date, DSM-5-TR severity grading, and an explicit attestation of no regular compensatory behaviors. This guide gives Psychiatric Mental Health Nurse Practitioners (PMHNP-BCs) a field-ready playbook for structuring every F50.81 encounter to satisfy clinical, coding, and payer requirements on the first submission. It introduces the discrete-data documentation model pioneered by the Scribing.io ICD-10 Documentation Library—an approach that eliminates the narrative-only gap responsible for the majority of BED claim recoupments and prior-auth rejections in outpatient psychiatric practice.

Scribing.io built this playbook because BED documentation failures follow a single, repeatable pattern: clinicians who know the DSM-5-TR inside out still lose revenue because their EHRs never prompt for the five discrete data points payers audit. The problem is structural, not educational. Every section below maps directly to a revenue-protection or patient-access outcome that PMHNP-BCs can implement on their next F50.81 encounter.

See our 2026 DSM-5-TR BED audit-defense workflow: real-time frequency capture, 12-week auto-tracking, severity calculator, and one-click prior-auth packet (including lisdexamfetamine) exported as discrete data to your EHR.

• Why F50.81 Documentation Fails: The Frequency Threshold Payers Enforce

• Beyond Frequency: The Discrete-Data Documentation Model Competitors Miss

• Scribing.io Clinical Logic: From Denied Prior Auth to First-Submission Approval

• Technical Reference: ICD-10 Documentation Standards for F50.81 and F50.2

• DSM-5-TR Criteria Crosswalk: Mapping Clinical Findings to ICD-10 Severity Specifiers

• PMHNP Workflow: Structuring the BED Encounter From Intake to Prior Authorization

• Differential Documentation: Distinguishing F50.81 From F50.2 and Comorbid Presentations

• Payer-Ready Output: Medical Necessity Language and Prior-Auth Packet Construction

Why F50.81 Documentation Fails: The Frequency Threshold Payers Enforce

The single most common reason a payer denies reimbursement or recoups payment for a binge eating disorder claim is the absence of explicit, quantified binge-episode frequency in the clinical note.

DSM-5-TR Diagnostic Criterion D for Binge Eating Disorder states:

The binge eating occurs, on average, at least once a week for three months.

This is not advisory language. It is a bright-line diagnostic gate. When a PMHNP documents "patient reports frequent binge episodes" or "binging since the spring," the note may satisfy a clinical intuition—but it fails the payer's audit checklist, which maps directly to DSM-5-TR criteria. The American Psychiatric Association's DSM-5-TR codified this threshold explicitly because sub-threshold binge eating has a different clinical trajectory, different treatment algorithms, and—critically—different coverage determinations.

What Payers Actually Look For in F50.81 Audits

Commercial payers and Medicare Administrative Contractors (MACs) evaluate F50.81 claims against documentation elements that map one-to-one to DSM-5-TR criteria. Here is what an audit reviewer's checklist looks like—and where narrative-only notes fail:

When even one of these elements is missing, the claim is vulnerable. When two or more are absent—which occurs in a significant majority of narrative-only BED notes based on published audit trends—the result is predictable: recoupment of evaluation-and-management charges, denial of add-on psychotherapy codes, and rejection of prior authorizations for BED-specific pharmacotherapy.

This is the Anchor Truth that every PMHNP must internalize: payers do not infer clinical reasoning. They audit against structured criteria. A note that reads like a clinical narrative but omits discrete, auditable data points will fail. The CMS MS-DRG Definitions Manual classifies F50.81 within PDX Collection 0645 alongside hundreds of other behavioral health codes—but provides zero guidance on the documentation architecture required to survive a post-payment review. That architectural gap is where revenue is lost.

Beyond Frequency: The Discrete-Data Documentation Model Competitors Miss

Most ICD-10 reference resources—including the CMS Definitions Manual and free code-lookup tools—provide code listings and hierarchical groupings. They tell you that F50.81 exists and which severity subcodes (F50.810 through F50.819) are available. What they never address is the documentation architecture required to defensibly assign those codes in a psychiatric outpatient setting.

This is the structural gap. Competitor resources list F50.810 through F50.819 as part of behavioral health code groupings alongside hundreds of other diagnoses. They provide no guidance on:

• What clinical data must appear in the note to support selection of F50.812 (severe) versus F50.811 (moderate)

• How to document the differential between F50.81 (BED) and F50.2 (bulimia nervosa)—a distinction that hinges entirely on the compensatory-behavior attestation

• What constitutes a payer-defensible baseline for medication prior authorization

• How severity is calculated and whether it requires explicit episode-count documentation or accepts qualitative descriptors

The Scribing.io Discrete-Data Approach

Rather than relying on narrative paragraphs that may or may not contain the critical data points, Scribing.io captures binge eating disorder as structured, discrete data fields. Beyond documenting "≥1x/week for 3 months," Scribing.io captures BED as discrete data: binge episodes/week + onset date + explicit "no compensatory behaviors" attestation, auto-calculates DSM-5-TR severity (mild/moderate/severe/extreme), and time-stamps a 12-week baseline/follow-up counter. This feeds payer-facing medical-necessity text and auto-fills prior-auth templates commonly required for BED pharmacotherapy (e.g., lisdexamfetamine) and higher levels of care—an EHR-structured gap most guides ignore when they rely on narrative notes alone.

This discrete-data model does what narrative documentation cannot: it creates auditable, machine-readable clinical evidence that feeds directly into payer-facing medical-necessity text and auto-populates prior-authorization templates. The insight competitors miss is that the documentation challenge for F50.81 is not a knowledge problem—most PMHNPs understand DSM-5-TR criteria—it is an EHR-structural problem. Free-text note fields do not prompt for the specific data payers audit. Structured fields do.

Scribing.io Clinical Logic: From Denied Prior Auth to First-Submission Approval

The Scenario

A PMHNP documents "frequent binges since spring" for a new patient and bills 99214 (established outpatient visit, moderate complexity) + 90837 (psychotherapy add-on, 53+ minutes). The note lacks a quantified binge count, a specific start date, a DSM-5-TR severity specifier, and an explicit statement of no compensatory behaviors.

The payer outcome: The insurer recoups $1,250 across the E/M and psychotherapy codes, citing insufficient documentation to support medical necessity for the level of service billed per AMA CPT guidelines. Separately, the lisdexamfetamine prior authorization is denied because the required baseline weekly binge count and 12-week history are absent from the clinical record. The FDA prescribing information for lisdexamfetamine specifies its indication for moderate-to-severe BED in adults—payers operationalize this by requiring documented severity meeting DSM-5-TR thresholds.

The PMHNP now faces a triple loss: revenue written off, medication access delayed for the patient, and administrative hours consumed by appeals.

How Scribing.io Prevents This Outcome: Step-by-Step Logic Breakdown

With Scribing.io's BED documentation module active, the same clinical encounter unfolds differently—not because the clinician's assessment changes, but because real-time structured prompts ensure every payer-required data point is captured during the session:

The Financial and Clinical Impact

The logic chain is unambiguous: the payer denial did not occur because the patient failed to meet BED criteria. It occurred because the documentation failed to prove the patient met BED criteria. Scribing.io closes that gap by converting clinical knowledge into structured, payer-auditable evidence at the point of care—before the claim is ever submitted.

Technical Reference: ICD-10 Documentation Standards for F50.81 and F50.2

ICD-10-CM provides two primary codes for episodic binge eating presentations. Selecting the wrong one—or documenting insufficiently to support either—triggers denials, recoupments, and downstream prior-auth failures. F50.81 - Binge eating disorder; F50.2 - Bulimia nervosa are the two codes PMHNPs must differentiate with surgical precision.

F50.81 — Binge Eating Disorder

F50.81 requires documentation of all five DSM-5-TR criteria (A through E) without evidence of regular compensatory behaviors. The CMS ICD-10-CM Official Coding Guidelines require that the code be supported by clinical documentation sufficient to assign it—meaning the five criteria must be traceable in the record. Subcodes at the 5th-character level specify severity:

• F50.810 — Mild (1–3 binge episodes per week)

• F50.811 — Moderate (4–7 binge episodes per week)

• F50.812 — Severe (8–13 binge episodes per week)

• F50.813 — Extreme (14+ binge episodes per week)

• F50.819 — Unspecified severity

Critical coding rule: F50.819 (unspecified) is a documentation failure, not a clinical decision. Every BED encounter should produce a weekly binge count that maps to a specific severity subcode. Payers increasingly reject F50.819 as evidence that the clinician did not quantify frequency—which cascades into denial of the entire encounter and any associated prior authorization.

F50.2 — Bulimia Nervosa

F50.2 applies when binge-eating episodes are followed by regular compensatory behaviors: self-induced vomiting, laxative or diuretic misuse, fasting, or excessive exercise. The compensatory-behavior attestation is therefore the single most important differential documentation element between F50.81 and F50.2. Scribing.io enforces this with a mandatory toggle field—the note cannot be finalized without a compensatory-behavior determination—eliminating the ambiguity that causes miscoding.

How Scribing.io Ensures Maximum Code Specificity

Scribing.io prevents F50.819 (unspecified) from ever being submitted by requiring the binge-frequency integer field to be populated before note finalization. The system then auto-maps the frequency to the correct severity subcode per DSM-5-TR thresholds. This is not a suggestion or a clinical decision support alert that can be dismissed—it is a hard stop in the documentation workflow. The result: every F50.81 claim leaves the practice with maximum ICD-10 specificity, eliminating a primary denial trigger. Research published in JAMA Psychiatry has consistently demonstrated that BED severity correlates with treatment response, making accurate severity coding a clinical imperative in addition to a billing one.

DSM-5-TR Criteria Crosswalk: Mapping Clinical Findings to ICD-10 Severity Specifiers

The DSM-5-TR defines BED across five criteria. Each criterion maps to a specific documentation requirement and, ultimately, to the ICD-10-CM code and subcode. This crosswalk eliminates ambiguity for PMHNPs documenting in real time.

Severity Specifier Mapping

DSM-5-TR severity is determined by weekly binge-episode frequency. This is the mapping that Scribing.io automates:

Note that the DSM-5-TR also permits severity to be "increased to reflect other symptoms and the degree of functional disability," which means a clinician may override the frequency-based severity upward based on clinical judgment—but the frequency-based threshold remains the minimum defensible documentation for payer purposes. Scribing.io accommodates clinician overrides while flagging the frequency-based calculation as the default for audit defense.

PMHNP Workflow: Structuring the BED Encounter From Intake to Prior Authorization

The typical PMHNP-BC encounter for BED spans intake assessment, diagnostic formulation, treatment planning, and—when pharmacotherapy is indicated—prior-authorization submission. Each phase has specific documentation requirements that, if missed, create downstream failures. This workflow standardizes the process.

Phase 1: Intake Assessment (First 15 Minutes)

1. Chief complaint capture: Document the patient's own words regarding eating behavior. Avoid paraphrasing into clinical jargon at this stage.

2. Binge frequency quantification: Ask directly: "How many times per week do you eat an amount of food that feels significantly larger than what most people would eat in a similar situation, during which you feel a loss of control?" Record the integer.

3. Onset date identification: Ask: "When did this pattern start?" Push past vague answers ("a few months ago") to a specific month and approximate date. Document the date.

4. Compensatory behavior screening: Systematically screen for self-induced vomiting, laxative use, diuretic use, fasting periods, and excessive exercise. Document each as present or absent. This is not optional—it is the differential gate between F50.81 and F50.2, and per NIMH epidemiological data, these conditions co-occur frequently enough that screening must be explicit.

Phase 2: Diagnostic Formulation (Minutes 15–25)

1. Criterion A verification: Confirm both elements—large amount of food in a discrete period AND sense of loss of control.

2. Criterion B checklist: Document ≥3 of 5 associated features. Scribing.io presents these as checkboxes with a minimum-3 validation.

3. Criterion C distress attestation: Elicit and document marked distress. Use the patient's own language plus your clinical assessment.

4. Criterion D auto-validation: Scribing.io calculates duration from onset date and confirms ≥12 weeks. If sub-threshold, the system flags "DSM-5-TR duration criterion not met" and prevents F50.81 from being assigned, instead suggesting "Other specified feeding or eating disorder" (F50.89) as appropriate.

5. Criterion E confirmation: Finalize the compensatory-behavior attestation. The system requires this before allowing F50.81 code assignment.

Phase 3: Treatment Planning and Code Assignment (Minutes 25–35)

1. Severity subcode assignment: Scribing.io auto-assigns based on documented frequency. Clinician reviews and confirms or overrides with documented justification.

2. Treatment plan documentation: If lisdexamfetamine is indicated, the system generates a treatment-rationale paragraph citing the FDA indication for moderate-to-severe BED, the patient's documented severity, and the inadequacy or contraindication of non-pharmacologic interventions alone.

3. CPT code alignment: The system validates that the documented encounter supports the selected E/M level (99213/99214/99215) and any add-on codes (90833/90836/90837) per AMA CPT coding standards.

Phase 4: Prior-Authorization Submission (Post-Encounter)

1. One-click packet generation: Scribing.io compiles the medical-necessity paragraph, all discrete diagnostic data, the treatment plan, and supporting documentation into a payer-specific prior-auth template.

2. Electronic submission: The packet exports as both PDF (for fax-based payers) and structured electronic data (for ePA-compatible payers).

3. 12-week tracking initiation: The system begins a time-stamped follow-up counter from the date of first BED assessment. Each subsequent encounter auto-populates the running duration and compares current binge frequency to baseline, generating treatment-response documentation that payers require for continued authorization.

Differential Documentation: Distinguishing F50.81 From F50.2 and Comorbid Presentations

The F50.81/F50.2 differential is the highest-stakes coding decision in eating disorder documentation. Get it wrong, and the payer denies for diagnostic inconsistency. Leave it ambiguous, and the auditor flags the claim. Here is how each scenario must be documented:

Comorbid Presentations

BED frequently co-occurs with major depressive disorder (F33.x), generalized anxiety disorder (F41.1), and obesity (E66.x). Each comorbidity should be documented with its own discrete data set and coded separately. Scribing.io's comorbidity module cross-references active diagnoses to flag potential coding conflicts—for example, alerting the clinician if both F50.81 and F50.2 are assigned simultaneously, which would indicate a documentation error since the two diagnoses are mutually exclusive by DSM-5-TR definition. Research in NIH/PubMed literature consistently demonstrates that comorbid documentation improves both clinical outcomes and reimbursement rates by establishing the full complexity of the patient's presentation.

Payer-Ready Output: Medical Necessity Language and Prior-Auth Packet Construction

The final output of every BED encounter should include a payer-ready medical necessity paragraph that can be dropped directly into a prior-authorization request, a letter of medical necessity, or an appeal. This paragraph must synthesize all discrete data into a concise, criterion-mapped statement. Here is the template Scribing.io generates:

Medical Necessity Statement — F50.81 Binge Eating Disorder

[Patient Name], a [age]-year-old [sex], meets full DSM-5-TR diagnostic criteria for Binge Eating Disorder, [Severity] ([ICD-10 subcode]). The patient reports [X] binge-eating episodes per week, characterized by consumption of objectively large amounts of food within a discrete period with a subjective sense of loss of control (Criterion A). Associated features include [list ≥3 of 5 Criterion B features] (Criterion B). The patient endorses marked distress regarding the binge eating, including [specific distress descriptors] (Criterion C). This pattern has been present since [onset date], a duration of [X] weeks, exceeding the DSM-5-TR minimum of 12 weeks (Criterion D). The patient denies self-induced vomiting, laxative or diuretic misuse, fasting, and excessive exercise as compensatory behaviors (Criterion E; rules out bulimia nervosa, F50.2).

[Medication name] is medically necessary because [clinical rationale: FDA indication, severity threshold, inadequate response to or contraindication for non-pharmacologic treatment alone]. Initiation of treatment is supported by the documented severity and duration of the disorder.

Prior-Auth Packet Components

Scribing.io's one-click export generates a complete prior-authorization packet containing:

1. Patient demographics and insurance information (auto-populated from EHR integration)

2. Prescriber information and NPI

3. Diagnosis code with severity subcode (F50.810–F50.813)

4. Medical necessity statement (auto-generated from discrete encounter data)

5. DSM-5-TR criteria attestation (structured, criterion-by-criterion)

6. Treatment history (prior interventions attempted, duration, outcomes)

7. Requested medication, dose, and duration

8. Supporting clinical documentation (encounter note with discrete data elements highlighted)

This packet satisfies the documentation requirements published by major pharmacy benefit managers and commercial payers for lisdexamfetamine prior authorization. The structured format—rather than a faxed narrative note—reduces payer processing time and eliminates the most common rejection reason: "insufficient clinical information to determine medical necessity."

Longitudinal Tracking: The 12-Week Follow-Up Counter

Payers do not just evaluate the initial prior authorization. Continued authorization at 6-month or 12-month intervals requires documented treatment response. Scribing.io's 12-week follow-up counter generates a longitudinal binge-frequency trend at each encounter:

This longitudinal dataset is the documentation that secures continued authorization. Without it, payers request "updated clinical information" at renewal—triggering the same administrative cycle that the initial structured approach was designed to prevent. Scribing.io generates this table automatically from encounter-over-encounter discrete data, exporting it as part of the renewal prior-auth packet.

The takeaway for every PMHNP-BC managing BED caseloads: documentation is not a back-office function. It is the clinical intervention that determines whether your patient gets their medication this week or in six weeks. The discrete-data model—structured fields, auto-calculated severity, mandatory compensatory-behavior attestation, and one-click prior-auth export—is the difference between a practice that writes off BED claims and one that collects in full on every encounter. Scribing.io makes that model operational from day one.