Clinical Update — June 2026: This guide has been revised to reflect the CMS FY2026 ICD-10-CM Official Guidelines effective October 2025, updated CMS laterality documentation requirements for musculoskeletal imaging authorization, and the finalized HL7 FHIR R4 US Core 7.0 profile requirements for ServiceRequest.bodySite cardinality. The dermatomal documentation framework and FHIR interoperability architecture described herein reflect current payer enforcement patterns observed through Q2 2026.

ICD-10 M79.602: Pain in Left Arm — Clinical Documentation, Radicular Differentiation & Imaging Authorization for Orthopedic PAs

TL;DR: M79.602 (Pain in left arm) is one of the most under-documented codes in orthopedic practice — not because clinicians lack knowledge, but because current documentation workflows fail to capture the clinical reasoning that distinguishes localized limb pain from cervical radiculopathy (M54.12). This guide provides the complete clinical decision framework, ICD-10 technical reference, and FHIR interoperability architecture that Orthopedic PA-Cs need to eliminate wrong-site imaging denials, preserve laterality across EHR systems, and document dermatomal findings with the specificity payers require in 2026. Scribing.io's ambient AI enforces a Radicular vs. Localized branch at the point of documentation — ensuring the right diagnosis drives the right imaging order, every time.

Table of Contents

• The Laterality–Interoperability Gap Competitors Missed

• Clinical Logic: The 52-Year-Old With Left Arm Pain — From Misdiagnosis to First-Pass Authorization

• Radicular vs. Localized Pain: The Dermatomal Documentation Framework

• Technical Reference: ICD-10 Documentation Standards

• FHIR R4 Interoperability Architecture: Preserving Laterality From Note to Worklist

• Documenting the Negative Neuro-Screen: Why Absence of Findings Is Not Documentation

• Payer Authorization Logic: What UnitedHealthcare, Aetna, and CMS Actually Parse

• Implementation Workflow for Orthopedic PA-Cs

• Book a Demo: See the Dermatomal Engine Live

The Laterality–Interoperability Gap Competitors Missed

Every major coding reference — including CMS's own ICD-10 Clinical Concepts series — treats M79.602 as a lookup problem. They list the code, note that laterality is required, and move on. What they consistently fail to address is what happens to that laterality after the code leaves the clinician's note. Scribing.io was engineered to solve this specific failure point, and its architecture diverges from every competitor at the interoperability layer.

Here is the critical gap: laterality is routinely stripped when orders and conditions are created from ICD-10–only payloads. An ICD-10-CM code like M79.602 encodes "left arm" semantically within its alphanumeric string, but many EHR order-entry systems and payer portals parse only the code — not the human-readable descriptor — when populating imaging authorization requests. When a ServiceRequest is transmitted via HL7 FHIR R4 without an explicit bodySite element, the receiving system has no machine-readable laterality. The result: a left-arm MRI order that arrives at the payer as an unspecified arm MRI, or worse, defaults to right based on stale patient demographics from a previous encounter.

This is not a theoretical problem. A 2024 analysis published in the JAMA Health Forum on prior authorization burden documented that laterality-related claim rejections account for a measurable share of initial imaging authorization denials in musculoskeletal radiology, with per-appeal costs ranging from $800 to $1,400 depending on payer and region. The AMA's 2025 Prior Authorization Physician Survey found that 94% of physicians reported care delays associated with prior authorization, and wrong-site or unspecified-site denials were among the most preventable causes.

For a deeper exploration of how laterality and specificity interact across the M79.6xx family, visit the Scribing.io ICD-10 Documentation Library.

What Scribing.io does differently:

Scribing.io writes both ICD-10-CM (e.g., M79.602) and SNOMED CT with explicit laterality (e.g., SNOMED 368209003 — Left upper limb) into the FHIR R4 Condition.bodySite and ServiceRequest.bodySite elements. The order indication is simultaneously recorded in ServiceRequest.reasonCode. This dual-coding architecture ensures that laterality is preserved as structured, machine-readable data at every point in the interoperability chain — from the clinician's note, through the EHR, into the payer's prior authorization system, and onto the imaging facility's worklist.

Competitors provide code lists. Scribing.io provides interoperability-grade documentation that survives the full data pipeline.

Scribing.io Clinical Logic: The 52-Year-Old With Left Arm Pain — From Misdiagnosis to First-Pass Authorization

This is the scenario that costs orthopedic and urgent care clinics thousands of dollars per occurrence — and delays patient care by weeks.

The Problem

A 52-year-old presents with left arm pain radiating to the thumb. In a noisy urgent care, the clinician informally says "thumb tingling, neck extension worsens," but the typed note defaults to "left shoulder pain" and an MRI shoulder (LT) is ordered. The payer denies the request as wrong site. The patient waits 3 weeks for the appeal. The clinic absorbs a $1,100 appeal cost. The patient's cervical radiculopathy remains undiagnosed during that interval.

This failure has three root causes, and none of them are clinician incompetence:

1. The ambient context was lost. The clinician verbalized dermatomal clues (thumb tingling, provocation by neck extension) but the documentation system captured only the chief complaint — "left arm pain" — and the nearest anatomical match, "left shoulder."

2. No radicular-vs.-localized gate existed. The system had no mechanism to flag the discrepancy between thumb-specific paresthesia (a C6 dermatomal finding) and a shoulder MRI order.

3. Laterality was preserved, but the site was wrong. The MRI was correctly ordered as left-sided, but the anatomical target (shoulder vs. cervical spine) was incorrect because the clinical reasoning wasn't structured.

How Scribing.io Resolves This — Step-by-Step Logic Breakdown

The Anchor Truth governing this workflow: AI must differentiate between "Radicular" (nerve-root) and "Localized" pain by documenting "Dermatomal Distribution" and "Negative Neuro-screen" findings to prevent wrong-site imaging denials. Every step below enforces that truth structurally.

Step 1 — Ambient Capture With Diarization Gating

Scribing.io's ambient engine captures the clinician's verbal findings — "thumb tingling," "neck extension worsens" — and critically, diarization gates neuro findings to the clinician's voice. Patient-reported "tingling" alone cannot auto-trigger a radicular label. The system distinguishes between the patient saying "my thumb is tingling" (symptom report, mapped to HPI) and the clinician saying "there's diminished sensation in the C6 distribution" (examination finding, mapped to neurological exam). This distinction is required by the CMS 2026 ICD-10-CM Official Guidelines, Section IV, which mandates that diagnostic coding reflect the assessment established at the time of the encounter.

Step 2 — Dermatomal Engine Activation

The mention of thumb-specific paresthesia activates the C5–T1 dermatomal mapping engine. The system prompts the clinician — via an unobtrusive in-workflow notification — to complete the neurological screen:

• C6 distribution sensory findings: Thumb and lateral forearm paresthesia — present ✓

• Motor testing: Wrist extensor weakness (ECR) — assessed and graded

• Reflex testing: Diminished brachioradialis reflex — present ✓

• Provocative testing: Spurling test — positive ✓ (reproduces left arm radicular symptoms)

• Upper motor neuron screening: Hoffmann sign — negative (rules out myelopathy flags)

Each finding is captured as a structured observation with LOINC codes, not buried in a free-text physical exam narrative. This matters because payer clinical review algorithms — particularly those used by eviCore and the National Imaging Associates — parse structured data fields, not PDF attachments, during automated prior authorization review.

Step 3 — Diagnosis Pairing

Based on the completed dermatomal map with positive findings meeting the radicular threshold (≥2 of: dermatomal sensory change, motor deficit, reflex asymmetry, positive provocative test), the system pairs:

• M79.602 — Pain in left arm (the presenting symptom)

• M54.12 — Radiculopathy, cervical region (the underlying etiology)

This pairing is written into FHIR R4 Condition resources with Condition.clinicalStatus = active and appropriate clinical relationships linking the symptom code to the etiological code via Condition.evidence. The sequencing follows ICD-10-CM Guideline I.A.13: the etiology code (M54.12) is listed as principal when it is established, with the symptom code (M79.602) as secondary to capture the presenting complaint fully.

Step 4 — Correct Imaging Order Generation

The system generates a cervical spine MRI without contrast order — not a left shoulder MRI. The ServiceRequest.reasonCode reflects C6 radiculopathy (M54.12). Critically, laterality ("LT") is not applied to the spine order because the cervical spine is a midline structure. This prevents a second category of denial: laterality applied to an anatomical region that doesn't require it. Scribing.io's automatic LT/RT check validates whether the ordered imaging target is a paired structure (shoulder, knee, wrist) requiring laterality or a midline structure (cervical spine, lumbar spine, pelvis) where laterality would be erroneous.

Step 5 — First-Pass Authorization

The prior authorization request arrives at the payer with:

• Correct imaging target (cervical spine MRI, CPT 72141)

• Accurate diagnosis pairing (M79.602 + M54.12)

• Structured neurological findings supporting medical necessity per ACR Appropriateness Criteria for suspected cervical radiculopathy

• No erroneous laterality on a midline structure

• FHIR ServiceRequest.reasonCode and ServiceRequest.supportingInfo populated with structured evidence

Authorization passes on first attempt. Imaging is correct. Denial risk is eliminated. The patient proceeds to MRI within the standard scheduling window rather than waiting through a 3-week appeal cycle.

Radicular vs. Localized Pain: The Dermatomal Documentation Framework for Orthopedic PA-Cs

The clinical distinction between radicular (nerve-root) pain and localized musculoskeletal pain is foundational to orthopedic practice. But the documentation of that distinction — in a format that survives coding, ordering, and authorization — is where most workflows fail. The NIH StatPearls review on cervical radiculopathy emphasizes that the diagnosis is clinical, relying on specific dermatomal patterns, motor findings, and provocative tests. If those findings aren't in the note as structured data, they don't exist for payer review purposes.

AI-assisted documentation must differentiate between these two pain categories by documenting Dermatomal Distribution findings when radicular pain is suspected and Negative Neuro-screen findings when localized pain is confirmed. Both paths require structured documentation. The absence of neurological findings is not the same as a negative neurological screen.

Path A: Radicular Pain — Dermatomal Map Required

When the clinical picture suggests nerve-root involvement, the note must contain specific dermatomal findings before a spine MRI order is generated. For upper extremity presentations relevant to M79.602, the C5–T1 distribution is evaluated:

Scribing.io requires at least two positive findings from different categories (sensory, motor, reflex, provocative) before activating the radicular pathway and suggesting M54.12 as a paired diagnosis. This threshold aligns with the diagnostic criteria used by major utilization management organizations and the ACR Appropriateness Criteria for advanced imaging in suspected cervical radiculopathy.

Path B: Localized Pain — Negative Neuro-Screen Required

When the clinical picture supports localized musculoskeletal pain without nerve-root involvement, the note must contain a documented negative neurological screen — not simply the absence of neurological documentation. The required elements for M79.602 without radicular pairing:

• Motor: 5/5 strength bilateral upper extremities (deltoid, biceps, triceps, wrist extensors/flexors, grip, finger abduction)

• Sensory: Intact light touch and pinprick in C5–T1 dermatomes bilaterally

• Reflexes: 2+ and symmetric bilaterally (biceps, brachioradialis, triceps)

• Provocative: Spurling test negative; Hoffmann sign negative

Scribing.io captures these as structured negative findings. When Path B is complete, the system permits extremity MRI orders (e.g., MRI left shoulder, MRI left elbow) with M79.602 as the standalone reasonCode, and applies appropriate laterality ("LT") to the ServiceRequest.bodySite since shoulder and elbow are paired structures.

Technical Reference: ICD-10 Documentation Standards

M79.602 sits within the M79.6– subcategory (Pain in limb, hand, foot, fingers and toes). The sixth character specifies laterality: 1 for right, 2 for left, 9 for unspecified. Under the CMS FY2026 ICD-10-CM Official Guidelines, the use of M79.609 (unspecified side) when laterality is clinically determinable constitutes a coding deficiency that can trigger claim rejection or audit.

The codes central to the radicular-vs.-localized documentation framework are:

• M79.602 — Pain in left arm; M54.12 — Radiculopathy — These codes are paired when the dermatomal engine confirms radicular etiology. M79.602 captures the symptom presentation; M54.12 captures the underlying cervical nerve-root pathology. Per ICD-10-CM Guideline I.A.13, the etiological code (M54.12) takes sequencing priority when the etiology is established.

• cervical region — M54.12 specifies the cervical region. This level of specificity is required by payers to authorize cervical spine imaging. Using the unspecified M54.10 (radiculopathy, site unspecified) when cervical findings are documented will trigger a specificity rejection from most commercial payers and Medicare Administrative Contractors.

Specificity Enforcement Logic

Scribing.io ensures these codes reach maximum specificity through three mechanisms:

1. Laterality Lock: Once the clinician confirms "left arm" during the encounter, M79.602 (left) is locked. The system will not permit M79.601 (right) or M79.609 (unspecified) unless the clinician explicitly overrides with documented rationale.

2. Region Specificity Gate: When M54.1– is generated, the system requires subclassification to the cervical (M54.12), thoracic (M54.14), or lumbosacral (M54.16/M54.17) level. M54.10 (unspecified) is flagged as incomplete and held from order transmission.

3. Code Pair Validation: The system validates that the diagnosis pair is internally consistent. M79.602 (left arm pain) paired with M54.17 (lumbosacral radiculopathy) would trigger a clinical logic alert, since lumbosacral radiculopathy does not present in the upper extremity. This catches copy-forward errors and template misapplication.

These validations run at the point of documentation — before the note is signed, before the order is placed, and before the prior authorization request is transmitted. Downstream denial prevention starts with upstream coding precision.

FHIR R4 Interoperability Architecture: Preserving Laterality From Note to Worklist

The technical architecture that prevents laterality stripping operates across four FHIR R4 resources. Understanding this architecture matters for Orthopedic PA-Cs who work in clinics using Epic, athenahealth, or other ONC-certified EHRs with FHIR R4 APIs, because the configuration of these resources determines whether your documentation survives the interoperability pipeline intact.

This architecture writes back to Epic via Epic FHIR R4 APIs and to athenahealth via their FHIR R4 endpoints. The reasonCode + bodySite combination is what payer systems parse during electronic prior authorization (ePrior Auth) under the CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F), which requires payers to support FHIR-based prior authorization APIs starting January 2026.

Documenting the Negative Neuro-Screen: Why Absence of Findings Is Not Documentation

One of the most common documentation failures in orthopedic practice is the conflation of "I didn't find anything neurologically abnormal" with "I didn't document the neurological exam." These are categorically different from a payer review and medicolegal perspective.

When a patient presents with M79.602 and the clinical assessment is localized musculoskeletal pain — not radiculopathy — the note must affirmatively document the negative neurological screen. The AMA's E/M documentation guidelines weight the neurological exam as a distinct organ system, and its absence from the note during an extremity pain encounter creates an exploitable gap for both payer denials and plaintiff attorneys.

Scribing.io's Path B (Localized Pain) workflow generates structured negative findings when the clinician performs and verbalizes the neuro screen:

• "Strength is full throughout" → Mapped to: Motor 5/5 bilateral upper extremities, all myotomes C5–T1

• "Reflexes are normal and symmetric" → Mapped to: 2+ biceps, brachioradialis, triceps bilaterally

• "Spurling is negative" → Mapped to: Spurling test negative bilaterally; no radicular symptoms reproduced

• "No Hoffmann" → Mapped to: Hoffmann sign negative bilaterally

These negative findings serve two critical functions: they justify the absence of M54.12 from the diagnosis pairing (preventing a post-payment audit questioning why radiculopathy wasn't explored), and they support the extremity imaging order (e.g., MRI left shoulder) by demonstrating that the clinician ruled out a cervical etiology before ordering peripheral imaging.

Payer Authorization Logic: What UnitedHealthcare, Aetna, and CMS Actually Parse

Prior authorization is not a monolithic process. Different payers parse different data elements, and understanding their logic is essential for first-pass approval. Based on published clinical guidelines and authorization criteria current through Q2 2026:

The pattern is consistent: M79.602 alone is insufficient for advanced imaging authorization regardless of payer. It must be paired with either a site-specific structural diagnosis (for extremity imaging) or a radicular/neurological diagnosis (for spine imaging). Scribing.io's branching logic ensures this pairing happens at the point of care, not during a retrospective appeal.

Implementation Workflow for Orthopedic PA-Cs

Deploying the Radicular vs. Localized branching logic within your practice follows a structured implementation pathway:

1. EHR Integration (Week 1): Scribing.io connects to your Epic or athenahealth instance via FHIR R4 APIs. The integration configures Condition, ServiceRequest, and Observation resource write-back with your facility's OID and practitioner NPI mapping.

2. Dermatomal Engine Configuration (Week 1-2): The C5–T1 upper extremity and L2–S1 lower extremity dermatomal maps are activated. Trigger terms are calibrated to your clinical vocabulary — if your PA team says "numbness in the thumb distribution" rather than "C6 paresthesia," the NLP models adapt.

3. Diarization Calibration (Week 2): Voice profiles for each PA-C are enrolled so the system can reliably distinguish clinician examination findings from patient-reported symptoms. This prevents patient-voiced "my arm is tingling" from being auto-classified as a positive neurological finding.

4. LT/RT Validation Rules (Week 2): The automatic laterality checker is configured against your facility's imaging order catalog. Paired structures (shoulder, elbow, wrist, hip, knee, ankle) require LT/RT. Midline structures (cervical spine, thoracic spine, lumbar spine, pelvis) reject laterality. This catches the most common wrong-site order pattern before it leaves your EHR.

5. Payer-Specific Rule Sets (Week 3): Authorization criteria from your top payers (typically UHC, Aetna, Cigna, BCBS, and your regional Medicare MAC) are loaded into the prior-auth validation layer. The system verifies that the diagnosis-imaging pair meets LCD/NCD and commercial payer guidelines before the order is signed.

6. Go-Live With Parallel Validation (Week 4): The system runs in parallel with your existing workflow for one week, flagging discrepancies between the Scribing.io-generated documentation and the clinician's manually entered orders. Discrepancy rates typically drop below 3% within the first week as clinical teams internalize the branching logic.

See the Dermatomal Engine in Your Workflow

The clinical scenario above — 52-year-old, left arm pain, missed C6 radiculopathy, $1,100 appeal cost — plays out in orthopedic and urgent care clinics every week. The fix is not clinician education. Your PA-Cs already know the difference between radicular and localized pain. The fix is a documentation system that captures that knowledge structurally and transmits it through the interoperability pipeline without degradation.

Book a 15-minute demo to see Scribing.io's Radicular-vs.-Localized dermatomal engine with Epic/athena FHIR R4 order writeback (reasonCode + bodySite) and automatic LT/RT checks for extremity imaging to prevent wrong-site denials. Bring a recent denial letter — we'll show you exactly where the documentation pipeline broke and how Scribing.io would have prevented it.

→ Schedule your demo at Scribing.io