Clinical Update — June 2026: This guide has been revised to incorporate the CMS FY 2026 ICD-10-CM coding guidance effective October 2025, updated PBM formulary logic for ocular antihistamines (alcaftadine, cetirizine ophthalmic, bepotastine), and the ONC HTI-2 Final Rule requirements for real-time prior authorization (FHIR Prior Authorization Support Implementation Guide v2.1.0). Step-therapy evidence packaging standards now reflect the NCPDP SCRIPT 2024071 ePA attachment requirements adopted by the three largest PBMs (CVS Caremark, Express Scripts, OptumRx) in Q1 2026.

ICD-10 H10.10: Atopic Conjunctivitis, Unspecified Eye — Why Laterality and Papillary Hypertrophy Documentation Determine PBM Approval

TL;DR: H10.10 (Atopic conjunctivitis, unspecified eye) is the most common reason PBMs deny prior authorizations for high-cost ocular antihistamines — not because the drug is inappropriate, but because the documentation fails to specify laterality and capture the discrete slit-lamp findings that rule out bacterial conjunctivitis. Upgrading to H10.13 — Atopic conjunctivitis, bilateral requires explicit documentation of bilateral presentation and papillary hypertrophy — the two evidentiary anchors PBMs use to distinguish allergic from bacterial disease. Scribing.io's Allergy Eye micro-ontology automates this capture at the point of dictation, packaging step-therapy evidence into ePA submissions so approvals return the same day instead of triggering denials, callbacks, and delayed treatment.

Table of Contents

• The Documentation Gap Competitors Miss — How PBM ePA Logic Hinges on Laterality and Slit-Lamp Findings

• Technical Reference: ICD-10 Documentation Standards for Atopic Conjunctivitis

• Scribing.io Clinical Logic — Handling the Denied PA That Should Never Have Been Denied

• Step-by-Step Logic Breakdown: From Dictation to Same-Day Approval

• Anchor Truth: Why Bilateral + Papillary Hypertrophy Is the Evidentiary Gate

• ePA Packaging: X12 278/275 and FHIR PAS Attachment Architecture

• Workflow Comparison: Manual PA vs. Scribing.io Automated Pipeline

• Implementation: Activating Allergy Eye Logic in Your Practice

The Documentation Gap Competitors Miss — How PBM ePA Logic Hinges on Laterality and Slit-Lamp Findings

Every major competitor resource for ICD-10 H10.10 — including CMS's own MS-DRG definitions manual — treats the code as a flat lookup: Acute atopic conjunctivitis, unspecified eye. The CMS page lists H10.10 alongside H10.11, H10.12, and H10.13 as coordinate codes in the H10.1x family, with no clinical guidance on when to select one over another, no discussion of downstream payer consequences, and zero insight into how PBM formulary logic processes these codes.

This is the gap that costs ophthalmology practices thousands of dollars in denied prior authorizations annually — and costs patients days or weeks of delayed therapy. Scribing.io exists to close it. The platform's Allergy Eye micro-ontology maps spoken clinical findings to the structured data that PBM adjudication engines actually parse, ensuring the documentation speaks the payer's language before it ever leaves the EHR.

What competitors explain: The difference between allergic and bacterial conjunctivitis at a textbook level. Code descriptions. Excludes1 and Includes notes.

What competitors miss entirely:

• PBM electronic prior authorization (ePA) adjudication engines parse the ICD-10 code before reviewing clinical attachments. An H10.10 submission signals to the PBM that the clinician did not confirm laterality — raising an automated flag that the diagnosis may be insufficiently documented. Per the AMA's 2025 Prior Authorization Physician Survey, 94% of physicians report care delays attributable to PA, and ophthalmic topical agents rank among the highest-denial drug categories.

• The two discrete clinical findings that PBMs require to distinguish allergic from bacterial conjunctivitis and authorize non-preferred ocular antihistamines are bilateral presentation and papillary hypertrophy with absence of mucopurulent discharge. These are the Anchor Truth findings.

• When an EHR defaults to "unspecified eye" because the clinician examined both eyes but never verbalized laterality, the entire ePA submission is undermined at its foundation — regardless of how thorough the clinical note may read in narrative form.

The insight is structural: the approval pathway for high-cost ocular antihistamines (alcaftadine, cetirizine ophthalmic, bepotastine) runs through a logic gate that checks for bilateral disease documentation + papillary hypertrophy + bacterial rule-out + step-therapy failure. Miss any one of these discrete data elements and the ePA returns denied — not because the treatment is wrong, but because the documentation didn't speak the PBM's language.

This article is the definitive clinical-library resource for understanding why H10.10 is a documentation trap, how to correctly document for H10.13 when bilateral disease is present, and how Scribing.io automates the entire capture-to-approval pipeline. For the complete ICD-10 documentation framework, see the Scribing.io ICD-10 Documentation Library.

Technical Reference: ICD-10 Documentation Standards for Atopic Conjunctivitis

Understanding the H10.1x code family requires precision about what each laterality extension communicates — clinically and administratively. The CMS ICD-10-CM Official Guidelines for Coding and Reporting (FY 2026) state in Section I.B.13 that laterality codes should reflect the side(s) affected, and that "unspecified" should only be used when clinical information is insufficient to determine laterality. For atopic conjunctivitis — a condition bilateral in approximately 95% of presentations per NIH/PubMed epidemiologic data — defaulting to unspecified is almost always a documentation failure, not a clinical reality.

Key Documentation Distinctions

H10.10 is not a "safe default." In ICD-10-CM convention, "unspecified" codes signal that the clinician either did not examine for laterality or did not document it. For atopic conjunctivitis — a condition that is bilateral in the vast majority of presentations — defaulting to H10.10 is a clinical documentation paradox: the disease is almost always bilateral, yet the code says the clinician didn't specify.

When a clinician uses H10.10, PBM algorithms reasonably infer one of two things:

1. The documentation is incomplete (most likely), or

2. The presentation is atypical enough that the clinician wasn't confident in laterality (raising questions about the diagnosis itself).

Neither inference supports PA approval.

The 4th-character laterality trap: Unlike many ophthalmic ICD-10 code families, the H10.1x series uses the 4th character for laterality (0 = unspecified, 1 = right, 2 = left, 3 = bilateral). There is no placeholder "X" or 7th-character extension. The laterality decision is baked directly into code selection, meaning EHR systems that auto-populate H10.10 when laterality is undocumented create an immediate downstream problem that cannot be fixed by addenda or modifiers after the ePA has already been submitted.

Relationship to Adjacent Codes — Differential Documentation

Clinicians — especially cornea/external disease specialists evaluating complex presentations — must understand how H10.1x intersects with related conjunctivitis codes. The JAMA Ophthalmology literature on allergic eye disease classification makes clear that papillary morphology and discharge character are the primary differentiators:

Scribing.io's micro-ontology maps each of these differentials in real time. When the clinician mentions "giant papillae" and "no contact lenses," the system weighs H10.13 over H10.41x and surfaces the distinction for one-tap confirmation. When the clinician says "stringy discharge" rather than "purulent discharge," the system registers the bacterial rule-out automatically.

Scribing.io Clinical Logic — Handling the Denied PA That Should Never Have Been Denied

A cornea specialist sees a 14-year-old with severe itching, giant papillae on the upper tarsal plate, stringy discharge, and no mucopurulent crusting. The clinician prescribes a non-preferred antihistamine/mast-cell stabilizer.

This is the exact scenario where documentation determines whether a patient starts therapy today or waits two weeks through a denial-appeal cycle.

Scenario A: Without Scribing.io — The Preventable Denial

1. Exam: The cornea specialist examines both eyes, notes giant papillae on the upper tarsal plate under slit-lamp, observes stringy mucoid discharge bilaterally, and confirms severe pruritus as the chief complaint. No mucopurulent discharge. No lid matting.

2. Dictation: "Fourteen-year-old with allergic conjunctivitis. Giant papillae upper tarsal plate. Stringy discharge. Significant itching. Plan: alcaftadine 0.25% one drop each eye daily."

3. EHR Processing: The ambient scribe captures the narrative. Because the clinician said "allergic conjunctivitis" without explicitly stating "bilateral," the EHR's code-suggestion algorithm offers H10.10 (unspecified eye) as the default. The clinician, mid-workflow, clicks accept.

4. The note reads well clinically — but the structured data tells a different story:

   • ICD-10: H10.10 (unspecified eye)

   • Papillary hypertrophy: Mentioned in narrative, but not captured as a discrete, queryable finding

   • Bacterial rule-out: Not explicitly stated

   • Step-therapy evidence: No dates of prior OTC ketotifen/olopatadine failure documented

5. ePA Submission: The practice's PA team submits the request. The PBM's adjudication engine receives a code that doesn't confirm bilateral disease, no discrete bacterial rule-out findings, no step-therapy failure dates. Result: Denied. Reason code: "Documentation does not support medical necessity for non-preferred agent. Please provide evidence of bilateral presentation, clinical differentiation from bacterial etiology, and prior OTC therapy failure."

6. Downstream impact: The denial triggers a callback to the patient's family. The specialist must either write a peer-to-peer appeal letter, schedule a re-visit to re-document, or switch to a less effective formulary agent. The AMA 2025 PA survey data indicates the average ophthalmology PA appeal takes 5–14 business days. The 14-year-old — in the middle of a school semester — waits.

Scenario B: With Scribing.io's Allergy Eye Micro-Ontology — Same-Day Approval

1. Exam: Identical clinical encounter.

2. Dictation: Same words: "Fourteen-year-old with allergic conjunctivitis. Giant papillae upper tarsal plate. Stringy discharge. Significant itching. Plan: alcaftadine."

3. Scribing.io Processing — Three Critical Interventions: (detailed in the next section)

Step-by-Step Logic Breakdown: From Dictation to Same-Day Approval

Here is the granular, real-time sequence of how Scribing.io resolves the documentation failure described above. Each step maps directly to the Anchor Truth logic: AI must document "Bilateral Presentation" and "Papillary Hypertrophy" to clinically rule out Bacterial Conjunctivitis, providing the evidentiary base required for PBM approval of high-cost ocular antihistamines.

Intervention 1 — Non-Verbalized Reasoning Detection: Laterality

Scribing.io's allergy eye micro-ontology detects that the clinician described findings (giant papillae, stringy discharge) and prescribed treatment for "each eye" — indicating both eyes were examined — but never explicitly stated "bilateral." The system's non-verbalized reasoning detector identifies the mismatch between the treatment plan (bilateral dosing) and the absence of a laterality statement in the dictation. Before note sign-off, a 2-second confirmation prompt appears:

"You examined both eyes and described tarsal papillary hypertrophy. Confirm bilateral and add bacterial rule-out?"

One tap: "Confirm."

This single interaction accomplishes three things simultaneously:

• Updates the ICD-10 code from H10.10 to H10.13 (bilateral)

• Inserts the laterality statement as a discrete, queryable data element — not just narrative text

• Triggers Intervention 2

Intervention 2 — Structured Finding Extraction: Papillary Hypertrophy and Bacterial Rule-Out

Upon bilateral confirmation, Scribing.io's ontology engine parses the dictation for the findings required by PBM adjudication logic and converts narrative language into discrete structured data:

The critical distinction: "stringy discharge" and "no mucopurulent crusting" existed in the clinician's narrative in Scenario A too. But they existed as free text. In Scenario B, Scribing.io converts them to structured findings with explicit present/absent flags. PBM adjudication engines — which process X12 275 attachment data — can parse structured findings. They cannot reliably NLP-extract clinical meaning from free-text notes.

Intervention 3 — Step-Therapy Evidence Packaging

Scribing.io's medication reconciliation module cross-references the patient's pharmacy claims history (via Surescripts or local EHR medication list) and the encounter history for prior OTC therapy documentation. For this 14-year-old, the system identifies:

• OTC ketotifen fumarate 0.025% used for 6 weeks (dates: March 15 – April 26, 2026) — documented in a prior visit note

• Patient/parent reported "minimal improvement" — captured in today's HPI

• No prior prescription antihistamine claims

Scribing.io auto-populates the step-therapy section of the ePA with:

• Drug tried: Ketotifen fumarate 0.025% (OTC)

• Duration: 6 weeks (exceeds typical PBM minimum of 2–4 weeks)

• Outcome: Inadequate response

• Source: EHR encounter note [date] + patient-reported history [today's date]

No staff member had to hunt through prior notes. No callback to the patient's parent to confirm OTC duration. The data was already in the system; Scribing.io surfaced and structured it.

Anchor Truth: Why Bilateral + Papillary Hypertrophy Is the Evidentiary Gate

PBM formulary committees construct their ePA logic trees based on clinical practice guidelines — primarily the AAO Preferred Practice Patterns for Conjunctivitis and the NIH/NEI classification of ocular allergy. The core diagnostic logic separating allergic from bacterial conjunctivitis distills to two binary questions:

1. Is the presentation bilateral? Bacterial conjunctivitis is typically unilateral at onset (may become bilateral secondarily). Allergic conjunctivitis is almost universally bilateral from the start. A bilateral code (H10.13) immediately shifts the pretest probability toward allergic etiology.

2. Is papillary hypertrophy present with mucoid (not mucopurulent) discharge? Papillary hypertrophy of the upper tarsal plate — especially giant papillae (≥1 mm) — in the absence of mucopurulent discharge is pathognomonic for allergic/atopic conjunctivitis. Bacterial conjunctivitis produces follicular reaction or papillary response with purulent discharge, not the stringy, ropy mucoid discharge of allergic disease.

When both conditions are met and documented as structured data, the PBM's logic gate opens. The ePA reviewer — human or algorithmic — can confirm: bilateral disease + papillary hypertrophy + absence of bacterial indicators = allergic etiology established. Non-preferred antihistamine justified after step-therapy failure.

When either condition is missing from the structured data, the gate stays closed — even if the clinical narrative makes the diagnosis obvious to any ophthalmologist reading the note. The CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F) mandates payer adoption of electronic PA by 2027, which means this structured-data dependency will only intensify.

ePA Packaging: X12 278/275 and FHIR PAS Attachment Architecture

Scribing.io does not stop at documentation correction. The platform packages the clinical evidence into the ePA transaction format the PBM requires:

The 275 attachment is where most manual PA workflows fail. Staff typically attach a PDF of the clinical note — which the PBM adjudication engine cannot parse for discrete findings. Scribing.io transmits structured, coded data elements that map directly to the PBM's logic gates. The difference between attaching a PDF and transmitting structured findings is the difference between a denial and a same-day approval.

Workflow Comparison: Manual PA vs. Scribing.io Automated Pipeline

Implementation: Activating Allergy Eye Logic in Your Practice

The Allergy Eye micro-ontology is part of Scribing.io's Ophthalmology Module and activates automatically when the platform detects an allergic conjunctivitis encounter context (chief complaint of itching/redness, ICD-10 H10.1x or H10.4x consideration, or prescription of an ocular antihistamine/mast-cell stabilizer). No additional configuration is required beyond the standard Scribing.io EHR integration.

What the Micro-Ontology Monitors For

• Laterality signals: "both eyes," "each eye," "OU," bilateral dosing instructions, or examination of both eyes without explicit laterality statement (triggers non-verbalized reasoning prompt)

• Papillary morphology: "papillae," "giant papillae," "cobblestoning," "papillary hypertrophy," "tarsal plate findings," "papillary reaction" — mapped to structured finding with size qualifier when stated

• Discharge character: "stringy," "ropy," "mucoid," "clear," "watery" → allergic indicators; "purulent," "mucopurulent," "yellow-green," "crusting," "matted lids" → bacterial indicators. The system documents both present and absent findings.

• Symptom hierarchy: "itching" / "pruritus" as predominant symptom vs. "pain" / "photophobia" / "foreign body sensation." Pruritus predominance is a discrete allergic indicator for PBM logic.

• Contact lens status: Documented to differentiate H10.13 (atopic) from H10.41x (giant papillary conjunctivitis associated with CL wear).

• Atopic history: Eczema, asthma, allergic rhinitis — cross-referenced from the problem list to strengthen allergic etiology documentation.

• Step-therapy history: OTC ketotifen, OTC olopatadine (Pataday), prescription olopatadine, prior steroid use and steroid intolerance — duration and outcome pulled from medication history.

EHR Integration Points

Scribing.io integrates via certified SMART on FHIR or HL7 v2 ADT/ORU interfaces with all major ophthalmic EHRs (Nextech, ModMed/EMA, Epic Ophthalmology, Compulink, DrChrono). The integration ensures that:

• ICD-10 code selection in the EHR reflects Scribing.io's laterality-confirmed output (H10.13 populates the encounter diagnosis, not H10.10)

• Structured findings flow into the EHR's discrete data fields, not just free-text note sections

• ePA transactions originate from within the EHR's prescribing workflow — no portal switching, no separate PA software

What This Means for the 14-Year-Old

The patient in our scenario leaves the clinic. By the time the family arrives at the pharmacy — often the same afternoon — the PA has already been approved. Alcaftadine 0.25% is dispensed. Therapy begins. No phone call from the pharmacy saying "your insurance denied this." No parent taking time off work for a re-visit. No two-week gap where the child suffers through severe pruritus at school while an appeal winds through a PBM queue.

That is what documentation precision delivers when it is automated at the point of care.

See it in action. Book a 15-minute demo to see Bilateral Logic + Papillary Hypertrophy prompts generate H10.13 automatically and push a one-click ePA (X12 278/275 or FHIR PAS) with step-therapy evidence to slash PBM denials.

Book Your Demo — Scribing.io