TL;DR — The 100% Rejection Fix

The problem in plain terms: Medicare Advantage payers reject PT claims (97110, 97530, 97112) when notes use vague descriptors like "strength improving" or "ROM better." CMS LCD L34428 explicitly requires that "functional limitations [be documented] in terms that are objective and measurable."

The structural fix at issue: Scribing.io captures MMT and ROM as discrete FHIR R4 Observations — ROM as Observation.valueQuantity (UCUM deg) with laterality/movement in Observation.component, and MMT as Observation.valueCodeableConcept bound to the MRC 0–5 scale. These write directly to Raintree's discrete flowsheet fields (or via DOM selector mapping), auto-populating objective deltas (65°→92° knee flexion, MRC 3/5→4/5 quad) into progress notes and claim support.

The measurable result achieved: Denials overturned on appeal; subsequent cycles pass prepay edits with zero rejections. Model your recovered revenue with the ROI Calculator.

• Jump to sections below:

• The LCD L34428 Compliance Gap

• FHIR R4 Observation Encoding Model

• Clinical Logic: Reversing 100% TKA Denials

• ICD-10 Documentation Standards

• Raintree Deployment Workflow

• Pricing, ROI & Next Steps

The LCD L34428 Compliance Gap: Objective Data as Structure

CLINICAL UPDATE 2026: Revised for new CMS CPT G2211 standards, SB 1120 compliance, and FHIR interoperability.

CMS Local Coverage Determination L34428 (Outpatient Physical Therapy), revised effective 05/18/2023 under Palmetto GBA jurisdiction, contains a single sentence that drives the majority of Medicare Advantage post-pay reviews for PT services.

"The physician or NPP and/or therapist must document the patient's functional limitations in terms that are objective and measurable." — CMS LCD L34428, General PT Guidelines §5

Here is the operational gap: the LCD — and the guidance it references at Pub. 100-02, Chapter 15, §220.3 — never operationalizes the standard. It tells you the standard but not the data structure required to survive an automated prepay edit.

Medicare Advantage utilization engines in 2026 no longer read narrative prose on first pass. Their utilization-management engines run NLP-based objectivity checks that flag notes lacking numeric values with attached units.

• "Strength improving" contains zero parseable objective tokens for an adjudication engine.

• "MRC 4/5 quadriceps, right" carries a coded scale value, a muscle, and laterality.

The Anchor Truth of this playbook: 100% of a claim batch can be rejected not because the therapy was unnecessary, but because the documentation was semantically unstructured. The solution is not "write better notes" — it is capturing data as discrete, coded observations at the point of care.

Scribing.io's Insight: MMT & ROM as FHIR R4 Observations

Every competitor and every payer document treats ROM and MMT as free text that a clinician types into a note. Medical AI Scribing at Scribing.io treats them as machine-verifiable clinical facts. This is the core information-gain differentiator.

The Encoding Model Itself

Because these are discrete Observation resources — not prose — they can be written to Raintree's discrete flowsheet fields. When a Raintree instance exposes FHIR endpoints, Clinical-Grade Scribing writes the Observation directly.

When endpoints are unavailable, Scribing.io falls back to DOM selector mapping — the same architecture documented in our Scribing.io Eclinicalworks V12 Chrome Extension Dom Mapping Reference and our Scribing.io Bypassing Modmed Ema Click Tax Chrome Dom Mapping Reference.

What the LCD Missed Here

The LCD requires "objective and measurable." A discrete Observation is inherently objective and measurable — the value, unit, scale binding, and laterality are structurally guaranteed to be present, or the resource fails validation.

The downstream consequence is direct: objective deltas (65°→92° knee flexion, MRC 3/5→4/5 quadriceps) are auto-inserted into progress notes and claim support, eliminating the "vague descriptor" rejection trigger.

Clinical Logic: Reversing 100% TKA Denials (97110, 97530, 97112)

This is the centerpiece scenario for any Clinical Operations Director managing Medicare Advantage exposure. Follow the exact clinical and data flow below.

The Presenting Problem Stated

A 74-year-old patient completes 12 PT visits following a right total knee arthroplasty (TKA), billed with CPT 97110 (therapeutic exercise), 97530 (therapeutic activities), and 97112 (neuromuscular re-education).

A Medicare Advantage payer rejects 100% of the claims and initiates a post-pay review because the Raintree notes state "strength improving" and "ROM better" without objective values.

The Ambient Clinical Intelligence Workflow

Why the Appeal Succeeds

The appeal packet no longer contains "ROM better." It contains a documented functional trajectory: a 27-degree gain in right knee flexion, a full MRC grade improvement in quadriceps strength, and a 4-rep gain on the 30-second sit-to-stand.

Each measure now carries unit, scale, laterality, and date. These satisfy LCD L34428's "objective and measurable" requirement at the structural level, so the payer's own adjudication logic can verify medical necessity.

The overturn is not persuasion; it is compliance. Subsequent cycles pass prepay edits because the data structure — not the prose — carries the evidentiary weight.

Technical Reference: ICD-10 Documentation Standards

Objective measures are only defensible when linked to a coded diagnosis that supports medical necessity. For post-surgical and impairment-based PT, two codes recur in the TKA and generalized-weakness pathways.

Coding note on specificity: M25.60 is an unspecified-joint code; where the record supports it, prefer a laterality-specific stiffness code. Scribing.io's Observation.component laterality encoding provides the evidentiary basis for specifying the affected joint, reducing unspecified-code use that can itself trigger payer scrutiny.

Raintree Deployment Workflow for Clinical Operations

Deployment falls into two integration tiers depending on how a given Raintree instance is provisioned. Both write the same discrete Observations to the flowsheet.

Rollout Sequence for Directors

1. Audit denial history first — isolate claims rejected for vague descriptor language.

2. Map Raintree flowsheet fields to the ROM and MMT Observation schema.

3. Pilot on one TKA cohort to validate prepay-edit pass rates.

4. Scale across all PT payers once zero-rejection cycles are confirmed.

Specialty configuration details are documented under our physical therapy resources at Scribing.io Specialties, with jurisdiction-specific rules under AI Scribe Laws.

Pricing, ROI & Next Steps

The revenue case is direct: a single 12-visit TKA episode rejected across 97110, 97530, and 97112 represents recoverable dollars that structured Observations protect on the first pass.

• Review plan tiers and terms at Scribing.io Pricing & Plans.

• Quantify recovered denial revenue with the AI Medical Scribe ROI Calculator.

For a Clinical Operations Director, the decision reduces to a compliance question. Structured, coded objective measures make LCD L34428 conformance a property of the data, not of the clinician's prose discipline.