STRATUM INTERNAL
PATTERN ANALYSIS
New denial KB patterns, SERI status, and corpus coverage gaps
13 NEW PATTERNS
TOS/DAU CLEARANCE FRAMEWORK — Updated 2026-04-27

Platform Ready: Patterns derived from public federal/regulatory sources (DOL, OIG, CMS, ERISA courts, TC-PUF, academic literature, Stratum prevention engine).

Internal Only: Patterns derived from CA/TX/NY state case-level corpus. Awaiting counsel opinion — legal review overdue from April 21 target.

11 of 13 patterns (85%) are platform-ready pending counsel confirmation.
NEW KB DENIAL PATTERNS — PENDING CORPUS ADDITION
KB-PATTERN-001
NQTL Parity Violation
HIGH PRIORITY REGULATORY PLATFORM READY
Non-Quantitative Treatment Limitation (NQTL) violations occur when payers apply more restrictive prior authorization requirements, clinical criteria, or step therapy protocols to behavioral health services than to comparable medical/surgical services. This is a federal MHPAEA violation category with strong enforcement precedent and high appeal success rates.
DOL has documented 135 enforcement actions citing NQTL violations across commercial health plans (2021–2024)
71 documented NQTL patterns identified across major commercial payers — prior auth requirements, step therapy, fail-first protocols, and geographic limitations applied asymmetrically to BH vs. medical/surgical
Appeal success rate: 70–86% when NQTL violation is correctly documented and cited — one of the highest-yield denial types available
CAA 2023 requires plans to submit comparative analyses to regulators; denial letters that cannot withstand NQTL scrutiny are increasingly being reversed at the internal appeal stage
Appeal Success
70–86%
DOL Actions
135
Patterns Documented
71
Corpus Status
Pending addition
Playbook
Not yet created
KB-PATTERN-002
Level of Care (LOC) Denial
HIGH PRIORITY CLINICAL CRITERIA PLATFORM READY
LOC denials occur when a payer refuses to authorize or continue a specific level of care (inpatient, residential, PHP, IOP, outpatient) claiming the patient's clinical needs do not meet the criteria for that level. Three sub-types require distinct documentation strategies: LOC Initiation (initial auth denied), Step-Down Failure (payer demands step-down before clinical stability), and Continued Stay (ongoing auth denied for active treatment episode).
LOC Initiation: Payer denies initial auth for requested level; clinical documentation must demonstrate LOCUS/ASAM criteria met at requested level — not just at a lower level
Step-Down Failure: Payer demands step-down from residential to PHP or IOP before patient is clinically stable; MHPAEA applies — payer cannot demand step-down faster than clinical criteria warrant
Continued Stay: Mid-episode auth denial; requires demonstrating active treatment response AND remaining risk factors that justify continued level
Withing the current corpus, LOC denials are under-documented — present in Optum and Anthem precedents but not formally categorized as a standalone type
Sub-types
3 (Initiation, Step-Down, Continued Stay)
Primary Criterion
LOCUS / ASAM
Corpus Status
Pending formal categorization
Playbook
Not yet created
KB-PATTERN-003
SUD Medical Necessity
MEDIUM PRIORITY CLINICAL CRITERIA INTERNAL ONLY
Substance Use Disorder (SUD) medical necessity denials specifically require ASAM criteria documentation — the American Society of Addiction Medicine multidimensional assessment framework. These denials differ from general medical necessity denials in that payers use ASAM as the accepted clinical standard, and appeals that do not reference ASAM dimensions explicitly have significantly lower success rates.
ASAM criteria cover 6 dimensions: acute intoxication/withdrawal, biomedical conditions, emotional/behavioral/cognitive conditions, readiness to change, relapse/continued use potential, recovery/living environment
Current TX corpus overturn rate for SUD medical necessity: ~46% — below other categories, suggesting systematic underuse of ASAM documentation in appeal letters
Most at-risk payer: Optum — enforces ASAM most rigorously; appeals without explicit ASAM mapping are almost never successful
Corpus Overturn
~46%
Primary Criterion
ASAM (6 dimensions)
Key Payers
Optum, Anthem, UHC
Corpus Status
Needs dedicated categorization
KB-PATTERN-004
Documentation Gap (Preventable Sub-type)
MEDIUM PRIORITY OPERATIONAL PLATFORM READY
A distinct sub-category of documentation denials: denials that result from preventable administrative failures rather than genuine clinical insufficiency. These include missing LOCUS forms not submitted with the claim, progress notes not attached, and outdated clinical assessments cited instead of current records. The existing DOC-001 playbook addresses post-denial recovery — this pattern addresses prevention.
Estimated 15% of behavioral health denials are preventable through documentation checklist compliance — missing LOCUS/clinical docs submitted at the wrong time or not at all
Magellan and Cigna show the highest concentration of preventable doc gap denials in this corpus — both require specific clinical form submission with claims
Prevention protocol: mandatory pre-submission checklist (LOCUS, GAF, progress notes, treatment plan sign-off) reduces preventable denials without increasing appeal burden
Preventable Denials
~15% of doc denials
Key Payers
Magellan, Cigna
Intervention Type
Pre-submission checklist
Corpus Status
Needs separation from DOC-001
KB-PATTERN-005
Phantom Network / Out-of-Network Denial
MEDIUM PRIORITY NETWORK ADEQUACY PLATFORM READY
Phantom network denials occur when a payer claims in-network BH providers are available, but the listed providers are not actually reachable — wrong phone numbers, full panels, no longer practicing, or not accepting the patient's specific plan. When a patient receives services OON because in-network providers are effectively inaccessible, the payer cannot deny at OON rates — this is a network adequacy violation.
EBSA (DOL) study found only 8% of listed behavioral health providers in Tennessee were reachable and accepting new patients — evidence that phantom networks are widespread in BH
Documentation for phantom network appeals: call log showing 3+ unsuccessful attempts to reach in-network providers, with date/time/outcome for each contact attempt
Network adequacy standards are set by state regulators for fully-insured plans; ERISA preemption limits state authority over self-funded plans but DOL has enforcement jurisdiction
No current corpus precedents categorized as phantom network — priority is to identify and add existing cases that meet this pattern
EBSA BH Reachability
8% (TN study)
Required Documentation
Call logs (3+ attempts)
Regulatory Authority
State + DOL/EBSA
Corpus Status
No precedents yet — gap
KB-SCHEMA-001
denial_stage Dimension
SCHEMA DATA MODEL PLATFORM READY
The corpus currently does not capture the service timeline dimension for denials — specifically whether a denial occurred before (pre-service) or after (post-service) care was delivered. This matters because appeal strategies differ substantially between the two: pre-service denials can trigger urgent/concurrent review rights, while post-service denials must meet substantive criteria for retroactive authorization. Adding denial_stage as a corpus dimension enables this stratification.
OIG found 18% of post-service denials for Medicare Advantage were inappropriate and would not have been denied under proper medical review — post-service denials are more contestable than payers imply
Pre-service denials trigger expedited (72-hour) review rights under ERISA if ongoing treatment is at issue — this right is lost if the denial is characterized post-service
Proposed schema: denial_stage: "pre-service" | "concurrent" | "post-service" | "retrospective"
OIG Post-Service Inappropriate Rate
18% (MA)
Pre-Service Expedited Window
72 hours (ERISA)
Implementation Priority
Platform Schema Change
Status
Proposed — awaiting engineering review
WEEK 2 SWEEP — NEW PATTERNS DETECTED (2026-04-27)
KB-PATTERN-007
ERISA Procedural Failure / Meaningful Dialogue
HIGH PRIORITY ERISA / PROCEDURAL PLATFORM READY
Payer failed to engage in "meaningful dialogue" as required by ERISA during the claims and appeals process. Four federal circuit court cases (10th Cir. 2023 ×2, 5th Cir. 2024, D. Utah 2025) establish this as a usable procedural reversal path against UHC/UBH regardless of the underlying clinical merit. This argument fires in parallel with clinical appeals — not as an alternative.
Source: behavioral_health_precedents_2026-04-03.json — 4 federal circuit court cases (UHC/UBH)
10th Cir. 2023 (×2), 5th Cir. 2024, D. Utah 2025 — consistent holding that payer's failure to engage claimant in substantive dialogue about denial basis is an ERISA procedural violation
Payer-specific: UHC / UBH only. Cross-jurisdictional — applicable across all states where these circuits have precedent
Use as parallel argument: raise procedural failure alongside clinical arguments in the same appeal letter — do not substitute one for the other
Payer Scope
UHC / UBH (all jurisdictions)
Circuit Record
10th Cir. 2023, 5th Cir. 2024, D. Utah 2025
Argument Type
Procedural — parallel to clinical
Corpus Status
Pending addition — HIGH
KB-PATTERN-008
Step Therapy / Fail-First NQTL
HIGH PRIORITY NQTL / REGULATORY PLATFORM READY
Sub-type of NQTL Parity Violation (KB-PATTERN-001) representing a specific enforcement track. BH-specific step therapy requirements are a per se MHPAEA violation per Colorado DOL guidance, and a Taft-Hartley settlement in Pennsylvania established the same principle for self-funded plans. Appeal argument: "No comparable step therapy requirement exists for medical inpatient."
Source: state_parity_compliance_tracker.json — Colorado DOL: BH-specific step therapy is per se MHPAEA violation; PA Taft-Hartley settlement
Colorado DOL enforcement guidance explicitly names fail-first BH protocols as NQTL violations when no equivalent exists for medical/surgical inpatient
PA Taft-Hartley settlement extends applicability to self-funded ERISA plans — important because DOL/EBSA has jurisdiction, not state regulators, for self-funded accounts
Appeal Success (Expected)
75–82%
Parent Type
NQTL Parity Violation (KB-PATTERN-001)
Key Jurisdictions
CO, PA (self-funded)
Corpus Status
Add as HIGH sub-type under NQTL
KB-PATTERN-009
Frequency/Duration — Acuity Justification Gate
MEDIUM PRIORITY CLINICAL CRITERIA PLATFORM READY
A 2-part evidence gate pattern currently misclassified under Benefit Limit / Frequency in the corpus. Requires two specific documentation elements to successfully overturn: (1) acuity_justification and (2) failure_to_respond documentation. Volume signal: 8.5M denials nationally PY2025 in this category per TC-PUF data.
Source: Prevention engine rule PR-D5 (rules-engine/prevention-engine.ts); TC-PUF volume 8.5M denials nationally PY2025
Two required evidence elements: acuity_justification (documenting clinical severity requiring the frequency) + failure_to_respond (documenting that lower-frequency treatment was tried and insufficient)
Currently misclassified — merging with Benefit Limit / Frequency obscures the specific 2-part gate requirement and leads to incomplete appeal letters
National Volume (PY2025)
8.5M denials (TC-PUF)
Evidence Gates
acuity_justification + failure_to_respond
Current Classification
Misclassified — Benefit Limit
Corpus Status
Add as standalone MEDIUM type
KB-PATTERN-010
Service Coverage Denial — Literature + Specialist Authority
MEDIUM PRIORITY COVERAGE PLATFORM READY
A distinct denial type that requires a different reversal path from Benefit Exclusion denials. Reversal requires two specific elements: peer-reviewed clinical literature establishing the service as medically accepted and a specialist authority statement supporting the specific treatment. National volume: 11M PY2025 in the "Services Excluded" TC-PUF category.
Source: Prevention engine rule PR-D4; TC-PUF "Services Excluded" category 11M PY2025
Reversal path differs from Benefit Exclusion: requires peer-reviewed clinical literature (establishing service as accepted standard) + specialist authority statement (treating or consulting specialist endorsement)
Conflating with Benefit Exclusion denials leads to under-resourced appeals — exclusion appeals argue coverage language, service coverage appeals argue clinical standard-of-care
National Volume (PY2025)
11M denials (TC-PUF)
Required Elements
Literature + Specialist statement
Distinction From
Benefit Exclusion (different argument)
Corpus Status
Add as distinct MEDIUM type
KB-PATTERN-011
Improvement Standard Misapplication
MEDIUM PRIORITY CLINICAL CRITERIA INTERNAL ONLY
Payer denies continued residential or inpatient care on the grounds that the patient is improving. The legal and clinical standard is settled: improvement during treatment does not equal "no longer medically necessary." A federal judge found a CA insurer unreasonably ignored medical evidence and misapplied improvement as a basis for denying continued care (CA DMHC precedent). Critical for active residential continued-stay denials.
Source: behavioral_health_precedents_2026-04-03.json — CA DMHC precedent: "Federal judge found insurer unreasonably ignored medical evidence and misapplied improvement as basis for denying continued care"
Core argument: improvement during treatment ≠ no longer medically necessary. The patient improving is evidence the current level of care is working — not evidence it should stop
Highest-risk scenario: residential continued-stay reviews mid-episode. Document remaining risk factors and stabilization goals that justify continued stay alongside clinical improvement notes
Precedent Source
CA DMHC / Federal court
Primary Scenario
Residential continued-stay denial
Core Argument
Improvement ≠ discharge-ready
Corpus Status
Add as MEDIUM type
KB-PATTERN-012
Reimbursement Rate Disparity / Inadequate Recruitment
MEDIUM PRIORITY NETWORK ADEQUACY PLATFORM READY
The network exists on paper but is structurally inadequate because BH reimbursement rates are too low to recruit or retain providers. The appeal argument shifts from directory accuracy to rate structure: the reimbursement rate differential itself is the NQTL — not individual directory errors. This is distinct from last week's Phantom Network pattern (KB-PATTERN-005), which focuses on inaccessible listed providers.
Source: state_parity_compliance_tracker.json — PA: BH claims paid at 88–98% Medicare vs. 120–123% for M/S; MD multi-plan enforcement action on rate disparity
Pennsylvania data: BH providers reimbursed at 88–98% of Medicare while M/S providers receive 120–123% of Medicare — rate gap is the structural cause of network failure
Maryland multi-plan enforcement established that rate disparity creating differential recruitment capacity is itself a parity violation — not just a business practice
Distinct from phantom network: this pattern is applicable even when providers are reachable — the argument is network structural inadequacy due to rates, not directory inaccuracy
PA Rate Gap
BH 88–98% vs. M/S 120–123% Medicare
Key Enforcement
MD multi-plan action
Distinction From
KB-PATTERN-005 (Phantom Network)
Corpus Status
Add as sub-type under Network patterns
KB-PATTERN-013
Pediatric BH Network Adequacy Disparity
MEDIUM PRIORITY NETWORK ADEQUACY PLATFORM READY
Pediatric-specific network adequacy disparity: standards for BH access are materially lower than for medical/surgical, creating a parity violation at the population level. Florida data: 85–90% pediatric BH access goal vs. 95% pediatric M/S access goal. Michigan data: 95% M/S within 10 miles vs. 85% BH within 30 miles — a 20-mile additional travel requirement for BH care. Sub-type of Network patterns, pediatric-specific.
Source: state_parity_compliance_tracker.json — FL: 85–90% pediatric BH access goal vs. 95% pediatric M/S; MI: 95% M/S within 10mi vs. 85% BH within 30mi
Florida disparity: pediatric BH network adequacy target (85–90%) is materially below pediatric M/S target (95%) — regulator-acknowledged gap in the same state market
Michigan disparity: BH access standard requires providers within 30 miles vs. 10 miles for M/S — 3× geographic travel burden for equivalent pediatric BH access
Appeal argument: cite state-specific adequacy standards and the M/S comparator standard in the same state. Geographic disparity in access standards is itself a parity violation
Appeal Success (Expected)
71–84%
Key States
FL, MI (state-documented)
Patient Scope
Pediatric BH only
Corpus Status
Add as MEDIUM sub-type, pediatric-specific
SERI STATUS — REGULATORY & ENFORCEMENT INTELLIGENCE
MHPAEA / CAA 2023 Enforcement
Active — Escalating
DOL and CMS enforcement of MHPAEA NQTLs is actively increasing under CAA 2023 requirements. Plans must now submit comparative analyses. Enforcement actions are up significantly — favorable regulatory environment for NQTL-based appeals through at least 2026.
WA State Parity (SB 5432)
Active — Strong Precedent
Washington SB 5432 (enacted 2023) requires parity compliance for all state-regulated plans, including explicit prohibition on step therapy for BH services unless equivalent requirements exist for medical/surgical. Strongest state parity law in the region — highly relevant to WA-based providers in the corpus.
OIG Medicare Advantage Audit
Monitoring — Ongoing
OIG's 2023 MA audit found 18% of post-service denial letters contained errors that did not meet CMS standards. CMS has issued corrective action requirements for several major MA plans. Watch for enforcement escalation in 2026 — may create new appeal leverage for MA BH denials.
Network Adequacy (BH)
Active — Regulatory Lag
State enforcement of BH network adequacy standards is inconsistent. DOL/EBSA has jurisdiction over self-funded plans but enforcement has been limited. The phantom network pattern is legally actionable but operationally requires sustained documentation. Monitor for state-level enforcement escalation.
RECOMMENDED KB ACTIONS — PRIORITY ORDER
ERISA Procedural Failure / Meaningful Dialogue
UHC/UBH-specific procedural reversal path. 4 federal circuit court cases. Use in parallel with clinical arguments on all UHC denials.
HIGH — Add as new KB type (payer-specific)
Step Therapy / Fail-First NQTL
Per se MHPAEA violation per CO DOL; PA Taft-Hartley settlement covers self-funded plans. Expected success 75–82%.
HIGH — Add as sub-type under NQTL Parity Violation
Frequency/Duration — Acuity Justification Gate
2-part evidence gate: acuity_justification + failure_to_respond. Currently misclassified under Benefit Limit / Frequency. 8.5M national volume PY2025.
MEDIUM — Add as standalone type with both evidence elements
Service Coverage Denial — Literature + Specialist Authority
Distinct from Benefit Exclusion. Requires peer-reviewed literature + specialist authority statement. 11M national volume PY2025.
MEDIUM — Add as distinct type (not Benefit Exclusion)
Improvement Standard Misapplication
Payer denies continued care because patient is improving. CA DMHC / federal precedent: improvement ≠ discharge-ready. Critical for residential continued-stay denials.
MEDIUM — Add as standalone type
Reimbursement Rate Disparity / Inadequate Recruitment
BH rate structure (88–98% Medicare) vs. M/S (120–123%) prevents provider recruitment. The rate differential itself is the NQTL. Distinct from Phantom Network.
MEDIUM — Add under Network patterns (rate disparity sub-type)
Pediatric BH Network Adequacy Disparity
FL: 85–90% BH vs. 95% M/S access goal. MI: 30mi BH vs. 10mi M/S access radius. Expected success 71–84%.
MEDIUM — Add under Network patterns (pediatric-specific sub-type)
CORPUS COVERAGE GAPS — WEEK 1 (2026-04-20)
Phantom Network OON
No precedents currently categorized. Present in case history but not formally captured. High-value pattern to add.
Action: identify + re-categorize existing cases
Level of Care Sub-types
LOC initiation, step-down, and continued stay denials are merged into general categories. Need formal separation.
Action: add denial_subtype field to corpus schema
NQTL Violation Type
Not yet a formal corpus category. Existing coverage gap denials may contain NQTL violations but are not tagged.
Action: audit coverage gap precedents for NQTL patterns
denial_stage Dimension
Pre-service vs. post-service not captured. Requires schema change before data can be backfilled.
Action: engineering ticket — add denial_stage field
SUD-Specific Precedents
SUD medical necessity denials are merged with general medical necessity. ASAM criteria is not tracked.
Action: add denial_condition: "SUD" | "MH" | "general" to schema
Optum & Magellan N Expansion
Both payers are below the 12-precedent confidence threshold. Strategy conclusions are unreliable at current N.
Action: prioritize case sourcing from Optum and Magellan accounts