VLADA · HEALTHCARE SPEND INTELLIGENCE

Know where your health plan is overpaying.

Vlada audits self-insured health plans against curated transparency data. Every finding links to a source file, a peer benchmark, and a remediation path.

FINDING 01 · 2026-04-14
Plan paid
$4,820
Peer median
$2,150
Multiplier
2.24×
CPT 74177 · NPI 1234567890
Source: UHC TiC MRF · 2026-Q1
Methodology: audit-v1 · §3.2
OVERPAYMENT FLAGGED
RESEARCH·EMPLOYER SPEND·2026

American self-insured employers spend $1.1T a year on healthcare. They cannot tell you where most of it goes.

67% of US covered workers are on self-insured employer health plans. Those plans cover 103 million Americans. Per-employee spending is projected to exceed $18,500 in 2026 — a 62% increase since 2017. Yet most employer plan sponsors cannot produce, on demand, a defensible account of what they paid for a given procedure, at a given provider, versus what peer plans paid.

$1.1T
Annual self-insured employer health spend in the US
KFF 2025 · CMS NHE 2024
67%
of US covered workers are in self-insured plans
KFF 2025 Employer Health Benefits Survey
$18,500
Projected 2026 healthcare cost per employee
Mercer Health Trends 2026
ERISA LITIGATION · 2026 OUTLOOK

22% of ERISA class actions in 2025 targeted health plans. The plaintiffs’ bar has a new target.

In February 2024, Cohen Milstein and Fairmark Partners filed Lewandowski v. Johnson & Johnson — the first major ERISA fiduciary-breach class action targeting a self-insured employer’s health plan. The complaint alleged J&J’s benefits committee failed to prudently monitor PBM drug pricing, citing a 498% markup on a generic MS medication.

The case was dismissed on Article III standing grounds and is on appeal to the Third Circuit. But in March 2026, Stern v. JPMorgan Chase partially survived a motion to dismiss on the merits — the first PBM-ERISA case to do so. Hecht v. Cigna settled for $6M in October 2025. Schlichter Bogard named Mercer, Gallagher, WTW, and Lockton as co-defendants in a December 2025 voluntary-benefits quartet. Consultants are now in the blast radius.

“The duty of prudence under ERISA § 404(a)(1)(B) requires plan fiduciaries to act with the care, skill, prudence, and diligence of an expert in the field.”

Fisher Phillips, “10 Best Practices for ERISA Health Plan Fiduciary Obligations,” Feb 2024
Lewandowski v. Johnson & Johnson
Dismissed on standing
Third Circuit appeal · 2026
D.N.J. · 3:24-cv-00671
Stern v. JPMorgan Chase
Partially survived MTD
Motion to dismiss · March 2026
S.D.N.Y. · 1:25-cv-02097
Hecht v. Cigna
Settled
$6.0M
Ghost-network theory · October 2025
S.D.N.Y.
RESEARCH·DATA QUALITY·NOV 2025

The federal government compelled every insurer in America to publish negotiated rates. Then they published garbage.

91.8%

of 3.1 billion negotiated rates published by US insurers are ghost rates — services providers virtually never perform.

Muhlestein, D., “Prevalence of Ghost Rates in Transparency Data,” Health Affairs Scholar 3(11), Nov 2025 · qxaf212 · Peterson Center on Healthcare

The Transparency in Coverage Rule (2022) requires every commercial payer to publish negotiated rates for every in-network provider and service. On paper, this is the largest public healthcare pricing dataset ever assembled.

In practice, the median insurer’s file is 84% unusable. At the provider-code level, 95.4% of published rates are for services that provider almost certainly never performs — a dermatologist quoted for cardiac bypass, a podiatrist quoted for neurosurgery.

Benchmarking an employer’s actual paid claims against raw transparency data produces statistical noise, not evidence. Every serious use of this data begins with curation.

FIG. 1 · GHOST-RATE PREVALENCE · 61 COMMERCIAL INSURERS · 2025
0481240–4545–5050–5555–6060–6565–7070–7575–8080–8585–9090–9595–100Ghost-rate prevalence by insurer (%)Insurers (n)
n = 61 insurers · Median: 84.3% · Range: 12.6%–97.3% · Source: Health Affairs Scholar, Nov 2025
THE PRODUCT · VLADA AUDIT AGENT

An AI agent that audits your health plan the way an ERISA attorney would — continuously, with evidence.

STEP 01

Connect your claims

Secure upload or a read-only TPA connection. Vlada audits against the largest curated pricing dataset in the US — ghost-filtered, entity-resolved, refreshed monthly.

STEP 02

Find the overpayments

Every paid claim is cross-referenced against peer-plan benchmarks, Medicare rates, and transparent-PBM comparators. Findings carry dollar impact, provider name, citation lineage, and severity.

STEP 03

Plan the remediation

TPA re-negotiation playbooks. RFP templates. Plan-design change models. Continuous monitoring proves whether the savings materialized.

VLADA AUDIT · SESSION PREVIEW · 2026-04-14
> show me where we overpaid in q1

Across 2,404 paid claims in Q1 2026, Vlada identified $3.1M in estimated overpayment concentrated in three categories:

1. Inpatient orthopedic surgery42% above peer median
Est. annual overpayment: $1.4M
Top driver: Facility Y, DRG 470
2. Outpatient imaging78% above peer median
Est. annual overpayment: $890K
Top driver: Facility Z, CPT 74177
3. Specialty infusion31% above peer median
Est. annual overpayment: $760K
Top driver: Drug X, JCode J1745

Every finding carries source lineage back to the relevant TiC MRF and peer-plan benchmark. Would you like to drill into one, or export a board-ready summary?

EVIDENCE · 2,404 CLAIMS · 61 INSURERS PEER SET · METHODOLOGY: AUDIT-V1 §3.2
METHODOLOGY · EVIDENCE LINEAGE

Every Vlada finding is designed to hold up under cross-examination.

If a plaintiff’s or defense attorney cannot reconstruct a Vlada finding from its lineage — from raw source file to benchmark to dollar impact — the finding isn’t defensible. We don’t surface it.

SPINE A · PAYER-SIDE (TiC 2022)
SPINE A · 01
TiC MRF
Per NPI · per plan
SHA-256 hash
SPINE A · 02
Ingestion
Schema 2.0
Row-level lineage
SPINE A · 03
Ghost filter
Taxonomy v1
FP <2%
SPINE B · HOSPITAL-SIDE (HPT 2021)
SPINE B · 01
HPT MRF
Per CCN · per payer
SHA-256 hash
SPINE B · 02
Ingestion
CMS v3 schema
Row-level lineage
SPINE B · 03
Ghost filter
Service-mix check
Taxonomy v1
KEYSTONE
Xref Bridge
NPI ↔ CCN ↔ system
Cross-spine validation
DOWNSTREAM · 01
Benchmark
Peer-plan percentiles
n ≥ 30 · HHI disclosed
DOWNSTREAM · 02
Finding
Dollar impact + severity
Citation lineage
METHODOLOGY

Source file hash

SHA-256 of the original MRF. Reproducible at any time.
Reference: audit-v1 §1.1
METHODOLOGY

Row-level lineage

Every rate links to a specific row ID in the source file.
Reference: audit-v1 §1.2
METHODOLOGY

Freshness metadata

Every response carries a last_updated timestamp and data vintage.
Reference: audit-v1 §1.3
METHODOLOGY

Peer distribution sample

Every benchmark discloses sample size, geographic scope, and ghost-filter version.
Reference: audit-v1 §3.2
BUYER PATHS · FOUR AUDIENCES

Vlada serves the people with a legal, professional, or technical need for defensible healthcare pricing evidence.

FOR PLAN SPONSORS

Audit one plan.

CFOs, General Counsel, Heads of Total Rewards at self-insured employers. Fulfill your fiduciary duty with evidence.

FOR BENEFITS CONSULTANTS

Deploy Vlada to your book.

Differentiate your advisory posture. Add a new tech-enabled revenue stream. Protect your own fiduciary exposure in the post-Schlichter Bogard environment.

FOR ERISA COUNSEL

Review the methodology.

Plaintiff-side and defense ERISA attorneys. Review Vlada's evidence pipeline against your discovery and admissibility standards.

FOR DEVELOPERS

Add Vlada to your agent.

Healthcare AI developers building RCM, prior-auth, navigation, or audit products. MCP server, REST API, source-cited responses.

EXAMPLE·AUDIT OUTPUT·2026-04-14

The output is a document your board can read and your counsel can defend.

FIG. A · SINGLE FINDING · SAMPLE
FINDING 07 · 2026-03-28
Plan paid
$82,400
Peer median
$43,800
Multiplier
1.88×
DRG 470 · Major joint replacement
Source: Aetna TiC MRF · 2026-Q1
Methodology: audit-v1 · §3.2
OVERPAYMENT FLAGGED
FIG. B · REMEDIATION SUMMARY · SAMPLE
Q1 2026 AUDIT · EMPLOYER ID VH-0042 · REVIEWED BY COUNSEL
ESTIMATED ANNUAL OVERPAYMENT
$3.1M
Inpatient orthopedic surgery
42% above peer median
$1.4M
Outpatient imaging
78% above peer median
$890K
Specialty infusion
31% above peer median
$760K
RECOMMENDED NEXT STEPS
  • Renegotiate DRG 470 bundle at Facility Y
  • RFP competing TPA for outpatient imaging
  • Evaluate transparent PBM for specialty infusion
Reviewed by Vlada Intelligence · Methodology audit-v1 · 2026-04-14
PLATFORM · MCP + API

Every healthcare AI agent should speak fluent negotiated rates.

Vlada ships an MCP server and a productized REST API. Add vlada-pricing to your agent and your product can cite healthcare pricing with source lineage, ghost-rate filtering, and entity resolution across the hospital–commercial payer boundary.

Registered in Anthropic’s MCP directory, published on Smithery, PulseMCP, and Glama. OAuth 2.1, usage-metered pricing, full OpenAPI spec.

shell
# Install in Claude Desktop, Cursor, or any MCP-native agent:
$ npx @vladahealth/mcp-install

# Or call the REST API directly:
$ curl https://api.vladahealth.com/v1/rates \
    -H "Authorization: Bearer $VLADA_API_KEY" \
    -d '{"npi": "1234567890", "code": "99213", "payer": "aetna"}'
TOOL SURFACE · V1
ToolPurposeReturns
get_negotiated_rateRate for NPI × code × payerRate + source lineage
compare_ratesDistribution for code in geographyPercentile distribution
find_overpaymentPaid claim vs. peer benchmarkFinding + remediation
list_providersSpecialty × geography filterProvider list
get_benchmarkCode benchmark by state / MSADistribution summary
get_hospital_chargemasterHospital chargemaster lookupFull chargemaster

One plan. One audit. One report you can show your board.

Vlada audits a single plan as a design-partner engagement. The output is a defensible audit report, delivered to you and your counsel, with no obligation beyond the review call.

Responses within one business day · BAA available on signed NDA