NOTICE — IDRE-235: Fortuna Arbitration's application for CMS certification as an Independent Dispute Resolution Entity (IDRE-235) is currently pending. Fortuna Arbitration is not yet authorized to adjudicate any CMS IDR cases at this time.
1

AI Pre-Processing

Arbitrus performs automated intake analysis within 3 business days of dispute initiation.

  • Conflict of interest screening across all parties
  • Federal IDR applicability determination
  • Compliance officer notification and case assignment
  • Anti-gaming safeguards — prompt injection detection, semantic consistency checks, diagnostic code manipulation detection

Fortuna Arbitration understands concerns surrounding AI technology in dispute resolution and implements various technical safeguards — including input validation pipelines, adversarial robustness testing, and human oversight at every stage — to ensure the integrity and reliability of its processes. Read more about AI safeguards in arbitration.

2

AI Driven Research

A three-stage classification system mirrors federal court analytical frameworks to produce an equity-adjusted award.

Arbitrus begins by reading the context of the submitted CPT or DRG code to establish the clinical baseline — what the code describes, the typical complexity of the service, and the standard reimbursement expectations associated with it.

The system then cross-applies this baseline to the specific facts of the case as presented by both parties, identifying where the actual clinical circumstances diverge from the code's typical parameters. This divergence analysis surfaces the additional circumstances that may warrant adjustment from the QPA.

Patient Acuity: Fortuna recommends that parties alleging patient acuity as an additional circumstance under 42 U.S.C. § 300gg-111(c)(5)(C)(ii) focus on presenting the medical facts objectively. Arbitrus evaluates acuity claims based on documented clinical indicators — ASA classification, complication severity, emergent versus elective designation, and deviation from the expected clinical course for the submitted code — rather than subjective characterizations.

Once the relevant additional circumstances have been identified, Arbitrus assigns a weight to each factor using probabilistic inference derived from federal court analysis of healthcare disputes and prior IDR determinations.

Rather than applying fixed coefficients, the system uses Bayesian posterior calculations that dynamically adjust based on the specific combination of circumstances present. This prevents any single factor — including the QPA — from mechanically dominating the analysis, consistent with the post-TMA III requirement that all statutory factors be weighed equally.

In the final stage, Arbitrus generates a holistic analysis that considers how the identified circumstances interact with one another — for example, how patient acuity compounds with provider teaching status and market concentration to affect the appropriate reimbursement level.

The system produces an equity-adjusted award figure and then applies the "baseball style" selection mandated by the statute: whichever party's final offer falls closest to the calculated fair value is selected as the determination.

Figure 1: Multiplicative Equity Framework for IDR Award Determination
Evidence Sources
  • Federal Case Law
  • Contract Interpretation
  • Group Cases
  • Pattern Extraction
Posterior Calculation Engine
  • Bayesian Inference
  • Statistical Weighting
  • Doctrinal Analysis
Weight Matrix
  • w1 = 1.80
  • w2 = 1.20
  • wn = 1.35
QPA Baseline
$2,450
Multiplicative Framework
FinalAward = QPA × (1 + Σ wici)
$2,450 × 4.2
Circumstances
  • c1 = 1
  • c2 = 1
  • cn = 1
$10,290
Final Award
Award to party with offer closest to $10,290

This figure is for illustrative purposes only and does not represent an actual IDR case or determination.

Process Component
Calculated Values
Final Output

Read more about the No Surprises Act, pending TMA litigation, and Fortuna's methodology.

Learn More — Gandall et al., Harvard JLPP (2025)

Disclosure: Advancing large language model (LLM) and machine learning (ML) technologies are continuously accelerating arbitration methodologies. Fortuna Arbitration's methodology is subject to change as these technologies evolve, in order to improve outcomes, accuracy, and consistency of determinations.

Features Analyzed

  • HHI market concentration
  • ASA scores
  • GME ratios
  • Network adequacy
  • DRG complexity

"Baseball style" arbitration — selects the party offer closest to the calculated fair value. Probabilistic fusion prevents QPA anchoring through statistical weighting of diverse evidence sources.

3

Human Decisionmaker & Certification

Every AI-generated determination undergoes mandatory expert review before issuance.

  • Healthcare billing expert reviews the AI output for accuracy and completeness
  • Case Supervisor (Certified Healthcare Adjuster) provides final sign-off
  • Determination issued to all parties with full reasoning and factor analysis
  • All appeals routed through the CMS IDR appeals process