← Case Library

Cigna (PXDX)

Q3
Healthcare · Shape forming · The Trust Breach
Researched as of April 2026
ShareLinkedInX
Shape Forming — Spiral proposed, not yet confirmed

Downgrading from the on-record Spiral, for which there is no supporting evidence at all: the narrative is a regulatory-failure and correction story (DMHC fine, congressional letters, ERISA litigation), not a digitally compounding or iteration-accelerating system — the opposite of what Spiral requires. Structurally the evidence concentrates in Q2/Q3 (breadth 2, colDominance 1.0, PXDX's 1.2-second average denial throughput across 300,000+ claims), which points toward Spear's concentrated-depth pattern, but the fit is imperfect since much of Q2 is external harm/oversight response rather than Cigna's own strategic surface, so Spear is proposed only as the closest available structural fit, not a confident positive characterization.

AdaptAugmentAssist
Q3Automator
Q6Optimiser
Q9Conductor
Q2Tactician
Q5Partner
Q8Explorer
Q1Operator
Q4Weaver
Q7Navigator
StructuredDynamicComplex
2/9 cells occupied
Shape Read

Cigna's PXDX system ran competently as an industrial code-matching engine: it compared submitted procedure codes against pre-associated diagnosis codes, routed mismatches into a denial queue, and issued denial letters at a throughput ProPublica calculated at roughly 1.2 seconds per claim across more than 300,000 denials in a two-month window. The problem was never whether the matching logic worked — it was that a medical-necessity adverse benefit determination is a decision California's Knox-Keene Act, ERISA fiduciary duty, and Cigna's own plan documents reserve to a licensed physician capable of evaluating the specific clinical file, not a batch sign-off appended to an automated queue. Cigna's medical-director review, alleged to take seconds per case, functioned as a cosmetic layer of authority over what was substantively an automated denial, and internal legal counsel cleared the design as being within a 'gray zone' without ever routing the question to an external regulator or clinical-oversight body. Correction only began once ProPublica and the Capitol Forum exposed the internal scorecards in March 2023, triggering congressional letters, a federal class action, and ultimately a California Department of Managed Health Care enforcement action in October 2025 that fined Cigna $500,000 and ordered a two-year look-back re-review of denied claims. The case stands as the clearest US private-sector instance of a rights-bearing decision relocated into an industrial-execution cell — a cell-choice failure, not an accuracy failure.

Trajectory: Remediation is running externally-in rather than internally chosen: DMHC-ordered policy refile and two-year claim re-review plus surviving ERISA and California UCL claims in Kisting-Leung are forcing Cigna toward substantive physician-level review rather than batch sign-off, with the outcome still contingent on ongoing litigation and monitoring.

Inside the company · not a cell

Nothing quoted yet. A rate or a saving will sit here when it is on the record. It does not light the grid.

Initiative Map
CellInitiativeConfidenceEvidence
Q2TacticianPre-PXDX retrospective claim review (pre-2010)Claims adjudication / utilization review✅Per ProPublica reporting, claims were 'entered into the insurance system, screened by a nurse and reviewed by a medical director'; processing a denial cost 'a few hundred dollars each,' making it cheaper for low-dollar claims to simply pay the bill before PXDX existed.As of April 2026
Q3AutomatorPXDX list creation (c. 2010–2014)Automated procedure-to-diagnosis code matching🔶The list reportedly grew to approximately 50 low-cost tests and procedures per Cigna's official statement; PXDX designer Alan Muney told ProPublica 'The PXDX stuff is not reviewed by a doc or nurse or anything like that'; an internal Cigna executive told ProPublica, 'We thought it might fall into a legal gray zone… We sent the idea to legal, and they sent it back saying it was OK.'As of April 2026
Q3AutomatorBatch denial workflow with medical-director sign-offClaim denial issuance🔶ProPublica reviewed an internal Cigna spreadsheet headed 'PxDx' listing per-medical-director denial totals; one anonymous former Cigna doctor said, 'We literally click and submit. It takes all of 10 seconds to do 50 at a time.' Three named medical directors did not respond to ProPublica's requests for comment.As of April 2026
Q3AutomatorTwo-month volume — 300,000 denials, 1.2-second averageClaim denial throughput🔶The same documents listed Dr. Cheryl Dopke at 121,000 denials in two months (roughly 60,000 in a single month), Dr. Richard Capek at 80,000+, and Dr. Paul Rossi at 63,000+.As of April 2026
Q3AutomatorPXDX scope — 'approximately 50 low-cost tests and procedures'Program scope definition🔶ProPublica stated explicitly that 'not all claims are processed through this review system'; Cigna has not disclosed the full PXDX procedure list or per-procedure denial rates publicly.As of April 2026
Q2TacticianPatient harm pattern — Van Terheyden vitamin D testIndividual claim harm / external review✅An external independent review by Medical Care Management Corporation (MCMC) reversed the denial seven months later, finding 'Testing was medically necessary and appropriate'; Cigna ultimately paid the bill after this external review.As of April 2026
Q2TacticianPatient harm pattern — Suzanne Kisting-Leung ovarian-cancer ultrasoundIndividual claim harm / named-plaintiff standing🔶Cigna submitted a declaration from Dr. Julie Kessel stating her claim was 'not denied through Cigna's PxDx review process'; the March 31, 2025 court order found that Kisting-Leung, along with Thornhill and Bredlow, lacked Article III standing on PXDX-specific claims because Cigna's evidence indicated their denials did not run through PXDX.As of April 2026
Q2TacticianPlan-document promise of medical-director reviewPlan-contract / ERISA fiduciary duty✅The March 31, 2025 court order found that 'interpreting plan language requiring a Medical Director to make a decision as allowing an algorithm to do so (even if a director pushes the button) could conflict with the plain language of the plan.'As of April 2026
Q3AutomatorCalifornia regulator finding (DMHC, October 2025)Regulatory enforcement✅DMHC found 'Cigna reviewed and denied claims without physicians conducting clinical reviews of the claims prior to issuing denials' and that 'the plan used a different review process than the policy it filed with the DMHC, which is a violation of the law'; Cigna agreed to pay the fine, re-review denied claims dating back two years, and revise and refile its policy.As of April 2026
Q2TacticianCalifornia Health & Safety Code § 1367.01(e) gateStatutory medical-necessity review requirement✅The court found plaintiffs adequately pleaded a § 1367.01(e) violation through PXDX use, and that the California UCL claim premised on it survives ERISA preemption via the savings clause.As of April 2026
Q2TacticianCongressional response — House Energy & Commerce, May 2023Federal legislative oversight✅House E&C sent a letter to Cigna's CEO requesting all PXDX policy memoranda and per-medical-director denial figures, citing the ~80% overturn rate on Cigna Medicare Advantage prior-authorization appeals as evidence that 'the PXDX review process is leading to policyholders paying out-of-pocket for medical care that should be covered.'As of April 2026
Q2TacticianCongressional response — Senate Permanent Subcommittee on InvestigationsFederal legislative oversight✅A May 17, 2023 Senate hearing, 'Examining Health Care Denials and Delays in Medicare Advantage,' was followed by an October 2024 Permanent Subcommittee on Investigations report, 'Refusal of Recovery: How Medicare Advantage Insurers Have Denied Patients Access to Post-Acute Care,' which focused on UnitedHealthcare, Humana, and CVS rather than Cigna directly but contextualizes the same cell-mismatch pattern.As of April 2026
Q2TacticianAccreditation-body inquiries — URAC and NCQAIndustry self-regulatory oversight◻️Both the Utilization Review Accreditation Commission (URAC) and the National Committee for Quality Assurance (NCQA) opened investigations into PXDX after the ProPublica reporting; neither body has publicly released findings, and both declined to respond in detail to ProPublica's questions about the investigations.As of April 2026
Q3AutomatorCigna public response — newsroom, July 27, 2023Corporate communications / legal defense positioning✅Cigna's position is consistent across multiple statements: post-treatment review only; approximately 50 procedures; roughly 99% of PXDX-reviewed claims auto-paid; less than 1% of total claim volume denied via PXDX; in-network providers should not bill patients on denial.As of April 2026
Q3AutomatorCigna public response — 'PxDx is not AI/algorithm'Corporate communications / technology labeling🔶The technical AI/non-AI distinction is orthogonal to the underlying cell-choice question: a deterministic rule-based matching engine that displaces a clinician's judgment on medical necessity is still an automated-execution cell regardless of whether it is labeled 'AI.'As of April 2026
Q3AutomatorSubsequent Cigna policy actions — E/M downcoding (October 2025)Adjacent automated payment-adjustment surface◻️Effective October 1, 2025, Cigna's new R49 policy automatically downcodes high-level E/M office-visit codes (99204–99205, 99214–99215, 99244–99245) when documentation does not support the billed complexity; Cigna says 'almost 99% of all in-network providers will not be affected,' and providers can submit documentation to restore the original level.As of April 2026
Strength
The underlying code-matching engine performed its narrow, intended function well — flagging procedure/diagnosis mismatches and accelerating routine claim throughput on a small slice (Cigna says roughly 50 procedures, under 1% of volume) of claims.
Fragility
The batch queue scaled to a throughput — over 300,000 denials in two months at an average of about 1.2 seconds per claim, with individual medical directors denying tens of thousands of claims each — that structurally could not accommodate the individualized clinical review the activity required.
Failure mode
California's DMHC found that Cigna denied claims as not medically necessary without physicians actually conducting clinical reviews before issuing those denials, and that the plan operated a different review process than the one it had filed with the regulator.
Governance ceiling
An internal legality concern was raised and resolved entirely inside Cigna's own legal department — reportedly cleared as being within a 'gray zone' — without ever engaging an external regulator, court, or clinical-oversight body, so the mismatch between the automated process and the statutory Knox-Keene physician-review requirement went uncorrected until ProPublica's 2023 investigation forced outside scrutiny.
← Back to the Case Library