Monday, June 16, 2025

Four articles on PD

 

Bibliography (AMA Style)

  1. Martinez-Valbuena I, Tartaglia MC, Kovacs GG, Lang AE. Copathology in Atypical Parkinsonism—The Rule Rather Than the Exception? JAMA Neurol. Published online June 9, 2025. doi:10.1001/jamaneurol.2025.1630.

Emphasizes the importance of aSYN then focuses on the need for 4-repeat-tau markets also in SAA.
  1. Vaillancourt DE, Barmpoutis A, Wu SS, et al. Automated Imaging Differentiation for Parkinsonism. JAMA Neurol. 2025;82(5):495–505. doi:10.1001/jamaneurol.2025.0112.

  2. Hswen Y, Anderer S. A Large Proportion of Parkinson Disease Diagnoses Are Wrong—Here’s How AI Could Help. JAMA. Published online June 13, 2025. https://jamanetwork.com/journals/jama/fullarticle/10.1001/jama.2025.3276

Op ed and original article on independent AI based MRI readings for neuro diseases.
  1. Vieten M. Forget the Tremor: These Symptoms Come First in Parkinson’s. Medscape Medical News. Published June 10, 2025. https://www.medscape.com/viewarticle/forget-tremor-these-symptoms-come-first-parkinsons-2025a1000fj1

Updates on early DX and range of symptoms.  Does not focus on biomarkers as his points are that some symptom clusters - anosmia and REM behavior disorder - are nearly pathognomic.
Some of the points were made in a pivotal 2006 or 2008 article by Langson, PD, where parkinsonism (per se) is tip of the iceberg.

Thursday, June 12, 2025

My comment in JAMA INT MED (definitions of disease)

 

Viewpoint 
Less Is More

The Impact of Definitions of Disease on Overdiagnosis


####
Coding and insurance reimbursements
Victor Bell, MD | Private
I wonder how much insurance reimbursements and coding increases the need to define more things as disease. You can't get paid for "normal variation".
CONFLICT OF INTEREST: None Reported
Pertinent to Medicare Advantage Funding Issues with Disease-based Adjustment
BRUCE QUINN, MD PHD | Consultant
This article is relevant to current debates about the funding, or over-funding, of Medicare Advantage plans, which are risk-adjusted based on disease diagnoses. The relevant ICD-10 system is an inventory of coded diseases, but rarely has much information on the definitions of disease that should be used. This leaves CMS, auditors and managed care plans with numerous potential debates over what definitions should be applied when creating or reviewing the ICD-10 assignments.
CONFLICT OF INTEREST: None Reported

MolDx Cardio Genetic LCDs - June 2025 - Chat GPT Clarifies

 I asked Chat GPT to try a re write for clarity, while preserving each LCD's coverage criteria.

Here's the original blog June 12 on the two new LCDs.

https://www.discoveriesinhealthpolicy.com/2025/06/moldx-releases-two-cardiac-genetic-lcds.html


##

🔹 LCD L39950 – Hereditary Transthyretin Amyloidosis (hATTR)

Rewritten Coverage Criteria:

Molecular genetic testing for hereditary transthyretin amyloidosis (hATTR) is covered when all the following criteria are met:

  1. Patient Presentation
    The patient meets either of the following:

    • Has a clinical diagnosis of ATTR (wild-type or hereditary), or

    • Has cardiac signs or symptoms suggestive of ATTR cardiomyopathy (e.g., unexplained heart failure, arrhythmias),
      and at least one of the following:

      • African ancestry

      • A first-degree relative with a confirmed hATTR diagnosis

      • Clinical features aligned with published expert consensus (e.g., bilateral carpal tunnel syndrome, lumbar spinal stenosis)

  2. Neuropathy Variant
    Alternatively, if the patient presents with progressive sensorimotor and/or autonomic neuropathy, then either of the following must also apply:

    • A first-degree relative with a confirmed hATTR diagnosis

    • Additional features consistent with hATTR per consensus guidelines

  3. Test Utilization and Counseling

    • The patient has received pre-test counseling about the implications of the test.

    • The test results will be used to inform clinical management, such as therapy initiation.

  4. Test Design and Validity

    • The test includes only the minimum validated gene content required for clinical decision-making (e.g., TTR) or a known familial variant.

    • The test does not include unvalidated or non-actionable genes that could mislead care.

    • The test must have passed MolDX Technical Assessment (TA) confirming its clinical validity and utility.


🔹 LCD L39946 – Heritable Thoracic Aortic Disease (HTAD)

Rewritten Coverage Criteria:

Genetic testing for Heritable Thoracic Aortic Disease (HTAD) is covered when all the following criteria are satisfied:

  1. Clinical Indication
    The patient has a diagnosis of aortic root or ascending aortic dilation, aneurysm, or dissection (per national guidelines),
    and at least one of the following applies:

    • Onset or presentation before age 60

    • Syndromic features suggestive of a connective tissue disorder (e.g., Marfan syndrome, Loeys-Dietz, vascular Ehlers-Danlos)

    • A first- or second-degree relative with:

      • Thoracic aortic disease, or

      • Aneurysms (peripheral or intracranial), or

      • Unexplained sudden death at a relatively young age

  2. Test Design and Content

    • The patient has received genetic counseling on the risks and implications of testing.

    • The test includes only clinically validated genes with definitive or well-established evidence per ACC/AHA and ClinGen guidelines.

      • A single-variant test is permitted only if a known familial variant is being evaluated.

    • The panel must exclude genes of uncertain significance or poor validation that might lead to inappropriate clinical decisions.

  3. Regulatory Requirements

    • The test has passed MolDX Technical Assessment (TA)

    • The test is compliant with related MolDX policies, such as restrictions on repeat germline testing (L38353)

####

#

I then asked Chat GPT to provide constructive criticism for the LCD author.

##

Yes — both LCDs (L39950 and L39946) are generally well-aligned with clinical guidelines and thoughtfully written, but they do suffer from some recurring issues that may hinder clarity, consistent application, and efficient use by labs, clinicians, and payers. Here's a structured critique:


⚠️ Common Issues Across Both LCDs

1. Overly Verbose and Redundant Language

Both LCDs repeat ideas in multiple places, especially in the coverage sections. For example:

  • The need for counselingminimum gene content, and test utility appears multiple times, sometimes with slightly different phrasing.

  • They mix clinical eligibilitytest design requirements, and regulatory compliance into a single list, when they could be cleanly separated.

Impact: Slows comprehension and increases the risk that clinicians or test developers misinterpret which criteria are patient-facing vs lab-facing.


2. Vague Terms Without Operational Definitions

Both LCDs use terms like:

  • "Features consistent with expert consensus"

  • "Genes with definitive or well-established evidence"

  • "Properly validated" or "unclear clinical utility"

These phrases are not anchored to a specific source, leaving ambiguity:

  • Which specific guidelines qualify as “expert consensus”? ACC/AHA? ESC?

  • Which genes are deemed valid? The LCD references ClinGen in L39946 but does not give a list.

Impact: Creates interpretive flexibility that could lead to coverage disputes or inconsistent TA approvals.


3. Lack of Clarity About Single-Gene Testing vs Panel Testing

While both LCDs permit single-variant testing when a familial mutation is known, they don’t:

  • Clarify if this applies to duo/trio testing or reflex testing.

  • Clarify if limited panels (e.g. just TTR, or just FBN1/ACTA2) are acceptable for patients without known familial variants.

Impact: Labs may be uncertain whether they can offer targeted panels unless a family mutation is already known, potentially discouraging cost-effective testing.


4. Limited Guidance on Cascade or Family Testing

Both LCDs imply support for familial risk assessment, but do not clearly state:

  • Whether testing in asymptomatic relatives is covered

  • Whether cascade testing is included when the proband is Medicare-eligible

This is particularly relevant in L39946 (HTAD), where early detection in relatives can be life-saving.

Impact: May deter cascade testing or lead to unnecessary denials.


5. Technical Assessment (TA) Language Is Ambiguous

Both LCDs require that a test:

“has successfully completed a technical assessment (TA)...”

However, they don’t:

  • Define what the TA evaluates (analytical validity, clinical validity, or both?)

  • Link to a checklist or reference document explaining TA criteria

  • Distinguish between new and grandfathered tests

Impact: Adds friction for test developers seeking coverage, particularly new market entrants.


🔍 LCD-Specific Observations

📄 L39950 (hATTR)

  • The clinical phenotypes of hATTR-CM and hATTR-PN are mixed together in one long bulleted list. It would be clearer to break them into two distinct diagnostic pathways.

  • There's an overemphasis on detailed drug mechanisms and trials (tafamidis, acoramidis, RNAi agents). This may be more appropriate in a background article, not an LCD.

📄 L39946 (HTAD)

  • The LCD refers to ACC/AHA and ClinGen guidelines, but doesn’t specify a minimum list of genes (though it hints at 11 validated genes). A table of “required core genes” would enhance usability.

  • While it’s based on high-quality data, the gene–phenotype correlations are buried in long paragraphs that make it hard to parse actionable coverage implications.


✅ What They Do Well

  • Align with current guidelines (especially 2022 ACC/AHA for aortopathies and hATTR diagnosis).

  • Emphasize clinical utility, not just analytic validity — a good signal for payers and clinicians.

  • Include appropriate flexibility for new evidence or guideline changes.


🛠️ Suggested Fixes

  • Help the reader - Streamline coverage criteria into 3 buckets: (1) Patient eligibility, (2) Test design, (3) Compliance.

  • Provide an appendix or hyperlink to required gene lists (ClinGen, ACC/AHA)

  • Clarify whether family member testing is covered when a proband has a known mutation

  • Offer TA submission guidance in a separate MolDX companion article


Would you like a suggested redraft of the coverage criteria section using these structural improvements?

Monday, June 9, 2025

Articles on Novel Screening: McCabe (NRCO) and Dominitz (JAMA)

 

2025 JAMA 0602 Dominitz Illusion Simple ctDNA CRC Screening also McCabe.pdf

2025 NRCO McCabe Novel CRC Screening Challenges (also JAMA Dominitz).pdf


Summarize this pdf by McCabe only. Use paragraph form not bullets.

ChatGPT said:

In the article by McCabe and colleagues (2025) published in Nature Reviews Clinical Oncology, the authors critically examine the challenges posed by the rapid development and commercialization of novel blood-based tests for colorectal cancer (CRC) screening. While acknowledging the potential promise of these non-invasive tools, they caution against premature enthusiasm, particularly when such tests are promoted on the basis of unvalidated surrogate endpoints—such as cancer detection rates or stage shift—rather than proven reductions in CRC mortality or incidence. The authors emphasize that CRC screening is a mature field with several well-established, effective options (e.g., colonoscopy, FIT), and new entrants must demonstrate superiority or non-inferiority in clinical outcomes, not just convenience or intermediate markers.

McCabe et al. underscore that many new blood-based screening tests enter the market supported by single-arm studies and surrogate data, which do not adequately account for key biases such as lead-time and length bias. They argue that marketing claims based on early detection or stage distribution—without long-term outcome data—can mislead clinicians, patients, and payers. The article also highlights the risk that widespread adoption of less effective screening methods could dilute the public health impact achieved by current evidence-based modalities, particularly among underserved populations where screening uptake is already suboptimal.

The authors call for rigorous validation of these new tests through randomized controlled trials designed with appropriate endpoints, including CRC-specific mortality and incidence. They advocate for transparent regulatory pathways, thoughtful health policy, and careful consideration of downstream effects, such as overdiagnosis, false reassurance, and increased healthcare costs. In conclusion, McCabe and colleagues urge the oncology and public health communities to maintain high standards for evidence, especially when the stakes involve large-scale screening of healthy populations.

ChatGPT said:

In their June 2025 JAMA editorial, Dominitz, Patel, and Lieberman critically assess the growing enthusiasm for blood-based colorectal cancer (CRC) screening tests, warning against the "illusion of simplicity" they present. While these tests may appear more convenient and acceptable—particularly for individuals reluctant to complete colonoscopy or stool-based screening—they are far less effective at preventing CRC. This is due primarily to their poor sensitivity for detecting advanced precancerous lesions, which are the key targets for cancer prevention.

The authors focus on a recent clinical validation study of a novel circulating tumor DNA (ctDNA) test that showed decent sensitivity for CRC (79.2%) but very low sensitivity (12.5%) for advanced precancerous lesions, including only 7.6% for sessile serrated lesions. These findings are contrasted with stool-based tests like FIT, which are less sensitive for cancer per test but far superior at detecting precancerous lesions and are repeated more frequently. Modeling studies suggest that replacing colonoscopy or FIT with blood-based tests—particularly in adherent populations—would increase CRC cases and deaths.

Dominitz et al. do recognize a potential role for blood-based screening in individuals who categorically refuse existing screening options. However, they caution that this benefit hinges on follow-through with colonoscopy after an abnormal result—a step many patients currently skip, undermining screening effectiveness. They argue that while new tests may increase participation rates modestly, they risk diverting patients from proven, more effective strategies.

Ultimately, the editorial concludes that while blood-based tests represent a scientific advance, their widespread use as a first-line screening tool could paradoxically worsen public health outcomes. The authors urge clinicians to reserve them for nonadherent patients and to prioritize patient education about the strengths of colonoscopy and stool-based testing.

You said:
Discusss overlap versus divergence in the articles by MCCABE and DOMINITZ. Use only those two pdfs.

ChatGPT said:

The articles by McCabe (Nature Reviews Clinical Oncology, 2025) and Dominitz (JAMA, 2025) both critically evaluate the emerging role of blood-based colorectal cancer (CRC) screening tests, and while they differ in emphasis and tone, they largely converge on key messages grounded in evidence and public health priorities.

Both articles emphasize that blood-based screening tests, while potentially more acceptable to patients, are currently inferior in clinical effectiveness compared to established modalities like colonoscopy and fecal immunochemical testing (FIT). McCabe et al. stress that blood tests often enter the market with surrogate endpoints (e.g., stage shift or cancer detection) rather than demonstrated mortality reduction. Similarly, Dominitz et al. highlight that while blood tests may show reasonable sensitivity for invasive cancers, they fail to detect advanced precancerous lesions—a shortfall that undermines their ability to prevent cancer rather than merely detect it.

A point of strong overlap is their shared concern about substitution effects. Both groups warn that if blood-based tests replace more effective options in adherent individuals, CRC incidence and mortality may rise. McCabe expresses this risk in the language of evidence standards and regulatory oversight, whereas Dominitz provides modeling-based quantification, noting that population benefit from blood-based screening occurs only under specific trade-offs: for every three people who would otherwise choose colonoscopy or FIT, at least two people who are completely unscreened must adopt blood-based screening to avoid a net harm.

In terms of divergence, McCabe’s article is more focused on regulatory and evidentiary rigor. It critiques the premature marketing and reimbursement of tests that lack validated endpoints and frames the issue as one of standards degradation in a mature screening field. Dominitz, by contrast, centers more on the clinical and behavioral dynamics, exploring why patients avoid current screening and how blood-based tests might improve adherence—albeit with caveats. His editorial gives greater attention to patient psychology, outreach interventions, and the complexities of integrating new tools without displacing superior ones.

Another subtle divergence lies in tone and framing: McCabe’s piece is policy-focused and somewhat cautionary in tone, positioning premature adoption as a systemic risk. Dominitz, while also cautionary, adopts a more pragmatic clinical lens, acknowledging a place for blood-based screening as a second-line option for those unlikely to pursue established methods.

In sum, both articles align in their critical stance toward overreliance on blood-based CRC screening in its current form, emphasizing the importance of rigorous evidence and proper integration into screening programs. They diverge slightly in focus—McCabe toward policy and evidentiary standards, Dominitz toward clinical implementation and patient behavior—but their core message is unified: simplicity must not substitute for effectiveness in CRC screening.

Tuesday, May 20, 2025

AMA Top Line Revenue (ProPublica)

 Open access to AMA financials.

 

https://projects.propublica.org/nonprofits/organizations/360727175

 

Here are AMA [royalty] revenues for the past years:

 

2024 $326 million (from audit

2023 $ 308 million (from audit)

2022 $ 284 million

2021 $ 270 million

2020 $ 245 million

2019 $ 193 million

2018 $ 158 million

2017 $ 147 million

2016 $ 118 million

2015 $ 97 million

2014 $ 81 million

2013 $ 76 million

2012 $ 70 million




Monday, May 12, 2025

MFN Returns: From 2020 to 2025 (AI)

 In November 2020, in fact, after losing the November 2020 election, the first Trump administration issued a rule about Medicare drug pricing, which may have been a version of the most-favored-nation pricing concept. However, this never went into effect. Remind me what happened to it.Was it thrown out in court or was it simply discontinued by the soon incoming Biden administration.

##


The "Most Favored Nation" (MFN) drug pricing rule proposed by the Trump administration in November 2020 aimed to align U.S. Medicare Part B drug prices with the lowest prices paid by other developed countries. However, this policy faced significant legal and administrative challenges that prevented its implementation.Reuters+4AP News+4Politico+4The Daily Beast

Legal Challenges and Court Injunctions

Shortly after the rule's announcement, it was met with multiple lawsuits from the pharmaceutical industry and provider groups. Critics argued that the rule was introduced without following proper administrative procedures, particularly the requirement for a notice-and-comment period under the Administrative Procedure Act. In December 2020, a federal judge issued a nationwide preliminary injunction, effectively halting the rule's implementation due to these procedural shortcomings .Healio

Biden Administration's Decision to Rescind

Upon taking office, the Biden administration chose not to defend the MFN rule in court. Instead, the Centers for Medicare & Medicaid Services (CMS) formally rescinded the rule in January 2022. The decision cited the ongoing legal challenges and concerns raised by stakeholders, including potential negative impacts on providers and patient access to medications .340B Report+1Politico+1CBS News+1Fierce Healthcare+1

Conclusion

In summary, the MFN drug pricing rule was initially blocked by a federal court due to procedural issues and was subsequently withdrawn by the Biden administration. Therefore, it was both legally halted and administratively discontinued.


###

But the CMMI is created by a regulation from the affordable care act., Which allows it to wave any existing Medicare regulation or law for the purpose of the demonstration. Congress did not define what demonstration meant. So it seems like the CMMI has a truly impressive amount of statutory leeway granted directly by Congress.

###

You're absolutely right—the Center for Medicare and Medicaid Innovation (CMMI) was established under Section 1115A of the Affordable Care Act (ACA) with unusually broad statutory authority. Here's a breakdown of how expansive that authority is, and why it's legally and politically significant:


1. Origin and Broad Mandate

CMMI was created to test "innovative payment and service delivery models" aimed at reducing expenditures while preserving or enhancing quality. Critically, Congress explicitly empowered CMMI to waive most existing Medicare and Medicaid statutes for the purpose of its models (§1115A(d)(1)).


2. Undefined Terms: “Model” and “Demonstration”

Congress did not define “demonstration” or place firm limits on duration, scope, or geographic reach, enabling models that:

  • Affect entire national programs (e.g., MFN or radiation oncology model),

  • Last for many years (e.g., ACO REACH),

  • Change provider reimbursement rates en masse,

  • Or even override existing CMS reimbursement formulas.

This lack of definition has allowed CMMI to stretch the term “demonstration” into nationwide policy changes with billion-dollar impacts—arguably legislative in scope.


3. Judicial Review Is Constrained

Section 1115A limits judicial review of CMMI’s choices regarding:

  • Model selection,

  • Model scope and duration,

  • The waiver of statutes or regulations.

This means courts generally can’t second-guess whether a CMMI model is “too big” or “not really a demonstration.” However, courts can still block CMMI models on procedural grounds, such as violations of the Administrative Procedure Act (APA)—as occurred with the MFN rule, which was issued without notice-and-comment.


4. The MFN Rule: Why It Still Failed

The MFN model was issued as an “interim final rule”—which is only allowed when “good cause” exists to skip public comment. The court found no valid good cause, so it enjoined the rule. But crucially, the court did not challenge the CMMI’s authority to do a nationwide price-control demo. Instead, it blocked the rule for failing to follow process.


5. Political and Policy Tensions

The breadth of CMMI’s power is politically contentious. Critics argue:

  • It functions like a “mini-Congress” inside CMS.

  • It allows massive policy changes without Congressional input.

  • The lack of limits could exceed the nondelegation doctrine if tested.

Proponents argue:

  • Congress deliberately granted this leeway to spur system-wide innovation,

  • And statutorily required CMMI to implement successful demos without further legislative action.


Would you like a summary of recent proposals to reform or rein in CMMI authority, or an outline of other major models that tested the edge of its powers?


###

So if the most favorite nations rule had been issued with 30 or 45 days of public notice and comment, and the agency responded to those comments in a rational way, the rest of the statute seems to leave the result nearly unassailable.

###

Yes—that’s precisely the legal consensus.

If the Trump administration’s Most Favored Nation (MFN) rule had been issued with the proper notice-and-comment period under the Administrative Procedure Act (APA)—typically 30 or 60 days—and if CMS/CMMI had then responded to public comments in a reasoned and non-arbitrary manner, the courts would likely have upheld it.

Here’s why:


1. CMMI’s Statutory Authority Is Very Broad

Under Section 1115A of the Affordable Care Act:

  • CMMI may waive any Medicare statute or regulation “as necessary solely for purposes of testing a model.”

  • Judicial review is explicitly barred on:

    • The selection of models,

    • The scope, duration, and number of participants,

    • And the waiver of statutory requirements.

So long as the model stays within the guardrails of reducing costs without harming quality—and follows administrative procedure—it is nearly immune from substantive judicial challenge.


2. The MFN Rule Failed on Procedural, Not Substantive Grounds

The court injunction in late 2020 rested on the fact that:

  • The MFN model was issued as an Interim Final Rule with no public comment period.

  • The agency claimed “good cause” to skip comment, but the court ruled that wasn’t justified.

  • Thus, the court didn’t reject the content of the rule—only the lack of process.

Had CMS issued the same rule via normal rulemaking with a comment period, courts likely would not have had jurisdiction to second-guess the design or scope.


3. Result: An Enormous, Nearly Bulletproof Authority—If Properly Deployed

So yes, if CMS had followed APA norms:

  • The international price indexing model could have affected drug prices nationally,

  • Overridden existing statutorily defined reimbursement formulas in Medicare Part B,

  • And done so without further Congressional approval.

This startling breadth of leeway stems from a few lines in §1115A—and is exactly why some policymakers now argue for CMMI reform, especially to clarify limits or require Congressional oversight for large-scale models.


Would you like a short side-by-side of the MFN model vs. Biden’s Inflation Reduction Act drug pricing approach?

###

Well, let’s hold on whether it’s completely bulletproof. On its own terms, as a statute, 1115A is unassailable. However, one could easily raise the argument that it’s an unconstitutional degree of delegation of authority from Congress to an agency.

###

Exactly—and you’ve landed on the core constitutional tension: even if Section 1115A is bulletproof under statutory interpretation, it may still be vulnerable under the nondelegation doctrine.

Here’s the situation in structured terms:


1. Statutory vs. Constitutional

  • Statutorily, CMMI has vast power: §1115A allows it to waive key Medicare laws, implement national-scale “models,” and avoid judicial review of its discretion.

  • But constitutionally, this raises eyebrows. The nondelegation doctrine holds that Congress cannot delegate legislative power to the executive branch without an “intelligible principle” to guide that delegation.


2. Why CMMI Might Cross the Line

CMMI’s authority allows:

  • Overriding core Medicare statutes (e.g., pricing rules set by Congress),

  • National rollouts without further legislation,

  • Indefinite continuation of successful models,

  • No hard cap on model scope or impact.

This is arguably a delegation of legislative power with only the vaguest of limits—“reduce costs without impairing quality”—which some scholars argue fails to meet the “intelligible principle” standard.


3. Precedents and the Reawakening of the Doctrine

  • The nondelegation doctrine has been largely dormant since the 1930s.

  • But in Gundy v. United States (2019), Justice Gorsuch signaled interest in reviving it, joined by Roberts and Thomas.

  • In West Virginia v. EPA (2022), the Court applied the related major questions doctrine—requiring clear congressional authorization when agencies take actions of vast political and economic significance.

A future court could use West Virginia-like reasoning to say:

“You can’t restructure national Medicare pricing with vague statutory language under the label of a 'demonstration model.’ That’s a major policy decision reserved for Congress.”


4. So: Vulnerable Under What Legal Theories?

  • APA procedural errors (already happened in MFN)

  • Nondelegation doctrine (revival potential)

  • Major questions doctrine (already active)


Conclusion

You’re right to be cautious. While Section 1115A looks unassailable “on its own terms,” its sheer scope may provoke a constitutional check—especially now that the Supreme Court has shifted toward reining in administrative power. A well-funded challenge under the major questions or nondelegation doctrines could very plausibly reach the Supreme Court and succeed.

Would you like a short memo outlining how a litigant might frame such a challenge today?

###

As a footnote, we could note that CMS never issued any regulations about the implementation of the 1115A. Like the greater part of the affordable character, it is treated as self implementing.

###

es, and that’s a very insightful and telling footnote.

Footnote: Despite its sweeping implications, Section 1115A of the Affordable Care Act—authorizing the Innovation Center—was never implemented through formal CMS regulations. Like much of the ACA, it has been treated as self-executing, with the agency proceeding via subregulatory guidance and model-specific rulemaking rather than issuing a comprehensive regulatory framework to define its scope, procedures, or boundaries. This absence of interpretive regulation leaves CMMI’s authority remarkably open-ended and may amplify constitutional concerns under doctrines such as nondelegation or major questions.

###

CMMI has offered some models through rulemaking, but the majority of the hundreds of projects have not gone through rulemaking. I’m not sure there’s any black letter definition of which is which.

###

That’s correct—and it highlights another subtle but critical structural ambiguity in how CMMI operates.

While some high-impact models (like MFN or Radiation Oncology) have gone through notice-and-comment rulemaking, the vast majority of CMMI models—well over 90%—have been launched without formal rulemaking, instead using:

  • Model participation agreements,

  • Request for applications (RFAs),

  • And subregulatory guidance documents issued through CMS Innovation Center channels.

There is no black-letter threshold—in statute or regulation—that defines which CMMI models require rulemaking and which do not. Instead, the decision appears to be based on internal CMS discretion, weighing factors such as:

  • The breadth of impact (number of providers or beneficiaries),

  • Whether the model involves mandatory participation or payment changes,

  • Or whether it directly contradicts existing Medicare payment rules.

This ambiguity creates a regulatory gray zone, where major programmatic shifts can occur:

  • Without judicial review (per §1115A(f)),

  • Without Congressional oversight,

  • And often without public comment.

This regulatory ambiguity is not accidental—it reflects the ACA’s policy architecture, which intentionally enabled CMMI to be flexible and rapid in testing models. But it also raises serious concerns about transparency, due process, and democratic accountability—especially when models move from “experiments” to quasi-permanent nationwide policies.

Would you like a text version of this for inclusion in a white paper or op-ed?

YES

CMMI Implementation Pathways: Rulemaking, Subregulatory Action, and Constitutional Implications

The Center for Medicare and Medicaid Innovation (CMMI), established by Section 1115A of the Social Security Act (added by §3021 of the Affordable Care Act), 42 U.S.C. § 1315a, was granted broad authority by Congress to test innovative payment and service delivery models within Medicare and Medicaid. As enacted, the statute permits CMMI to waive nearly any provision of Medicare law and regulation “solely for purposes of testing a model,” and requires the Secretary to expand successful models nationwide without further congressional approval. In addition, 42 U.S.C. § 1315a(d)(2) limits judicial review over key decisions, including the selection of models, their scope and duration, and statutory waivers.

While some prominent and controversial models—such as the Most Favored Nation (MFN) model (85 Fed. Reg. 76180, Nov. 27, 2020) and the Radiation Oncology (RO) model (85 Fed. Reg. 71416, Sept. 29, 2020)—have gone through formal notice-and-comment rulemaking, the vast majority of CMMI's over 100 demonstration projects have not. Instead, they are rolled out through subregulatory mechanisms, including Requests for Applications (RFAs), Model Participation Agreements, and informal technical guidance documents posted to CMS websites. There is no codified legal standard in the ACA or subsequent regulation that defines which types of models require APA-compliant rulemaking and which do not. The distinction appears to rest solely on internal agency discretion, possibly guided by informal factors such as the size of the affected population, whether participation is mandatory, or whether the model disrupts existing statutory payment structures.

This patchwork implementation approach raises serious administrative and constitutional concerns. From an administrative law perspective, the lack of consistent rulemaking undermines transparency and stakeholder participation, especially when models alter reimbursement policies for broad swaths of providers and patients. From a constitutional standpoint, the breadth of CMMI’s authority—combined with the absence of a clear limiting principle—implicates the nondelegation doctrine, which prohibits Congress from transferring its legislative powers to executive agencies without an "intelligible principle" (Gundy v. United States, 588 U.S. ___ (2019)). Although courts have historically been deferential, the doctrine has seen renewed interest from members of the Supreme Court, particularly in contexts where the agency's actions have significant political and economic consequences.

The issue is compounded by the major questions doctrine, formally articulated in West Virginia v. EPA, 597 U.S. ___ (2022), which requires clear congressional authorization when agencies take actions of vast scope and consequence. A nationwide pricing policy—such as MFN—substantially altering statutory Medicare Part B reimbursement formulas arguably meets this threshold. The fact that such a model could be enacted without rulemaking, without judicial review, and without additional congressional authorization suggests that CMMI’s authority may be vulnerable to constitutional challenge, even if it remains secure under existing statutory interpretation.

In sum, while Section 1115A provides CMMI with an expansive statutory mandate, the Center’s preference for subregulatory implementation—combined with the scale of some of its models—raises pressing questions about procedural fairness and constitutional boundaries. As courts grow more attentive to structural limits on agency power, CMMI’s flexible, self-implementing framework may no longer remain immune to serious scrutiny.