vbcc wedn x 3

 

The titles below follow the agenda. Sidebar items are paraphrases, with transcript timestamps for reference. “Brooks” means CMS official John Brooks, distinct from Friday panelist Gabriel Brooks.

CMS Innovative Initiatives: Providers on OCM, EOM, and What’s Up Next

The provider panel presented a sharp disagreement over whether EOM advances oncology value-based care or constrains it. Several participants argued that narrowing eligibility to seven cancers and concentrating on active treatment creates financial volatility, excludes valuable decisions not to treat, and limits investment across the full cancer journey. Ron Kline defended population-level accountability: practices must evaluate aggregate performance rather than fixate on individual losing cases, and excluding drugs would remove major opportunities for savings. Yet the discussion exposed how poorly claims capture clinical reality—stage, tumor biology, treatment intent, and whether someone is receiving palliation or simply completing survivorship follow-up. Biomarker results buried in incompatible PDF reports were a particularly concrete measurement obstacle. Participants praised ePROs, social-needs services, navigation, and commercial arrangements combining total-cost accountability with quality requirements, while emphasizing their infrastructure costs. The unresolved dispute was whether better voluntary economics could attract practices or whether mandatory participation would be necessary. For Friday’s panel, the central lesson is that richer clinical measurement could address important blind spots, but financial incentives, population definitions, and implementation support remain equally consequential. 20261007 Panel All About CMMI O…

CMS Leadership Discussion with Abe Sutton

Abe Sutton described CMMI as CMS’s research-and-development arm for payment, emphasizing evidence, scalability, and willingness to abandon disappointing models—including a kidney model he helped design. Home dialysis improved nationally, but mandated model participants did not improve significantly more than comparison areas: favorable trends alone did not demonstrate model success. He explained certification as evidence of lower spending without worse quality, or better quality without higher spending, while noting that certification does not automatically compel nationwide expansion. He discussed joint-replacement expansion alongside continued testing of TEAM, defended WISeR’s technology-assisted review of selected services, and promised an accelerated early evaluation rather than waiting for the usual reporting cycle. Drug-policy discussions covered GLP-1 access, GLOBE’s international-reference approach, and GENEROUS’s pooled Medicaid rebate and coverage arrangements. Sutton welcomed outside proposals that change incentives across a clinical category or market, rather than obtain special treatment for one product. For oncology, he was open to new arrangements and potentially helpful legal flexibilities, but rejected simply increasing payments without accountability. He also acknowledged that gaps between models can dismantle staffing and infrastructure that practices have already built. 20261007 Joe grogan interviews …

CMS Leadership Discussion with John Brooks

John Brooks framed value-based care as aligning treatment incentives with the care patients would want, rather than rewarding visit volume or the margin on a particular drug. His broad policy discussion covered Part D stabilization after the IRA, Medicare Advantage’s benefits and unresolved payment and quality concerns, prior-authorization standardization, and greater transparency around 340B. He stressed that CMS cannot simply compensate providers for statutory reimbursement changes enacted by Congress. For AI, his position was simultaneously optimistic and skeptical: current applications can increase spending by amplifying existing incentives, while appropriately designed outcome-based payment could make technology reduce costs and expand access. He highlighted ACCESS as a promising framework for paying for measurable clinical improvement rather than adding another billable service. Outcomes-based drug contracts, he argued, primarily allocate uncertainty about therapeutic effectiveness between manufacturers and payers. Brooks welcomed validated data and proposals jointly supported by providers, payers, and manufacturers. His memorable warning was that consultants can sell companies months of expensive work pursuing a bespoke CMMI demonstration, only for CMS to reject it: the agency wants structural solutions that can work across a market. 20261007 joe grogan intvu brook…

Sidebar: 20 Takeaways for Friday’s AI and Oncology Measurement Panel

1. A narrower model can produce noisier results. EOM’s restricted cancer mix and focus on early treatment can make practice-level performance swing with a few expensive cases, even when those fluctuations wash out nationally. Better measurement must distinguish performance from case-mix volatility. (CMMI provider panel; 8:18–10:24)
2. Some valuable care never enters the denominator. A lengthy discussion leading a patient to decline burdensome treatment can save money and improve goal-concordant care, yet fail to trigger a treatment-based episode. This directly echoes the “missing denominator” issue in your background notes. (CMMI provider panel; 10:24–11:26, 44:21–46:13)
3. What two clinicians understand immediately may be invisible in claims. Kline recalled trying to distinguish methotrexate for rheumatologic disease from cancer treatment, and palliation from survivorship follow-up. AI’s potential contribution is extracting the clinical context needed to interpret the transaction. (CMMI provider panel; 13:37–16:58)
4. The biomarker report is there—but the measurement system cannot readily use it. Kline described opening a laboratory PDF at Johns Hopkins to discover which biomarkers had been tested, then observed: “There’s no way you can do quality measures with that.” A concrete challenge for AI extraction and interoperability. (CMMI provider panel; 36:14–38:10)
5. ePROs and social-needs screening create services, not merely data. Henschel credited these requirements with timely interventions and investments in meals, transportation, social work, and palliative care. Their value depends on the response infrastructure; collecting information alone does not deliver the benefit. (CMMI provider panel; 33:01–35:36)
6. Quality can be an actual condition of payment. Florida Cancer Specialists described a longstanding total-cost arrangement in which meeting a quality gate unlocks shared savings. The practice itself proposed the gate. Friday’s question: which AI-enabled measures would be trustworthy enough to play that role? (CMMI provider panel; 20:23–21:15)
7. An oncologist cannot accept a pen—but can earn thousands in drug margin. Thurmes illustrated the incentive mismatch with Minnesota’s strict gift restrictions. His point was that conscientious physicians still work inside an economic structure that rewards drug revenue. Measurement reform must confront that structure. (CMMI provider panel; 24:08–26:05)
8. Low participation has competing explanations—and competing remedies. Kline argued that voluntary models struggle against financially attractive fee-for-service and may require mandatory participation. Mehring countered that a model benefiting all stakeholders would attract participants voluntarily. Better measures alone will not settle this dispute. (CMMI provider panel; 37:22–42:40)
9. The expensive consultant engagement ends with “no.” Brooks joked that consultants can convince companies CMMI will create a special model for them; months of paid work culminate in a meeting where CMMI says it has no interest. His requirement: a replicable, structural proposal. (Brooks; 39:27–40:33)
10. AI can make an inefficient payment system more expensive. Brooks’s near-term assessment was that AI has been cost additive: helping the system do more can amplify its existing inefficiencies. An AI tool’s productivity gains therefore do not automatically translate into payer savings. (Brooks; 41:35–43:20)
11. Pay for the clinical improvement the technology produces. Brooks highlighted ACCESS and the example of lowering HbA1c. For oncology, the corresponding challenge is to identify meaningful, auditable outcomes that could become reimbursement targets, rather than simply adding payment for using an algorithm. (Brooks; 2:31–3:44, 41:35–43:20)
12. Physician compensation should not depend on how expensive the prescribed drug is. Brooks called the existing arrangement a deeply unsatisfactory system, while acknowledging how difficult it is to change. New oncology measures will operate within those prescribing incentives unless payment design changes too. (Brooks; 33:13–35:16)
13. A drug contract reallocates risk; it does not change the molecule’s behavior. Brooks distinguished provider incentives from outcomes-based pharmaceutical agreements. The latter distribute uncertainty about effectiveness between payer and manufacturer—making reliable outcome ascertainment central to whether the contract works. (Brooks; 37:30–38:48)
14. Bring validated data and a solution the stakeholders developed together. Brooks said joint proposals from providers, payers, and manufacturers cut through competing requests for money. For Friday’s panel, a shared clinical measure with demonstrated usefulness would be more persuasive than another constituency’s wish list. (Brooks; 53:00–54:42)
15. An improving metric does not prove the payment model worked. Sutton’s kidney example was explicit: home dialysis rose nationally, but model areas did not outperform comparison areas significantly. This reinforces your speaker notes’ emphasis on a credible counterfactual when evaluating AI or payment interventions. (Sutton; 5:11–6:30)
16. CMMI’s success test has two routes. Sutton described certification as improved quality without higher costs, or lower costs without reduced quality. That distinction matters for oncology: a useful AI intervention need not demonstrate both better outcomes and lower spending simultaneously. (Sutton; 10:55–11:23)
17. Evaluation can arrive too late to guide decisions. Sutton contrasted the usual annual evaluation cycle with WISeR’s planned early snapshot using six months of experience and three months of claims runout. Friday’s discussion could distinguish rapid operational feedback from mature evidence sufficient to judge success. (Sutton; 22:47–24:47)
18. A gap between models can erase the investment the first model created. Sutton described kidney programs losing the ability to support staff during a transition gap. Oncology measurement infrastructure—including ePRO support and clinical-data capabilities—also needs a financing path that survives model transitions. (Sutton; 44:05–46:11)
19. CMMI wants an open market, not a privileged product. Sutton described ACCESS as a reimbursement structure in which multiple companies can compete, rather than a demonstration tailored to one solution. An oncology proposal should define the clinical problem and payment framework broadly enough for multiple approaches. (Sutton; 38:52–41:24)
20. The door is open for another oncology model—with conditions. Sutton welcomed proposals that improve incentives and support patient care, while rejecting a simple payment increase without downside accountability. He also expressed willingness to consider legal flexibilities for contracting, within statutory purposes. (Sutton; 42:17–43:05, 47:26–48:06)

 

LONG FORMAT SUMMARIES

 

CMS and Oncology Value Based Care Session Summaries

Prepared for Bruce Quinn | October 7, 2026

These three sessions examine oncology payment models, CMS policy priorities, and the clinical information needed to evaluate value. Their recurring concern is how to connect meaningful patient outcomes with payment incentives while preserving practices' ability to deliver care.

CMS Innovative Initiatives: OCM, EOM, and What’s Up Next

Wednesday, October 7, 2026 | Day One | Session 15 | 11:00–11:50 a.m.

Location: Olympic Suite 1, 10th Floor

Moderator: Maddi Davidson, Managing Director of Market Access, Avalere Health

Panelists:

Paul Thurmes, MD, President and Medical Oncologist, Minnesota Oncology

Gabrielle Rocque, MD, MS, Chief Medical Officer, Atlas Oncology Partners; Associate Professor, University of Alabama at Birmingham

Rhonda Henschel, MBA, SVP, Payer Lifecycle and Value Optimization, McKesson

Ron Kline, MD, Retired CMO, Quality Measurement and Value-based Incentives Group, CCSQ/CMS

Kiana Mehring, MBA, PPMC, LION, VP of Payer Strategy & Revenue Cycle, Florida Cancer Specialists & Research Institute

Participants and central disagreement

Moderated by Maddi Davidson, this session brought together Paul Thurmes, Gabrielle Rocque, Rhonda Henschel, Ron Kline, and Kiana Mehring. The discussion exposed a fundamental disagreement about EOM: several practice leaders saw its narrower design as a retreat from OCM, while Kline defended the logic of population-level financial accountability. Participants agreed that modern oncology has changed dramatically, but differed over whether existing payment models recognize those changes adequately.

Thurmes emphasized that immunotherapies and targeted treatments complicate judgments about benefit, treatment sequence, and when to stop. Kline maintained that the definition of value remains outcomes relative to cost, including survival and quality of life. Mehring cautioned that programs have nevertheless drifted toward treating reduced expenditure as the principal evidence of value.

The population and denominator problem

Rocque argued for encompassing the whole cancer population, from diagnosis through survivorship and end of life. Restricting EOM to seven cancers and early treatment concentrates drug spending and reduces the population available to absorb financial variation. Mehring said Florida Cancer Specialists' modeling predicted losses under EOM, but a favorable result when the broader OCM cancer mix was restored. Henschel reported substantial swings across practices and performance periods within the US Oncology Network.

Kline responded that practices must assess their aggregate results, accepting gains on some patients and losses on others. He recalled repeated complaints about losses on oral myeloma therapies without equivalent attention to gains elsewhere. The disagreement concerned whether EOM's overall financial design was viable, as well as practitioners' tendency to focus on unfavorable subgroups.

Thurmes identified another omission: patients who decline systemic treatment after extensive discussion may receive valuable supportive care without entering a treatment-triggered episode. An audience member similarly argued that observation or delayed treatment can save substantial spending without earning model credit. Recognizing these patients reliably became a recurring challenge.

Claims and biomarker information

Kline described the gulf between clinical understanding and claims analysis. Two clinicians can readily distinguish methotrexate for cancer from methotrexate for rheumatologic disease; a claims-based program has much less information. Likewise, a breast-cancer diagnosis can represent active advanced disease, palliation, or long-term survivorship. He criticized legacy coding's ability to specify anatomical location more readily than clinically decisive tumor biology.

Mehring argued that practices and payers could exchange stage and other clinical data to improve their models. Participants differed over whether CMS was sufficiently responsive to practice feedback: Henschel said CMMI had been accessible, but stakeholders often reached different conclusions because they examined different data.

Biomarker reporting supplied a particularly vivid example. Thurmes described inconsistent testing platforms and uncertainty about repeat testing. Kline recalled opening laboratory PDFs at Johns Hopkins to determine which biomarkers had been assessed. Incompatible reporting formats obstruct routine quality measurement even when the underlying clinical information exists.

Drugs and financial incentives

Kline opposed removing drugs from total-cost accountability because biosimilars, generics, and equivalent guideline-supported options offer important savings opportunities. Rocque agreed that drugs belong in the calculation, while warning that supportive-care interventions cannot solve broader pharmaceutical pricing problems.

Thurmes illustrated the prescribing incentive problem: Minnesota restrictions prevented him from accepting even a small industry gift, yet practices could receive substantial drug margins. Henschel warned that simultaneous reimbursement changes, IRA implementation, and overlapping reporting requirements make participation harder and obscure which policy causes which result.

Services worth preserving

Henschel strongly supported ePROs and health-related social-needs screening, citing earlier intervention and investments in transportation, meals, social work, and palliative care. These services require substantial financial support. Once established, practices may resist dismantling them, but that does not mean new participants can build them without funding.

Mehring described a longstanding commercial total-cost arrangement with a quality gate controlling access to shared savings. Rocque's organization instead assumed downside risk while insulating participating practices. These examples demonstrated different ways to support care transformation.

Participation and relevance to Friday

Kline argued that attractive fee-for-service economics limit voluntary enrollment and that mandatory models may be necessary. Mehring countered that mutually beneficial models can attract participation while saving payers money. For Friday's panel, the session supplies concrete measurement targets: treatment intent, decisions not to treat, tumor biology, symptoms, and the full disease trajectory. It also makes clear that better data must be accompanied by workable incentives and infrastructure financing.

 

 

CMS Leadership Discussion with Abe Sutton

Wednesday, October 7, 2026 | Day One | Session 17 | 1:00–1:50 p.m.

Location: Lounge, 9th Floor

Moderator: Joe Grogan, JD, Fellow, USC Schaeffer Center; Former Assistant to the President, Director of the Domestic Policy Council

Featured speaker: Abe Sutton, JD, Deputy Administrator & Director, CMS Innovation Center, Centers for Medicare & Medicaid Services

CMMI's role and the administration's approach

Interviewed by Joe Grogan, Abe Sutton described CMMI as CMS's research-and-development arm for payment. Its task is to test whether different incentives can improve care and make public spending more efficient. He contrasted initial ambivalence about an Affordable Care Act institution during the first Trump administration with the current administration's willingness to use its authority actively. The emphasis was on evidence-driven decisions rather than preserving models because an administration had created them.

Sutton said the team reviewed the existing portfolio and canceled four models, three originating in the first Trump administration. Career staff supplied recommendations. One canceled initiative was a kidney model Sutton had helped develop, making the discussion an unusually direct acknowledgment that a favored policy had not met expectations.

Improvement versus attributable improvement

The kidney example illustrated his evaluation standard. National home-dialysis rates rose from approximately 11% when the model was designed to approximately 16%. However, mandated model areas did not show a statistically significant improvement over comparison areas. Other policy changes, technology, and evolving clinical practice apparently contributed to the national trend.

Sutton argued that static performance targets could consequently reward improvement that the model itself had not generated. The relevant question was whether the payment intervention changed behavior beyond what would otherwise have occurred. This is especially pertinent to evaluating AI in settings where clinical practice and technology are already improving.

Certification and nationwide expansion

Sutton described certification as demonstrating better quality without higher spending, or lower spending without worse quality. CMS's actuarial and clinical assessments inform the secretary's decision. Certification permits consideration of expansion but does not automatically require it.

He used joint replacement to explain why CMS might expand a successful approach while continuing to test an alternative. CJR and TEAM use different episode durations and designs; future evidence could favor the newer framework. He also noted that a model's relationship to other payment programs matters. Apparent savings may be less attractive if inconsistent benchmarks create opportunities for arbitrage between programs.

WISeR and technology-assisted review

A substantial portion of the interview defended WISeR, which applies technology-assisted review to selected services in six states. Sutton emphasized that the model enforces existing coverage standards rather than introducing new ones. Providers can submit for prior authorization or face prepayment review; he reported that most chose authorization.

He described participant incentives intended to discourage inappropriate denials: repeat denials do not generate repeated rewards, eventual approval can eliminate a claimed saving, and excessive reversals can trigger penalties. He said denials require physician review and argued that rapid approvals and accessible clinical discussion distinguish the process from familiar authorization frustrations.

Grogan pressed him about congressional opposition and reported implementation friction. Sutton acknowledged early problems, particularly in Washington State, but defended the model as protecting patients from inappropriate care as well as taxpayers from waste. He promised an early evaluation snapshot using six months of experience and three months of runout. His favorable assessment remained an account of implementation, ahead of that evaluation.

Drug access and purchasing arrangements

Sutton described balancing broader GLP-1 access against substantial near-term costs, referring to a bridge demonstration and an announced price of $245 per month. On GLOBE, he explained using the IRA's inflation-rebate structure to test international price references. He contrasted its manufacturer-focused design with earlier approaches that also contemplated changing physician payment.

GENEROUS, in Medicaid, offers standardized supplemental rebates alongside standardized coverage terms. States can compare that offer with their existing arrangements. Sutton said manufacturer participation exceeded initial expectations and presented pooled contracting as a way to reduce the burden of negotiating separately with many states. He drew a parallel with the cell-and-gene-therapy model.

What oncology proposals could succeed

Sutton welcomed outside ideas backed by persuasive data, including proposals from academic researchers. He wanted approaches that could reshape a clinical category or create a competitive market. A request to bypass ordinary coverage review for one device would not fit that purpose.

For oncology, he was open to new arrangements but rejected simply increasing payments without meaningful accountability. He also indicated willingness to consider contracting flexibilities within statutory limits. Finally, he warned that gaps between models can destroy staffing investments and make subsequent recruitment harder. For Friday, his remarks connect clinical measurement to causal evaluation, scalable payment design, and continuity of the infrastructure needed to act on the data.

 

 

CMS Leadership Discussion with John Brooks

Wednesday, October 7, 2026 | Day One | Session 21 | 2:00–2:50 p.m.

Location: Lounge, 9th Floor

Moderator: Joe Grogan, JD, Fellow, USC Schaeffer Center; Former Assistant to the President, Director of the Domestic Policy Council

Featured speaker: John Brooks, JD, MBA, Chief Policy and Regulatory Officer, Deputy Administrator, Centers for Medicare & Medicaid Services

Patient interests and payment incentives

Interviewed by Joe Grogan, John Brooks framed value-based care around the treatment incentives patients would want their clinicians to face. He questioned a system in which drug margins influence prescribing and visit-based payment can reward additional utilization. His goal was to align payment with patient benefit while recognizing that changing an established reimbursement structure is difficult.

Brooks saw new opportunities in digital therapeutics and AI that were less developed during his previous government service. Such tools could extend access at low marginal cost, but their economic effect depends on payment design. He described the administration as having established a policy direction toward outcomes, with substantial implementation and evaluation still ahead.

Part D and Medicare Advantage

Brooks characterized Part D as adapting to the IRA's redesigned benefit. His assessment combined lower beneficiary financial exposure, higher taxpayer costs, fewer plan choices, and a more stable market. He defended reducing additional premium-stabilization support after CMS examined plan bids, rather than making that decision without market evidence. He also identified a future premium-cap issue that would require congressional attention.

On Medicare Advantage, he acknowledged a crisis of confidence involving quality measurement, coding intensity, and selection. Nevertheless, he emphasized its additional benefits and appeal to beneficiaries seeking affordable coverage. He described collaboration between CMS and MedPAC to reconcile coding-related payment estimates, noting that different data years and assumptions explained some apparent disagreement. The comparison remained complicated by differences between the enrolled populations.

Prior authorization and data exchange

Brooks described insurer commitments to reduce and harmonize services subject to authorization, standardize documentation, and improve electronic processing. He reported an initial reduction in the number of procedures requiring authorization, while emphasizing continued work toward faster decisions.

Large health systems and insurers can sometimes achieve nearly instantaneous processing because they possess the necessary resources. Smaller practices need access to comparable capabilities. Interoperability, investment, and longstanding distrust between parties remain obstacles. Brooks's objective was timely exchange of the information necessary to establish that a patient should receive the requested care.

340B and provider reimbursement

Brooks defended examining hospital acquisition costs for 340B drugs and reconsidering reimbursement accordingly. He emphasized the beneficiary as well as the taxpayer: coinsurance calculated from reimbursement can sometimes exceed the institution's own drug acquisition cost. He described transparency and avoidance of duplicate discounts as important goals and credited CMS's implementation of the Medicare transaction facilitator.

He was equally direct about oncology compensation. Physician reimbursement should not depend on how expensive a prescribed drug is, but the agency's ability to repair that structure is constrained by statute. CMS cannot simply replace revenue reductions Congress has embedded in law. He hoped pressure from forthcoming changes would create legislative interest in a better arrangement. In audience discussion, he also acknowledged that reimbursement disparities and administrative burdens can encourage consolidation.

AI and outcomes-based reimbursement

Brooks's AI position combined optimism with skepticism. He said experience so far had largely been cost additive: improving the efficiency of individual tasks can amplify the existing system's spending incentives. Technology's potential to reduce costs and expand access would require an appropriate reimbursement framework.

He highlighted ACCESS as a promising example of paying for measurable clinical improvement, using reduced HbA1c as an illustration. CMS and CMMI could then work together on a payment structure tied to those outcomes. He also discussed reducing the gap between FDA approval and Medicare coverage, while stressing the need for evidence relevant to Medicare's coverage standard.

For pharmaceutical outcomes contracts, Brooks drew a useful distinction. Provider payment can change behavior; a drug contract principally allocates uncertainty about effectiveness between payer and manufacturer. He supported facilitating such arrangements while resisting designs that merely redistribute best-price discounts.

Supply chains and proposals for CMS

Asked about China, Brooks separated dependence on pharmaceutical starting materials from competition in innovation. He was concerned about supply-chain resilience and the US clinical-trial environment, while expecting the country to remain an attractive pharmaceutical market. His warning was that decades of optimizing procurement for cost had reduced resilience.

Brooks's wry warning was that consultants can sell companies months of expensive preparation for a bespoke CMMI model, only to arrive at CMS and be told the agency has no interest. He wanted structural approaches that could serve a broader market.

He urged stakeholders to bring data CMS can validate and proposals developed jointly by providers, payers, and manufacturers. Shared acknowledgment of a problem and evidence that a solution works are more persuasive than competing requests for money. For Friday, his challenge is to show which clinical outcomes AI improves, how those improvements can be measured reliably, and why the resulting payment arrangement would improve patient care rather than simply add spending.

Sources

Agenda: AVBCC 2026 Agenda 10.2.docx, Wednesday October 7 sessions 15, 17 and 21.

Provider panel: 20261007 Panel All About CMMI OCM EOM etc etc.docx.

Sutton interview: 20261007 Joe grogan interviews CMMI leader Abe Sutton.docx.

Brooks interview: 20261007 joe grogan intvu brooks Ctr Medicare CM.docx.

 

CAP Discusses Distributed Genomics, Digital Pathology

 

AI on CAP on DIGITAL PATHOLOGY and REMOTE BIOINFORMATICS

 

https://chatgpt.com/g/g-p-699fbe1fc2c8819189cf64e8233f3a55-danaher-dig-pathol/c/6a8f3daa-126c-83e8-ae47-b699f7b1c832

 

https://chatgpt.com/share/6ab35108-06e8-83e8-b0f2-4c09492ebb45

 

Yes. The CAP record is unusually clear—and it directly conflicts with CMS’s categorical premise that these computational-pathology analyses “do not require” a CLIA laboratory and may be performed by “any non-regulated entity.”

Strongest CAP authority

The most useful document is CAP’s February 23, 2026 response to the HHS Health Sector AI RFI. It is recent, formal, addressed to HHS, and explicit:

• CAP identifies itself as a CLIA-deemed accrediting organization that accredits more than 8,300 laboratories.
• CAP states that its accreditation checklists require laboratory-director validation or verification before patient testing, “including AI and machine learning algorithms.”
• CAP describes FDA regulation and CLIA oversight as “complementary and necessary” for clinical AI.
• CAP says laboratories must validate or verify local AI performance under CLIA before implementation.
• Later in the letter, CAP states directly that CLIA requires pathologists, acting as laboratory directors, to assess AI systems before local implementation.
• CAP recommends using its accreditation and proficiency-testing infrastructure to evaluate, monitor, and safely deploy AI tools.

This is not an inference from general laboratory principles. CAP expressly places clinical AI under CLIA laboratory-director oversight and CAP accreditation requirements.

CAP’s actual inspection checklist

CAP’s 2025 accreditation checklist edition is described in detail in the February 2026 CAP TODAY article, “Digital path practices reflected in latest checklist changes”.

The Laboratory General Checklist now contains a section expressly titled:

Digital Pathology Including Remote Data Assessment

CAP explains that it revised this section specifically to encompass developments in digital algorithms and artificial intelligence. Its definition expressly includes pathologists’ review and diagnostic interpretation of tissue using digital pathology and whole-slide images. It also ties the activity to the laboratory’s CLIA address and explains when a remote interpreting location becomes a separate referral laboratory.

Page two of the CAP article identifies particularly useful checklist requirements:

• Inspectors sample digital-pathology policies and procedures.
• Inspectors sample reports generated through digital pathology.
• Inspectors review digital-pathology validation and verification records.
• GEN.50630 requires the laboratory to validate or verify digital-pathology systems used for clinical diagnostic purposes, with approval by the laboratory director or a qualified designee.
• GEN.52860 requires digital-pathology services to be included in the laboratory’s quality-management system.
• CAP inspectors may examine scan-failure rates, image-quality criteria, risk-mitigation procedures, discordance reconciliation, and other ongoing performance measures.

That is probably the best operational evidence for your letter. CAP is not merely saying that a pathologist should use good judgment when purchasing software. CAP has placed clinical digital pathology inside the inspection checklist used to determine continuing laboratory accreditation.

CAP’s specific AI-validation guidance

CAP’s educational guidance, “How to Validate AI Algorithms in Anatomic Pathology”, is also remarkably direct:

• Both CLIA and CAP require validation of a new test, device, or diagnostic aid before reporting patient results.
• CAP applies its whole-slide-imaging validation principles to image-analysis algorithms.
• CAP states that any image-analysis or image-recognition system—FDA-authorized or not—must be validated before clinical use.
• Validation occurs in the laboratory, using cases representative of the laboratory’s intended clinical use and patient population.
• The medical director establishes acceptance criteria and approves implementation.
• CAP even supplies suggested report language for a non-FDA-authorized AI system treated as a laboratory-developed test.

Although this is educational guidance rather than the checklist itself, it demonstrates how CAP interprets and operationalizes CLIA and CAP requirements for H&E-based AI.

Second formal CAP policy letter

CAP made essentially the same point in its December 1, 2025 comments to FDA on measuring and evaluating AI performance.

CAP states that:

• CAP is a federally deemed CLIA laboratory accrediting organization.
• CAP checklist validation and verification requirements expressly include AI and machine-learning algorithms.
• AI represents a new component of pathologists’ responsibilities as CLIA laboratory directors and section directors.
• CLIA and CAP impose quality requirements beyond the manufacturer’s FDA-authorized operating instructions.
• Laboratories should establish controls, performance metrics, corrective actions, and procedures addressing AI performance drift.

Again, CAP treats AI applied in clinical pathology as part of the laboratory test system—not as an unregulated computer transaction detached from the laboratory.

CMS’s own CLIA guidance

There is also an apparent inconsistency within CMS itself. In its May 11, 2023 CLIA Post-PHE Guidance, CMS treated the remote review of digital laboratory data, results, and images as testing performed under a primary laboratory’s CLIA certificate:

• The remote work must fall within the specialties or subspecialties on the primary laboratory’s certificate.
• The laboratory director remains responsible.
• Records are retained by the primary laboratory.
• Survey findings and enforcement actions attach to that laboratory’s CLIA certificate.
• Digital proficiency-testing review is treated as occurring under the primary site’s CLIA number.

This does not decide every possible algorithm-only service, but it refutes any general theory that converting tissue information into digital images automatically removes subsequent clinical interpretation from the CLIA framework.

The precise conflict with the OPPS proposal

CMS states in the CY 2027 OPPS proposed rule that stand-alone algorithmic analyses are separate from a CLIA laboratory’s examination of human material. CMS then concludes that they:

• do not require CLIA-regulated laboratory services or entities;
• can be performed by any “non-regulated entity” possessing the software;
• are “other diagnostic tests,” rather than clinical diagnostic laboratory tests; and
• should therefore be removed from the CLFS and paid as SaMS under OPPS.

For a purely secondary calculation performed on a finalized laboratory result, CMS’s distinction may sometimes be defensible. It is much less persuasive for computational pathology applied to an H&E whole-slide image. The WSI is a direct digital representation of the patient’s tissue specimen, and the algorithm produces a patient-specific diagnostic, prognostic, or predictive result. CAP expressly subjects that clinical process to laboratory validation, laboratory-director oversight, quality management, reporting requirements, and inspection.

Payment classification and regulatory classification need not always be identical. But CMS’s proposal does not merely select a different payment system. It makes the broader factual and regulatory assertion that no CLIA entity is needed. That is where the collision becomes unavoidable.

Suggested comment-letter language

CMS’s proposed treatment of computational pathology as Software as a Medical Service creates a direct conflict with the existing federal laboratory-quality framework. CMS states that these algorithmic analyses do not require a CLIA-certified laboratory and may be performed by an otherwise unregulated entity. In contrast, the College of American Pathologists—the nation’s largest federally deemed CLIA accrediting organization—expressly applies its laboratory accreditation requirements to artificial intelligence and machine-learning algorithms used for patient testing. CAP requires laboratory-director validation or verification before implementation, and its 2025 Laboratory General Checklist includes a dedicated section entitled “Digital Pathology Including Remote Data Assessment.” CAP inspectors review digital-pathology policies, patient reports, validation records, and quality-management activities; checklist requirements GEN.50630 and GEN.52860 specifically address laboratory validation, laboratory-director approval, and continuing quality management of digital-pathology systems.

This conflict is especially pronounced for computational-pathology tests that apply an algorithm to an H&E whole-slide image to produce a patient-specific diagnostic, prognostic, or predictive result. Such a service is not simply an arithmetic calculation performed on a previously reported laboratory value. It is the clinical examination and interpretation of a digital representation of human tissue. CAP’s current accreditation framework treats that activity as patient testing conducted under the responsibility of the CLIA laboratory director.

CMS should therefore withdraw or qualify its categorical conclusion that SaMS analyses performed on laboratory-test data require no CLIA-regulated entity. At minimum, CMS should clarify that classification and payment of a service under OPPS does not displace otherwise applicable CLIA requirements; distinguish purely secondary analysis of finalized data from computational examination of whole-slide tissue images; and consult with CMS’s CLIA program and federally deemed accrediting organizations before finalizing the regulatory status of computational-pathology services.

One terminology refinement: CAP is not universally “CLIA’s inspection body.” The technically exact phrase is “a CMS-approved, federally deemed CLIA accrediting organization.” CAP states that, for CAP-accredited laboratories, a CAP inspection occurs in lieu of a CMS inspection. That wording will be harder for CMS to evade.

 

Prosigna Messy Journey

 You could not find 81520 in the 2017 PAMA data because it was not yet the Prosigna code. The test was reported under 0008M during the PAMA data-collection period. CPT 81520 replaced 0008M effective January 1, 2018.

Prosigna pricing chronology

Year	Code	Medicare CLFS rate	What happened
2016	0008M	$3,419.42	New Prosigna MAAA code; priced identically to 81519, confirming the Oncotype DX crosswalk
2017	0008M	$3,443.36	Inflation-adjusted pre-PAMA national limitation amount
2018	81520	$3,099.02	0008M replaced by 81520; first PAMA reduction, capped at 10%
2019	81520	$2,789.12	Second 10% reduction
2020	81520	$2,510.21	Third 10% reduction
2021–2023	81520	$2,510.21	Further reductions postponed by Congress

The arithmetic is exact:

• $3,443.36 × 90% = $3,099.02

• $3,099.02 × 90% = $2,789.12

• $2,789.12 × 90% = $2,510.21

The remarkable answer: its PAMA target was $900

CMS’s original 2018 PAMA calculation table, under 0008M, reports:

• 2017 CLFS rate: $3,443.36

• Private-payor weighted median: $900.00

• Nominal reduction: 73.86%

• 2018 phase-in rate: $3,099.02

• 2019 phase-in rate: $2,789.12

• 2020 phase-in rate: $2,510.21

Thus, $3,099 and $2,510 were not the PAMA target prices. They were transitional prices descending toward an underlying PAMA weighted median of only $900. Congressional freezes interrupted the descent after 2020.

Bottom line

Your recollection of the structure was right but the code changed in the middle:

Prosigna was originally priced as 0008M, crosswalked to Oncotype DX code 81519. Its 2017 rate was $3,443.36. PAMA produced a startling weighted-median target of only $900, but statutory phase-in protection limited the reductions to 10% annually. When 0008M became 81520 in 2018, the price history transferred with it.

The roughly $3,800 figure does not appear as a national CLFS amount in the CMS files I checked; it may have been an earlier MolDX/local allowed amount, list price, or commercial reimbursement figure. The documented national sequence begins at $3,419.42 in 2016, rising to $3,443.36 in 2017.

Official CMS materials: 2016 CLFS file, 2017 CLFS file, 2018 CLFS file, and 2018 PAMA payment-rate files.

https://chatgpt.com/share/6ab1c101-ed14-83e8-8850-9f0f890648f0

ICER Consulting

 https://www.linkedin.com/safety/go/?url=https%3A%2F%2Fisa%2Eicer%2Eorg%2F&urlhash=N6u9&mt=OYjUSiWO94LMLLwQgpEHXv_9iOq4e0O0lGlRX-V4bI0Kz4IydbaEfU784oUx14FnDv9pThXaK5xg5bRfKKOazbqatiJ9nPJBy6O_XXu-PYHsuIv9p5_p0kiA&isSdui=true

 The LinkedIn “safety” wrapper is obscuring a very simple direct address:

ICER Scientific Advice: https://isa.icer.org/

ICER formally launched this consulting/scientific-advice service in March 2026. It is aimed at life-science companies and advises on clinical-trial design and evidence-generation strategy, especially evidence likely to matter in later HTA/value assessment. ICER explicitly says the consulting work is kept separate from its independent assessment process.

If that subdomain still refuses to load for you, the ICER announcement describing the service is here: ICER Scientific Advice announcement

https://icer.org/news-insights/press-releases/icer-scientific-advice-offering-2026/?utm_source=chatgpt.com

Institute for Clinical and Economic Review Launches Program to Accelerate the Consideration of Value into Drug Development Programs

– ICER will leverage its health technology assessment (HTA) expertise to encourage better evidence development in clinical trial programs –

BOSTON, March 25, 2026 – The Institute for Clinical and Economic Review (ICER) today formally launched ICER’s Scientific Advice, a new offering to advise companies on evidence generation strategies that will support a more comprehensive assessment of clinical effectiveness and value.   

“Today, ICER joins its HTA colleagues across the world who have for years provided scientific advice on clinical trials. ICER’s Scientific Advice will address a key concern that current clinical trials fail to capture the elements of value that matter to key stakeholders.  Identifying the right kind of evidence needed from clinical trial programs will ensure that value assessments are a foresight, not an afterthought,” said ICER President and CEO Sarah K. Emond. “Our goal is to improve the evidence that is developed by companies during clinical development to ensure value assessors have the information they need to do a fair assessment at the time of FDA approval.”

Drawing on ICER’s experience having conducted over 100 value assessments geared towards clinicians, payers, and patients, the program will advise on trial designs that successfully capture:

• Clinically meaningful, patient-centered outcomes
• Appropriate surrogate outcomes that predict clinical outcomes
• Data collection outside of the trial environment
• Data that can help inform payer decisions

Based on potential trial results and analogous conditions, ICER will show potential evidence ratings and realistic price ranges, giving a clear view of what different outcomes could mean for a future commercial strategy.  One of ICER Scientific Advice’s key goals is to help companies prioritize their clinical assets so the therapies that ultimately receive FDA approval are those with the greatest potential to meaningfully improve patients’ lives.  ICER will also help provide consideration of what outcomes patients prioritize through structured focus groups and advise on recruiting diverse patient populations.

ICER Scientific Advice is a completely separate workstream from ICER’s independent and publicly available evidence assessments. Working with ISA does not impact the likelihood that ICER will conduct a health technology assessment of a company’s product in the future, nor will it impact the independence of our assessment should it be selected through our topic selection process.

For more information on the program and its offerings, visit isa.icer.org.

Undertanding CAP vis-a-vis CMS OPPS and PFS Rules on Digital Pathology

 To: NOTE TO FILE

From: Bruce Quinn, MD, PhD
Date: August 28, 2026
Re: A Priority Issue for CAP Engagement—CMS, SaMS, CLIA, and Computational Pathology

Executive summary

What are current issues involving the College of American Pathologists that might warrant policy engagement. 

AI recommends focusing on one unusually important and timely issue: CMS’s proposal to remove certain computational pathology tests from the clinical laboratory category and reclassify them as Software as a Medical Service, or SaMS.

This proposal unexpectedly challenges two leadership roles that CAP has been building for years. 

• First, CAP has helped shape the coding architecture for digital and computational pathology through the AMA CPT process, including Category III digital pathology codes, pathology-specific algorithm codes, the evolving Clinically Meaningful Algorithmic Analyses framework, and CPT Appendix S for artificial intelligence. 
• Second, CAP is the nation’s most influential real-world implementer of CLIA requirements. CLIA regulations change slowly; CAP’s accreditation checklists, committees, inspectors, and laboratory teams continually translate those regulations into workable standards for new technologies—including digital pathology, AI validation, remote interpretation, and distributed services involving more than one CLIA-certified laboratory.

CMS now proposes, in both the CY 2027 OPPS and PFS rules, that certain algorithmic analyses of laboratory data are not clinical laboratory tests at all. CMS would move them from the CLFS into the much broader statutory category of “other diagnostic tests” and pay them as SaMS. For computational pathology companies such as ArteraAI, Valar Labs, Paige, and others, the proposal could separate the billable algorithmic test from the CLIA laboratory framework in which CAP, CPT, developers, and laboratories have been placing it.

CMS is not formally declaring that every aspect of digital pathology lies outside CLIA. Its immediate proposal concerns stand-alone algorithmic analyses whose CPT descriptors do not name a conventional laboratory method. But the agency’s reasoning could reach directly into clinical computational pathology, where an algorithm examines an H&E whole-slide image and generates a new patient-specific diagnostic, prognostic, predictive, or treatment-selection result.

CAP has built leadership through CPT

CAP has been highly active in establishing the coding pathway for digital pathology. It worked with the AMA CPT Editorial Panel to create 13 Category III digital pathology add-on codes for 2023 and another 30 codes for 2024. CAP explicitly described those codes as an “on-ramp” for artificial intelligence and as a pathway toward eventual Category I recognition and national payment.[1]

That work fits within a wider CPT architecture for algorithmic services:

• Multianalyte Assays with Algorithmic Analyses established that a laboratory test may combine laboratory measurements with an algorithm to produce a patient-specific clinical result.

• Category III codes provide an entry point for new computational pathology tests and generate utilization data needed for later Category I consideration.

• CPT Appendix S classifies AI-enabled medical services as assistive, augmentative, or autonomous. The AMA substantially revised Appendix S in 2026 to clarify what constitutes a clinically meaningful AI output.[2]

• The AMA is developing a possible new framework, tentatively called Clinically Meaningful Algorithmic Analyses, for algorithms that produce medically actionable outputs even when no physician is directly involved at the point of service.[3]

CAP’s February 2026 comments to HHS described the AMA CPT Editorial Panel as a key external governing body for AI adoption. CAP urged HHS to work with CPT to create a clear and clinically appropriate coding framework for AI services.[4] CAP therefore sees coding not merely as billing mechanics, but as part of the governance and adoption pathway for digital pathology.

CAP has also built digital pathology leadership through CLIA accreditation

CAP’s second role is arguably even more important. CAP is a federally deemed CLIA accrediting organization, and a CAP inspection substitutes for a CMS inspection in CAP-accredited laboratories. CAP accredits more than 8,000 laboratories and updates its checklists annually, while the underlying CLIA regulations may remain substantially unchanged for years or decades.[5]

In practice, CAP continually builds the leading edge of “CLIA in real life.” This includes questions that the original regulations could not have anticipated:

• How should a laboratory validate a whole-slide-imaging system?

• Who is responsible for validating an AI algorithm before it is used for patient testing?

• How should performance drift and software updates be monitored?

• How should the laboratory maintain positive patient identification across scanners, image-management systems, algorithms, and reports?

• What happens when slides are created at one CLIA laboratory, scanned at another site, analyzed by a third entity, and interpreted by a pathologist working under yet another CLIA certificate?

• When is a remote site part of the primary laboratory, and when does it become a separate referral laboratory?

• Which laboratory director is accountable for validation, quality management, records, and the final patient result?

CAP’s 2025 Laboratory General Checklist now contains a section specifically titled “Digital Pathology Including Remote Data Assessment.” CAP inspectors may review digital pathology policies, patient reports, validation records, image-matching procedures, scan-failure rates, image-quality criteria, discordance reconciliation, and risk-mitigation procedures. Checklist requirement GEN.50630 requires laboratory validation or verification and laboratory-director approval of digital pathology systems used for clinical purposes. GEN.52860 places digital pathology within the laboratory’s continuing quality-management system.[6]

CAP’s formal policy statements are equally clear. In February 2026, CAP told HHS that its accreditation checklists require laboratory directors to validate or verify new tests and methods before patient testing, expressly “including AI and machine learning algorithms.” CAP described FDA review and CLIA oversight as complementary and necessary, and stated that CLIA requires laboratory directors to assess AI systems before local implementation.[4]

In December 2025, CAP similarly told FDA that AI had become an important new element of the pathologist’s responsibilities as a CLIA laboratory director or section director. CAP emphasized that CLIA and CAP requirements extend beyond FDA authorization and manufacturer instructions to include local validation, controls, performance monitoring, corrective action, and management of changes that could affect patient results.[7]

The CMS proposal moves in the opposite direction from CAP

Section 1861(s)(3) of the Social Security Act contains three broad diagnostic categories reflecting the world of 1965:

1. Diagnostic X-ray tests;

2. Diagnostic laboratory tests; and

3. Other diagnostic tests.

Modern radiology fits largely within the first category. Clinical laboratory medicine and anatomic pathology have historically occupied the second. The third is a broad residual category for diagnostic services that do not fit the first two.

CMS now proposes to extract certain computational tests from the second category and place them in the third. Under the OPPS proposal, CMS would remove ten existing codes from the CLFS and assign them to New Technology APCs as SaMS. Under the PFS proposal, CMS would remove the same codes from the CLFS and contractor price them. CMS also proposes that future codes describing stand-alone algorithmic analyses generally follow the SaMS pathway.[8,9]

CMS gives several reasons:

• The algorithm operates on data generated by a prior laboratory test rather than directly on physical human material.

• The analysis is entirely computer based and, in CMS’s view, can be performed by any unregulated entity possessing the software.

• Algorithmic analysis of laboratory data should be treated consistently with algorithmic analysis of radiology images.

• CLFS crosswalking and gapfilling do not work well for proprietary algorithms whose development costs are difficult to observe.

• The CLFS lacks beneficiary cost sharing and budget neutrality, creating payment and program-integrity concerns.

Some of these are legitimate payment questions. They do not establish that a patient-specific computational pathology test lies outside CLIA.

An H&E whole-slide image is not merely a finalized laboratory value. It is a high-resolution digital representation of the patient’s tissue specimen. An algorithm examining morphology, tissue architecture, spatial relationships, or tumor microenvironment features is computationally examining information derived directly from human tissue. Its output may depend on fixation, staining, slide preparation, scanner characteristics, image resolution, software configuration, patient identification, local workflow, algorithm version, and performance drift. These are exactly the variables CAP addresses through laboratory validation, quality management, and inspection.

CMS’s own CLIA program has previously treated clinical review of digital images as work performed under a CLIA certificate. Its 2023 remote-review guidance left the primary CLIA laboratory and laboratory director responsible for digital review, documentation, reporting, and survey findings.[10]

The statutory payment question

Section 1834A makes the CLFS/PAMA methodology the default for a separately payable clinical diagnostic laboratory test. New CDLTs are assigned CLFS payment through crosswalking or gapfilling. CMS can carve some CLIA-regulated services out of separate CLFS payment—for example, hospital laboratory tests bundled into an OPPS payment, anatomic pathology services paid under the PFS, or tests incorporated into inpatient payments.[11]

But there is an important distinction between creating a defined payment exception and declaring that a clinical test is no longer a laboratory test. If CMS wants to establish a special payment pathway for computational pathology, it should identify the relevant authority and preserve applicable CLIA requirements. Dissatisfaction with PAMA ratesetting, cost transparency, coinsurance, or budget neutrality does not itself convert a CLIA test into an unregulated “other diagnostic test.”

Why CAP should care—and where a consultant could engage [THIS IS LOOK AND FEEL AI DRAFT ONLY]

This is not only a reimbursement dispute. It potentially removes an emerging field from two domains in which CAP has invested substantial institutional authority: laboratory coding and CLIA accreditation.

If computational pathology becomes simply an unregulated SaMS transaction:

• CAP’s validation and quality-management requirements could become optional or unclear.

• Responsibility could become fragmented among the scanning laboratory, algorithm developer, interpreting pathologist, hospital, and billing entity.

• CPT may demand CLIA documentation for a code that CMS subsequently declares requires no CLIA entity.

• Laboratories may be uncertain which organization owns the test result and which laboratory director is accountable.

• Software vendors may lack a workable Medicare supplier-enrollment and billing category.

• CMS could create a regulatory gap precisely as computational pathology begins generating increasingly consequential prognostic and treatment-selection results.

For example, the digital pathology industry has a constructive interest in helping CAP address this issue. The objective need not be to reject SaMS as a concept. Rather, engagement could encourage CAP to propose a clinically credible boundary between:

• A genuinely secondary calculation performed on finalized data, which may reasonably be treated as SaMS; and

• Computational examination of a whole-slide image or other specimen-derived data to create a new clinical result, which should remain subject to CLIA laboratory oversight.

Specific outside consultant engagement topics could include: [AI SUGGESTIONS ONLY - NOT CONFIRMED]

• Coordinated CAP comments on both the OPPS and PFS rules;

• Preservation of CLIA oversight even if CMS chooses a different payment method;

• A clear framework for distributed computational pathology involving multiple CLIA laboratories;

• Allocation of responsibility among the slide-producing laboratory, scanning site, algorithm provider, interpreting pathologist, and reporting laboratory;

• Validation and revalidation standards across different scanners, stains, sites, and software versions;

• A workable pathway from Category III coding to Category I recognition and national payment; and

• Alignment among CAP accreditation, CPT coding, FDA regulation, and Medicare supplier enrollment.

This is a natural issue for CAP leadership. CAP has already asserted that clinical AI belongs within the responsibilities of the CLIA laboratory director. CMS has now proposed a payment policy premised on the opposite conclusion. Policy experts can help CAP defend an important quality role while supporting a practical, innovation-friendly framework for the distributed digital pathology systems that are now emerging.

References

1. College of American Pathologists. How to Use the New Digital Pathology CPT Codes; see also Digital Pathology Codes.

2. American Medical Association. CPT Appendix S: Taxonomy for Artificial Intelligence in Medical Services and Procedures.

3. American Medical Association. CPT Codes Offer the Language to Report AI-Enabled Health Services.

4. College of American Pathologists. Response to the HHS Health Sector AI Request for Information, February 23, 2026.

5. College of American Pathologists. CAP Laboratory Accreditation Program.

6. CAP TODAY. Digital Path Practices Reflected in Latest Checklist Changes and page two, February 2026.

7. College of American Pathologists. Response to FDA on Measuring and Evaluating AI-Enabled Medical Device Performance, December 1, 2025.

8. Centers for Medicare & Medicaid Services. CY 2027 OPPS Proposed Rule, 91 Fed. Reg. 41734, especially 41918–41927.

9. Centers for Medicare & Medicaid Services. CY 2027 PFS Proposed Rule, 91 Fed. Reg. 43842, especially 43910–43912.

10. Centers for Medicare & Medicaid Services. CLIA Post-Public Health Emergency Guidance, May 11, 2023.

11. Social Security Act §1834A, 42 U.S.C. §1395m-1. Official statutory text.

12.