Business Economics and Pathology: Going Beyond the Basics

 The June 2026 issue of CAP TODAY features an article on pathology business economics by Dr Thomas Mezzetti.  He cites a 2025 AMA study on health insurance consolidation, and 2024 white paper from CAP adverse impacts of private payors.  Chat gpT offered to read both and bring out similarities and insights.

Available as PDF white paper.

 

CAP’s 2024 payors white paper and AMA’s 2025 health insurance competition report describe two levels of the same problem.  CAP shows how insurer interference reaches into pathology: narrow networks, steering, prior authorization, coding edits, take-it-or-leave-it contracts, and disruption of locally integrated diagnostic care. AMA supplies the market-structure explanation: most commercial and Medicare Advantage markets are highly concentrated, giving insurers monopsony power over physicians and hospitals.  Together, the papers show that payer power is not merely a reimbursement issue but a governance issue. Insurers increasingly control the “rules layer” of health care: networks, documentation, medical necessity, claims edits, appeals, and payment timing. Physicians are weakened by fragmentation; pathologists are especially exposed because their work is essential but often invisible and easily mischaracterized as commodity testing. This white paper then builds beyond summary to offer an original industry-strategy argument. Payer “friction” is often taken for granted as the familiar background noise of health care administration — denials, prior authorization, coding edits, appeals, and related burdens. Viewed through the Nobel Prize-winning work of Ronald Coase and Oliver Williamson, however, this friction can be understood as a manifestation of transaction costs that, under asymmetric market power, may become instruments of control.

###

The 2027 OPPS Proposed Rule: CMS Wrestles Actively with Software as a Medical Service (SaMS)

 Header:  The OPPS proposed rule for 2027 is released.  After a couple years of patchwork solutions, CMS adopts broad interim measues for software-dominant services.  CMS also promised a more worked-out plan for next year.

##

Find the proposed rule here:

https://www.federalregister.gov/public-inspection/2026-13656/medicare-program-hospital-outpatient-prospective-payment-and-ambulatory-surgical-center-payment

##

CMS Takes the Next Step on Paying for Medical Software:

From RFI to Rulemaking

In July 2025, CMS asked the public how Medicare should think about reimbursement for software-based medical services. It did so in both major Medicare payment worlds: the Physician Fee Schedule, where the issue is practice expense and Part B professional services, and OPPS, where the issue is hospital outpatient payment. In the PFS proposal, CMS specifically asked for comment on “Software as a Service,” noting that software algorithms and AI do not fit comfortably into older practice-expense categories built around clinical labor, supplies, and equipment.

December widened the aperture. In December 2025, HHS issued a broader RFI on accelerating the adoption and use of AI in clinical care. One of its sections focused directly on reimbursement, noting that government fee-for-service payment can be slow in recognizing high-value interventions. (HHS.gov) (Federal Register)

That December HHS RFI was not limited to Medicare coding minutiae. It asked a larger policy question: if AI-enabled tools can improve care, reduce friction, or avoid downstream costs, how should federal health programs avoid becoming the bottleneck?

ACCESS as the Parallel Universe

CMMI is already testing a different answer. The ACCESS model — Advancing Chronic Care with Effective, Scalable Solutions — tests outcome-aligned payments in Original Medicare for technology-supported chronic care. CMS describes ACCESS as a model that uses predictable payments, with full payment tied to measurable health goals rather than simply paying for enumerated activities. (CMS) (CMS)

That matters because ACCESS is, in effect, one policy laboratory for paying software-enabled care. Vendors and care organizations can receive recurring payments for chronic disease interventions, but they must meet defined outcome measures to avoid reduced payments. This is very different from the OPPS problem, where CMS is still trying to decide whether a software output should sit in a clinical APC, a new technology APC, a packaged status indicator, or something else entirely.

SaaS Becomes SaMS

In the new OPPS proposal, CMS says it has been evaluating a comprehensive SaMS payment approach for several years, has sought input through several comment solicitations, and wants a payment strategy aligned with quality, health improvement, cost reduction, and system strengthening.

A new name for a Medicare problem. CMS now proposes to retire “SaaS” and use “Software as a Medical Service,” or SaMS. The reason is partly linguistic and partly policy-driven. “SaaS” already means ordinary cloud-based software in the technology world. CMS wants a Medicare term for software-based technologies that support clinical decision-making through algorithmic analysis, including diagnostic or clinical functionality.

The distinction is important. CMS is not talking here about ordinary hospital IT, billing software, EHR modules, or cloud subscriptions. It is talking about clinical software that produces medical information — for example, algorithmic analysis of images, risk models, computer-aided detection, and other outputs that assist diagnosis or treatment planning.

Why OPPS Has Trouble With Software

The old system prices tangible inputs. CMS candidly states the core problem. Medicare Part B payment systems, including OPPS, were built mainly to pay for services with material resources: staff time, equipment, supplies, rooms, and procedural overhead. SaMS is different. Its value may lie in a proprietary algorithm, a trained model, scalable non-material infrastructure, licensing terms, and clinical performance.

That creates two awkward questions. First, how should Medicare value a service when the “cost” is not a scanner, a catheter, or a technician’s minutes, but a licensed algorithm? Second, how should Medicare handle hospital acquisition models that may involve subscriptions, enterprise licenses, per-use charges, or “per-click” fees?

CMS also flags program integrity. A per-click software fee can look like a cost input, a royalty, a utilization incentive, or all three, depending on how the business arrangement is structured.

The 2027 Interim Policy: A Holding Pen

New technology APCs for now. For CY 2027, CMS proposes an interim policy. It would designate 36 HCPCS codes as SaMS services. For 21 separately paid SaMS codes currently assigned to clinical APCs, CMS would move them into new technology APCs that approximate their CY 2026 payment rates.

This is not yet a full value-based methodology. It is more like a structured holding pen. CMS is saying that the clinical APC architecture does not adequately accommodate SaMS, but CMS is not yet ready to announce a final long-term payment model.

• SIDEBAR:  Table 61 shows 36 HCPCS codes that will be reclassifed as SaMS.

The New O1 Status Indicator

A label with payment consequences. CMS also proposes a new OPPS status indicator: “O1,” defined as Software as a Medical Service, paid under OPPS with separate APC payment. Functionally, O1 would behave like status indicator S, meaning separate payment and no multiple-procedure discounting.

CMS also asks whether a status indicator behaving more like T — subject to multiple-procedure discounting — would be more appropriate. That is a small technical question with large implications. If SaMS tools are frequently used alongside imaging, procedures, or other diagnostic services, multiple-procedure discounting could materially change the business model.

Packaging Is Not Forgotten

CMS avoids sudden disruption. Some SaMS codes are currently conditionally packaged under Q1. CMS considered more aggressive packaging, including unconditional packaging under status indicator N, but concluded that such a move could interrupt patient access. For 2027, CMS proposes to maintain current clinical APC and status-indicator assignments for SaMS codes that are already conditionally packaged.

That is classic CMS incrementalism. The agency is not declaring that every algorithm deserves separate payment. But it is also not forcing all software into packaging before it has a durable framework.

The Table Is the Policy

Table 61 is the map. The proposal includes a detailed Table 61 listing the HCPCS codes CMS proposes to treat as SaMS. The list includes automated retinal imaging analysis, coronary FFR derived from CT angiography, fracture-risk software, concussion eye-movement analysis, ECG algorithmic risk assessment, brain MRI quantitative analysis, cardiac arrhythmia simulation, prostate cancer estimation mapping, perivascular fat cardiac risk assessment, 3D anatomical segmentation, and other software-derived clinical analyses.

The diversity of the list is the point. SaMS is not one clinical specialty. It is a payment category cutting across ophthalmology, cardiology, radiology, neurology, orthopedics, oncology, and procedural planning.

The Laboratory Pivot

The next problem is lab algorithms. CMS proposes that primarily software-driven or algorithmic lab tests are no longer classed as lab tests, but as non-CLFS software medical services.  This will certainly be controversial.  CMS proposed to remove about a dozen codes from the CLFS and plant them instead on the OPPS code list.  For now, prices would be crosswalked from the most recent CLFS price.

Why This Matters

CMS is separating the specimen from the software. This is the conceptual move. The original lab work may be regulated, coded, and paid as a laboratory test. But subsequent algorithmic interpretation of already-generated data may be a different thing. It may be proprietary. It may be scalable. It may not map cleanly onto CLFS, PFS, or OPPS conventions.

For laboratories, hospitals, AI companies, and genomics firms, this is a major signal. CMS is beginning to build vocabulary and payment architecture for software-derived medical information, even when that information is downstream from a conventional diagnostic input.

The Policy Direction

Not value-based yet, but not business-as-usual. The 2027 OPPS proposal does not create a grand unified theory of medical software payment. It does something more modest and more practical. It renames the category, creates a SaMS-specific status indicator, moves many separately paid services into new technology APCs, preserves some packaging arrangements, and asks for further comment.

Seen alongside the July 2025 RFIs, the December HHS AI RFI, and the ACCESS model, the direction is clear. Federal policy is no longer treating medical software as a novelty at the edge of the payment system. CMS is now trying to decide whether SaMS should be paid like a diagnostic test, a procedure, a hospital resource, a software license, or an outcome-linked intervention.

For 2027, the answer is still transitional. But the transition itself is the story.

Run, Don’t Walk: AMA CPT Big Changes for Software-Dependent Services

Run, Don’t Walk: AMA CPT Big Changes for Software-Dependent Services

I wrote a long blog on Appendix S a few days ago, here's the concise version I used as a mini-post on Linked In.

###

AMA has published the 2027 version of “Appendix S,” its greatly-revised policies for software-intensive services. 

• But there’s more! AMA is also rolling out BIG changes and never-before-seen requirements in the CPT new code process.
• Apply quickly to understand and track these changes before the 9/2026 CPT meeting in Minneapolis.

See the AMA CPT AI press release:
https://lnkd.in/gufqGaNj

See the new 2027 Appendix S for AI:
https://lnkd.in/gZ-55ggj

Go to AMA CPT “Smart App”, being sure you click at top for “interested Party Portal.” (May require email registration.)
https://lnkd.in/gx-YVr8P

Now, click on Tab 94 for IP (Interested party) access.

This is the lab version of the CPT application, but I believe the general-code version opens for comment by July 13.

See "green" text: There are A LOT of changes proposed for evaluating and documenting AI or other software during the CPT application process – including a lot of new applicant work and new committee review responsibilities.

I expect by mid-July CMS will release the same changes for the "general purpose" code change process.  AMA CPT may also soon release information on the new CMAA algorithmic services code category - that will be either now July-August or this winter December-January.

#artificialintelligence #amacpt #coding #cpt #appendixS

CMS Drops a Bomb on top of AI Lab Policy (OPPS Rule)

HEADER: CMS Proposes Big Shift for Software-intensive Lab Services

 See pages 457-462 of the new Hospital Outpatient rulemaking for CY2027:

https://www.federalregister.gov/public-inspection/2026-13656/medicare-program-hospital-outpatient-prospective-payment-and-ambulatory-surgical-center-payment

###

What was Wet Lab and Dry Lab is now... Wet Lab and Not-A-Lab?  ?!?

###

News in Brief

[ChatGPT]

CMS has proposed a major CY2027 OPPS policy shift for “software as a medical service” analyses performed on prior laboratory test data. In recent years, lab AI and lab-plus-algorithm services have often been channeled through MAAA, PLA, and CLFS coding/payment pathways. The new OPPS proposal partially reverses that direction. 

CMS argues that once the original laboratory work is complete, downstream algorithmic analysis of stored data—such as genomic, transcriptomic, or digital pathology data—may no longer be a clinical diagnostic laboratory test at all.

CMS proposes to move 10 such HCPCS codes out of the CLFS and into OPPS New Technology APCs, with a new status indicator “O1” for separately paid software-as-a-medical-service items. The agency emphasizes that these analyses may be performed by non-CLIA software entities, creating a new Medicare boundary between the laboratory test and later AI/software interpretation.

[I'm not sure that they can be performed by non-CLIA entities, but I'll leave that to CLIA law experts.]

Deeper Dive

CMS’s CY2027 OPPS proposed rule includes a compact but potentially far-reaching section on what it calls “SaMS analyses performed on laboratory tests.” SaMS stands for software as a medical service. The target is not ordinary hospital software, and not simply AI used inside a lab workflow. Rather, CMS is focused on algorithmic analyses performed downstream of a prior laboratory test, especially where the raw or processed data already exist and the later service is essentially computational.

CMS gives the example of genomic sequencing. The sequencing itself may be performed once by a CLIA-certified laboratory. But after that, the same underlying sequence data can be reanalyzed many times by different algorithms to produce diagnostic, prognostic, risk, or treatment-related information. CMS’s key point is that this later algorithmic step may not require a CLIA laboratory at all. A non-CLIA software entity with the relevant program could perform the computation. That distinction drives the policy proposal.

This is why the proposal feels like a flip-flop from the recent coding and payment trajectory for lab AI. Over the past several years, the ecosystem has treated many lab-plus-algorithm products as laboratory services. AMA created MAAA codes for multi-analyte assays with algorithmic analysis. PLA codes have also been used for proprietary lab and algorithmic services. CMS has then often placed these items on the Clinical Laboratory Fee Schedule, using crosswalk or gapfill logic. In practical terms, the system has been willing to treat “lab data plus algorithm” as a lab test.

Now CMS is drawing a sharper line. The agency says it does not believe these secondary algorithmic analyses should be considered clinical diagnostic laboratory tests or priced under the CLFS when they do not require CLIA-regulated laboratory performance. CMS instead characterizes them as “other diagnostic tests” under Medicare law, not “diagnostic laboratory tests.” That is a significant conceptual move.

CMS also flags payment-policy concerns. 

• The CLFS generally lacks beneficiary cost-sharing and is not budget-neutral in the same way OPPS is. 
• CMS also says it often receives limited transparent cost information for algorithmic components because companies consider the details proprietary. 
• Historically, CMS may have relied on visible lab methods in code descriptors—NGS, RT-PCR, methylation, and so on—but CMS now says that comparison is not appropriate when the service at issue is entirely computer-based.

For CY2027, CMS proposes to assign 10 existing HCPCS codes to OPPS New Technology APCs, choosing APC bands that approximate current CY2026 CLFS payment levels. These include oncology and digital pathology AI services, transcriptomic or genomic algorithmic analyses, tumor-response prediction, recurrence scoring, MSI/HRD assessment, and comparator exome analysis. The proposal also creates a new status indicator, “O1,” meaning software as a medical service, paid separately under OPPS.

The larger issue is not just these 10 codes. CMS also proposes that future SaMS analyses performed on lab-test data should go to New Technology APCs under OPPS rather than the CLFS. If finalized, this could reshape the business model for lab-adjacent AI, blur the old PLA/MAAA assumptions, and force companies to think much harder about whether their product is a laboratory test, a software diagnostic, or both.

##

##

SIDEBAR: The Road to Today...

The CMS OPPS announcement was inevitable once AMA CPT last winter began issuing lab-AI codes as Category III codes. CMS handles every Category III code in its OPPS rulemaking, so the AMA CPT maneuver back in February,  loaded the gun for this OPPS proposal now in July.

SIDEBAR: Odd Code List Table

CMS lists 9 PLA codes plus CPT code 81416 (comparator exome).  I would think they'd leave off wet lab exome 81416, but could have included 81417 (exome re-analysis dry lab) and 81427 (genome re-analysis dry lab.)

SIDEBAR: Re-examining Genomes

Timely to this discussion: New high-bandwidth approaches to reviewing genomes/exomes at future dates.  See See Welland et al here (24 June Nat Med), see briefing here, coverage at Genomeweb here.

SIDEBAR: Typo  493.21, 493.2

CMS refers erroneously to "CFR 493.21," which does not exist.  What CMS is saying there - in awful syntax - is that it thinks there are things that are performed on lab tests, which are software rather than CLIA lab services (as define at CLIA 493.2).  These "things" will be called, by CMS, SaMS or Software as a Medical Service.

 There's no claim by CMS that SaMS is a CLIA term, at 493.2, 493.21 (sic), or anywhere else. 

What was Wet Lab and Dry Lab is now... Wet Lab and Not-A-Lab?

Changing Codes Table 62. Two parts.  Click to enlarge.

Lighthouse Lab Services / July 22 / Webinar on Qui Tam Developments

Back on April 23, I posted an AI-driven analysis of a new qui tam court case involving, "MD Labs."

Lighthouse Lab Services will host a webinar on the topic on July 22, with some of the original attorneys involved.  (Some of them also spoke at Executive War College a month back.)

Register here:

https://www.linkedin.com/events/7477359203842482176/

A New Go-To Essay on the Value of CDx (from Hannah Mamuszka)

Hannah Mamuszka, a cofounder of consultancy Alva10, has released a Go-To Essay on the Value of CDx, particularly in a world where drugs are increasingly expensive.

This caught my attention in part because I was able to co-author an article this spring about how diagnostics get more and more cost-effective as drug costs rise (Kansal et al., 2026).  

Mamuszka discusses a wide range of emerging drug/test examples, and discusses some marketplace disincentives to better HEOR. 

Find Mamuszka's essay at Substack here:

https://substack.com/home/post/p-196116530

Here’s a ~200-word summary / Chat GPT:

Hannah Mamuszka argues that the national debate over drug prices misses half the problem: many drugs are prescribed to patients who will never respond to them. Modern drug development often targets specific biological pathways, yet prescribing still frequently proceeds by trial and error. TNF inhibitors, for example, may work well for patients whose disease is TNF-driven, but response rates are often only around 30–40%. Similar issues occur across oncology, autoimmune disease, cardiology, psychiatry, and other fields.

The remedy, she argues, is broader use of diagnostics that predict either drug response or serious toxicity. Pharmacogenomic tests can identify patients at risk from drugs such as fluoropyrimidines, thiopurines, abacavir, clopidogrel, and antidepressants. Response-prediction tests can determine whether a patient’s disease biology matches a drug’s mechanism, as already accepted in many areas of oncology.

The obstacle is not scientific feasibility, but market design. Pharma benefits from broad drug markets; PBMs may favor drugs with higher rebates; insurers often resist paying for tests. But employers and patients may increasingly see the value of diagnostics that prevent months of ineffective, costly, or harmful treatment. The core message: precision diagnostics are not a luxury; they are a way to stop paying for predictable failure.

Humor: AI News Article About My Appendix S Blog

Header: All About a New Blog about Appendix S

June 27, 2026 (Los Angeles)

By: Chat GPT 5.5

In a new post at Discoveries in Health Policy, medical innovation and reimbursement expert Bruce Quinn uses an unusual but effective format: he asks two advanced AI systems to analyze the AMA’s revised 2027 CPT Appendix S, then compares their conclusions. The result is a layered review of one of the more important technical-policy documents now shaping how artificial intelligence may be described in medical coding.

The central finding is that AMA’s revised Appendix S is a real improvement over the prior version. The newer taxonomy moves away from vague language about “work performed by the machine” and instead focuses on software outputs, clinical meaningfulness, and the role those outputs play in care. The familiar three categories remain: assistive, augmentative, and autonomous. But the 2027 version gives these categories sharper practical meaning.

Both AI reviews identify the same unresolved problem. Appendix S is titled as a taxonomy for artificial intelligence, yet it expressly declines to define “AI.” That may be defensible, since CPT generally codes medical services rather than underlying technologies. But it creates an odd threshold issue: if “AI” is undefined, when exactly must an applicant use the AI appendix?

In an interview, Quinn provided this example. An applicant feels his service is not AI, and doesn't refer to Appendix S.  At review, several panelists feel the service is AI, and can't proceed without conformity to Appendix S.  But now 3 or 4 panelist voice their agreement with the applicant, and say the service falls outside of Appendix S.  More panelists join each side of the debate, offering diverse rationales for either position.  How is this paradox to be resolved, with no AMA definition of AI?

The stronger critique concerns decision logic. The revised Appendix S is written in prose and table form, but it does not provide a flow chart or formal classification rules. Quinn’s post emphasizes that this matters because ambiguous prose can hide contradictions. A logic tree would have forced the AMA to write an equally logical appendix - and decide whether classification turns on the type of software output, the physician’s role, or the action that follows.

The Claude analysis adds a particularly useful point: Appendix S may be blending two axes—output type and physician involvement—while treating them as one. That could create edge cases and contradictions for developers, payers, and CPT applicants.

Overall, the post is a thoughtful, technically informed critique: Appendix S is improving, but it still reads more like a negotiated taxonomy than a fully tested classification machine.

___

See also: Is an ICD an AI?

See also:  AMA press release on New Appendix S:

https://www.ama-assn.org/practice-management/cpt/cpt-editorial-panel-strengthens-ai-taxonomy-keep-pace-tech

Appendix S from A to Z. Two AI's Review New Appendix S for 2027

 Two AI's Review the AMA's New Artificial Intelligence Coding Rules "Appendix S 2027"

See AMA"s press release about early-release of a new Appendix S on AMA CPT AI Policy.

https://www.ama-assn.org/practice-management/cpt/cpt-editorial-panel-strengthens-ai-taxonomy-keep-pace-tech

As of June 27, I believe the 2027 version of Appendix S is here:

https://www.ama-assn.org/practice-management/cpt/cpt-appendix-s-taxonomy-artificial-intelligence-medical-services-procedures

And the 2025-and-earlier version still online here;

https://www.ama-assn.org/system/files/cpt-appendix-s.pdf

The rest of this blog is a comparison of New and Old "Appendix S," first by Chat GPT then by Claude Opus.

###

CHAT GPT 5.5 followed by CLAUDE OPUS 4.8

###

AMA’s 2027 Appendix S improves the AI taxonomy by focusing on software outputs, clinical meaningfulness, and the assistive/augmentative/autonomous distinction. But it still oddly refuses to define “AI,” and it lacks formal decision rules or a flow chart. Without logic-tree testing, the prose may hide ambiguities, contradictions, and edge-case classification problems for future CPT applicants and reviewers.

AMA CPT: Coding Policy EARLY RELEASES: Cat I, Cat III, PLA, APP S AI, ADMIN MAAA

 AMA has released a website that displays "early release" updates in policy in five areas:

• Category I Codes
• Category III Codes
• PLA Codes
• Appendix S: The "AI Taxonomy"
• Administrative MAAA codes

Find them all here:

https://www.ama-assn.org/practice-management/cpt

Early Release

The Category I, Category III, PLA, and administrative MAAA code sets all include some form of early-release codes. Category III, PLA, and administrative MAAA codes are early release by definition: Category III codes are released twice a year, and PLA codes are released quarterly. Category I codes are rarely released early, with vaccines being one exception.

Guest AI Blog: Radiology is Starting to Write Reports. Pathology Should Pay Attention.

 

Radiology AI Is Starting to Write Reports. Pathology Should Pay Attention.

[Claud Opus 4.8]

Radiology AI is crossing the line from flagging findings to drafting the report. STAT's recent coverage of Cognita, Aidoc, RSNA 2025, and two new FDA breakthrough designations is, for digital pathology, a preview of the next three years — including the awkward part nobody markets: a draft report doesn't save the physician's time so much as it relocates that time into editing.

See also Jerome Avondo at Linked In.  Jonathan Govette at Linked In.

---

For a decade, AI in medical imaging has meant a spot detector. Flag the pneumothorax. Count the positive cells. Measure the tumor. Highlight the suspicious region. Useful, narrow, and bounded — and in every case the physician still writes the report.

That line is now being crossed. A narrow tool says "possible pneumothorax." A report-drafting model says "small right apical pneumothorax, new since prior study," and drops that language into a preliminary report for a radiologist to review, edit, and sign. The first is an alert. The second is an attempted clinical work product. The difference is the whole story.

Three STAT articles by Katie Palmer trace the move. Worth reading as a set, because they form an arc that pathology is likely to walk in turn: a marquee acquisition, a field-wide reality check, and the FDA beginning to engage.

Pathology Codes for September 2026 Minneapolis Meeting

The next AMA CPT meeting is in Minneapolis, September 17-19, 2026.

The registration isn't open yet, but will be found here:

https://www.ama-assn.org/membership/events/cpt-editorial-panel-meeting

Lab tests follow an early comment cycle, from now June 26 to July 13.

Note that comments are due noon central, ten pacific, on July 13.

(Regular, non-lab code list to be released July 10).

See the lab code agenda here:

https://www.ama-assn.org/system/files/september-2026-lab-agenda.pdf

The lab code list is pretty simple (note, there will also be 20-plus PLA codes.)

Tab 38, revise bone marrow biopsy, 88305-7.

Tab 39, deletet 81520, breast cancer recurrence score. (This is a legacy Prosigna score, a new NGS based Prosigna code is coming.)

Tab 40, revisions to blood bank services.

Tab 41, 87xxx for adenovirus.

Tab 68, Cat III for digitized prostate slides.  "X672T"

Tab 94, Revise pathology code application to accomodate AI issues.

##

Note: The agenda PDF has a link that sends you to a comment portal.  You will find yourself at "CPT Smart App" and near the top you have to click the option for "Interested Party Portal."

As of noon PT, 6/26, the Agenda PDF for labs was posted but no comment options had yet appeared on the 'CPT Smart App."  

Very Brief Blog: Price Analysis of Wild 2017 PAMA Data

 We're gearing up for PAMA cycle II, CMS will release millions of lines of data in August or September 2026, in preparation for the new CLFS fee schedule in 2027.

Here's a look back at some of the wild data results in 2017.

https://www.discoveriesinhealthpolicy.com/2017/09/pama-raw-data-provides-unprecedented.html

Here's one panel from the 2017 blog.

Eg BRCA pricing ran from $1000 to $3220 (without mentioning outliers).  I suspect the peak around $3000 was legacy 2014-forward Myriad BRCA sole source pricing, and the range around $1700-2000 was newer entrants contracting to a fraction like 70% or 80% of the CLFS at the time.

###

AI in My Daily Life

I knew you could still get some of the 2017 data (not line items less than 10 units) at data.cms.gov, but Google wasn't finding it.  Chat GPT was able to find it in a second.

###

ME:
Cms.data.gov still has a web page that takes you to cloud data for 2017 PAMA CLFS pricing, but it's hard to find the link. Help??

###

Chat GPT quickly answers:

###

Yes — CMS hides the important link a bit.

The CMS CLFS landing page still has it near the bottom under “For more information” as:

“View the CLFS Applicable Raw Data File (from 2017 reporting)”
https://data.cms.gov/provider-characteristics/hospitals-and-other-facilities/medicare-clinical-laboratory-fee-schedule-private-payer-rates-and-volumes

CMS describes the current CLFS as based on PAMA private-payor-rate reporting and says the 2017 reporting used the January 1, 2016–June 30, 2016 data collection period; that same period still drives 2020–2026 CLFS rates because later statutory delays postponed new reporting effects. (CMS)

The easier direct CSV link, [Excel] from the federal data.gov catalog, is:

https://data.cms.gov/sites/default/files/2021-12/CLFS%20Applicable%20Information%202018%20Raw%20Data%20File-updated%2012152021.csv

Data.gov also lists the API endpoint:

https://data.cms.gov/data-api/v1/dataset/0e57f57d-0acc-4c9c-8f8c-973e3f4a3c4b/data

The dataset title is “Medicare Clinical Laboratory Fee Schedule Private Payer Rates and Volumes.” Data.gov describes it as raw data from applicable laboratories reporting private payor rates and associated volumes for CLFS tests, and it lists the CSV download URL above. (catalog.data.gov)

The Wide Dispersal of Payment Rates for...Anatomic Pathology

From time to time, there are articles about the wide dispersal of prices in US healthcare.  (One patient and health plan pays $1000, another patient and health plan pays $6000).  We saw that wide dispersal of pricing in the 2016 clin lab payment data released in mid 2017 by CMS as part of PAMA market rates for lab test pricing.  By August 2026, we'll have a huge update of 1H2025 clin lab prices for PAMA-II.

Here's a new case study, provided by Jason Nakka at LinkedIn.  He had access to data on contracted pathology biopsy rates (CPT 88305) from Cigna.  As he writes,

"A slide prepared in Washington State can be read by a pathologist in Iowa, for three times the price.   The 88305 read pays median $17 in Oklahoma, and $71 in Nebraska, a 4.2X spread."

Find his dollarized article here.  Find his earlier ratio-based article here.  

See a commercial consultancy/vender for price data PriceMedic.Com.

click to enlarge

He remarks, "the slide travels..."   Obviously that slide, once prepared on glass, can travel even faster by digital pathology.

##

Sidebar

I asked Chat GPT to find 3 articles on the same topic, wide price variability (in other clinical domains).  I quote below but haven't validated myself.

##

Here are three strong comparators for your pathology-price-variation blog. I’d use Oakes 2024 as the closest “John with BCBS vs Tom with United” example, Whaley 2025 as the broad Transparency in Coverage / TiC-era national frame, and Cooper 2019 as the classic economics anchor.  TiC = federally required payer machine readableprice files.

1) Cooper Z, Craig SV, Gaynor M, Van Reenen J. “The Price Ain’t Right? Hospital Prices and Health Spending on the Privately Insured.” Quarterly Journal of Economics. 2019;134(1):51–107. doi:10.1093/qje/qjy020.

Full link: https://doi.org/10.1093/qje/qjy020
Yale page: https://isps.yale.edu/research/publications/isps15-027

This is the classic article showing that, for privately insured patients, prices—not utilization—drive much of the geographic variation in spending, unlike Medicare. It reports threefold variation in private spending across HRRs and very large price dispersion for relatively homogeneous services, including lower-limb MRI prices varying twelvefold nationally and about twofold within HRRs. (isps.yale.edu)

___

2) Oakes AH, Ikard M, Patton C, et al. “Understanding Variation in Negotiated Rates Using Novel Health Plan Price Transparency Data.” JAMA Health Forum. 2024;5(9):e243020. doi:10.1001/jamahealthforum.2024.3020.

Full link: https://jamanetwork.com/journals/jama-health-forum/fullarticle/2823759
PubMed: https://pubmed.ncbi.nlm.nih.gov/39302670/

This is probably the best direct analogy to your “same service, different payer” point. Using Transparency in Coverage data, the authors compared UnitedHealthcare and Blue Cross Blue Shield of Illinois rates for hip/knee replacement and found substantial variation within the same market and even within the same hospital across different payers; the PDF reports a mean same-hospital BCBSIL-vs-UHC difference of $5,178, about 20% of the mean negotiated rate. (JAMA Network)

___

3) Whaley C, Radhakrishnan N, Richards M, Simon K, Chartock B. “Understanding Health Care Price Variation: Evidence from Transparency-in-Coverage Data.” Health Affairs Scholar. 2025;3(2):qxaf011. doi:10.1093/haschl/qxaf011.

Full link: https://academic.oup.com/healthaffairsscholar/article/3/2/qxaf011/7965202
PMC: https://pmc.ncbi.nlm.nih.gov/articles/PMC11798183/

This is the broadest current national Transparency in Coverage - TiC-based article, documenting price variation across geography, payer, and provider. A nice quotable example for your blog: the mean facility fee for a foot X-ray was $86 at Anthem vs $190 at UnitedHealth, and the authors found weak consistency across inpatient and outpatient price levels—supporting your point that commercial prices are negotiated artifacts, not natural reflections of service cost. (academic.oup.com)

Digital Pathology Association/DPA Issues Two Linked Documents on AI Validation and Reimbursement

Digital Pathology Association Issues Two Linked Documents

on AI Validation and Reimbursement

1. Recommendation Statement for the Validation, Implementation, and Clinical Application of Artificial Intelligence Within a Clinical Laboratory from the Digital Pathology Association

https://journals.sagepub.com/doi/10.1177/2993091X261455975

2. Full Speed Ahead: Shaping the Future of Digital Pathology Reimbursement in 2026
https://digitalpathologyassociation.org/blog/full-speed-ahead-shaping-the-future-of-digital-pathology-reimbursement-in-2026

[chat GPT 5.5]

The Digital Pathology Association (DPA) has recently issued two important and closely linked documents. The first is a formal recommendation statement, published in AI in Precision Oncology, Silberman et al., on validation, implementation, and clinical use of AI in digital pathology. The paper addresses analytical validation, clinical validation, scanner and algorithm performance, pathologist oversight, clinical utility, and real-world reliability.  (I covered this 27-page paper in detail earlier.)

The second is an April 6, 2026 DPA Reimbursement Task Force webpage outlining the association’s 2026 policy agenda for digital pathology reimbursement. It frames reimbursement and market access as central barriers to adoption, and states that DPA’s goal is to define and shape the pathway to reimbursement for digital pathology solutions.

The linkage between the two is explicit:

DPA presents the Silberman validation paper not only as a clinical roadmap for laboratories, but also as an evidence base for reimbursement. In DPA’s formulation, payers will not pay simply because AI is interesting or efficient. The field must show that digital pathology AI is safe, validated, clinically meaningful, and important to patient care.

The April 6 policy page is especially notable because it signals a more organized reimbursement strategy. 

DPA says it has partnered with McDermott+ to monitor and influence CMS, AMA, Congress, and other policy venues. The stated goal is to promote positions supporting coding, coverage, payment, and adoption of digital pathology technologies and infrastructure through official comment and direct agency engagement.

DPA writes, "We are no longer just reacting to policy - we are proactively influencing and shaping the future reimbursement framework. 

Take home lesson: 

For readers following digital pathology, this burst of activity is worth watching. The validation paper supplies the scientific and clinical framework; the reimbursement page describes the policy machinery DPA intends to build around it. Together, they suggest that DPA is trying to move digital pathology from promising technology toward a reimbursable, policy-recognized clinical infrastructure.

_________________________

Sidebar: Reimbursement Task Force
https://digitalpathologyassociation.org/reimbursement-task-force

_____

Sidebar: Conference

The DPA "Pathology Visions" conference will be October 16-18 in San Diego.

_____

Sidebar: CMS and Software Valuation

Last summer, CMS issued RFIs (requests for information) in both Physician/PFS and Outpatient Hospital/OPPS rules.  Responses were due by September; CMS merely acknowledged the receipt of numerous responses in the November final rules.  We may see more in July 2026 policymaking for PFS and OPPS.

Last December, HHS issued an RFI for clinical AI which asked about (1) regulation, (2) R&D, and (30 Reimbursement.  HHS held a webinar yesterday (June 25) highlighting its activities in some of these areas, but leaving "reimbursement" out of scope. 

_____

Sidebar: The DPA in 2023 and in 2026

DPA has had policy articles twice in 2023, then none in 2024 or 2025, and now this new reimbursement program of activity in 2026.  

Chat GTP compares the 2023 and 2026 blog here.  Concisely it concluded:

DPA’s October 2023 reimbursement article was a sophisticated but technical map of possible payment pathways, centered on the then-new Category III slide digitization codes but extending into CMS authority, HCPCS alternatives, OPPS software pa

 

yment, and the need for valuation data.

The April 2026 article is different. It is shorter, less technical, and more strategic. Rather than explaining the maze, it announces that DPA is building the machinery to navigate it: a validation/evidence statement for AI, a more explicit reimbursement mission, and a policy partnership with McDermott+ to engage CMS, AMA, Congress, and other decision-makers. In that sense, the 2026 piece reads less like a reimbursement tutorial and more like an organizational reboot.

The June 25 HHS Update on AI in Clinical Care: Meeting Summary

In December 2025, HHS announced an RFI on how it could improve access to AI for clinical care.  See webpage here, Fed Reg here.  (90 FR 60108, 12/23/2025).

On June 25, HHS held an hour-long webinar to update the public on progress.  While the three arms of the program are regulation, research and development, and reimbursement, they opened by announcing this webinar would include regulation and R&D (not reimbursement).

See coverage at Healthcare Dive here.  HCD frames the webinar as it was advertised, an update based on the December RFI and January public feedback.  But I think it was really an direct update on a half-dozen various active projects and demos underway at  HHS.  

Claude Opus summarizes the transcript.

###

HHS Hosts a "One HHS" Clinical-AI Webinar — A Landscape Tour, Not a Comment Readout

A working record of the June 2026 HHS webinar tied to the December 23, 2025 Request for Information, "Accelerating the Adoption and Use of Artificial Intelligence as Part of Clinical Care" (RIN 0955-AA13).

---

Summary  

In June 2026, HHS held a public webinar tied to its December 23, 2025 Request for Information on accelerating AI in clinical care (RIN 0955-AA13). Rather than reporting comment feedback, seven officials sketched a landscape of departmental AI activity under Secretary Kennedy's "One HHS" banner. The RFI names three levers—regulation, reimbursement, and research & development—but the webinar addressed only regulation and R&D; reimbursement was noted to be out-of-scope, for today.

Speakers positioned AI as a "third lever" alongside lifestyle and care infrastructure within the Make America Healthy Again agenda, emphasizing cost growth, chronic-disease management, agentic clinical AI (ARPA-H's ADVOCATE), caregiving and aging (ACL), and risk-proportionate FDA oversight.