PAMA: The Secret Data Views that CMS Never Displays (for GSP Genomic Sequencing Procedures)

Medicare genomic sequencing pricing under PAMA is based on anywhere from 200 to 275,000 claims for the "GSP" service type.   For Exome and Genome, a large body of claims were paid at about half the Medicare CLFS rate.

Note that none of this data is Medicare data; it's private payer data...the behavior of private payers

###

Goodbye to a myth?

Are CMS CLFS prices a benchmark that strongly drive the market?   Well, the data shows that for some  codes (from 81408 to 81425) the actual market prices can be skewed or centered quite a distance from the CLFS benchmark at the time.

###

What's Up?

Medicare's public raw data files for PAMA 2026 provide interesting data views.  Here, I reveal some more, using simple commands and Chat GPT 6.

Case Study: GeneDx (ticker WGS)

The information is relevant. Over the past year, GeneDx has been battling reimbursement issues, with share price varying in that time from around $33 to around $170.  On one day in May, the share price dropped 50%.   

PAMA Prices Trend Down for Exome, Genome

Not helping is the downward trend in the latest PAMA data for genomic procedures, including 81415 exome and 81425 genome.   Here's some data; I include the $2000 (CLFS) rare gene code 81408.

In each case below, the current CMS CLFS price is in the fourth column.  But a lot of the reported payments are cheaper: most of the claims for 81415 (exome, middle row) and 81425 (genome, bottom row) are in the second and third columns.  Those are claims at $1400-2800 [second column] and $2800-4200 [third column.] 

Note that exome, but not genome, has a significant bump to-the-right (blue bar in column 6).  We showed earlier that such behavior, high-paying plans, can really drive free cash flow and profit.  

Star - CLFS price

The skew to the left is even more prominent for 81408, where many payments populate the far left bar ($0-550, while CMS price, $2000).

So if my market depended on the price of Genome being $5000 or nearby, and the largest bucket was column 2, $1400-2800... eeek.

##

Also in the 814nn series, GSP or Genomic Sequencing Procedures, CMS let us count how many services or claims were reported for about 36 GSP codes.   Here is a sample screen shot:

Here's the full table:

Here's the Tier 2 claims, 81400 to 81408 as a bar chart.

Devil in the Details; Understanding GeneDx's Earning Shocks via Reimbursement

 

GeneDx Earnings Calls Show Why Reimbursement Determines the Economics of Genomic Testing

GeneDx’s May earnings shock (down 50%) exposed the financial consequences of a rapid shift toward genome sequencing. Its August and November calls trace the response: manage test mix, expand coverage, and improve collections. The latest results show progress, alongside a practical tradeoff between frictionless ordering and the documentation needed to secure payment.

ChatGPT 6 (Los Angeles, November 10, 2026). GeneDx’s November earnings discussion offers a detailed account of a problem that extends across advanced diagnostics: clinical demand can grow faster than the machinery that converts testing into payment. A laboratory can perform more tests, deliver valuable answers, and still collect less revenue per case than its business plan anticipated.

GenomeWeb captured the latest development in its November 10 headline, “GeneDx Shares Drop as Firm Starts Prioritizing Reimbursement Over Testing Volume.” The company is collecting more documentation before genome testing begins, accepting some delays and lower near-term volume expectations in exchange for a better prospect of reimbursement.

The stock reaction deserves perspective. Shares fell from approximately $65.72 to $54.83 on November 10, about 16.6%, while the five-day chart shows a decline from roughly $69 to $55. But the shares had also traded near $100 only a few months earlier, and their 52-week range was approximately $32 to $171. The latest decline sits within a much larger history of volatility.

The more informative story begins in May. That quarter exposed the economic problem; August described the repair strategy; November showed both progress and the friction involved in carrying it out.

May exposed the consequences of the genome transition

The striking early-May discontinuity in GeneDx’s stock chart followed its May 4 first-quarter report. On May 5, shares fell approximately 49%, closing at $34.51. The company reduced annual revenue guidance from $540–555 million to $475–490 million, a $65 million reduction at the midpoint. investors.com

The May earnings call explained that the shift toward genome had depressed blended average reimbursement. Management attributed $36 million of the annual revenue revision to reimbursement effects associated with product mix, $11 million to lower volume contributions from expansion markets, and $18 million to noncore businesses.

An important qualification accompanied that explanation. Management said contracted prices and collection rates within the respective products and channels had not materially changed in Q1. The immediate surprise was the composition of the business: a greater share of testing was moving into services with less favorable realized economics. The Motley Fool

This distinction matters. A laboratory does not need to experience an explicit payer price cut for its average revenue per test to fall. A change in what clinicians order—and the settings and payer populations in which they order it—can produce that result.

August laid out the repair strategy

By the August 3 second-quarter earnings call, GeneDx was describing a coordinated effort to improve reimbursement. Management emphasized test mix, coverage expansion, payer-specific workflows, and technology investment.

The underlying economics challenged an intuitive assumption that a broader assay should generate more revenue. GeneDx reported that genome’s contracted rates were lower than exome’s, while genome cost nearly twice as much to produce, largely because of reagents.

Collections added another constraint. Only about 32% of outpatient genome claims were being paid, versus 43% a year earlier. Management described the earlier acceptance of volume ahead of coverage as a market-development investment.

The company began moderating the transition. Genome’s share of outpatient insurance-based testing fell from nearly 40% in Q1 to 32% in Q2, with more emphasis on exome and reflex offerings. Genome remained the long-term destination, but the company was adapting the pace of adoption to the economics of delivering it. The Motley Fool

GeneDx reported Q2 revenue of $114.4 million, adjusted gross margin of 70%, and adjusted net income of $0.4 million. The improvement established a baseline for the next phase of the reimbursement work. ir.genedx.com

November made the operational tradeoff explicit

The November 9 third-quarter earnings call described what happens when a laboratory begins enforcing more of the payer’s documentation requirements before it performs the test.

GeneDx aligned genome workflows to obtain additional information from clinicians before laboratory testing started. This can improve the likelihood that the eventual claim satisfies the payer’s requirements. It can also extend the interval between receipt of an order and the start of testing.

Management reported cancellations, downgrades, and delayed starts. Accessions grew 28% year over year, while completed exome and genome result volume grew 25%. The company lowered its full-year expectations for exome and genome volume growth from at least 30% to at least 28%, and revenue growth for those services from at least 20% to at least 17%. GeneDx, LLC

Those changes illustrate why revenue cycle management is a clinical workflow issue as well as a billing function. The information needed for payment often originates in the ordering clinician’s office. If it is missing, the laboratory must obtain it, proceed with reimbursement uncertainty, or alter the testing pathway.

A process that is easy to order can generate strong demand while leaving substantial work unresolved downstream. Moving that work earlier makes the obstacle visible to clinicians and patients before testing proceeds.

What the denial analysis revealed

President Mark Gardner supplied one of the most consequential details of the November call. GeneDx’s analysis classified 62% of exome and genome denials as administrative or procedural. The remaining 38% involved disputes over clinical necessity or patients outside payer coverage policies.

The distinction identifies two different kinds of work. Administrative denials may respond to better documentation, authorization procedures, claim preparation, and follow-up. Coverage and medical-necessity problems may require new evidence, policy changes, or a different way of financing access.

The percentages describe the composition of denials. They do not mean that 62% of all tests were denied, or that correcting administrative processes will automatically recover 62% of lost revenue. Management’s assessment identifies the principal area of opportunity; the financial outcome still depends on execution.

For market access teams, the finding is particularly useful. Obtaining a favorable policy is only one part of commercialization. The laboratory must also demonstrate that a particular patient meets that policy’s requirements and present the claim in the form the payer will accept.

The financial results show progress with limits

The quarter’s results indicate improvement, although the year-over-year comparisons remain more difficult than the sequential ones.

Measure	Q2 2026	Q3 2026
Total revenue	$114 million	$123 million
Exome and genome revenue	$100 million	$110 million
Exome and genome tests resulted	30,785	32,134
Blended average reimbursement per test	$3,258	$3,436
Genome share of outpatient insurance volume	32%	26%
Outpatient genome claims paid	32%	36%
Adjusted gross margin	70%	72%
Adjusted net income	$0.4 million	$8.6 million

Sources: GeneDx quarterly disclosures and earnings calls; November 10 GenomeWeb reporting. www.fool.com

Blended reimbursement rose approximately 5% sequentially, helped by better genome collections and the shift toward exome. However, it remained below the approximately $3,800 reported in Q3 2025. Exome and genome volume increased 25% year over year, while revenue from those services increased 12%.

The product mix is important to interpreting the improvement. A higher blended reimbursement figure can reflect better collections, a larger share of better-reimbursed tests, or both. It should not automatically be read as an increase in payer prices.

Nor should the company’s losses be interpreted as evidence that every genome test loses money. GeneDx reported a Q3 GAAP net loss of $12.3 million while generating positive adjusted net income. Company profitability, product gross margins, and the economics of an individual claim are different measures.

Coverage must become operational before it becomes cash

The latest call also distinguished formal coverage from payment in practice. Management reported that 82% of outpatient genome volume was submitted to payers with positive coverage policies, up from 67% in Q2. Yet the genome collection rate was 36%; exome’s was approximately 32%.

These measures have different definitions and should not be subtracted to calculate a denial rate. Together, however, they show why a coverage announcement does not immediately resolve reimbursement.

Management attributed part of the genome collection improvement to plans following Carelon’s coverage policy. Medi-Cal implementation remained more difficult, with a larger benefit expected in 2027. Payers must translate coverage decisions into operating procedures, while laboratories must satisfy the patient-specific requirements. www.fool.com

GeneDx is investing in technology and artificial intelligence to make these processes easier to navigate. Such systems may help obtain and organize information, but they cannot by themselves create coverage for an excluded patient or establish clinical utility where a payer remains unconvinced.

The company’s appointment of Nicole Maassen as chief patient access officer, overseeing market access and government affairs, places these responsibilities prominently within its leadership structure.

PAMA adds a second reimbursement challenge

The collection problem is occurring alongside a prospective reduction in the reference prices used by many payer contracts.

The October 12 analysis of exome and genome pricing reported CMS’s preliminary PAMA median targets of $3,680.60 for exome, CPT 81415, and $3,018.72 for genome, CPT 81425. Genome’s ultimate target is approximately 18% below exome’s.

The reduction limits delay the crossover in actual Medicare payment:

Rate or projection	Exome 81415	Genome 81425
2026 CLFS rate	$4,780	$5,031
Preliminary capped 2027 payment	$4,063	$4,276
Projected 2028 payment	$3,680	$3,635
Preliminary ultimate median target	$3,680	$3,018

The 2028 figures are projections assuming unchanged preliminary medians and reduction limits, rather than separately announced CMS rates. The reference-price inversion begins in 2028 under those assumptions. www.discoveriesinhealthpolicy.com

The business implication reaches beyond direct Medicare claims. On the November call, CFO Kevin Feeley said approximately 45% of company revenue was exposed to PAMA through CLFS-linked contracts. Management anticipated a mid-single-digit headwind to total revenue in 2027, while expecting collection improvements to offset or exceed it. www.fool.com

Those are separate forces. Improving the percentage of tests paid can raise revenue even while the payment benchmark declines. Conversely, growing test volume does not guarantee revenue growth if collections weaken or the reference rate falls.

The next stage is a test of execution

GeneDx’s three earnings discussions describe an evolving response to the same economic challenge. May revealed the consequences of a faster genome transition. August established the repair strategy. November demonstrated an initial improvement in reimbursement and margins, while exposing the delays and ordering friction that the strategy can create.

The company retained its full-year revenue range of $475–490 million and raised its adjusted net income outlook to approximately $12 million, even as it reduced exome and genome growth expectations. That combination explains the tension facing investors: improving economics per test can coexist with a more cautious near-term growth outlook. GeneDx, LLC

The next evidence will come from sustained collections, the time required to move orders into testing, and the extent to which reimbursement improvements persist as genome utilization rises again. For the wider diagnostics sector, GeneDx provides a concrete example of how clinical adoption, coverage policy, laboratory operations, and payment must advance together for scientific capability to become a durable business.

##

Note:  CMS PAMA data tell us nothing about how much GeneDx gets paid; that's hidden.  But PAMA data does show that in 2025, large numbers of commercial claims for exome 81415 and genome 81425 were getting paid at around half the CLFS price.  Here.

---

Chris Klomp: #2 at HHS, Update on his Views and Priorities for CMS

HEADER.  Chris Klomp is about to become the official #2 at HHS - the Deputy Secretary.  How does Klomp think Medicare should pay for innovation? His recent WSJ interview and two Senate hearings offer revealing answers: reward outcomes, make clinical AI reduce costs, and align FDA and CMS evidence expectations earlier. But the details matter: RAPID excludes laboratory tests, and senators challenged AI-assisted prior authorization. Our deep-dive AI review examines and connects all three appearances to explain an increasingly influential HHS official’s priorities—and what they could mean for diagnostics, biotechnology, and reimbursement.

###
AI CORNER
###

In March, this blog featured several articles about Chris Klomp, a senior CMS policymaker who was being moved to a #2 position under RFK Jr at HHS.  A Senate committee recently moved his nomination for Deputy Secretary forward.

What's new for him?  WSJ just published a 25-minute interview between Klomp and Liz Whyte.  Find it here.   We cover that first.  Then, in a second section, we provide highlights from his Senate appearances in September.

Across all three appearances, Klomp returns to several consistent themes:

• Reward outcomes rather than activity. The hearings emphasize shared savings and appropriate preventive spending; the WSJ interview extends this argument to clinical AI and explicitly cites ACCESS.
• Accelerate innovation while preserving standards. September 15 highlights FDA–CMS evidence alignment; September 16 emphasizes inspections and development timelines; WSJ broadens this into faster trials and biotechnology competitiveness.
• Lower drug costs without weakening American biotechnology. He repeatedly presents these as simultaneous obligations.
• Modernize government technology. The hearings focus on aging Medicare systems; WSJ emphasizes interoperable data as essential infrastructure for clinical AI.
• Manage through trust, competence and accountability. These leadership principles recur almost verbatim.

The strongest continuity is his belief that better incentives and better execution can improve access while restraining costs. The WSJ interview develops that philosophy more expansively; the hearings expose the difficult implementation questions.

###

AI Guest Faculty: Lessons from 1870s for Growth of Genomics Today

 What can the rise of Germany’s synthetic-dye industry in the 1800s teach us about American genomics today?

Johann Peter Murmann’s fascinating historical research, followed by a 2016 computer simulation, offers an unexpected starting point.

My new white paper explores how scientific talent, capital markets, and Medicare reimbursement shape industrial winners and losers. One intriguing question: Would genomics have evolved differently if Medicare coverage had been fragmented among 20 independent regional payers?

A historical comparison—and a proposed AI-assisted experiment.

###

###

Executive Summary

Why do some genomics companies become industry leaders while others, sometimes equally innovative, struggle or disappear? The answer may lie as much in institutions and financing as in scientific discovery.

This white paper draws an unexpected historical comparison with the nineteenth-century German synthetic-dye industry. In his 2003 book, Johann Peter Murmann examined how universities, scientific talent, patent laws, and industrial competition helped Germany achieve global dominance. In 2016, Murmann and Thomas Brenner used computer simulations to test these historical explanations, finding that universities' responsiveness to industry's demand for chemists was particularly consequential.

The paper proposes applying a similar approach to the evolution of American genomics, with special attention to Medicare reimbursement and MolDX.

One largely overlooked institutional feature deserves scrutiny: a centralized laboratory can receive traditional Medicare reimbursement for testing beneficiaries nationwide through its own Medicare contractor. What if Medicare had instead been divided among dozens of independent regional payers, each making separate coverage decisions?

Such fragmentation might have profoundly altered the industry's financing, growth, and competitive landscape.

The paper outlines an AI-assisted research program combining company histories, reimbursement decisions, financial data, and computational simulations to investigate these possibilities.

The objective is not merely to explain which companies succeeded, but to understand how America's institutional environment helped shape their success.

###

DOWNLOAD THE 10p PDF WHITE PAPER... HERE.

Three Companies: Investor Calls for Tempus AI, Caris, and Natera, Quarter to Quarter

Tempus, Caris, and Natera see enormous opportunity in genomics—but reimbursement determines the pace. Our new AI-generated white paper compares November earnings calls with prior-quarter expectations, examining MolDX coverage, payer pricing, clinical evidence, and promises still awaiting payment.

Here's a summary:

• Tempus AI, Caris, and Natera report strong demand for genomic testing, but reimbursement determines how quickly new products become sustainable businesses. Comparing November earnings calls with the prior quarter reveals distinct strategies. 
• Tempus anticipates substantial oncology pricing gains while pacing MRD expansion around coverage.
• Caris expects higher liquid-biopsy reimbursement to help finance new monitoring and detection products.
• Natera is converting clinical evidence into broader coverage, including a new MolDX indication for liver cancer surveillance. Several earlier expectations remain unresolved, particularly payment for emerging applications and Japanese commercialization.
• [Steve Chapman steps down as CEO - here.] 
• Across the companies, the key distinction is between technical performance, clinical utility, coverage, and actual collections. Competitive advantage increasingly depends on generating persuasive evidence and financing the long interval before reliable payment.

Find the 15-page report in the cloud  HERE.

x

x

 #

Why a Strong Diagnostic Test May Add Little "New" Information

[Chat GPT 6]

My colleague Luis Cano recently posted a useful example of a question that diagnostic developers should ask much earlier: What does the new technology add to what we already know?

His post described a pancreatic neuroendocrine tumor preprint involving 1,438 patients and approximately 30,000 histology images. An AI model extracted prognostic information from routine H&E slides.  

Yet, as Cano reported, a model using established clinicopathologic information achieved a C-index of 0.875; adding the AI increased it to 0.883. [Sivasubramanian]

The AI could predict recurrence. The clinical information could already predict recurrence very well. Those findings can coexist.

My comment on his post drew on a lesson I remember from MBA statistics. Given an outcome and ten candidate variables, variables A, B, and C might predict the outcome extremely well. Then variables D, E, and F might do almost as well. That does not mean we have discovered two separate reservoirs of predictive power that we can simply add together. We may have found two ways of measuring much of the same thing.

For molecular diagnostics, digital pathology, and AI, this distinction can separate an impressive publication from a compelling case for clinical adoption.

The arithmetic works but the attribution changes

We expect addition to be commutative: 3 + 5  =  5 + 3. It is tempting to assume that combining predictors works the same way.

An important qualification makes the lesson precise. Ordinary correlation is symmetric: the correlation of X with Y equals the correlation of Y with X. And if we fit the same ordinary regression model to the same data with the same predictors, simply rearranging the columns does not change its final predictions.

• What depends on order, is the incremental credit we assign to each source of information.

“How well does this test predict recurrence?” and “How much does this test improve prediction after we use the pathology report?” are different questions.

The first measures performance in isolation. The second measures added value against an existing baseline. A test can perform very well on the first and make only a modest contribution on the second.

Here, “clinical first” means choosing the clinical model as the benchmark. 

The Genomics Consulting Book Guide to CLAUDE

If this looks familiar, I posted at Linked In  about November 1, but not here at the DIHP blog.

They Infest Linked In

Lately, every day Linked In has been offered me "McKinsey Guide to CLAUDE," "Consultant's Guide to CLAUDE."  Usually the exciting-sounding offers dead end at a registration page and sign in with your email.

Steve Jobs Got His Best Ideas Walking, Too

Walking home from lunch with some consultant colleagues, it occurred to me that my own Claude account could probably write me a guide to strategy consulting in ten minutes. It did. And honestly, it’s pretty good.

As I read, I kept thinking: “I can use THIS to help Client A today—and THAT with my deck for Client B tomorrow.” So I’m sharing it here. Free. No registration or email required.

What You Get (for free)

It’s a 44-page booklet in two parts.

The first 17 pages introduce 21 consulting skills—from framing a problem and auditing assumptions to pricing strategy, competitive war-gaming, and writing a decision memo. Each includes a reusable instruction you can paste into Claude, plus a worked example involving a fictional genomics company facing major strategic decisions.

Pages 18–44 start over and take that same company through a detailed consulting engagement. You see how the skills connect, how one finding changes the next decision, and how the analysis becomes a recommendation for the board..

It’s a little “MBA in a Box”—with a genomics startup as the classroom.

DOWNLOAD IT HERE.

----

In the case study, nearly every fact is labeled with its origin.  S, sourced from a "data room" (a list of key facts), D (derived from such facts, such as revenue - cost = profit [D]).  Or A, assumed, which you'll find why in that sentence or two. 

___

This source offers a booklet of 21 Claude Consulting Skills based on email registration and temporary access.  I'm dubious how great it is.  I bet you can do better.  Please use Fable t5.1 medium to simulate this booklet (see attached) and its 21 topics where Claude Fable can simulate real consulting teams and outputs.  The attachment is the cover page - simulate the booklet.

This was my initial prompt, with some screen shots of covers. It came back with a book, 16pp,  7500 words.  It offered to expand with a detailed business school type case study; I said yes. Claude then basically doubled its length which produced the 44 page work product.  Claude and I spent perhaps an hour with back-and-forth refinements.  

The case study, which Claude made up quickly while writing its deliverable.Meridian Genomics is a fictional 120-person CLIA lab in Phoenix with a tumor-informed MRD assay for colorectal cancer. It has a PLA code but no Medicare coverage, $14 million in revenue, a $6.2 million annual loss, and $22 million in cash. The board wants a three-year plan and has mostly decided what it should say: get MolDX coverage. Management’s plan is to submit in early 2027, expect coverage a year later at about the $1,900 rate the dominant national competitor reportedly receives, and triple commercial volume by 2029.

The first twist comes from arithmetic on the client’s own numbers. Meridian’s fully loaded cost per test is $2,100. Its commercial tests bring in $1,522, so each one loses $578. Coverage at $1,900 would still lose about $200 per test. The board’s strategy would turn a flat, money-losing channel into a growing, money-losing channel. The real business is biopharma. Trial sponsors using the assay as an endpoint account for 68% of revenue, grew 40% without any sales effort, and earn about $694 per test. The board had been treating that business as a bridge to coverage.

The second twist is in the profit pool. The money in MRD isn’t in the test fee. It sits in trial services, meaning documentation, logistics and regulatory support, and in licensing longitudinal data. Meridian already does trial services but has never charged for them as a product. The national competitor’s low-margin covered volume turns out to be a way of building the data asset.

The recommendation is to make trial services a named business line with per-program pricing, data-rights clauses and regulatory-grade documentation, and to cut cost per test toward $1,400. Meridian would prepare the coverage dossier but submit it only after passing gates the board sets: cost at or below $1,500, a full sponsor pipeline, and a realistic MolDX timeline. In the base case this is worth about +$19 million in NPV, compared with roughly −$2 million for cost-then-coverage. The downside is mild under this plan and could eat most of the balance sheet under the coverage-first plan.

The third twist comes from the war game. It breaks an earlier assumption that the competitor would take 12 to 18 months to respond. In the simulation it offers sponsors free assays in exchange for data rights within a year. Meridian can’t beat free on price, so it retreats to registrational trials, where documentation and exclusivity are worth more than a discount. That revision then flows back into the segmentation and the roadmap.

 

CMS Pulls Algorithmic Lab Tests Off the CLFS: The CY2027 SaMS Rules, Decoded

Wednesday, November 4, 2026

AI CORNER [New replacement article, Claude, issued Nov 9]

OPPS final rule:

https://www.federalregister.gov/documents/2026/11/06/2026-22739/medicare-program-hospital-outpatient-prospective-payment-and-ambulatory-surgical-center-payment 

PFS final rule:

https://www.federalregister.gov/documents/2026/11/09/2026-22743/medicare-and-medicaid-programs-cy-2027-payment-policies-under-the-physician-fee-schedule-and-other 

Read here or download as 12p PDF report.

_ _ _

A Big Change - Suddenly It's Here

The CY2027 OPPS and PFS final rules contain a policy that looks like housekeeping. It isn't. CMS has decided that a stand-alone algorithm run on lab data that already exists is not a clinical laboratory test at all. That covers a genome re-analysis, an AI read of a digitized H&E slide, and a classifier applied to an RNA transcriptome. Each is now an "other diagnostic test," a cousin of the AI-read CT scan, and starting January 1 it gets paid like one.

For 2027 the dollars barely move. Almost everything else does: who pays coinsurance, which date-of-service rule applies (none of the lab ones), how new codes get priced (nobody quite knows), and whether a code that includes staining a slide counts as "software" (it depends). Here's the tour.

The Short Version

CMS finalized 14 CPT codes (up from 10 proposed) that leave the Clinical Laboratory Fee Schedule. Under the OPPS they move to new technology APCs with a brand-new status indicator, O1 ("Software as a Medical Service"), which flags them as software and pays them separately. Under the PFS they get national pricing with procedure status A (active). Payment is crosswalked from 2026 CLFS rates and rounded into the fixed new technology payment bands, so the prices are familiar.

The surroundings are not. These services now carry ordinary Part B cost-sharing. They escape the laboratory 14-day date-of-service rule. And they sit inside payment systems built on budget neutrality and, perhaps eventually, packaging. CMS corrected its code list after comment (two codes out, two in, one long-deleted code dropped, two moved up a payment band), but it gave up nothing on principle. It has said where future algorithmic lab codes will be paid, but not how they will be priced when they arrive.

There is also a true philosophical oddity underneath the rulemaking for SaMS. Physical work does not disqualify a service from the SaMS category (it is almost welcomed), unless that physical work happens to be laboratory work. More on that below. 

Why Some Genomics Labs Reach Financial "Escape Velocity" - But Many Don't

Recently, Merck KGaA annnounced it was acquiring BioTechne for over $11 billion dollars.  BioTechne's management has been strong - that's up from a $5B valuation about ten years ago.  So the C-level team knew what it was doing.

But, if you look through a genomics lens, along the way the same BioTechne C-suite bought Exosome Dx for $325M (2018) and sold it  for $15M (2025).   Why was it so difficult to forecast the value of the Exosome business - by twenty-fold (5%)?

We discuss in a new white paper why the financials of the genomics industry can be so noisy, and often for reasons that will be missed by outsiders.

Our paper includs a humous appendix of real statements from genomic SEC documents.  Like, 'We may get paid rarely, if ever."

Find the white paper in the cloud HERE.

Puzzling PAMA Pricing #2: Labs & Losses Deep Dive

 You've heard about how PAMA affects different groups - big labs, small labs, physician labs, and the labs of different medical specialties.

Here, I used the assumption that PAMA pricing was pushed all the way - that tomorrow's prices were set at the actual medians found in PAMA data.   

My assumption was, we'd see a normal or Gaussian distribution around the average price cut.  But no.  For the whole population of providers (with >$50,000 Medicare lab income), the distribution of losses is very irregular.

I divided these into four groups:

1. Group 1, Winners (N=82+80)
2. Group 2, Average (N=1758)
3. Group 3, Losers (N=984)
4. Group 4, Big Losers (N=15+123)

Here's the view, click to enlarge:

###

AI CORNER

###

Four Populations, One Fee Schedule:
Reading the Fully-Phased PAMA Impact

Nov 3, 2026 · @Bruce Quinn

Introduction

Suppose PAMA's private-payer medians were applied to Medicare lab pricing in one stroke, with no 15%-a-year phase-in and no further congressional delays. The intuitive answer is "everyone takes about a 16% haircut," which is the aggregate figure CMS has reported. The intuitive answer is wrong. Line up every provider with at least $50,000 in 2026 Medicare lab revenue (4,855 NPIs, about $5.8 billion in baseline payments), sort them by fully-phased percent change, and you do not get a bell curve. You get two tall peaks, a small cluster hugging zero, and a tail of providers losing 40% or more.

The $920 million aggregate cut works out to 15.8%, almost exactly what CMS said. But that average describes almost nobody. The histogram is a distribution of code portfolios, and code portfolios sort by specialty with remarkable discipline. This post walks through four groups descriptively, then asks, as a microeconomist would, what is really going on in each.

One housekeeping note: the rows are NPIs, not consolidated companies. That turns out to be a feature. The same corporation can land in three groups under three NPIs, and that tells you something.

Orientation: The Shape of the Distribution

The histogram uses fifteen four-point bins from −46% to +6%, with edge bars catching anything beyond. Two spikes dominate. The first, at −18% to −14%, holds 1,758 providers, over a third of the population. The second, at −30% to −26%, holds 994, about a fifth. Between them sit three modest bins (310, 256, 724). To the right of the main peak the counts collapse to 71, 82, and 80 in the three bins nearest zero. To the left of the second peak the tail runs 290, 150, 123, and finally 15. Two providers actually gain.

We carved out four clusters for discussion. They are not exhaustive; roughly 1,800 providers fall in the connective tissue between them.

Group 1, Winners (N=162): −6% to +2%. Essentially untouched.

Group 2, Average (N=1,758): −18% to −14%. Close to the headline CMS number.

Group 3, Losers (N=994): −30% to −26%. Nearly double the headline.

Group 4, Big Losers (N=138): −46% to −38%. Since the statutory 15%-per-year phase-in compounds to only −38.6% after three years, this group is defined by cuts that take four or more years to fully unwind.

The dollars are as lopsided as the counts, in the opposite direction. Group 1 holds $1.1 billion in baseline, 19% of the total, but absorbs $6 million of the cut, under 1% of the pain. Group 2 holds $2.4 billion and absorbs $398 million, almost exactly proportional. Group 3 holds $720 million but absorbs $200 million, 22% of the pain on 12% of the dollars. Group 4 is a rounding error in aggregate, $21 million, though not to the practices inside it.

Group 1: Commercial and Medicare Already Agreed

The 162 providers nearest zero split cleanly by bin.

The 80 in the −2% to +2% bin are the proprietary molecular diagnostics industry. Several sit at exactly 0.00%.

The adjacent 82 at −6% to −2% have a different character: 34 clinical labs, but also 29 ophthalmology and optometry practices, which is every eye-care provider in the eligible population, nearly all at a modeled change of −2.00% to two decimal places. That is not twenty-nine negotiations converging. That is one code with a −2% adjustment and a specialty that bills little else on the CLFS besides 83861. 

This bin also holds academic reference labs (Mayo, Washington University, Johns Hopkins) and specialty molecular firms (Exagen, Biodesix, Interpace, Assurex) that are lightly exposed to a commoditized code or two, so they fall short of 0% change.

Two providers gain outright: Caris at +44% and Pacific Edge at +33%, both proprietary molecular tests. It's rare but when the private median exceeds the Medicare rate, PAMA raises the rate, and the increase carries no phase-in cap.

Group 2: Mainstream Medicine Plus the Duopoly

The main peak is the clinical mainstream. Internal medicine (474) and family practice (247) make up 41% of the group; hematology-oncology and medical oncology add 275; endocrinology, rheumatology, nephrology, and cardiology add 272. These are practices running routine chemistry, hematology, and endocrine panels.

The 354 clinical labs are where the money is. Labcorp's North Carolina NPI ($208 million, −17.7%) and New Jersey NPI ($146 million, −16.0%), Quest's New Jersey ($113 million, −16.5%) and Florida ($106 million, −16.9%) entities, Unilab, Sonora Quest, and a parade of other Labcorp and Quest state entities all land within a point or two of −16.5%. The national labs are, by construction, the "average" cut, and I argue below that this is a mechanism rather than a coincidence.

What is absent is as telling as what is present. Zero pain management or anesthesiology providers. Only 28 nurse practitioners and 19 urologists out of 1,758.

Group 3: A Specialty Roster - It's Not a Random Draw

The second peak looks nothing like the first. Nurse practitioners (197) and physician assistants (90) make up 29%. Urology contributes 131, most of the eligible urologists. Pain management, interventional pain, and anesthesiology contribute 192. Podiatry contributes 38 of the 41 podiatrists in the study.

The 241 clinical labs are a distinctive cohort too: HealthTrackRx ($49 million, −27.9%), CAP Diagnostics, Sierra Genomics, Genetic Technological Innovations, Patients Choice of Indiana, and a long list of Texas, California, Florida, and Arizona LLCs named Lone Star Tox Screen, RealTox, QualiTox, DrugScan, Equaltox, Dominion. Many bunch at −29% to −30%, the group's left edge. A pathologist recognizes the toxicology and PCR-panel sector without being told.

One corporate detail deserves a pause. A Labcorp NPI in North Carolina sits in Group 3 at −29.2% on $9.6 million, while the Labcorp flagship sits in Group 2 at −17.7% and a small Labcorp NPI in New York sits in Group 1 at −5%. Quest has an Oregon NPI in Group 1 at −6%. BioReference has a Texas NPI at 0.0% and its New Jersey flagship at −19.7%. Same company, same contracting department, same leverage, three groups. Whatever is sorting providers, it is not bargaining power.

Group 4: Small Dollars, One-Code Portfolios

The 138 providers losing 38% or more total only $21 million in baseline, but they lose roughly 40 cents on the dollar. Nurse practitioners (41), physician assistants (22), pain and interventional pain (32), anesthesiology (16), and PM&R (9) make up 87% of the group, mostly individual practitioners with $50,000 to $150,000 baselines in Texas, Oklahoma, Michigan, Florida, and New Jersey.

The exceptions are a dozen clinical labs, two of them surprising: Invitae ($753,000 under this NPI, −42.9%) and Baylor Miraca Genetics ($709,000, −39.9%) are germline genetics labs, not tox shops.  Potential suspects: 81412 Ashkenazy panel fell from $2448 to $1591, and 81443   pan ethnic carrier panel fell from the same $2448 to even lower new rate, $1461.  

The 1,800 in Between

The unclustered providers include the biggest single losers in dollars: Aegis ($33 million, −36%) and Millennium ($21 million, −35%), just outside the Big Loser cutoff. Myriad ($34 million, −25%), Silverpath, and Genview sit just right of the Loser peak. Prince Laboratories ($61 million, −9.7%) and Reliable Results ($33 million, −9%) sit between the Average peak and the Winners. This is where the large, diversified, multi-code operators live; their position is a weighted blend of the four archetypes.

What Is Going On: A Microeconomist's Reading

PAMA sets each code's price at the volume-weighted median of private-payer payments for that code. Specialty never enters the formula. The specialty sorting is therefore a second-order effect: specialties run characteristic code portfolios, and the codes face characteristically different commercial markets. The question is why those markets differ so much.

The Winners: When You Are the Only Seller, the Median Is Your List Price

For a sole-source proprietary test, the "private-payer median" is not the output of a market. It is the output of one lab's contracting book. If the lab simply does not below its Medicare rate, and that rate was itself set by crosswalk, gapfill, or ADLT pricing at a level the lab proposed, the median comes back at exactly the Medicare rate and the change is exactly 0.0%. The Medicare rate holds for 2026; the same rate except paid by Medicare Advantage commercial plans rather than the MAC, pegs the rate for 2027.

The eye-care cluster at −2.00% is the same story in a different key: a single dominant code, 83861, few suppliers, a commercial price that barely moved.  Ophthalmologists receive $7M in Medicare lab payments, of which $6.2M is 88361, microfluidic tear osmolarity analysis used to diagnose and evaluate dry eye disease.  It's about $22 and fell 2%. It has no CMS controls (no articles or LCDs). Two providers in California, both optometrists, reaped $273,113 and $256,540 with 8 tests per patient and about 1500 patients each.  (Small chart at bottom.  Subset data here $7M, all national providers at bottom, $10M).

The Average Group: The Duopoly's Contracts Become Everyone's Price

For routine chemistry, hematology, and endocrine codes, the volume-weighted median is dominated by Labcorp and Quest because they bill the overwhelming majority of commercial volume. Their discounted contract rates are the median, and the median becomes the Medicare rate. The national labs land at the average cut almost tautologically. In a sense they wrote it.

The 1,200 physician practices in Group 2 are price-takers in a negotiation they never attended. An internist with an in-office analyzer did not discount CPT 80053 to Aetna; Quest did. But the internist inherits Quest's discount through Medicare. This externality is the heart of PAMA's design. Diversified across dozens of commodity codes, these practices get the blended commodity result, about −16%.

The Losers: A Prisoner's Dilemma in Which Medicare Was the Prize

Toxicology, PCR infectious-disease panels, and the urine, wound, and nail-fungus molecular codes that dominate pain, urology, podiatry, and advanced-practice billing face a fragmented commercial market with hundreds of small labs competing for the same specimens. Payers have managed these codes aggressively for a decade through lab benefit managers, prior authorization, and narrow networks.

The structure is a classic prisoner's dilemma. Each lab, chasing network access or referral flow, discounts. Each discount is individually rational. Collectively, the discounting drags the median down 28% to 30%, and PAMA then imports that median into Medicare, which for many of these labs was the best-paying customer they had. The commercial price war destroyed the Medicare price. Nobody intended it; nobody could have prevented it alone.

A selection effect compounds this. PAMA counts only claims actually paid by private payers. An out-of-network tox lab that bills $800, gets denied, and writes it off contributes nothing. The in-network labs that accepted $120 contribute everything. The median is a median of those who capitulated.

The practitioners in Group 3 are riding the same codes. A pain practice billing definitive drug testing under its own NPI, or a urology group running PCR panels in-house, has a one- or two-code CLFS portfolio. Their result is simply the code's result.

The Big Losers: Concentration Risk

Group 4 is Group 3 without diversification. An NP with $80,000 in Medicare lab revenue has it in one or two codes, and if those were the codes where the commercial market collapsed most completely, the practice gets the full cliff.

The genetics labs are the puzzle. Invitae and Baylor Miraca at −40% or worse does not fit the tox story. Either these NPIs carry a narrow set of panel or cytogenetic codes that high-volume competitors undercut dramatically, or the corporate flagships sit elsewhere and these are satellites with unrepresentative portfolios. Either way, the lesson matches the pain practices: in a per-code median system, a concentrated portfolio is an undiversified bet on one commercial market.

The Timing Game

We do know:  Because cuts are capped at 15% per year, the difference between a −28% lab and a −43% lab does not appear in payment until 2029 or 2030. That is three to four years in which every actor in the losing groups has the incentive and the time to respond: migrate toward codes with smaller adjustments, restructure which NPI bills what, shift volume to the physician fee schedule where possible, exit the worst codes, consolidate, or lobby for another delay. Goodhart's Law [when a measure is treated like 'the target] applies in full. The fully-phased histogram projects what happens if everyone holds still, and in the losing groups nobody will.

Conclusion

The fully-phased PAMA distribution is not a bell curve with a mean of −16%. It is four economic regimes on one axis. The proprietary molecular sector controls its own median and sits at zero. Mainstream medicine inherits the national labs' discounts and sits at the headline average. The toxicology and PCR-panel sector, with the specialties that bill those codes, competed its commercial prices down and now watches Medicare follow, near −28%. A thin tail of one-code portfolios absorbs 40% or more.

The most useful fact in the data may be the corporate one: the same company occupies three groups under three NPIs. PAMA does not sort providers by size, leverage, or sophistication. It sorts them by code, and the codes sort by the structure of the commercial market each lives in. 

That is a very different policy from the one Congress thought it wrote in 2014, when the premise was that Medicare overpaid for lab tests across the board. What it built instead is a transmission line that imports each commercial sub-market's competitive structure, be that monopoly and price war alike, into the Medicare fee schedule, one code at a time.

[claude fable, using output #2]

###

83861.  $22, PAMA change -2%.  All providers 2024, 1715 rows (NPIs).  Services 491,298 for 181,113 patients, 2.7 per patient.  5 providers paid >$100,000 (up to   $  263,113).   The top 2 were optometrists in Maryville and Yuba City CA (7-8 tests per patient, for about 1500 patients.) 

The following charts PAMA raw data for price reports.  

Puzzling PAMA Pricing Data #1. Profiles of Six More Genetic Tests.

 In several blogs we've gone past the median price discounts in various lab services.  We've actually profiled the range of  reported PAMA raw price information - the shape of the marketplace.

We showed that some codes, which have a right-sided isolated peak of high prices, can have as much as half of profit coming from those sky-high outliers. Here.

Here, we do price profiles for six additional genetic tests.  These are:

• 81162, BRCA 1-2
• 81432, BRCA Panel
• 81241, Factor V (one of the most common genes)
• 81412, Ashkenazi Carrier Panel
• 81419, Epilepsy panel, a recent boom-code in FL, TX
• 81443, Pan-ethnic carrier panel

Profiles are based on actual observed PAMA medians, that is, what would be fully-phased in prices.

Figures, click to enlarge.

I don't know why the profiles vary so much, but it's like snowflakes, no two alike.

AMA CPT Now Releases Editorial Panel "Meeting Minutes"

Header: For AMA CPT Meetings, AMA will now released full minutes (100-pages-plus.)

###

The AMA has two major reimbursement-related committees: the CPT Editorial Panel and the Relative Value Update Committee- the RUC.

For years, detailed meeting minutes of the RUC have been online.  In contrast, AMA CPT posted only an abbreviated "Summary of Panel Actions" - e.g. code applications that were voted in, rejected, or postponed.

Now, AMA CPT posts both the "Panel Actions" and also detailed "Minutes" (over 100 pages) for CPT meetings.

###

The updated web page for editorial actions says, "To provide transparency in the decision-making process, meeting minutes and voting results, along with the summary documents, are included on this page for the last five years."  However, as of now, I only see meeting minutes for September 2026 (Minneapolis CPT).

###

Find the CPT page here:

https://www.ama-assn.org/about/cpt-editorial-panel/summary-cpt-editorial-panel-actions-meeting-minutes

Find the RUC page here:

https://www.ama-assn.org/about/rvs-update-committee-ruc/ruc-recommendations-minutes-voting

###

Very Brief Blog: Pennsyvania Passes Colorectal BIomarker Coverage State Law

 Here's a mini post to guide you to a press release at Guardant.  

Pennsylvania has passed a biomarker law, HB 1123, aligning CRC coverage with ACS aand NCCN guidelines.

https://investors.guardanthealth.com/press-releases/press-releases/2026/Guardant-Health-Applauds-New-Law-Ensuring-Access-to-Colorectal-Cancer-Screening-Innovations-with-No-Cost-Sharing-for-Pennsylvanians/default.aspx

Signed by governor October 22; see more at state Sen. Devlin Robinson.  

See the home page for HB 1123.  The house floor vote was 187 to 16. 

Look for a gold star banner which I think is the final version of PN4076.  Fiscal notes show no fiscal impact on the state (because it affects private insurance not state budget insurance.)

When Websites Go Down; McDermott Plus

In an adjacent blog, I asked Claude to find publication dates for PFS and OPPS rules.  It sourced the data from different sites.  Sometimes it used McDermott plus, where you might find snippets of archives but the main site went down when the entity closed around September 1.

I noted my own Discoveries in Health Policy has had 8M views and 3500 articles in over ten years.

Here's what Claude commented:

Confirmed: Law360/MLex reports that McDermott Will & Schulte is closing its healthcare lobbying and consulting shop after a dozen years. That's a real loss for this corner of the field. 

The McDermott+ rule summaries were the fastest, most reliable day-of digests of PFS and OPPS releases, and their OPPS/PFS data dashboard was the only free tool I know of that let you track APC and RVU payment rates across years without rebuilding it from the addenda yourself. Their "Healthcare Preview for the Week" was also one of the few places that tracked the regulatory calendar prospectively rather than just reporting what had already dropped.

The pattern is worth naming. Law-firm and consultancy thought capital lives on a corporate website whose existence depends on the business unit surviving a partnership vote, and the content has no archival home of its own. When the unit closes, the content isn't moved to a successor page; it's just unplugged. The same thing happened to Health Policy Alternatives' public material, to much of the Avalere grey literature, and to a fair bit of what the Advisory Board published before the Optum acquisition. The Wayback Machine catches some of it, but WordPress ?p= URLs and PDF attachments are exactly what it captures least reliably.

When Will CMS PRS and OPPS Final Rules Appear? Today? Next Week? CLAUDE knows.

 Annual "final rules" for Physician Fee Schedule and Outpatient Prospective Payment System (PFS and HOPPS or OPPS) come out around November 1, always after "market close."  November 1, so they are 60 days ahead of the implementation date of January 1, 2027.

I asked Claude for a look-back and here's what I got.  BTW the query took about a minute.  The rules are named for the upcoming calendar year. Today on October 30, 2026, we're awaiting the rule for CY27.

Short answer: November 1 and November 2 each appear about half the time.  Usually, both rules on same day.    

And: Claude, 1 minute, Chat GPT 4 minues.  All answers matched.

#### CLAUDE