We're in the run-up to Visions 2026, the annual digital pathology conference in San Diego this weekend. Find it here.
DPA has released 3 policy documents on AI and software-intensive reimbursement.
Find the new October Fask Force blog here.
Find the DPA CMS PFS comment letter here.
Find the DPA CMS OPPS comment letter here.
They're summarized individually then holistically below. (Chat GPT 6). For "Holistic," I stir into the mix their April 2026 blog about new initiatives, plans, and zeal re: reimbursement policy. DPA April.
###
1. October blog — DPA Reimbursement Task Force
30-word summary
The DPA warns that Medicare’s proposed reclassification of diagnostic algorithms threatens payment predictability and laboratory billing, with implications for investment, commercialization, and access extending well beyond the nominal reimbursement rate.
Extended summary
The Digital Pathology Association’s October 10 blog presents Medicare’s proposed 2027 payment changes as a consequential restructuring of reimbursement for diagnostic software. According to the post, CMS proposes removing ten software-only diagnostic algorithms, including several digital pathology laboratory-developed tests, from the Clinical Laboratory Fee Schedule (CLFS). Payment would instead move to Medicare contractor pricing under the Physician Fee Schedule or temporary hospital outpatient payment categories. CMS’s terminology, “Software as a Medical Service” (SaMS), provides the umbrella for this approach.
DPA emphasizes three risks: less transparent rate setting, geographic payment variation, and uncertainty about laboratories’ ability to bill directly. Its defense of the CLFS rests on the system’s established crosswalk and gapfill procedures, opportunities for public participation, and market-based payment updates. Contractor pricing, by comparison, could leave laboratories uncertain about both the amount and timing of reimbursement.
The association advances four positions: retain diagnostic algorithms on the laboratory fee schedule; postpone reclassification while federal agencies reconsider oversight of data-only facilities; establish national prices if services move; and preserve direct laboratory billing without arbitrary reductions for multiple distinct algorithms applied to one specimen.
From a business perspective, the central issue is the predictability of the revenue model. A developer evaluating a diagnostic product needs to estimate who will purchase it, who will submit the claim, what payment will follow, and how long collection will take. Uncertainty across those variables can weaken an investment case even before CMS reduces any published rate. Requiring hospital billing could also change the laboratory’s commercial relationship, making hospital contracting and purchasing decisions more influential in adoption.
The post is therefore best understood as an accessible statement of DPA’s advocacy priorities, rather than a detailed economic assessment. It does not quantify potential payment losses, establish that every affected product would fare worse, or demonstrate that CLFS prices necessarily measure clinical value. Its strongest strategic message is that payment classification can reshape the market for diagnostic innovation. For companies developing algorithms—and equipment suppliers whose growth depends on their adoption—the final policy’s operational details may matter as much as its headline reimbursement provisions.
2. Physician Fee Schedule letter — September 13, 2026
30-word summary
DPA’s physician payment letter defends laboratory classification for diagnostic algorithms, seeks national pricing and direct billing protections, and supports targeted coding reforms while resisting disruptive changes to established reimbursement infrastructure.
Extended summary
DPA’s Physician Fee Schedule letter challenges CMS’s proposal to reclassify algorithm-only laboratory services as “other diagnostic tests,” remove them from the CLFS, and establish payment through local Medicare contractors. Its foundational argument is that computational analysis remains laboratory testing when it depends on information derived from human specimens. Separating the analysis in time or location from the original laboratory procedure should not determine its payment classification. DPA also argues that reclassification is premature while CMS is considering CLIA oversight of data-only facilities.
The letter defends the CLFS as an established framework for pricing proprietary algorithms. It suggests improving transparency through confidential submission of algorithm characteristics, development expenditures, and other relevant information. DPA contrasts private-payer-based CLFS pricing with physician fee schedule resource accounting, which it considers poorly suited to research-intensive software. Economically, however, negotiated market prices and demonstrated clinical value are different concepts; the letter’s case is stronger as an argument for an appropriate pricing framework than as proof that existing prices represent value.
Its fallback requests are concrete. Laboratories should retain direct billing rights, with immunohistochemistry and in situ hybridization offered as precedents for laboratory billing under the PFS. If CMS instead requires enrollment as an Independent Diagnostic Testing Facility, DPA requests at least a year’s delay. It also seeks national pricing derived from the proposed New Technology APC bands, rather than fragmented contractor decisions.
The letter separately supports retaining AMA’s coding infrastructure while improving stakeholder representation, transparency, consistency, and participation in pathology coding deliberations. It rejects ICD-10-PCS as a suitable replacement in its current form. On duplicate testing, DPA distinguishes unnecessary repetition from clinically appropriate serial or complementary testing and urges CMS to address ordering decisions rather than impose the determination on performing laboratories.
The business theme is transition risk across an interconnected reimbursement system. Coding, provider enrollment, pricing, and claims submission must work together before a service can generate dependable revenue. DPA’s strategy combines institutional continuity with targeted reform: preserve functioning infrastructure, improve its weaknesses, and avoid forcing laboratories to rebuild billing operations during a short implementation window. For developers, these provisions affect launch timing, working capital, and the cost of maintaining Medicare access.
3. Hospital Outpatient Prospective Payment System letter
30-word summary
DPA’s outpatient letter challenges moving laboratory algorithms into hospital payment, highlighting billing eligibility, pricing methodology, multiple algorithm discounts, and service dates as determinants of commercial viability and patient access nationwide.
Extended summary
DPA’s OPPS letter examines CMS’s proposed transfer of algorithm-only laboratory services from the CLFS to separately payable New Technology Ambulatory Payment Classifications (APCs). It supports the concept of “Software as a Medical Service” terminology but cautions against treating laboratory-developed algorithms, FDA-authorized laboratory software, and imaging software as interchangeable. It also requests coordination with the ongoing CLIA review before payment policy settles questions about the status of computational laboratory services.
The preferred position is to retain algorithm-only LDTs on the CLFS. DPA argues that these analyses depend on human specimens and remain laboratory services despite using digital inputs. It favors established laboratory pricing processes over hospital cost-and-charge methodologies, which may poorly represent the economics of developing and validating proprietary algorithms.
The letter nevertheless offers a differentiated fallback. For laboratory software devices that other provider types may perform, DPA supports interim New Technology APC placement. Existing CLFS prices should inform APC assignments for previously priced services; hospital invoice prices should inform assignments for new services.
Direct billing is a central concern. Because OPPS pays hospital outpatient departments, independent laboratories could lose their ability to bill directly and instead have to contract with hospitals. From a business perspective, that would alter the sales channel and bargaining relationship: laboratories would depend more heavily on hospital procurement, contract terms, and payment practices. Preserving a nominal Medicare rate would not necessarily preserve the laboratory’s realized revenue or margin.
DPA also opposes automatic discounts for multiple algorithms applied to the same data, emphasizing independent development, validation, and clinical purposes. The economic distinction deserves care: separate development costs do not establish that shared operational efficiencies are impossible. A persuasive payment policy would distinguish demonstrable shared costs from the value and costs of distinct analyses.
Finally, DPA requests that, if reclassified, these services use the report-return date as their date of service, supporting billing after outpatient discharge. It also seeks faster, more predictable APC application decisions.
The letter thus addresses the mechanics of commercialization as much as reimbursement levels. Billing eligibility, service dates, discounts, and administrative timelines collectively determine whether a technically reimbursable product can sustain a workable business model.
#####
PULLING IT ALL TOGETHER
and adding also their April 2026 v"big viewpoint' blog.
####
Read together, these four documents reveal a field making progress in organizing its advocacy while confronting a harder question: what, exactly, is the economic product that digital pathology expects Medicare to purchase? April’s ambitions encompass digital pathology infrastructure, clinical implementation, and AI adoption. The later letters concentrate on protecting reimbursement for particular algorithmic diagnostic services. Those objectives overlap, but success on the second does not automatically finance the first.
The shift from April’s “Full Speed Ahead” to October’s defensive posture is revealing. April announces two institutional investments: a recommendation statement on clinical AI and an experienced Washington advisory partner. The subsequent letters demonstrate the value of the latter through detailed, operationally informed advocacy. Defending an existing payment pathway against disruption is meaningful work. Nevertheless, the documents establish a stronger case that DPA has developed an advocacy capability than that the industry has developed a comprehensive reimbursement strategy. These are different stages of progress.
The strongest arguments concern commercial continuity. The letters recognize that a payment rate is useful only when an eligible entity can submit a payable claim. Questions about laboratory enrollment, direct billing, service dates, and implementation periods therefore reach the foundations of the business. A company can survive an imperfect reimbursement methodology more readily than a period during which nobody knows how it can collect payment.
The business consequences extend beyond administrative expense. Moving a laboratory from direct Medicare billing into hospital contracting changes who controls customer access and negotiates the laboratory’s compensation. Even an unchanged Medicare payment could produce a lower laboratory margin, longer collection times, or slower adoption. Conversely, hospitals might favor services that fit their procurement and clinical workflows over otherwise attractive products requiring bespoke contracts. Payment classification can influence market structure without CMS explicitly choosing commercial winners.
DPA is particularly persuasive in exposing these transition risks. Its broader claim that computational services belong on the CLFS requires more careful handling. The letters themselves cite laboratory services already paid under the PFS, including immunohistochemistry. Consequently, showing that an algorithm is a laboratory service—or that its provider requires CLIA certification—does not by itself settle the appropriate payment system. Regulatory identity and payment classification are related questions, but they are not interchangeable. The stronger economic defense of CLFS placement concerns the nature of the diagnostic service, the suitability of its pricing mechanisms, and the costs of disrupting an established pathway.
April’s discussion of evidence also deserves a sharper distinction. A recommendation statement can improve validation practices, establish professional expectations, and make clinical implementation more credible. Those are valuable contributions. But a statement describing how AI should be validated is not itself evidence that a particular product changes management, improves outcomes, or merits a particular price. The April blog sometimes compresses that distance when describing the pending statement as the evidence base for future payment requests.
The missing bridge is product-specific execution. Who will fund the studies, select the clinical indications, recruit participating laboratories, establish meaningful comparators, and pursue coverage decisions? These four documents do not answer that question. That does not establish that such work is absent elsewhere; it means the published strategy should not be mistaken for a funded development program. An association can create standards and improve the policy environment. Individual sponsors still need to turn those advantages into reimbursable products.
The letters’ economic arguments likewise mix several distinct concepts: development expenditure, service delivery cost, negotiated market price, and clinical value. High development costs explain why a company needs substantial revenue across its installed base. They do not independently establish what Medicare should pay for each analysis. Private-payer prices provide market evidence, but they are not direct measurements of patient benefit. The argument would be more durable if these were presented as complementary considerations rather than interchangeable justifications.
This distinction becomes especially important in the opposition to multiple-algorithm discounts. DPA reasonably argues that independently developed tests answering different clinical questions should not be treated as redundant simply because they use the same image. But its categorical assertion that additional algorithms generate no marginal efficiencies is vulnerable. Shared image preparation, data transfer, integration, reporting, and commercial infrastructure could produce efficiencies even when each algorithm required separate validation. A stronger position would demand evidence of actual shared costs before imposing a discount, while protecting payment for distinct diagnostic contributions.
Looking ahead, the largest strategic tension may be between proprietary tests and integrated software platforms. The letters largely defend a model in which a discrete algorithm generates a discrete billable result. That model can support valuable diagnostics. Yet an increasingly capable platform could combine several predictions, quality checks, measurements, and reporting functions within one workflow. If each incremental function requires a separate reimbursement campaign, commercial integration could advance faster than payment policy can accommodate.
Companies therefore need to distinguish among three economic propositions: a diagnostic test purchased for its clinical information; a workflow tool purchased for productivity or capacity; and infrastructure purchased to enable a broader service. Some products combine them, but the purchasing rationale remains important. A tool that substantially reduces interpretation time may have a compelling business case through operating savings. A separately ordered prognostic test may depend much more heavily on coverage and a defined payment. Scanner adoption may depend on the combined economics of both. “Digital pathology needs reimbursement” is too broad to guide investment across these categories.
For equipment suppliers, that creates an additional question: does a policy victory expand the overall digital pathology market, or mainly protect revenue for a limited group of proprietary tests? Both can be worthwhile, but they offer different returns to the ecosystem. A scanner manufacturer’s interest in widespread adoption will not always coincide perfectly with an algorithm developer’s interest in preserving payment for each individual analysis.
The next stage should make April’s promise of measurable progress concrete: named clinical applications, accountable sponsors, evidence budgets, coding and coverage milestones, and explicit assumptions about who purchases and bills each service. The later letters provide capable protection against foreseeable policy disruption. The larger opportunity is to connect that advocacy to an executable commercial program—one that can explain which products become viable, for which customers, on what evidence, and over what period.
####
