I was looking through comments to the summer PFS and OPPS rules (blog here), and I noticed that PathAI had an interesting comment on SaaS valuation, a topic that CMS had requested advice about.
Find PathAI's comment online here: * I summarize PathAI's comments below, but this will be imperfect, so see their full-length comment.
https://www.regulations.gov/comment/CMS-2025-0304-13277
See also two exciting new applications of digital pathology, via a David Braxton essay, here.
For CAP's position on SaaS, see its letter to CMS, page 14, here.
PathAI Letter
PathAI opens by noting comments were solicited on both PFS and OPPS rules It focuses on PFS, but suggests these principles are consistent with OPPS policy.
CMS asked 6 questions, which I abbreviate. (1) What factors should Medicare consider for rates for SaaS? (2) What's the experience with risk-based payment for SaaS? (3) Have risk-based payments reflected the value of SaaS? (4) How to avoid limitations of RVU PE methods. (5) How should physician work be valued? (6) How do SaaS technologies impact chronic disease?
(1) Factors for valuing SaaS?
Consider the full service, licensing fees, capital equipment, ancillary staff, etc.
Consider new codes, which CMS can create (aka HCPCS Level II).
(a) New code for use of Assistive AI which "assists" pathologist in his own diagnosis.
(b) Code when AI provides information the pathologist cannot.
(c) Code when AI is a "companion diagnostic."
(d) Add on codes for each mono IHC and/or each multi IHC staining.
CMS should consider image acquisition costs, data storage, New staining codes should reflect physician management and use of AI (not yesterday's interpretation time). CMS should be proactive in creating new codes. CMS should learn proactively from experience when single gene and gene panel codes were introduced. "Adopt a single code for AI, regardless of the number of algorithms run on a single patient."
(2,3) Risk based models.
The Oncology Care Model OCM began with chemotherapy, and after diagnosis. This effectively excluded diagnostics from consideration. Avoid this mistake. Include 'high value testing" in models like future OCM.
(4) Avoiding RVU PE problems.
Continue to crosswalk some PFS code values based on OPPS valuation; in fact streamline this.
(5) Valuing Physician Work
PathAI recommends several valuations for physician worked, pegged to existing pathology professional component codes (-26 payments), using 88361 $40, 80503 $26, 88361 ($40) for code types a,b,c as shown earlier.
(6) Chronic disease?
SaaS in digital pathology provides more accurate diagnostics, such as in prostate cancer (ArteraAI, IBEX, etc).
_____
* There are 4 PathAI PDFs, which seem identical, two 9p, two 10p.
_________________________________________________________
##
AI CORNER: PathAI and Artera Letters
##
I asked Chat GPT 5 to summarize the Artera (here) and the PathAI letters separately, then compare. I post the result below "as is."
##
Review of the Artera Letter (CMS-2025-0304-12910)
Artera’s letter to CMS is polished and policy-oriented, emphasizing continuity and precedent in developing Medicare payment for Software as a Medical Device (SaMD) rather than reinventing new methodologies. The company situates itself as a leader in AI-driven digital pathology, highlighting its FDA de novo–authorized ArteraAI Prostate Test as a pioneering prognostic software tool. The central argument is that CMS should extend the existing OPPS and PFS frameworks used for software services such as HeartFlow’s FFR-CT and Cleerly’s plaque analysis to new AI pathology tools.
Artera organizes its submission around principles and precedents, not CMS’s numbered questions. Its tone is confident, with an emphasis on regulatory and economic realism. The letter urges CMS to:
-
Use SaMD terminology (implying higher regulatory rigor) instead of “SaaS.”
-
Establish five-year new technology payment eligibility and remove low-volume barriers.
-
Create separate APC categories for SaMD and support physician reimbursement for interpretation.
-
Recognize the capital and workflow costs of digitizing pathology.
It then provides detailed examples from prior CMS payment rules, presenting tables that crosswalk OPPS and PFS rates for software technologies. Artera’s logic is straightforward: once a SaMD achieves a new-technology APC, that payment rate should inform its PFS rate. The letter is technical but clearly written, targeting CMS economists and policymakers. It avoids aspirational rhetoric, framing its requests as extensions of established CMS logic rather than new departures.
Review of the PathAI Letter (CMS-2025-0304-13277)
PathAI’s submission is more discursive and structured explicitly around CMS’s six focus questions. It offers both policy recommendations and educational background, explaining why digital pathology differs from radiology in data size, infrastructure, and cost. PathAI argues that CMS must design new HCPCS code families to differentiate among types of AI services—assistive, prognostic, companion diagnostic, and multiplex IHC add-ons—and to anticipate future scaling challenges.
The letter is notably didactic: it explains the physical steps and hidden costs of digitizing slides, data storage, and computing, using these to justify inclusion of digitization costs in reimbursement. PathAI also draws analogies to genetic testing history, warning CMS that without proactive coding rules, AI billing could mirror the early “explosion” of gene-by-gene billing. It recommends that CMS establish panel-like codes now to prevent runaway unit billing later.
In addition to payment mechanics, PathAI discusses risk-based payment models, arguing that current oncology models neglect diagnostic contributions such as digital pathology. It proposes a voluntary demonstration model to reward high-value diagnostics and reduce hospital investment risk. Finally, it gives examples (Paige, Ibex, ArteraAI) to show that AI can guide prognosis and therapy, underscoring that reimbursement, not science, is the main bottleneck. The tone is academic yet advocacy-oriented, seeking to educate and persuade through analogy and evidence.
Comparison and Contrast
Both letters converge on the same policy goal—predictable, fair payment for AI-enabled pathology services within CMS frameworks—but they diverge sharply in approach and rhetorical stance. Artera speaks as a manufacturer with a single validated product seeking parity with earlier SaMD precedents. Its argument is concise, precedent-based, and aimed at CMS rule-writers. PathAI, by contrast, writes as a thought leader and ecosystem advocate, treating the issue as a category-wide structural reform for pathology SaaS.
Artera’s letter assumes CMS familiarity with OPPS logic and argues that consistency with HeartFlow-style payment suffices; PathAI assumes CMS uncertainty and builds a case from first principles, elaborating cost components and proposing an entire HCPCS taxonomy. Artera frames the issue as one of regulatory alignment (“use SaMD, crosswalk to APC 1511, pay ~$1450”), while PathAI frames it as systems design (“create new codes, capture digitization costs, prevent coding distortions”).
Stylistically, Artera is terse, corporate, and precedent-minded—like a legal brief. PathAI is expansive, analytic, and pedagogical—like a policy white paper. Both advocate inclusion of professional-component payment for physicians, but Artera’s version stresses parity and precedent, whereas PathAI details distinct levels of physician engagement with different AI modalities.
In summary, Artera offers CMS a template to adopt immediately, grounded in prior agency decisions; PathAI offers CMS a blueprint to govern the next decade of AI pathology reimbursement. Together, they illustrate two complementary poles of industry advocacy—incremental institutionalization versus proactive framework-building—in shaping how AI services will be valued under Medicare.
###
BONUS - Compare CAP letter with the two company letters.
####
The College of American Pathologists (CAP) takes a markedly different tone and role compared to both Artera and PathAI. While all three letters address CMS’s solicitation on payment policy for Software as a Service (SaaS) and artificial intelligence (AI) under the Physician Fee Schedule (PFS), their positions reflect fundamentally different institutional vantage points: CAP speaks as the governing professional body for pathologists, PathAI as an industry innovator, and Artera as a regulated commercial test developer.
CAP Letter: A Professional Governance Perspective
The CAP’s section on SaaS (Section 6 of its broader PFS comment letter) is brief, measured, and institutional. Its guiding principle is caution, not acceleration. CAP warns CMS not to rush into defining payment structures for AI but to “proceed with an abundance of caution rather than a sense of urgency.” This posture contrasts sharply with the advocacy tone of both Artera and PathAI, who press CMS to act swiftly in establishing equitable reimbursement pathways.
Rather than proposing specific payment rates or models, CAP’s central message is that CMS should anchor all SaaS policymaking within the AMA CPT Editorial Panel’s ongoing work, particularly through the Digital Medicine Payment Advisory Group (DMPAG) and the Digital Medicine Coding Committee (DMCC). CAP’s emphasis is procedural: CMS should not create ad hoc payment mechanisms outside established CPT governance channels. It repeatedly invokes the AMA’s Appendix S taxonomy (assistive, augmentative, autonomous AI) as the proper framework to guide valuation and classification.
In essence, CAP’s approach is regulatory harmonization and process discipline. It views the question of SaaS payment as primarily one of codification and definitional alignment with AMA and FDA initiatives. CAP does not advocate for specific OPPS crosswalks, rate levels, or technology categories, and it avoids commercial examples. Instead, it seeks predictability and professional legitimacy, ensuring that any AI payment policy continues to recognize the physician’s role, even in semi-autonomous or algorithmic workflows.
Contrast with Artera
Artera’s letter is a technocratic policy brief written from the standpoint of a company seeking operational reimbursement clarity for an existing product. It offers detailed payment precedents, citing specific CPT and APC codes (HeartFlow, Cleerly, Optellum) and recommends pegging SaMD payment to established OPPS rates (around $950–$1,450). Artera’s emphasis is continuity of precedent and speed of implementation: CMS “does not have to reinvent the wheel.”
In contrast, CAP’s emphasis is governance, not precedent. CAP tells CMS to work within the AMA’s deliberative committees; Artera tells CMS to act based on past OPPS practice. Where Artera is instrumental (how to price today), CAP is architectural (how to define tomorrow). Artera advocates for physician reimbursement for test interpretation, but its rationale is tied to cost realism and policy symmetry, not to professional autonomy or coding ethics. CAP, on the other hand, frames valuation through the taxonomy of physician work categories and the preservation of professional recognition within an evolving AI ecosystem.
In short: Artera is pragmatic and operational; CAP is institutional and normative.
Contrast with PathAI
PathAI’s letter is the most elaborate of the three and most focused on systemic innovation and code architecture. Like CAP, PathAI discusses the need for proper coding categories (including new HCPCS codes), but its motivation is opposite: to accelerate CMS’s adoption of AI payment structure. It describes the physical and computational infrastructure of digital pathology in depth, drawing analogies to the history of genetic testing, and urges CMS to proactively design payment codes to avoid future distortions.
Where CAP urges restraint, PathAI urges foresight. CAP warns CMS to defer to AMA governance; PathAI sees CMS as the primary policy engine to create and shape the payment framework. CAP positions itself within the AMA process, while PathAI positions itself within the innovation economy. Both recognize the importance of defining physician work, but CAP focuses on codifying its boundaries (assistive vs. autonomous), while PathAI focuses on expanding reimbursement recognition for the hybrid roles pathologists now play.
PathAI’s letter is technocratic but forward-looking, with examples of prostate and lung cancer AI. CAP’s letter is procedural and conservative, deliberately abstract. PathAI argues that lack of payment is the barrier; CAP argues that lack of definitional clarity is the barrier.
Synthesis
The triangular contrast among the three letters reveals three complementary but distinct philosophies shaping the future of AI reimbursement:
-
Artera: “We have a validated SaMD today—use existing CMS mechanisms to pay for it.”
-
PathAI: “AI pathology is transforming medicine—create a scalable coding and payment structure now.”
-
CAP: “AI payment policy must evolve through established professional and coding governance—move carefully.”
Artera speaks in the language of precedent and policy continuity, PathAI in the language of innovation and system design, and CAP in the language of professional oversight and procedural legitimacy.
Ultimately, CAP’s letter functions as a counterweight to the commercial urgency in Artera and PathAI’s submissions. It prioritizes consistency, codification, and deliberative process over rapid expansion, implicitly reminding CMS that AI reimbursement must remain anchored in physician-defined frameworks rather than in vendor-defined models or ad hoc payment precedents.
