In both the summer 2025 Physician Fee Schedule Proposed Rule (entry point here) and Outpatient Proposed Rule (entry point here), CMS requested public comment on how it should price Software as a Service.
Here's an AI writeup of the four-page OPPS Federal Register solicitation. [I'll add a link to Fed Reg pages after 7-17].
[Outpatient inspection copy, 213-216, PFS inspection copy, slightly different, 319-322]
Comments are due in 60 days (around September 15).
##
AI CORNER
##
OPPS SAAS INFORMATION REQUEST
CMS is actively soliciting public comment on how to establish payment policy for Software as a Service (SaaS) technologies under the Outpatient Prospective Payment System (OPPS). This request arises because of the growing use of SaaS tools, including AI-based software, in clinical decision-making and the lack of a clear, consistent payment methodology across settings.
Key Areas Where CMS Seeks Input:
-
Factors for Payment Rates
What should Medicare consider when setting SaaS payment rates? Should costs include acquisition, maintenance, development, or other factors? -
Appropriate APC Assignment
Should SaaS be paid through existing clinical APCs, new technology APCs, or new dedicated APCs for software? -
Cost Assessment Methods
How should CMS evaluate hospital acquisition costs for SaaS? What benchmarks or methodologies are appropriate? -
Data Sources for Rate-Setting
Given limited claims data, what alternative sources could CMS use to establish realistic payment levels? Examples might include hospital invoices, market research, or cost modeling. -
Discrepancy in Costs
Why are Medicare claims data often showing lower geometric mean costs for SaaS than manufacturers’ stated prices? Is this due to reporting practices, acquisition models, or other factors? -
Value and Efficiency
How should CMS factor in SaaS efficiencies (e.g., operational savings, faster diagnostics) in hospital outpatient and ASC settings? -
Quality and Efficacy
How can Medicare payment policies reflect the clinical value, quality, and efficacy of SaaS technologies?
Context:
Historically, CMS packaged SaaS into payments for the primary clinical service (e.g., MRI image processing). More recently, separate payments have been made under New Technology APCs or clinical APCs. However, CMS lacks a dedicated policy framework, and stakeholders argue this inconsistency can impede patient access and market adoption.
CMS highlights challenges like non-transparent R&D costs, lack of comparator services, and limited claims data as barriers to creating fair and predictable payment rates.
Parallel Rulemaking:
CMS is also soliciting similar comments on SaaS under the Physician Fee Schedule (PFS).
HOW IS THE PFS SOLICITATION DIFFERENT THAN THE OPPS ONE?
####
Key Differences Between OPPS and PFS SaaS
Comment Solicitations (CY 2026 Proposed Rules)
Outpatient Prospective Payment System (OPPS) Solicitation:
Under OPPS, CMS is primarily focused on how to pay hospitals and ASCs for SaaS technologies used in outpatient settings. CMS emphasizes challenges around:
-
Assigning SaaS to APCs (existing vs. new)
-
Assessing hospital acquisition costs (given lack of transparency and data)
-
Aligning payment with value and efficiency
-
Establishing methodologies that reflect true costs and value, given the lack of cost comparators and limited claims data
The OPPS questions center on the hospital setting, hospital purchasing behavior, and how to model SaaS costs fairly in the APC payment system.
Physician Fee Schedule (PFS) Solicitation:
Under the PFS, CMS focuses more on how SaaS costs fit into the existing Practice Expense (PE) methodology, which relies on outdated data not designed for software-heavy services. CMS raises concerns about:
-
How to treat software and AI licensing as direct vs. indirect costs
-
The fact that indirect cost allocations (from the AMA PPI survey) do not capture SaaS well
-
Whether to continue the practice of crosswalking from OPPS payment levels
-
How to capture physician work value related to interpreting SaaS outputs
-
How SaaS fits into value-based and risk-based care models, especially for chronic disease management
The PFS solicitation asks broader questions about future pricing strategies for SaaS, including potential alignment with OPPS data, and reflects a deeper concern with how SaaS will integrate into evolving models of physician practice and primary care delivery.
Summary of the Difference:
-
OPPS = Facility-based perspective: Focused on how hospitals purchase and use SaaS, how APCs should classify and price it, and concerns over acquisition costs and claims data.
-
PFS = Physician-based perspective: Focused on the practice expense methodology, the physician's role in interpreting SaaS, and how SaaS integrates into physician payment models, particularly risk-based care and chronic disease management.
If you'd like, I can prepare a comparative memo summarizing these differences for stakeholders or clients.
####
HUMOR
Here is a comment by comedian Lewis Black.)
MOCK COMMENT LETTER TO CMS:
If you'd like, I can help you draft a formal comment letter addressing these points.
YES, Provide a consolidated single response letter to both solicitations, from a hypothetical organization that provides software and AI in both physician office and hospital outpatient settings.
####
####
Thank you for clarifying. Below is a more detailed, narrative-style consolidated comment letter, written in the voice of a sophisticated industry stakeholder. It includes more context, nuance, and specific recommendations to reflect the complexity of the issues CMS has raised in both the OPPS and PFS proposed rules. This version moves beyond brief bullet points and provides substantive rationale for each point, while staying focused and professional.
[Your Organization Letterhead]
[Your Organization Name]
[Address]
[City, State ZIP]
[Date]
SUBMITTED ELECTRONICALLY VIA REGULATIONS.GOV
Centers for Medicare & Medicaid Services (CMS)
Department of Health and Human Services
Attn: CMS-XXXX-P (CY 2026 OPPS Proposed Rule)
Attn: CMS-XXXX-P (CY 2026 PFS Proposed Rule)
RE: Comments on Proposed Policies for Software as a Service (SaaS) and AI-Based Technologies in the CY 2026 OPPS and PFS Rules
Dear Administrator Dr. Oz:
[Organization Name] appreciates the opportunity to provide comments on the Centers for Medicare & Medicaid Services’ (CMS’s) solicitation regarding payment policies for Software as a Service (SaaS), including AI-enabled technologies, in the CY 2026 Hospital Outpatient Prospective Payment System (OPPS) and Physician Fee Schedule (PFS) proposed rules.
Our membership provides advanced clinical software and AI-based solutions widely adopted in both physician office and hospital outpatient settings, so we are uniquely positioned to comment on the challenges and opportunities in creating consistent, equitable, and forward-looking Medicare payment policies for these technologies.
We commend CMS for recognizing the policy gaps that have emerged as the healthcare industry shifts from equipment-centered models toward software-centric clinical innovations. However, we also want to emphasize the significant risks to innovation and patient access if CMS continues without a clear, coherent, and sustainable framework for these services across Medicare’s payment systems.
I. Comments on OPPS Solicitation
CMS’s questions under OPPS rightly recognize the structural problem: the current APC-based system was never designed to accommodate SaaS-based technologies whose cost structures and clinical impacts differ fundamentally from traditional devices or equipment. Without reform, CMS risks undermining access to services that hospitals increasingly rely on to deliver efficient, evidence-based care.
Challenges Under OPPS
Hospitals face genuine difficulty aligning acquisition and reimbursement for SaaS products. Unlike traditional devices, SaaS costs often appear as operational expenses—licensing, service fees, updates, cybersecurity protections—rather than capital assets tied to individual procedures. Yet CMS payment through OPPS remains deeply rooted in equipment-centric models. Hospitals are left uncertain about whether and how these costs will be recognized in Medicare payment.
Further, SaaS often delivers value that is distributed across multiple episodes of care or service lines, making direct attribution to a single APC problematic. For example, an AI-driven cardiac risk assessment platform may support imaging, outpatient consults, and chronic care management, but current CMS frameworks force hospitals to seek reimbursement through fragmented, ill-fitting APCs.
Additionally, CMS’s reliance on sparse claims data for setting payment rates creates circularity: innovative technologies, by definition, lack historical volume; low payment leads to underutilization, which in turn depresses future payment updates.
Recommended Solutions
We encourage CMS to:
Establish distinct, dedicated APCs (or an analogous mechanism) for SaaS technologies. These could reflect tiers of complexity and clinical impact rather than attempting to fit novel services into legacy APCs designed for devices or physical interventions. This would create predictability for hospitals and technology developers alike.
Consider a hybrid valuation model that blends acquisition cost benchmarking with assessments of clinical value and efficiency gains. Software’s impact often lies in reducing unnecessary services (e.g., avoidable imaging) or improving diagnostic accuracy. Payment should not be solely backward-looking but reflect prospective value to Medicare beneficiaries.
Provide clear guidance to hospitals on cost reporting for SaaS expenses. Absent this clarity, hospitals may inconsistently report these costs, further distorting claims data used in rate-setting. CMS could issue standardized methodologies for accounting for SaaS in cost reports and claims submissions.
Coordinate with PFS policy to ensure alignment. Hospitals and physicians often use the same SaaS platforms; inconsistent valuation undermines adoption and confuses purchasers.
II. Comments on PFS Solicitation
Under the Physician Fee Schedule, CMS faces related but distinct challenges. The Practice Expense (PE) methodology is outdated and structurally incapable of capturing SaaS technologies accurately. The PE allocation methodology assumes that expenses are either clinical labor, medical supplies, or equipment—categories which do not reflect the operational realities of modern software platforms.
Challenges Under PFS
CMS’s current practice of treating SaaS and AI expenses as indirect costs fails to recognize that for many services, software is not incidental—it is integral to the provision of care. For example, AI-based diagnostic algorithms may function as the primary tool through which physicians interpret patient data. Treating these costs as part of overhead ignores their direct connection to clinical activity.
Further, the AMA’s PPI survey data is insufficient for modeling SaaS costs at the granularity required for precise valuation. Crosswalking to OPPS (when available) is an imperfect but necessary interim step, though it risks perpetuating the structural mismatch described above.
Lastly, CMS’s questions about chronic disease management rightly acknowledge that SaaS is deeply entwined with emerging care models—virtual care, remote monitoring, team-based approaches—that do not fit legacy fee-for-service assumptions. Without modernized PE valuation, physician adoption of high-value tools will lag.
Recommended Solutions
Revise the PE methodology to treat SaaS as a direct cost when it is necessary for service delivery. This would reflect reality: clinicians increasingly rely on licensed software for tasks central to diagnosis and treatment.
Develop a forward-looking, transparent methodology for SaaS valuation within PFS. This could include:
Reference to OPPS payment rates where aligned.
Consideration of licensing models (e.g., per-use vs. enterprise agreements).
Recognition of recurring costs for updates and cybersecurity.
Incorporation of efficiency and quality metrics where measurable.
Acknowledge SaaS as foundational to chronic disease management models. CMS’s movement toward advanced primary care and risk-based models will increasingly rely on AI, predictive analytics, and data integration platforms. Payment policies must not inadvertently discourage adoption of these tools.
Engage industry stakeholders in structured dialogue to refine policies. We recommend convening a technical expert panel that includes SaaS developers, health economists, physician groups, and hospital representatives to help CMS develop durable, future-proof solutions.
III. Cross-System Considerations
We urge CMS to recognize the need for harmonization between OPPS and PFS policies to avoid perverse incentives and confusion. SaaS tools do not inherently change function when used in a hospital outpatient department versus a physician’s office; payment policy should reflect this reality.
Further, CMS’s concern about fiscal responsibility is valid, but the long-term value of SaaS lies precisely in its ability to prevent higher downstream costs through improved diagnostics, reduced unnecessary care, and better chronic disease management. Payment models must evolve to support these outcomes, not undermine them.
Conclusion
We applaud CMS’s willingness to confront these emerging challenges. However, inaction or piecemeal approaches will create fragmentation, stifle innovation, and limit beneficiary access to technologies that support more accurate, efficient, and equitable care.
We look forward to continued engagement with CMS and stand ready to provide further input as policies develop.
Sincerely,
[Name]
[Title]
[Organization Name]
In a Linked In Comment, Dr Lennerz recommend readers see two AMA resources. CMAA Clinically Meaningful Algorithmic Analysis. AMA Appendix S here. AMA Digital Medicine Workgroup here.
