Dear Administrator Oz:

Thank you for the opportunity to provide comments on the Centers for Medicare & Medicaid Services’ (CMS) proposed rule concerning the Calendar Year (CY) 2027 Medicare Physician Fee Schedule (PFS) and Other Changes to Part B Payment and Coverage Policies (CMS-1848-P). The proposals and requests for information contained in the rule raise several fundamental questions about how Medicare determines the price of physician services, how those valuations should change as clinical practice evolves, and how better information can improve the accuracy and efficiency of federal health programs. As the Director of Health Care Policy at the American Action Forum, I support policies that increase access to affordable, quality health care. Policies that promote such access are largely underpinned by free market principles, competition-based economics, and innovation.

The PFS attempts to translate an enormous and continually changing range of physician services into administratively determined payments. That challenge has become more difficult as sites of care, practice ownership, technology, and clinical workflows have evolved while portions of the payment methodology continue to rely on data and assumptions developed many years ago. CMS appropriately uses this proposed rule to reconsider several fundamental components of that system, including practice expense methodology, the process for coding and valuing physician services, and the accuracy of global surgical payments, and also seeking ways to improve the information available to Medicare through greater interoperability and more reliable identification of 340B claims.

I broadly support the objective underlying these efforts and encourage CMS to move forward with several of the policies as proposed. Medicare payments should more closely reflect the resources required to furnish care and should adjust when those resources or patterns of care change. Where CMS can obtain better empirical evidence, improve transparency, or correct information deficiencies, it should do so before relying on broad administrative adjustments intended to approximate efficiencies. Several elements of the proposed rule move policy in precisely that direction. Others would benefit from applying the same evidentiary standard. The recommendations below reflect that distinction.

Modernizing Practice Expense Methodology and Site-of-Service Payment

CMS proposes to continue a multi-year effort to move PFS practice expense (PE) methodology away from its reliance on specialty-level data that, in some cases, date to 2007 or earlier. The agency proposes phasing out the methodological step that effectively anchors aggregate specialty PE relative value units to those historical estimates and replacing it with a stabilizer intended to mitigate short-term volatility. CMS also seeks comment on whether the existing facility/non-facility site-of-service differential remains appropriate, particularly for physicians employed by hospitals, health systems, or other entities.

The underlying goal is appropriate. The resources necessary to operate a physician practice have changed considerably since many of the inputs underlying the existing methodology were collected. Labor costs, technology, administrative functions, practice ownership, and the settings in which physicians furnish services have all evolved. Continuing to preserve historical relationships among specialties solely because those relationships are embedded in the existing methodology can gradually move reimbursement farther from actual resource use.

CMS should proceed toward a methodology that relies to the greatest extent practicable on current, objective, routinely updated, and auditable cost information. Survey-based estimates can remain useful where better evidence is unavailable, but they should not be afforded presumptive weight when more reliable information is available. A payment system responsible for allocating substantial federal resources should be able to update its assumptions as underlying costs and clinical workflows change.

The same principle should guide CMS’ reconsideration of the facility/non-facility distinction. Medicare should reimburse resources where those resources are expended. An independent physician practice that bears the cost of office space, equipment, staff, information technology, and other overhead faces a different cost structure from a physician whose hospital or health system furnishes some of those resources and receives payment for them elsewhere. The payment methodology should recognize that distinction rather than automatically attributing equivalent indirect expenses to both arrangements.

Payment rules should also avoid tilting the competitive field toward one organizational model. When clinically comparable services can be furnished safely in physician offices, ambulatory facilities, or hospital outpatient departments, Medicare should not create an artificial competitive advantage merely because one provider operates within a higher-paid institutional setting. Payment differences should be justified by differences in the resources necessary to furnish care, not by ownership status alone.

This is particularly important given the continued consolidation of physician practices into larger hospital and health-system organizations. Medicare payment policy should not reinforce consolidation by making otherwise comparable services more financially attractive when delivered through one organizational structure rather than another.

CMS should therefore continue its transition toward contemporary PE data and reconsider assumptions that may duplicate costs already reimbursed through facility payments. At the same time, significant methodology changes should be phased in sufficiently to avoid abrupt disruptions, particularly for independent practices that genuinely incur the costs the PFS is intended to recognize.

Reconsidering the CPT and RUC Framework

CMS’ request for information regarding Current Procedural Terminology (CPT) and the American Medical Association/Specialty Society Relative Value Scale Update Committee (RUC) raises broader questions about the infrastructure through which physician services are defined and valued. The current framework has substantial benefits – a common coding vocabulary facilitates billing and administration across public and private payers, while the RUC provides CMS with organized clinical input on the time and intensity associated with physician services. Those benefits, however, do not mean the existing process should be the exclusive or presumptive source of information used to establish Medicare payment.

There is substantial evidence supporting CMS’ concern about survey-based valuation. A Government Accountability Office (GAO) review found that RUC surveys used for payment year 2015 had a median of just 52 responses and a median response rate of 2.2 percent; 23 of 231 surveys had fewer than 30 respondents. GAO concluded that low response rates, small samples, broad variation in responses, and potential financial conflicts could undermine the accuracy of recommendations based on those surveys.

More recent evidence offers useful context rather than a reason to disregard those concerns. A September 2026 JAMA Network Open study examined 2,173 RUC recommendations covering 2004 through 2025. CMS accepted the RUC valuation without adjustment in 66.1 percent of cases. When CMS did alter a RUC valuation, 62.2 percent of those changes were downward and 37.8 percent were upward. These findings demonstrate that CMS exercises meaningful independent judgment, but they also underscore how central RUC recommendations remain to the initial valuation framework.

The appropriate response is neither to preserve the existing process unchanged nor to replace one centralized valuation apparatus with another. CPT provides substantial value as a common language, and wholesale replacement would impose significant transition costs on clinicians, payers, health systems, vendors, and other participants that have built administrative systems around it. Similarly, clinical expertise is indispensable when assessing physician work. Claims data alone cannot measure the intensity or complexity of every service.

CMS should instead broaden the sources of information against which CPT/RUC recommendations are evaluated. Where reliable empirical information exists—including observed service times, claims patterns, electronic health record data, actual acquisition or labor costs, and other independently verifiable sources—it should supplement and, where appropriate, supersede estimates derived from small surveys. The agency should also improve transparency regarding the evidence supporting its acceptance or rejection of recommended values.

This would introduce greater discipline into an inherently administrative pricing system. Medicare should not depend on a single private body, a single survey methodology, or a single government estimate when competing sources of evidence can better inform valuation. Greater contestability among data inputs can help expose outdated assumptions and reduce the likelihood that incumbent payment relationships persist simply because they have historically been embedded in the fee schedule. Such an approach would preserve the useful standardization of the existing system while making the valuation process more empirically grounded and less dependent on any one source.

Revalue Global Surgical Packages Using Observed Postoperative Care

CMS is also considering how to improve the accuracy of payments for procedures with 10- and 90-day global surgical periods. Under these packages, Medicare payment incorporates an assumed amount of postoperative care whether or not each assumed visit ultimately occurs. CMS proposes pausing the current Medicare Access and CHIP Reauthorization Act-required postoperative data collection while it evaluates the information gathered to date and weighs alternative data sources and future revaluation strategies. CMS states that several years of collected information indicate postoperative visits included in global payments frequently are not occurring.

Here, there is meaningful evidence that the existing payment assumptions warrant reconsideration. The Department of Health and Human Services Office of Inspector General (OIG) recently examined CMS’ postoperative visit data and found that the number of visits provided was generally lower than the number incorporated into global surgical valuations. In its sample, the global fee did not reflect the number of postoperative visits provided in 91 of 105 cases. OIG estimated that aligning payment with actual utilization within its sampling frame would have reduced Medicare spending by approximately $5.7 million and beneficiary cost sharing by approximately $1.7 million.

At the same time, OIG found inaccuracies in CMS’ underlying postoperative reporting data in 45 of the 105 cases reviewed. That finding does not support maintaining payment assumptions that appear inconsistent with clinical practice. It does, however, caution against treating the existing data collection as sufficiently precise to support indiscriminate reductions across all global packages.

CMS should move toward evidence-based revaluation of global surgical packages. If a procedure is priced on the assumption that a certain amount of postoperative physician work will routinely occur, but sufficiently reliable evidence demonstrates that substantially less care is typically furnished, continuing to pay for the historical assumption would itself perpetuate a payment inaccuracy.

This is an appropriate application of market discipline within the PFS. Administrative payment should not preserve compensation for resources that are no longer routinely used simply because they were historically incorporated into the valuation. At the same time, CMS should not assume that every procedure has experienced the same change in resource use. Revaluation should therefore reflect differences among procedures and global periods. Utilization patterns may vary materially between 10- and 90-day packages and among individual surgical services. CMS should use the most granular reliable evidence available rather than apply a single arithmetic reduction across global services.

The agency should also improve its methodology for measuring postoperative utilization. Targeted reporting, claims and encounter information, representative audits, electronic clinical information where appropriate, or some combination of these sources may ultimately produce more reliable information at lower administrative cost than the existing reporting requirement.

CMS Should Not Finalize the Proposed 50-percent Reduction for Same-Day E/M and Global Services

CMS proposes a markedly different approach when a separately identifiable office or outpatient evaluation and management (E/M) service is furnished on the same day as a procedure with a 0-, 10-, or 90-day global period. Under the proposal, the highest-valued affected service would receive full payment, while other qualifying E/M or procedural services furnished on the same day would receive 50 percent of the otherwise applicable payment. CMS argues that efficiencies likely exist when these services are furnished together and that the current methodology therefore may duplicate payment. The existence of some overlapping resources is plausible. It does not establish that 50 percent of the lower-valued service is duplicative.

This proposal provides an important contrast with the global-surgery revaluation discussed above. In the global-surgery context, CMS has collected several years of utilization information, independent researchers and OIG have examined the data, and there is observable evidence that assumed postoperative services often are not furnished. Even with that body of evidence, CMS is appropriately seeking additional information about how best to revalue the affected services.

For same-day E/M services, CMS proposes a 50-percent payment reduction without a comparable empirical demonstration that half of the resources incorporated into the affected service are duplicative. CMS considered a similar policy in the CY 2019 PFS rulemaking but ultimately did not finalize it. The renewed proposal should require more than the reasonable intuition that efficiencies sometimes occur when multiple services are furnished during a single encounter.

Separately billable E/M services furnished on the same day as a procedure must already meet requirements demonstrating that the E/M service is significant and separately identifiable. If CMS believes particular resources included in the valuation of the procedure nevertheless overlap with resources incorporated into that E/M service, it should identify the overlapping components and quantify their value.

Administrative price reductions should follow evidence of reduced resource use, not substitute for it. Setting payment at 50 percent because CMS believes efficiencies are likely risks replacing measurement with centrally determined approximation. That is particularly difficult to justify when the proposed rule elsewhere emphasizes the need to replace historical or subjective assumptions with better empirical evidence.

A broad same-day reduction could also create unintended incentives. If Medicare reimburses two clinically appropriate services differently depending primarily on whether they occur on the same or separate calendar days, practices may have a financial incentive to separate care across encounters. Such behavior would inconvenience beneficiaries and potentially add utilization without producing a corresponding clinical benefit.

CMS should not finalize the 50-percent reduction as proposed. Where the agency identifies credible evidence of resource duplication, it should use the misvalued-code and rate-setting processes to determine which resources overlap and make targeted adjustments supported by that analysis. If further study demonstrates systematic efficiencies across specific families of services, CMS could subsequently propose an appropriately calibrated multiple-procedure policy.

Address Duplicate Testing by Correcting the Information Problem First

CMS’ request for information regarding duplicate laboratory tests, imaging, result sharing, and interoperability identifies a different form of inefficiency. The agency notes that diagnostic results frequently remain siloed within the electronic health record system in which they were generated, leaving clinicians elsewhere unable to determine that a test has already occurred or to retrieve its results. CMS correctly identifies several potential consequences: duplicative testing, increased Medicare expenditures, delays in care, and unnecessary radiation exposure.

The nature of the problem should inform the policy response. If a physician orders a repeat test because information from the original test is inaccessible, the immediate failure is not principally a payment failure; it is an information failure. Denying or reducing payment for the subsequent service does not make the earlier result accessible to the treating clinician.

CMS should therefore prioritize making existing laboratory and imaging results reliably discoverable and usable across care settings. Better information can allow clinicians themselves to avoid unnecessary utilization without requiring Medicare to prescribe the appropriate number of tests through increasingly rigid payment controls.

This also has competitive implications. Closed or poorly interoperable information systems can favor large vertically integrated organizations whose clinicians share a common electronic infrastructure, while independent clinicians or competing health systems may face greater difficulty obtaining prior results. Effective interoperability can reduce that structural advantage by allowing information to follow a patient rather than remain tied to an organization. It requires more than technical compliance with interoperability standards. Information exchange must function within ordinary clinical workflows so that physicians can reasonably determine whether relevant testing has already occurred and retrieve results in time to inform treatment.

Payment controls may eventually have a role. Once clinicians can reliably access prior information, CMS would have a stronger basis for identifying circumstances in which repetition is unlikely to be medically necessary and could explore narrowly targeted claims edits, frequency limits, or other mechanisms.

Any such policy would require substantial clinical exceptions. Repeat testing may be appropriate when a patient’s condition has changed; when serial testing is necessary to monitor disease progression or treatment; when prior imaging or laboratory information is inadequate; when sufficient time has elapsed that the previous result no longer answers the clinical question; or when urgent circumstances make obtaining an outside result impracticable.

CMS should also be cautious about policies that place the financial consequences of fragmented health information disproportionately on individual clinicians or laboratories that do not control the underlying exchange infrastructure. Improve access to existing information; allow better information to discipline unnecessary utilization; measure what duplication remains; and only then design payment policies targeted to the residual problem. Interoperability should be treated as part of the solution rather than using payment policy to compensate for its absence.

Finalize Claims-level Reporting for the Medicare Part D 340B Repository While Coordinating Federal Data Standards

CMS proposes requiring Medicare providers and suppliers participating in the 340B Drug Pricing Program to submit specified information for covered Part D 340B claims to the Medicare Part D Claims Data 340B Repository beginning with claims with dates of service on or after January 1, 2027, opening the repository for voluntary submissions beginning October 1, 2026.

CMS should finalize the proposed claims-level reporting requirement. The need for more precise transactional data has grown alongside the size and complexity of 340B. The Health Resources and Services Administration (HRSA) reports that the program contained 15,249 covered entities and 49,214 associated sites as of April 2026, while purchases at discounted 340B prices reached approximately $100 billion in 2025. The proliferation of participating sites, contract-pharmacy arrangements, and interactions with other federal drug-payment programs increases the importance of being able to determine accurately which individual transactions involve 340B drugs.

The immediate purpose of the CMS repository is appropriately narrower. Federal law requires CMS to exclude 340B units from Part D inflation-rebate calculations, and CMS currently relies on a claims-based methodology to identify those units. CMS has indicated that repository data initially will be used analytically to determine whether covered-entity submissions can more accurately identify 340B units.

Reliable claims-level identification is preferable to having CMS infer 340B status where more precise information can be obtained from the participants directly involved in the transaction. Better transactional information improves accountability while reducing the need for broad assumptions about which claims may or may not involve 340B drugs. Data elements including the covered entity’s 340B identifier and claim-specific information such as the date of service, prescription reference number, fill number, National Drug Code, and dispensing pharmacy National Provider Identifier are all routinely collected, and burden should be appropriately minimal for covered entities.

The repository should also improve the informational foundation on which participants interact. Disputes become more difficult to resolve when manufacturers, covered entities, pharmacies, and federal agencies operate from different or incomplete representations of the same transaction. Standardized claims information can reduce that uncertainty without predetermining the outcome of broader policy disagreements over the appropriate design of 340B.

CMS should, however, consider the repository in the context of parallel changes occurring across 340B administration. HRSA’s revised 340B Rebate Model Pilot Program, scheduled to begin January 1, 2027 for approved manufacturer plans, likewise relies on transaction-level information to administer rebates and address concerns such as duplicate discounts and diversion. HRSA has emphasized that the considerable growth and complexity of 340B have created oversight challenges that were less pronounced when the program was smaller.

CMS and HRSA should therefore coordinate their data standards from the outset. This does not require CMS to repurpose information collected for the Part D Inflation Rebate Program for unrelated regulatory purposes. Federal agencies, however, could avoid requiring covered entities to produce several incompatible versions of substantially similar claims information.

Where legally and operationally feasible, CMS and HRSA should use common definitions, compatible submission formats, consistent identifiers, and overlapping data fields. The agencies should also permit appropriately structured third-party submission arrangements while retaining clear responsibility for the accuracy of information provided on behalf of covered entities.

A claims repository will not resolve every policy dispute surrounding 340B, nor should CMS attempt to make it do so through this rulemaking. It can, however, establish something increasingly essential to sound program administration: a more reliable factual record of when 340B transactions actually occur. Greater transparency and more precise information can improve federal rebate calculations while reducing reliance on estimates, administrative assumptions, and preventable disputes among market participants.

Conclusion

The CY 2027 PFS proposed rule contains several promising efforts to improve the accuracy of Medicare payment and program administration. Moving practice-expense methodology toward contemporary cost information, introducing additional empirical evidence into physician-service valuation, reassessing global surgical packages using observed utilization, improving the exchange of existing diagnostic information, and establishing claims-level 340B reporting all move toward decisions based on better information rather than inherited assumptions.

Those reforms can also improve market discipline within a system that necessarily relies on administered prices. Medicare payment policy should avoid favoring particular ownership structures, preserve meaningful competition among sites of care, and respond when observed resource use no longer matches historical payment assumptions. Greater transparency and more contestable sources of valuation information can further reduce the risk that outdated arrangements persist simply because they are embedded in the existing system.

That same standard should apply when CMS seeks to reduce payment. The evidence indicating that global surgical packages incorporate postoperative services that frequently are not furnished provides a legitimate basis for careful revaluation. The proposed 50-percent reduction for separately identifiable E/M services does not yet rest on comparable evidence. Likewise, payment denials are an incomplete response to duplicate testing when clinicians remain unable to access results that Medicare has already paid to produce.

Medicare will necessarily remain an administered-price system, but CMS can limit the distortions that administration creates. Payment should be neutral among competing organizational models and responsive to empirical changes in the resources necessary to furnish care. Better data, transparent valuation, and competition among sources of information offer a more durable path to payment accuracy than blunt reductions or rules that inadvertently entrench incumbent delivery models.

Thank you for the opportunity to provide input on the Calendar Year 2027 Medicare Physician Fee Schedule and Other Changes to Part B Payment and Coverage Policies (CMS-1848-P). I appreciate the agency’s consideration of these comments and would be happy to work with you and agency staff on policies and reforms that advance competition, affordability, and value in Medicare.