AHA Comments on CMS FY 2027 Inpatient Rehab Hospital Proposed Payment Rule

June 1, 2026

The Honorable Mehmet Oz, M.D.
Administrator
Centers for Medicare & Medicaid Services
7500 Security Boulevard
Baltimore, MD 21244-1850

Re: Medicare Program; Inpatient Rehabilitation Facility Prospective Payment System for Federal Fiscal Year 2027 and Updates to the IRF Quality Reporting Program; 91 Fed. Reg. 17,195 (April 6, 2026).

Dear Administrator Oz:

On behalf of our nearly 5,000 member hospitals, health systems and other healthcare organizations, including approximately 900 inpatient rehabilitation facilities (IRFs); our clinician partners — more than 270,000 affiliated physicians, 2 million nurses and other caregivers — and the 43,000 healthcare leaders who belong to our professional membership groups, the American Hospital Association (AHA) appreciates the opportunity to comment on the Centers for Medicare & Medicaid Services’ (CMS’) fiscal year (FY) 2027 IRF prospective payment system (PPS) proposed rule.

IRFs play a critical role in the continuum of care for Medicare beneficiaries. They care for patients with complex conditions who are undergoing an intensive course of rehabilitation while being monitored and treated by a rehabilitation physician. This provides the optimal opportunity for maximum functional recovery following a serious injury or illness. The treatment provided by IRFs pays dividends down the road, allowing patients to return to their previous lives to the maximum extent possible.

As such, the AHA urges CMS to strengthen the FY 2027 IRF payment update by revisiting the market basket forecast and working with Congress to reduce the magnitude of the productivity adjustment. Current market basket increases, especially when reduced by the productivity adjustment, do not reflect hospitals’ rising labor, drug, supply and administrative costs. The AHA also urges CMS to revise its proposed coverage and documentation changes that would impose rigid timing and paperwork requirements for therapies, preadmission screenings and interdisciplinary team meetings. In addition, the AHA has serious concerns about the agency’s Request for Information (RFI) on broad IRF PPS payment reforms, which it modeled after skilled-nursing facility (SNF) payment policies. Overall, the AHA urges CMS to preserve clinical flexibility and pursue more targeted, IRF-specific policy refinements.

Our detailed comments follow.

MARKET BASKET UPDATE

The AHA remains concerned that CMS’ proposed annual market basket update of 3.2% is not keeping pace with real-world cost growth. In recent years, CMS’ market basket forecasts have consistently come in below broader inflation, let alone medical inflation, which has exceeded growth in the overall economy. Layered on top of that, the productivity adjustment, proposed to be 0.8 percentage points for FY 2027, further erodes the update, leaving Medicare payments increasingly out of sync with the cost of care. We urge CMS to revisit its market basket forecast and to work with Congress to reduce the magnitude of the productivity adjustment, as well as consider their combined effect on reimbursements for hospitals.

Rising Costs of Care Continue to Strain Hospitals and Health Systems 

Hospitals, including IRFs, continue to face sustained inflationary pressures. As detailed in our comments on the FY 2026 IRF proposed rule, inflation has continuously pushed up labor, drug, supply and other core operating costs. A recent AHA report found that total hospital expenses increased by 7.5% in 2025 alone.¹ Much of this increase reflects labor costs, which CMS notes account for nearly three-quarters of the IRF market basket, and an AHA analysis found that workforce costs rose by 5.6% in 2025.² Further, advertised salaries for registered nurses have averaged 5.5% growth over the last two years — more than double the rate of inflation.³ These dynamics have an especially pronounced effect on IRFs given the labor-intensive rehabilitation, nursing and close medical supervision required for IRF patients. Finally, the AHA has expressed concern that recent actions, such as changes to federal student loan limits that exclude nurses and other clinicians from enhanced borrowing limits, will exacerbate workforce shortages, which contribute to higher costs for labor.

Cost pressures, however, extend well beyond labor. Like other hospitals, IRFs are increasingly caring for sicker and more complex patients, requiring additional and more costly drugs and supplies, and these costs also continue to climb. An AHA analysis showed that in 2025, supply costs rose 9.9%, while drug costs rose a staggering 13.6%.⁴ In addition, a report from the Department of Health and Human Services (HHS) report found that list prices for nearly 2,000 drugs increased by an average of 15.2% from 2017 through 2023 — outpacing general inflation.⁵ These cost challenges strain IRFs, which, like other hospitals, must be prepared to provide treatment for a wide range of conditions and comorbidities while patients undergo their intensive course of rehabilitation.

Hospitals also are absorbing escalating administrative costs that are not reflected in payment updates. In particular, most Medicare Advantage (MA) plans require prior authorization for IRF admissions, and IRFs must devote substantial time and resources to navigating these processes. The HHS Office of Inspector General found that many post-acute care prior authorization requests were denied inappropriately, requiring IRFs and other hospitals to expend significant resources appealing erroneous denials.⁶ In addition, a 2024 Premier study found that hospitals spend just under $20 billion annually appealing denials.⁷ Indeed, the AHA estimated that about 6.5% of total hospital employment was for administrative staff dedicated to functions relating to claim denials and prior authorization.⁸ Since plans do not reimburse these administrative expenses, IRFs must absorb them while caring for a growing share of MA patients.

Viewed collectively, these cost increases for staffing, drugs and other essential supplies and services are placing significant strain across the healthcare continuum. They also are forcing hospitals, including IRFs, to redirect resources that otherwise could be used to support patient care, adopt new technologies and make other efficiency-enhancing investments. The reality that IRF market basket increases are at most half, and sometimes less than half, of the cost increases observed by hospitals adds to the serious strain. In addition, and as discussed further below, these same pressures also amplify the negative impact of the productivity adjustment by limiting hospitals’ ability to fund the very investments that can drive operational efficiencies.

The Productivity Adjustment Exacerbates Insufficient Market Basket Updates

Under the Affordable Care Act, the IRF PPS payment update is reduced each year by a productivity factor equal to the 10-year moving average of changes in the annual economy-wide, private nonfarm business total factor productivity (TFP). The private nonfarm business TFP is intended to reflect gains from new technologies, economies of scale, business acumen, managerial skill and changes in production. As such, it effectively assumes that the hospital field can achieve productivity gains comparable to those realized by private nonfarm businesses. However, as discussed in more detail below and in our comments on last year’s proposed rule,⁹ hospitals and the broader healthcare field cannot mirror these gains. As a result, it is not an appropriate or reliable proxy for hospital productivity. Therefore, we ask CMS to work with Congress to reduce the magnitude of the productivity adjustment.

A core problem is that the productivity construct embedded in the private nonfarm business TFP is a poor fit for measuring hospital productivity. TFP outputs are measured based on the total quantity and prices of goods and services produced in private nonfarm businesses. In industries that sell tangible products, outputs can often be measured in relatively straightforward and standardized ways. Hospital outputs, however, do not operate in the same manner. For example, hospital “quantity,” such as volume of visits or procedures, is not necessarily an appropriate proxy for output; it may instead reflect the underlying disease burden in a community. More hospital volume — i.e., more quantity — does not equate to higher productivity in the way it can for private nonfarm businesses.

Further, hospitals often cannot adjust prices per unit of service in response to changes in demand or quality in the way private nonfarm businesses can. Much of hospitals’ reimbursement is paid through fixed payment systems, such as the IRF PPS, which limits providers’ ability to alter prices. Similar constraints apply in the commercial market: Hospitals and health systems do not unilaterally set their rates, and prices for commercially insured patients are established through negotiations that frequently lock in rates for multiple years. Accordingly, applying a TFP output framework based on quantity and prices — as experienced in the private sector — to hospitals is problematic because that output function does not translate to the hospital field.

In addition, hospitals also differ from many private nonfarm industries because hospital services are inherently labor-intensive. As discussed further in the report referenced above, economic literature has long recognized that sustained productivity gains are difficult to achieve in labor-intensive service industries because labor cannot be scaled or automated in the same way as in other sectors. In this respect, hospitals are more comparable to fields such as education and social assistance, which tend to experience lower total factor productivity rates. For example, Bureau of Labor Statistics data show rates ranging from -0.4 for educational services to -0.1 for social assistance, compared with 1.9 to 4.9 for industries such as mining, oil and gas, information, and professional services.

CMS itself has acknowledged that hospitals are not positioned to achieve productivity gains comparable to the broader economy over the long run. Specifically, CMS found that hospitals can achieve productivity gains equal to only one-third of those seen in the private nonfarm business sector.¹⁰ Accordingly, using the private nonfarm business TFP to adjust the market basket is inappropriate.

Finally, we continue to find it especially troubling that the productivity adjustment appears to be applied only when it reduces Medicare payments. For example, in FY 2021, the 10-year moving average growth of the productivity factor forecasted was ‑0.1%. CMS acknowledged that subtracting a negative growth factor from the hospital market basket would have increased it by 0.1 percentage point. However, the agency set the productivity factor at 0, stating that it is required to reduce — not increase — the hospital market basket by changes in economy-wide productivity.¹¹ Put simply, the agency uses the productivity factor only when it lowers Medicare spending.

The cumulative, compounding effect of these annual reductions — coupled with the asymmetric treatment of periods of declining economy-wide productivity — has widened the gap between payments and the cost of providing services, leaving hospitals increasingly underfunded. In light of this, the AHA continues to have serious concerns about the proposed productivity cut, particularly given the extraordinary pressures under which hospitals and health systems continue to operate.

PROPOSED REVISIONS TO IRF COVERAGE REQUIREMENTS

CMS is proposing to revise multiple IRF coverage and documentation requirements, including changes it says are intended to promote earlier initiation of therapy, additional preadmission screening documentation and a more defined timeframe for the initial Interdisciplinary Team (IDT) meeting. The AHA shares CMS’ goal of the provision of high-quality, patient-centered rehabilitation and agrees that care should be well coordinated across disciplines. However, we are concerned that CMS has not identified a clear policy problem that its proposals are seeking to address. IRFs already operate under detailed Medicare coverage criteria that require, among other things, a comprehensive preadmission screening, defined intervals of physician visits, a minimum threshold of multiple therapies, weekly team meetings and an individualized, interdisciplinary plan of care. Yet, the agency proposes to add additional, time-based prescriptive mandates and new documentation elements for which the goal is not defined.  In addition, in practice, these types of requirements often reduce the ability of rehabilitation physicians and therapists to tailor care to a patient’s clinical status and evolving needs. In addition, adding such an administrative burden to clinicians’ workloads contributes to burnout. Therefore, we urge CMS to pursue policies that support timely and coordinated care while preserving reasonable clinical flexibility and to avoid creating coverage conditions that could lead to inadvertent noncompliance for reasons unrelated to quality or patient outcomes.

Initiation Of Therapies Within 36 Hours from Midnight on the Day of Admission. CMS is proposing to revise 42 C.F.R. § 412.622(a)(3)(ii) to clarify that all therapy treatments and/or therapy evaluations must begin no later than 36 hours after midnight on the day of IRF admission, and that claims would not be considered reasonable and necessary if this criterion is not met. However, we are concerned that an inflexible requirement that all therapies begin within 36 hours would create access and compliance issues for clinically appropriate care.

In many cases, the need for a specific therapy, such as speech-language pathology, may not become clear until after the patient is assessed in the IRF setting. Or, for example, orthotics or prosthetics therapy may not be deemed appropriate until the patient has completed initial physical or occupational therapy or a custom device has been fabricated. Despite this, the proposed regulatory text states that “[a]ll required therapy treatments and/or therapy evaluations ordered must begin no later than 36 hours from midnight the day of admission to the IRF.” While it’s possible that the inclusion of the word “ordered” would allow for therapies not yet ordered to be initiated beyond the 36-hour timeframe, this is far from unambiguous.

To avoid the unintended consequence of precluding IRFs from being able to order and provide therapies for which the need only arises later in a patient’s stay, we urge CMS to revise its proposed regulation text. Specifically, we urge the agency to explicitly state in the text that the 36-hour requirement applies only to those therapies that are ordered upon admission, and that subsequently ordered therapies may begin after that timeframe when clinically appropriate. Doing so would preserve the intent of timely rehabilitation while ensuring the policy aligns with physician ordering requirements and real-world clinical practice in IRFs.

In addition to the above clarification, we urge CMS to revise the regulatory text in a way that preserves the clinical flexibility of rehabilitation physicians to utilize their training and experience to order additional therapies at any time during the stay and not only at predefined points in time, such as during interdisciplinary team meetings, for example. Rehabilitation physicians often round on patients daily, conducting ongoing assessments and using their judgment to adjust treatment plans. Consistent with Medicare rules on physician-directed treatment in hospitals, rehabilitation physicians must also be afforded the ability to order subsequent treatments for their patients at any time.

Documentation of the Patient’s Current Functional Status in the Preadmission Screening. CMS is proposing to revise 42 C.F.R. § 412.622(a)(4)(i)(B) to require that the IRF add patients’ current functional status to the preadmission screening. This would be in addition to the current requirement to document patients’ level of function prior to the event or condition that led to the IRF stay.

While we understand CMS’ interest in ensuring that IRFs have sufficient information to develop an appropriate plan of care, the proposed requirement raises significant concerns unless CMS provides clear parameters. Because the preadmission screening is completed prior to the IRF admission decision — and often while the patient is in an acute care hospital — IRFs may need to rely on information from the prior facility’s clinical record (for example, recent nursing or therapy documentation) to add the required information to the screening. In other words, they would not be able to get this information from an IRF-performed functional assessment. Without clear guidance on acceptable sources and level of detail, the requirement could create inconsistent expectations, increase administrative burden and raise the risk that routine clinical variability between a preadmission description based on the acute care hospital records and the patient’s function at IRF admission could be cited during audit activities.

Accordingly, we urge CMS to clarify that “current functional status” may be documented using information from the acute care hospital record (and other clinically reliable sources available before admission), rather than requiring an IRF to conduct a new, detailed functional assessment prior to transfer. In addition, we urge CMS to issue robust education and subregulatory guidance on what it expects to see in the preadmission screening, including clarification that a general description of functional status is sufficient and that minor discrepancies between preadmission documentation and admission findings are expected given clinical change and the timing of assessment. This approach would support care coordination and appropriate planning without imposing new documentation burdens that detract from patient care.

Timing of the Initial and Subsequent Interdisciplinary Team Meetings. CMS is proposing to revise 42 C.F.R. § 412.622(a)(5)(ii) to require that the initial IDT meeting occur on or before the fourth day after midnight on the date of admission, and to clarify that subsequent IDT meetings occur at least once per week after the date of the prior IDT meeting. The AHA and our members support strong interdisciplinary coordination. However, we urge CMS not to finalize this policy change. We are concerned that creating a less flexible coverage condition would increase the burden without commensurate benefit. Interdisciplinary communication in IRFs occurs through multiple mechanisms, not only the IDT meeting. For example, it also occurs through real-time clinician-to-clinician discussions, review of evaluation and treatment notes, and ongoing updates to the plan of care. In addition, the existing requirements already require the development of an individualized overall plan of care by day 4, with input from the IDT. All of these mechanisms facilitate interdisciplinary care. However, CMS’ proposal only focuses on and, as such, overemphasizes the importance of the weekly IDT meeting.

Further, CMS has not demonstrated that establishing a new, accelerated timing requirement would improve patient outcomes beyond what IRFs already achieve through existing interdisciplinary processes. Before adopting a more prescriptive requirement, CMS should first consider whether less burdensome approaches could achieve its objectives without creating arbitrary rules. 

If CMS nonetheless finalizes this proposal, we urge the agency to address more clearly and consistently describe the timing standard. The preamble, proposed regulatory text and accompanying diagram do not appear to use the same counting method to define the “fourth day.” For example, the preamble gives an example of a Tuesday admission labeled as “Day Zero,” and states that Friday would be the deadline for the initial team meeting. However, the diagram seems to indicate that Saturday would be the latest permissible date. Before moving forward, CMS should provide clearer and more consistent examples, including when day 0 is initiated and examples of the accompanying deadline for the conference. We also ask it to confirm whether the same counting convention would apply to related IRF requirements such as the plan of care. It may also be helpful for CMS to delay any new timing standard for at least one year so providers, Medicare contractors and the agency can align on clear operational guidance before compliance and audit expectations attach. In addition, should CMS move forward, we urge it to permit reasonable flexibility where a patient’s documented circumstances make a formal meeting by day 4 impracticable or inadvisable, so long as interdisciplinary input to the plan of care is documented and ongoing coordination occurs. This would ensure that the policy supports coordinated care without creating avoidable denials or administrative burden.

RFI: POTENTIAL FUTURE IRF PPS PAYMENT REFORM

In this RFI, CMS seeks feedback on potential future payment reforms that would replace or revise key elements of the current IRF PPS patient classification system. Specifically, CMS asks whether it should adopt SNF patient-driven payment model (PDPM)-based clinical categories to classify IRF patients for payment purposes. It also asks whether it should use a revised comorbidity scoring and binning methodology in place of the current tier structure.

The AHA urges CMS not to move forward with a wholesale replacement of the current IRF clinical classification methodology based on the SNF PDPM framework. We are concerned that CMS has not identified a clear policy problem that would justify such a significant restructuring of the IRF PPS. The RFI repeatedly references “modernization” and closer alignment with other post-acute care settings, but modernization is not an end unto itself. If CMS were to pursue this type of fundamental restructuring, it should first explain why the current IRF-specific structure cannot be improved through narrower refinements and provide stronger IRF-specific evidence that any proposed alternative would better predict costs without weakening the connection between payment and patients’ rehabilitation needs. Until CMS can make that showing, the agency should retain the current IRF-focused approach rather than pursuing alignment for alignment’s sake. As written, the proposal would replace an IRF-specific system with one built for a different provider type, patient population and payment structure. This risks numerous unintended consequences, including a payment system that is less, not more, suited for accurate reimbursement based upon patient characteristics and resource use.

IRFs are very different from SNFs. They are licensed hospitals that furnish intensive rehabilitation services under hospital conditions of participation, physician oversight requirements and coverage rules that differ substantially from those applicable to SNFs.

In examining how to more closely align post-acute payment systems, the Medicare Payment Advisory Commission has emphasized that, although there is overlap across post-acute care settings, such an approach must account for major differences in payment structure, regulatory requirements, cost structure, and, of course, the actual services provided to Medicare beneficiaries.¹² CMS itself acknowledges in the rule’s technical memorandum that IRFs and SNFs differ in therapy intensity, admission pathways and payment design. Yet the agency still does not explain why a per-diem payment model developed for SNFs and their distinct patient population and treatment needs should serve as the foundation for a per-discharge hospital payment model for a different patient population with much different treatment needs.

Indeed, CMS appears to begin with the premise that the SNF PDPM structure is inherently superior. But the fact that CMS is separately seeking comment in the SNF rule on PDPM case-mix creep underscores that the SNF model itself remains subject to important questions. CMS should not import those vulnerabilities into the IRF PPS without a far more robust evidentiary basis.

In addition, the agency does not seem to have explored whether more targeted refinements to the IRF PPS would achieve its goals. For example, in the FY 2020 IRF PPS final rule, CMS undertook a targeted modernization of the IRF PPS. Specifically, CMS finalized a significant update to the patient assessment and grouping methodology used in IRF payment by removing the Functional Independence Measure instrument from the IRF Patient Assessment Instrument (PAI) and shifting the case-mix system to use standardized functional status items (Section GG) collected on the IRF-PAI.¹³ In doing so, CMS modernized core elements of the IRF PPS without attempting to import a payment system designed for an entirely different setting.

Finally, the AHA will continue to evaluate technical issues related to the potential reform concepts described in the RFI. Given the complexity of the potential changes to primary diagnosis classifications and comorbidity adjustments, CMS should proceed carefully and engage further with stakeholders before advancing any reform concepts through proposed rulemaking.

Potential Changes to IRF Patient Clinical Classifications. The AHA is concerned that the patient classification changes under consideration could weaken payment accuracy. As discussed, the approach CMS describes appears to rely heavily on SNF-based categories and prior acute care diagnoses that may not accurately capture IRF patients’ rehabilitation needs, functional impairments or expected resource use. It is unclear how SNF-based clinical categories are better suited to IRF patients than the current IRF-specific categories, and the agency does not provide evidence to suggest they would be.

The AHA is particularly concerned by the apparent reliance on the principal diagnosis from a prior inpatient stay as the starting point for designing IRF clinical categories. For many IRF patients, the diagnosis that drove the acute care hospitalization is not the same as the reason the patient requires intensive rehabilitation. A beneficiary may be hospitalized for an acute cardiac event, major surgery, infection or other precipitating condition, but the need for IRF care may arise from the resulting functional loss, neurologic impairment, debility, orthopedic complication or other rehabilitation need that emerges during or after the acute stay. In those circumstances, the prior inpatient principal diagnosis describes only why the patient entered the acute care hospital, not why the patient requires intensive rehabilitation services following the acute hospital stay.

In addition, from a coding perspective, this approach is fundamentally misaligned with established ICD-10-CM coding conventions. Under the ICD-10-CM Official Guidelines for Coding and Reporting (Section II), the principal diagnosis is defined as the condition chiefly responsible for occasioning the admission to that particular episode of care. By definition, this concept is encounter-specific and does not carry forward across care settings.

Our concerns about the development of the clinical categories are amplified by the category structure itself. The technical memorandum shows that the original SNF PDPM clinical categories did not fit the IRF population cleanly and had to be split and reworked because certain categories were too broad for IRF patients.¹⁴ They required substantial retrofitting to avoid clinically incoherent groupings. This is telling and indicates that the SNF-derived framework does not naturally reflect the case mix of the IRF setting. It also is a warning sign that the proposed direction is misguided. IRF patients should not be forced into categories designed for a different setting merely because those categories already exist.

The AHA also is concerned about CMS’ stated interest in moving toward more “diagnosis-driven” grouping methods. The Social Security Act requires IRF case-mix groups to account for impairment, age, comorbidities and functional capability.¹⁵ Diagnosis is therefore only one part of the payment picture, not a substitute for the broader statutory framework that requires incorporating functional capability, as the current payment system does. To the extent CMS is considering a future structure that elevates diagnosis to the exclusion of impairment or functional status, that approach would neither be consistent with the governing statute nor reflect the reality that functional capability is central to rehabilitation resource needs. Whatever refinements CMS considers in the future, the agency must preserve a prominent role for both function and impairment in IRF payment classification.

Potential Changes to IRF PPS Comorbidities. CMS also seeks feedback on whether to replace the current IRF PPS comorbidity tier structure with a new additive scoring methodology that would assign points for selected comorbidities and group patients into score bins for payment purposes. CMS indicates that this approach is informed by the SNF PDPM non-therapy ancillary (NTA) methodology but modified for the IRF setting. However, we urge CMS not to adopt a new comorbidity scoring methodology unless and until the agency provides a more robust justification, additional methodological detail and stronger evidence that the proposed approach would substantially improve IRF payment accuracy.

The AHA is concerned that CMS has not yet shown that the proposed comorbidity scoring approach is sufficiently transparent, stable or IRF-specific to justify replacing the current tier methodology. The technical memorandum itself acknowledges that CMS began from the SNF PDPM NTA framework but concluded that it did not adequately capture the clinical complexity of IRF patients. Thus, it required the creation of a new IRF-specific list and several methodological departures from the SNF approach. This is another warning sign that the proposed direction is misguided. That is, the fact that a SNF-based methodology must be substantially reworked before it can be applied to IRFs suggests that it may not be an appropriate starting point for IRF payment changes.

We also are concerned that the memorandum indicates that condition selection relied in part on clinician review and iterative judgment calls regarding overlap, relevance and coding reliability. While clinical input is important, CMS should not move forward without providing stakeholders far greater transparency regarding how conditions were retained, excluded, redefined and weighted.

Finally, we are not persuaded that the modest improvement in model fit described in the technical memorandum, standing alone, is sufficient to support a major change in comorbidity payment policy. Even if an additive score may better capture the cumulative effect of multiple conditions in some cases, CMS has not yet demonstrated that the proposed score bins would improve payment accuracy in a way that is reliable.

WAGE INDEX POLICIES AND REQUEST FOR INFORMATION ON AN IRF-SPECIFIC WAGE INDEX 

For FY 2027, CMS proposes to continue to use the inpatient PPS wage index to adjust IRF payments, specifically employing the relevant pre-floor, pre-reclassification inpatient hospital wage index. However, the agency also solicits comments on whether it should consider using alternative data sources, such as IRF cost reports or Bureau of Labor Statistics (BLS) wage data, to construct an IRF-specific wage index for potential use in future years. 

The AHA has conducted extensive policy work on the wage index. It is a difficult issue without a consensus solution. For example, using IRF cost reports to construct an IRF-specific wage index would be very burdensome, both to IRFs and to CMS itself. It also would create a system that is circular and self-perpetuating. Specifically, using only IRF data in setting the wage index would mean that IRFs could influence their own wage index values. This could lead to a problem where IRFs with low wage indices may be unable to increase wages to become competitive in the labor market. 

Policymakers also have considered using BLS data to calculate the wage index. The AHA and our members have examined these data closely and found that, while their collection and use may be significantly less burdensome, they have critical shortcomings. For example, BLS data exclude the cost of benefits. However, benefits are an important component of the wage index because the portion of total compensation attributable to benefits varies systematically. If benefits were excluded, the wage index would be understated in areas where benefits account for a greater portion of compensation; it would similarly be overstated in areas where they account for a lower portion. Therefore, any adjustments made to include benefit costs would have to be market-specific. In addition, if hospital-specific benefit information is to be added, it would have to be collected on CMS’ Medicare cost report. Yet doing so would add regulatory burden as well as some degree of circularity back into the system.  

In addition, BLS data are derived from voluntary surveys and a sample of employers. Estimates using a sampling methodology like the BLS approach would be less reliable than using the entire universe of cost reports. CMS’ current process in calculating the inpatient PPS wage indices, and therefore presumably any new process for IRF wage indices, allows for extensive public scrutiny of the data while the BLS approach does not. Unlike CMS’ public process for review and correction of cost report wage data, BLS has a strict confidentiality policy. Hospitals would thus be unable to verify the accuracy of the data. 

BLS data also include data on the wages of healthcare workers employed in all industries. For example, healthcare sector data from hospitals, physician practices, skilled-nursing facilities, ambulatory surgical centers, home health agencies and hospices are all included. Yet, IRFs differ from the universe of all employers in terms of the wage levels necessary to recruit and retain qualified employees, the percentage of compensation paid in benefits, the likelihood of unionization and other factors that affect compensation rates. 

Finally, we urge CMS to consider whether it would need to create a system for reclassifications of IRFs to a different labor market. Labor markets cannot realistically be defined as hard boundaries, and certain adjustments to the wage index may be necessary to accurately capture differences in labor costs across IRFs. However, our members have expressed concern that the number of reclassifications and exceptions permitted under the current inpatient PPS is complex and confusing.

IRF QUALITY REPORTING PROGRAM

Beginning with the FY 2029 IRF QRP, CMS proposes that IRFs must complete their data submissions and make corrections to their IRF-PAI assessment data as necessary no later than the 15th day of the second month after the end of the calendar quarter. If this day falls on a Friday, weekend, or Federal holiday, the date would be delayed until 11:59 p.m. EST on the next business day. Currently, IRFs have approximately 4.5 months (135 days) after the end of each quarter to submit IRF QRP data; this proposal would allow for approximately 45 days for data submission for both assessment-based data and data submitted through the National Healthcare Safety Network.

AHA’s members work hard to meet the existing data submission deadlines, and, as CMS notes in its analysis, 99.08% of all IRF-PAI assessments were submitted to CMS within a 45-day timeframe, and 88.5% of all IRFs submitted NHSN data within a 45-day timeframe. We believe that the shortened timeframe for data submission is likely sufficient; however, to ensure consistency and clarity, we suggest that CMS instead consider a clearer deadline. For example, requiring submission by the “last business day of the month,” rather than the “15th day of the month,” would be simpler for facilities to interpret. In addition, we urge CMS to monitor patterns of data submission leading up to the new deadline and provide adequate notice and assistance to facilities approaching the deadline without having submitted data; for example, reminding them of their options for extraordinary circumstances exceptions.

We appreciate your consideration of these issues. Please contact me if you have questions, or feel free to have a member of your team contact Jonathan Gold, AHA’s senior associate director for policy, at (202) 626-2368 or jgold@aha.org.

Sincerely,

/s/

Ashley Thompson
Senior Vice President
Public Policy Analysis and Development

__________

¹ AHA. (March 2026). The Cost of Caring: Challenges Facing America’s Hospitals as They Care for Patients in 2026 (https://www.aha.org/costsofcaring).
² Id.
³ Id.
⁴ Id.
⁵ ASPE. (October 2023). Changes in the List Prices of Prescription Drugs, 2017-2023 (https://aspe.hhs.gov/reports/changes-list-prices-prescription-drugs).
⁶ HHS OIG. (April 2022). Some Medicare Advantage Organization Denials of Prior Authorization Requests Raise Concerns About Beneficiary Access to Medically Necessary Care (https://oig.hhs.gov/reports/all/2022/some-medicare-advantage-organization-denials-of-prior-authorization-requests-raise-concerns-about-beneficiary-access-to-medically-necessary-care/).
⁷ Premier. (March 2024). Private Payers Retain Profits by Refusing or Delaying Legitimate Medical Claims (https://premierinc.com/newsroom/blog/trend-alert-private-payers-retain-profits-by-refusing-or-delaying-legitimate-medical-claims).
⁸ AHA. (March 2026). The Cost of Caring: Challenges Facing America’s Hospitals as They Care for Patients in 2026 (https://www.aha.org/costsofcaring).
⁹ https://www.aha.org/system/files/media/file/2025/06/aha-comments-on-cms-inpatient-rehabilitation-facility-fy-2026-proposed-payment-rule-letter-6-10-2025.pdf
¹⁰ Centers for Medicare and Medicaid Services. (February 2016). Hospital Multifactor Productivity: An Updated Presentation of Two Methodologies (https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/ReportsTrustFunds/Downloads/ProductivityMemo2016.pdf).
¹¹ 85 Fed. Reg. 58797 (Sept. 18, 2020).
¹² Medicare Payment Advisory Commission, (June 2023). Report to the Congress: Medicare and the Health Care Delivery System, ch. 10, at 416–18, 429–33.
¹³ 84 Fed. Reg. 39054, 39057, 39059-60, 39064-67 (Aug. 8, 2019). Medicare Program; Inpatient Rehabilitation Facility Prospective Payment System for Federal Fiscal Year 2020 and Updates to the IRF Quality Reporting Program.
¹⁴ Acumen, LLC. (Aug. 8, 2019). Inpatient Rehabilitation Facility Prospective Payment System Reform: Primary Diagnosis Technical Memo, at 3–5, 8, 10.
¹⁵ Social Security Act § 1886(j)(2)(a).