RESPONSES PROVIDED BY CLEVERLEY + ASSOCIATES July 21, 2025

Recently, CMS released a Request for Information regarding the accuracy and completeness of hospital machine-readable files (MRFs). According to the request, “To meet the President’s Executive Order to ensure compliance with the transparent reporting of complete, accurate, and meaningful data, CMS seeks public input on whether and how CMS can improve hospital price transparency (HPT) compliance and enforcement processes to ensure that the pricing information in the machine-readable file (MRF) is accurate and complete.” The RFI specifically posed six questions related to the accuracy and completeness of data required in hospital MRFs. You can read the complete RFI here (link).

Cleverley + Associates is a recognized leader in hospital pricing and price transparency. We have helped hundreds of hospitals and health systems understand and comply with federal and state guidelines related to price transparency mandates. We submitted our comments to CMS in response to this RFI and posted these comments here! You can also download the pdf here!

QUESTION ONE RESPONSE

Should CMS specifically define the terms “accuracy of data” and “completeness of data” in the context of HPT requirements, and, if yes, then how?

CLEVERLEY + ASSOCIATES QUESTION ONE COMMENT

First, we appreciate the opportunity to comment on these important questions. We fully support healthcare price transparency and are committed to helping achieve more clarity and simplicity for patients. We thank CMS for continually reviewing and collaborating with stakeholders to achieve these common goals.

To specifically address the first question, we do not believe definitions of “accuracy of data” and “completeness of data” need to be further defined. With the addition of the attestation statement in the file schema released on July 1, 2024, hospital administrators understand that all required data elements must be provided. We do believe, however, that CMS could continue to clarify how these data elements are derived. We will comment in a future question on how industry alignment on the data sources and methods to extract required data elements would address the perception that MRF data is inaccurate and/or incomplete.

QUESTION TWO RESPONSE

What are your concerns about the accuracy and completeness of the HPT MRF data? Please be as specific as possible.

CLEVERLEY + ASSOCIATES QUESTION TWO COMMENT

We believe that a critical reason for MRF data being perceived as inaccurate or incomplete is driven from a lack of alignment on the data sources and methods used to extract the required MRF data elements. Hospital administrators are utilizing data within the hospital billing environment to construct the MRF but can derive the values from different sources.

First, industry stakeholders must agree that standard gross charges and payer specific negotiated charges are separate and derived from different data sources. Standard gross charge information should be defined as information from the hospital’s charge description master (CDM) and any additional drug/supply modules. Payer specific negotiated charges would come from the hospital’s contract management system and claims billing environment.

Perceptions of inaccuracy and incompleteness appear to be focused on the payer specific negotiated charge where the greatest amount of complexity resides. CMS appropriately recognized this complexity in adding the “Charge Method”, “Algorithm”, and “Estimated Allowed Amount” variables in the current MRF schema to provide additional context for the provided values. The challenge hospitals face is how to select the appropriate charge method.

There has been, however, confusion in how to leverage the charge method and associated fields:

We appreciate and agree with the definition CMS provided in the CY24 OPPS Final Rule to appropriately describe these situations where multiple methods and types occur to derive the payer specific negotiated charge: algorithm.

CMS also appropriately concluded “in the interest of reducing burden and complexity of files, we will allow hospitals provide a description of the algorithm, rather than attempting to insert the specific algorithm itself in the MRF” (CY24 OPPS Final Rule). Common algorithm logic is seen below to emphasize the prudency of this conclusion:

CLEVERLEY + ASSOCIATES QUESTION TWO COMMENT, CONTINUED

We provide this context as we believe the main cause for variation within the payer specific negotiated charge is how hospitals are trying to derive a compliant value. The hospital can attempt to derive a portion of the algorithm contents to conform to the limited set of defined charge methods (per diem, case rate, fee schedule, percentage of charge) or it can leverage the estimated allowed amount to completely account for all the algorithm elements.

The challenge with the first method is that these values do not fully account for all charge methods and algorithm contents needed to determine the payer specific negotiated charge. Further, attempting to provide this content isn’t feasible because of the incredible administrative undertaking and inability to conform to the file schema. We understand that some may believe that payment can be simplified in these basic charge method categories, however, in the thousands of hospital-payer contracts we have analyzed and modeled, we have yet to see where one of these methods are used without additional terms and conditions (absent an entirely percentage of charge contract).

In sum, the two primary ways hospitals are deriving the payer specific negotiated charge (illustrated on the following page) are limiting the usefulness of the HPT data and leading to a perception that it is incomplete and/or inaccurate. In reality, both are accurate and compliant as they are leveraging the data elements, definitions, and schema structure. However, only the other/algorithm/estimated allowed amount can truly be considered “accurate” AND “complete” as all payer specific negotiated charge methods and algorithm logic have been accounted for in the MRF values.

QUESTION THREE RESPONSE

Do concerns about accuracy and completeness of the MRF data affect your ability to use hospital pricing information effectively? For example, are there additional data elements that could be added, or others modified, to improve your ability to use the data? Please provide examples.

CLEVERLEY + ASSOCIATES QUESTION THREE COMMENT

Per our response in question two, the utility of the MRF data is limited because the payer specific negotiated charge is not consistently derived among hospitals. We do not believe additional data elements are needed, but contend that the answer to addressing the accuracy, completeness, and utility of the HPT data is found in having all hospitals (and payers) report payer specific negotiated amounts using the existing estimated allowed amount.

Some may contend that another path could be to require all algorithm elements from the contract management system to be provided. There are two key objections to this, though:

1)The administrative burden to compile this information would be prohibitively high. Further, it is difficult to imagine the required effort to create a uniform file schema to account for this complexity of thousands of variables and conditions within a single MRF.

2)Most striking, even if the first point could somehow be solved, developers and researchers leveraging this massive database would then also require patient claims data from the hospital to understand how that hospital’s treatment patterns create the final payer specific negotiated charge. So, why not use the estimated allowed amount where the application of the hospital’s unique contract management database has already been applied to patient claims to create the payer specific negotiated charge?

In sum, exclusively using the Estimated Allowed Amount for the payer specific negotiated charge would:

• Satisfy the “letter” and the “spirit” of the transparency requirements to convey a complete picture of payer specific negotiated charge that also captures the hospital’s typical treatment pathways
• Increase the utility of these values as they could be directly compared Essentially, the estimated allowed amount is indifferent to all the numerous contracting methods and values across hospitals and payers creating an ability to benchmark across disparate rates, methods, and conditions
• Address the concerns of inaccurate or incomplete data by providing a standardized definition and data source for both hospitals and payers.

Several final comments regarding the use of the estimated allowed amount:

1)First, defined this way, hospitals and payers could then leverage a straight-forward grouping logic to report payer specific negotiated charges by a single inpatient and outpatient code type (MSDRG for inpatient and primary APC/HCPCS for outpatient, as example). This would permit consistency across HPT and TIC files, dramatically reduce file size, and permit seamless comparison across hospitals and payers regardless of differing contract structures, rates, conditions, and methods.

2)Second, this definition would also permit the elimination of the “Charge Method” variables as these values do not “completely” describe any payer specific negotiated charge (percentage of charge could be the only exception where there are still some contracts that can be entirely POC).

3)While not a new field, we would support the creation of a national payer code database to be able to provide more comparability for the payer variable.

4)Finally, we offer a suggestion to change the name from “estimated allowed amount” to simply “allowed amount.” Because the values are derived using billed claims the results are not based on “estimates” but “actual” patient encounters. We believe the use of the term “estimate” could cause confusion among the public that the results are somehow inaccurate.

QUESTION FOUR RESPONSE

Are there external sources of information that may be leveraged to evaluate the accuracy and completeness of the data in the MRF? If so, please identify those sources and how they can be used.

Per our responses in questions two and three, if estimated allowed amount (or simply, “allowed amount”) became the definition for the payer specific negotiated charge then average historic reimbursement could become the single source of truth for this variable.

CMS could then define how to derive average historic reimbursement from a claims source – 837/835 claims data, as primary example – for a given time period and under certain conditions to permit appropriate and accurate reporting. We do express concern over recent guidance on May 22, 2025, to instruct hospitals to create this average from as little as one claim (not technically an average) or in the absence of claims data (not feasible to create historic reimbursement where no historic claims exist). We do understand the reason for the guidance to reduce the number of reported nine 9s in hospital MRFs which may have resulted from hospitals implementing previous CMS guidance to restrict values where HIPAA thresholds of less than eleven claims were present. We believe CMS could amend that guidance to using at least two claims in order to maintain an average (minimum for an arithmetic mean) and thereby creating more values in the MRF. Null or single claim volume instances should not be viewed as problematic as their exclusion in the file would not dramatically impact the vast majority of services that patients at that hospital would be interested in seeing. In fact, excluded cases could help inform patients that the treatment is not common at that hospital.

In sum, using 837/835 claims data to derive the payer specific negotiated charge could create alignment between hospital and payer MRF files to permit accuracy, completeness, consistency, and comparability.

QUESTION FIVE RESPONSE

What specific suggestions do you have for improving the HPT compliance and enforcement processes to ensure that the hospital pricing data is accurate, complete, and meaningful? For example, are there any changes that CMS should consider making to the CMS validator tool, which is available to hospitals to help ensure they are complying with HPT requirements, so as to improve accuracy and completeness?

CLEVERLEY + ASSOCIATES QUESTION FIVE COMMENT

We commend CMS for releasing the online validator tool to help hospitals and interested stakeholders understand a hospital’s adherence to the compliant file schema. As mentioned in previous responses, solidifying a data source and methodology to derive all elements for the payer specific negotiated charge would improve the accuracy, completeness, and meaningfulness of the data. We believe that the estimated allowed amount (or simply “allowed amount”) remains the most viable option to accomplish these goals.

QUESTION SIX RESPONSE

Do you have any other suggestions for CMS to help improve the overall quality of the MRF data?

CLEVERLEY + ASSOCIATES QUESTION SIX COMMENT

We appreciate the opportunity to share our responses to the previous questions and would welcome an opportunity to discuss further. Thank you for creating this RFI as a means to promote dialogue on this important topic.