Considerations for federal agencies tasked with improving health plan price transparency data 

By Stacey Pogue and Nadia Stovicek

For people who have private health insurance, the prices paid for health care are generally set through negotiations between insurers or employer health plans, typically through their third-party administrators (collectively referred to here as “payers”) and in-network health care providers. These negotiated rates were long considered proprietary until federal Transparency in Coverage (TiC) rules required payers to publicly post them free of charge starting in 2022. TiC rules require payers to post price information in two formats. The first is a consumer-friendly web tool meant to help patients see upfront costs and shop for care. The second format is through detailed machine-readable files (MRFs). While not intended to be directly accessed by consumers, MRFs are nonetheless intended to benefit the public. The TiC rule envisions that MRFs would be accessed by users including researchers, policymakers, state and federal regulators, employers, and app developers, who would leverage the data to deliver “more targeted oversight, better regulations, market reforms to ensure healthy competition, improved benefit designs, and more consumer-friendly price negotiations.”

But the reality has not lived up to the vision. The MRFs available under current guidance and enforcement have proven challenging to access and use, greatly limiting the ability of intended audiences to gain insights from the data and take actions that benefit consumers and markets. However, a renewed focus under the Trump administration aims to improve TiC data.

This blog post catalogs known issues with TiC MRFs raised by independent researchers in published analyses and commentaries, explains the directives to improve the data in a recent executive order, and explores the thorny issues that limit access to publicly available TiC data. 

Data Issues Limit Progress Toward Transparency in Coverage

Since TiC requirements for MRFs took effect in July 2022, stakeholders have identified many issues that make the data hard to access, analyze, and draw meaningful conclusions from. For example, the KFF-Peterson Health System Tracker recently released a detailed analysis that illustrates common structural issues with the TiC data. 

Below, we organized issues raised in analyses and commentaries from independent researchers into five themes: availability, accessibility, standardization, quality, and utility. We assigned specific data issues to just one theme below for simplicity, though several issues could fit into more than one category.

Availability. Data required by the rule are not or may not be posted by payers. 

• Federal agencies that oversee TiC requirements have not indicated that they have assessed payer compliance with posting files, nor announced any mechanism to do so, through audits or other means. In contrast, federal agencies and other organizations have done much more to assess compliance with parallel hospital price transparency requirements. 
• Federal agencies have not implemented requirements for a prescription drug MRF. Payers are waiting on the agency to release needed technical specifications. 

Accessibility. Issues with locating the data and obtaining it, either directly or through commercial data vendors.

• Data files can be hard to find. They are hosted on each insurer’s or health plan sponsor’s website, with no central repository for either the data or links to files on payers’ websites.
• One of the most commonly cited issues with the data files is their massive size, which significantly limits access. Payers post a staggering volume of data that can only be directly accessed by entities with specialized and expensive computer processing infrastructure and highly skilled data engineers and programmers. For most stakeholders, the MRFs are simply inaccessible. 
• Several structural features of the data reporting requirements unnecessarily inflate file sizes:
  • Use of an inefficient file format and file structures.
  • Significant redundancy in the data. One analysis found that almost half of price files posted by six large insurers were duplicates. 
  • A large amount of irrelevant data or “ghost rates,” contracted rates from providers who do not perform a specific health care service (e.g., the rate for a cardiology service billed by a podiatrist, or vice versa). One analysis found that among 34,000 Colorado providers shown in the TiC data with a negotiated rate for a colonoscopy, only 300, or about 1%, had actually submitted a claim for a colonoscopy in roughly the two prior years. 
• The cost to license TiC data is a barrier for all but the best-resourced stakeholders. Because few entities can access the raw data directly, many researchers and stakeholders who want access buy the data from commercial vendors that specialize in importing, organizing, and cleaning the massive TiC MRFs. 
• The index file lacks information on which providers or services are in which specific subfiles, requiring users to open, possibly, thousands of huge subfiles to find needed information.

Standardization. Issues with variation allowed in the federal schema (technical specifications for reporting data). 

• Variation is allowed in how payers structure the files that connect rates to a specific provider, and some payers use a structure that makes it significantly harder to collect and process the data. Lack of standardized file labels and file types adds additional challenges. 
• The lack of standardized conventions for the use of numerical provider identifiers makes it hard tolink TiC data to other datasets, a basic step needed to fully leverage TiC data and understand cost drivers:
• The TiC schema allows payers to identify covered items and services using common billing code types, like Current Procedural Terminology (CPT codes), or unique payer custom codes. Payer use of custom codes, sometimes in place of common billing codes for common services, makes it challenging to compare across payers. 

Quality. Issues with payers posting data in an incomplete or inaccurate manner.

Utility. Issues that limit the usefulness of TiC data.

• Some specific data points lack enough context to be meaningful:
• The TiC dataset as a whole lacks certain relevant information that would add substantial value:

Renewed Federal Agency Attention Could Help Address Issues

The Trump Administration issued an executive order in February that directs the Departments of Health and Human Services, Labor, and Treasury (collectively, the “tri-agencies”) to “rapidly implement and enforce” TiC and parallel hospital price transparency rules to make more meaningful price information available. 

The order directs the tri-agencies to undertake the three enumerated actions below within 90 days (by May 26, 2025). Depending on the tri-agencies’ interpretation and priorities, they could address many of the known TiC data issues under the banner of the executive order’s directives, as shown below: 

1. Require that “actual prices of items and services, not estimates” are posted;

The tri-agencies could address some utility-related issues affecting rates set as a percentage of billed charges and prices per dosage unit for physician-administered drugs.

2. Issue guidance or proposed rules to ensure price data is “standardized and easily comparable across hospitals and health plans,” and

The tri-agencies could further standardize file formats, file structures, conventions for the use of provider identifiers, and the use of custom billing codes.

3. Issue guidance or proposed rules to increase enforcement and improve compliance with the rules.

The tri-agencies could address some availability- and quality-related issues by releasing the needed schema for prescription drug price reporting and developing a system to assess payer compliance issues that affect data quality.

TiC issues categorized above as related to accessibility would not necessarily lend themselves to fixes through the executive order’s focus on actual prices, standardization, and compliance. If steps to improve access—such as using a more efficient file format and structure, removing ghost rates, preventing data redundancy, augmenting index files, and centralizing either the data itself or links to it—are not part of the initial executive order response, then they will need to be part of a longer-term effort in order for TiC data to be as accessible and impactful as envisioned. The TiC rule preamble contrasts TiC data that is “available to the public free of charge” with an example of a proprietary commercial claims dataset that is “costly to purchase” for researchers at $45,000 a year. Yet, anecdotes from several researchers place the cost to license TiC data in the same ballpark, which often renders data cost-prohibitive as implemented, despite the rule’s intention.

Looking ahead

TiC data straddle the line between public and proprietary. By law, insurers must post them publicly and free of charge, yet due to several data issues, few entities outside of commercial data vendors have the costly infrastructure and expertise needed to access the data, which they can parse, organize, and sell as proprietary. Commercial data vendors have greatly increased access to TiC data—there would be very little access without them. Yet access challenges remain for researchers, state and federal regulators, policymakers, and employers—the entities that the TiC rule envisions will leverage the data to benefit consumers.

Despite challenges, researchers are starting to share new insights and tools made possible by TiC data. But given that even research institutions with the resources and expertise to use TiC data still struggle to access and analyze it, it could be quite a while before TiC data can be translated into actionable information for policymakers, regulators, employers, and consumers. With the recent executive order, the tri-agencies have the opportunity to shorten that window, should they choose to do so.