The Price Tool Hospitals Use Against You

Introduction

The federal rule that was supposed to let you compare hospital prices before your procedure is being used, nine times out of ten, by your hospital to demand higher rates from your insurer. The company that makes that possible has a name and an executive who said it out loud. Turquoise Health aggregates the price files the government forced hospitals to publish, then sells them back to hospitals and health plans. Marcus Dorstel, an executive there, told NPR what the data is actually for: "I think nine times out of 10 you will hear them say that the price transparency data is a vital piece of the contract negotiation now."

A rule pitched as consumer protection became a competitive-intelligence product, and the people selling the product will tell you so if you ask.

How a Shopping Tool Became a Negotiation Weapon

Walk the timeline. In 2019, a Trump executive order directed HHS to make hospitals post their prices so patients could, in his words, "go online and compare all of the hospitals and the doctors and the prices." CMS turned that into the Hospital Price Transparency Rule, effective January 1, 2021. The format CMS chose was the machine-readable file, or MRF: every hospital had to publish a giant data file listing every standard charge and every negotiated rate. A companion rule did the same to insurers starting in July 2022.

The format choice is where it all turned. A machine-readable file is not a price list you read. A large academic medical center's file routinely tops 500 MB and runs 2 to 5 million rows, with 50 to 150 distinct negotiated rates attached to a single procedure code. Cigna's insurer file goes further, which I'll get to. These are documents built for software to parse, not for a person scheduling a knee replacement to open on a laptop. So the transparency mandate didn't fail by accident. It was built in a format that locked patients out from day one and handed the industry a benchmarking tool it never had before.

A market appeared to fill the gap. By 2022, Turquoise Health and a handful of competitors were collecting those files from thousands of hospitals, cleaning them up, and selling the aggregated data to the only customers who could use it: hospital systems and insurers preparing for contract fights with each other. By 2026 Turquoise had raised $40 million in venture funding and signed more than 280 enterprise clients, including ten of the top 25 health systems and four of the five national insurers. None of those customers are patients. They're the institutions on both sides of your bill.

What the Files Actually Say

Start with Cigna, because its own compliance page makes the point better than I could. The insurer's machine-readable file, the page warns, "may be as large as one Terabyte (TB) in file size, which has significant system requirements for use." Then this:

"The Table of Contents is not intended to be opened on personal workstations."

That is the company telling you, in writing, on the page that exists to satisfy the transparency rule, that the transparency file cannot be opened on your computer. A terabyte is roughly a thousand high-end laptops' worth of storage in one file. It exists because the government required it so you could shop for care, and you cannot open it, though the hospital's contract negotiator can.

The government's own watchdog reached the same conclusion from the other direction. The GAO's October 2024 report to the House Energy and Commerce Committee found flatly that "patients rarely use hospital shoppable service pricing information." The report's table of who actually uses the data lists insurers, third-party administrators, employers, and data vendors as the real users, with patients handled separately through "patient-friendly" estimator tools, not the raw files. Stakeholders told the GAO that the data vendors who "process, standardize, and aggregate hospital pricing data" had become the de facto gatekeepers of information that was supposed to be public and free.

The GAO also caught the quiet part in CMS's own rulemaking. The agency's 2024 final rule language conceded the price data "could help health plans negotiate prices more effectively." That's CMS naming the intended user, and the intended user is the health plan, not you. The Congressional Budget Office had already predicted the patient side would amount to almost nothing, estimating that transparency would produce "very small" price reductions because hospital prices are "primarily the result of relative market power," not information gaps.

And the industry uses what the GAO described. Turquoise's own blog walks hospitals through using competitor rates in "managed care negotiations." Its case study of Huntsville Hospital, an 800-bed system in Alabama, describes the hospital "scrapping its previous negotiation strategy" after seeing what rivals were getting paid. Blue Cross Blue Shield of Minnesota's Eric Hoag told NPR the insurer uses the data to stay "competitive, or, you know, more than competitive against other health plans." Hospitals and insurers now walk into the room with the same dataset, the one that was supposed to discipline prices, and use it to position against each other.

The Enforcement That Mostly Isn't

If the rule were really aimed at patients, you'd expect CMS to enforce the parts patients need. Look at the numbers and a different priority shows up.

Through June 2025, CMS reviewed about half of the roughly 7,098 hospitals subject to the rule. Of those reviewed, 65% received at least one warning notice or corrective action request, totaling 2,291 warning notices and 1,331 correction requests. Warnings are routine. The consequences almost never follow: in four and a half years, CMS charged exactly 27 monetary penalties, and hospitals appealed 21 of them. As of mid-2025, 288 hospitals still had open cases, and 264 that had cleared their cases later slid back into noncompliance and got fresh warnings.

The reason the warnings rarely become fines is structural. The average CMS case takes 195 to 220 days to run from warning to resolution, and penalties only start counting from the date of the first warning, not from when a hospital first broke the rule. So a hospital can ignore the requirement, wait for the warning, comply just enough to close the case, and pay nothing. Enforcement did pick up recently: CMS fined 10 hospitals in 2025 after fining 3 in 2024, with penalties ranging from $32,301 to $309,738. But four of the ten hospitals fined in 2025 had 30 or fewer beds. Since enforcement began in 2022, only two hospitals with more than 550 beds have ever been fined, despite those being exactly the systems with the most market power and the biggest bills.

Who Benefits

Three groups get something real, and patients aren't among them.

Hospital systems get competitive intelligence they never used to have. Before the rule, a hospital knew only its own negotiated rates. Now it can see, in detail, that the system across town gets paid 15% more by the same insurer for the same procedure, and it walks into the next negotiation demanding to match it. That pushes rates up. Huntsville Hospital scrapped its strategy after looking at the data, and it did not scrap it to charge less.

Insurers get the mirror-image benefit: a benchmarking tool to pressure their own networks and, in at least one documented case, a litigation weapon. The GAO report footnotes a suit in which an employer used hospital pricing data against its own third-party administrator for allegedly failing to pass along negotiated discounts. The data cuts in whatever direction the institution holding it needs.

Then there's the layer that turned a regulatory mandate into a revenue stream. Turquoise Health and its competitors exist because the files are unusable in raw form, so somebody has to aggregate them. Turquoise's products are named "Market Intelligence," "Contract Intelligence," and "Precision Contracting." Its pricing is enterprise-only, scoped per customer. CEO Chris Severn confirmed to MedCity News in March 2026 that providers use the data "primarily for benchmarking and contract negotiations." Investors put $40 million behind this because it's a validated business, not a side effect. The public data the government promised patients became a product sold back to the institutions that bill them.

Why Shopping Was Never Going to Work

There's a layer underneath all of this. Even if the files were perfectly readable, the premise was broken. Patients don't shop for care on price, because the system doesn't work that way.

A 2021 study in the Journal of Health Economics by Cooper, Chernew, and colleagues looked at lower-limb MRI scans, about as close to a commodity as medicine gets. The median patient in their data had 16 MRI providers within a 30-minute drive of home. On the way to the appointment they actually kept, the average patient drove past six cheaper providers. What explained where people ended up getting scanned? The referring physician accounted for 52% of the variation in price paid. What the patient owed out of pocket, and who the patient was, each accounted for less than 1%.

People go where their doctor sends them. The "go online and compare prices" model assumes a patient behaves like a shopper comparing flights, when the actual decision is made by a referral the patient doesn't price-check and often can't. So the file-size problem and the billing-code problem are real, but they sit on top of a deeper one: the consumer-shopping theory of healthcare pricing fails before anyone tries to open the file.

The Bottom Line

This is the rare policy failure that's bipartisan and structural rather than a story of one bad actor. Trump's 2019 order started it, the Biden administration expanded enforcement and raised penalties, and Trump's February 2025 executive order doubled down, with a White House fact sheet citing a 2023 economic analysis projecting as much as $80 billion in savings. Every version sold the same promise to patients while keeping the format that delivered the benefit to the industry instead. HHS Secretary Robert F. Kennedy Jr. now says the 2026 rule "closes the loopholes hospitals exploit to hide real prices," which is its own admission that five years of the rule didn't deliver what was promised.

The open question is whether the 2026 changes, which require actual payment data and a named executive's attestation that the files are accurate, do anything for the patient who still can't open a terabyte file, when the deeper problem is that she follows her doctor's referral regardless of price. If the rule keeps producing data that only Turquoise's customers can use, the next round of "transparency" reform is just a better-stocked weapon for the same negotiation. Read Dorstel's quote again and ask who the data was ever really for.

Full breakdown, every source linked, at {{ghost_url}}.