Back to Insights
TrustWorks On Call Newsletter Header

Please Price Transparently

October 8, 2026

Welcome to TrustWorks On Call, here with your healthcare business and strategy 411 for the week. If you enjoy our work, please consider forwarding it along to a friend and encouraging them to subscribe.

This week, we go Beyond the Whiteboard to examine the wave of leadership departures at Epic, before Dialing In on why AI-powered billing gains are a pocket of profitability that will not last. But first the news, starting with the Trump administration’s latest effort to make price transparency a thing.

Behind the Headlines

Unpacking the forces driving healthcare's biggest stories.

1. Trump admin finalizes payer price transparency rule, warns hospitals.

  • On Monday, the Departments of Health and Human Services, Labor, and Treasury finalized a rule overhauling the Transparency in Coverage requirements.
  • The rule lets insurers drop “ghost” rates for services a provider would never perform (such as surgery rates for a psychologist), requires rates to be reported as dollar amounts where possible, lowers the out-of-network disclosure threshold from 20 claims to 11, and requires a named executive to certify the data’s accuracy.
  • The same day, Federal Trade Commission (FTC) Chairman Andrew Ferguson sent warning letters to two dozen unnamed hospital systems, stating that missing or incomplete price disclosures could violate federal consumer protection law.
TrustWorks Take: This administration is doing what it takes for price transparency to catch on by targeting the primary complaint about the first round of insurer data: it was too bloated and messy for anyone but dedicated analysts to use. Cleaner files are an important step toward high-information healthcare shopping, but the Congressional Budget Office expects this to produce only a small downward pressure on healthcare prices. Even with full compliance, the impact of price transparency will be limited by the number of healthcare shoppers—patients in theory, but mostly large employers in practice—willing and able to act on this information. The most direct path to healthcare savings is for employers to use these data to steer away from the highest-cost providers in their networks, but that is not easy to do with the most popular and dominant health system in town. 
 
For hospitals, the price of noncompliance has been low but is set to grow. The Centers for Medicare and Medicaid Services (CMS) has fined only 28 hospitals since 2021, despite one report finding that more than half of hospitals are still not fully compliant. However, FTC Chairman Andrew Ferguson said that he considers the CMS rule merely a “regulatory floor,” with the FTC adding its own enforcement powers on top. Health systems treating price transparency as a box-checking exercise should audit their postings now and prepare to defend their prices to payers, employers, and other consumers. But until the FTC and CMS get serious about enforcement with real monetary penalties, the price of compliance is still likely to be greater than noncompliance.
 

2. Utah approves autonomous clinical AI pilots.

  • Last Friday, Utah’s Office of Artificial Intelligence Policy approved three new healthcare AI pilots through its regulatory “AI sandbox,” which allows the state to waive regulations for AI developers, including a Nolla Health app that will prescribe topical acne medications based on patient photos and an August AI system that will renew prescriptions for chronic conditions.
  • Physicians will initially review every AI prescription before it is sent, but with regulators’ written approval, later phases could move to review after the fact.
  • The pilots follow a Doctronic pilot for AI medication renewals, which the state’s medical licensing board unsuccessfully asked regulators to suspend.
TrustWorks Take: While the Food and Drug Administration (FDA) figures out how to regulate clinical AI without quashing it, Utah is charging ahead into a new frontier, becoming the first state to allow a pilot for AI prescribing. Acne is a reasonable starting point given its relatively standardized treatment protocols and the ability to assess many cases visually, although even seemingly straightforward diagnoses carry clinical risks. Utah was careful here in how it structured the phased-in approach, but the state caused controversy by not informing its own medical board about the Doctronic pilot before it launched, which the board feared would put Utahns at risk. If autonomous AI will practice medicine in their states, physicians should be helping to set the guardrails, rather than reading about them after the fact.
 
Only real-world data can tell us how autonomous AI will perform with actual patients, who present atypical symptoms, submit blurry photos, or leave details out of their medication histories—all things (human) physicians know how to handle. Otherwise, evaluating an AI model relies too much on benchmarks, which developers can game even without meaning to. AI tools will make mistakes, but people make mistakes too. This pilot is an opportunity to build a system that detects those mistakes and learns from them, so they can be minimized before AI tools are allowed to treat patients with true autonomy. 
 

3. HRSA picks drugs and drugmakers for 340B rebate pilot.

  • Last Thursday, the Health Resources and Services Administration (HRSA) announced its selection of 21 drugs from 10 drugmakers, including Pfizer, AbbVie, Merck, and AstraZeneca, participating in its revised 340B Rebate Model Pilot Program, which launches January 1, 2027.
  • Instead of receiving 340B discounts upfront, covered entities will have up to 45 days from dispensing to submit claims-level data to manufacturers to receive rebates.
  • HRSA’s first rebate pilot was blocked in court ahead of its planned January 2026 launch; the American Hospital Association (AHA) said it is “considering all available options” to stop the revised version.
TrustWorks Take: The rebate model is going to be a headache for providers. Covered entities will have to front the full cost of these drugs, submit claims through manufacturer-run systems, and wait to be paid. Providers are rightly worried about an administrative mess of denied claims and delayed or missing payments, with safety-net hospitals bearing the brunt. The pilot’s scope is limited—the eligible drugs represented less than 5.5 percent of 340B sales in 2025—but could serve as a template should HRSA choose to expand.
 
A reform like this has felt like an inevitable response to the 340B program growing with almost no friction to surpass $100B in discounted purchases. The rebate model is a deliberate attempt to slow it down. The challenge for HRSA (and CMS) is to rein in the program in a balanced fashion, improving program integrity without undermining the safety-net providers the program was intended to support. They also have not necessarily addressed the underlying questions about how program savings, which are intended for charity care, are generated and used. Health systems should prepare for greater administrative complexity, increased working capital requirements, and potentially smaller realized 340B margins.
 

Beyond the Whiteboard

Visualizing key trends from the healthcare industry

Epic Presses On After Leadership Exits
Epic is at a massive strategic juncture, launching tools for health systems to build their own AI agents and a generative AI model trained on data from over 320M patients, as it tries to place itself at the center of health system AI solutions. So far, the electronic health records (EHR) company is succeeding, having cultivated an “Epic-first” mindset in 71 percent of executives at systems using Epic. That means they prefer to incorporate Epic solutions over any third-party vendor, even if that means waiting for Epic to roll out its solution. However, the famously insular company is suddenly more vulnerable than ever amid succession challenges. President Sumit Rana, long seen as CEO Judy Faulkner’s heir apparent, left in August, followed by a string of senior leaders across R&D, AI, and cloud infrastructure. Seth Howard, executive vice president of R&D, is seen as a leading contender to succeed Faulkner, but no heir has been named. 
 

Epic’s dominance is largely unquestioned today, but there is a real risk of overplaying its hand, especially with so many experienced leaders gone. The company faces an FTC antitrust probe, newly enforced information blocking fines of up to $1M, and regulators who see it as a gatekeeper of patient data. Asking customers to bet their AI strategies on Epic raises the stakes of every misstep, at a moment when some clients already feel they cannot keep up. Some health system leaders were relieved when Faulkner said AI rollouts would be delayed over cybersecurity issues. Epic has also changed its measuring stick of success: it increasingly sees itself as an AI company competing with the likes of OpenAI and Anthropic, rather than just an EHR company. (At least those leaving Epic are not warning that their former employer might destroy the world on the way out.)

Dialing In

Sharing insights from our work with clients

AI Coding Profits Will Not Last
"We're finally seeing a return on an AI investment.” A health system finance executive was describing improvements in revenue capture following the implementation of AI-assisted clinical documentation and coding tools. He could quantifiably demonstrate how the investment produced more complete documentation, fewer missed diagnoses, and higher reimbursement on some encounters. Reimbursement gains are only one benefit of better documentation, but they are the most contentious, having become the primary front in the budding AI coding war. Blue Cross Blue Shield estimates that AI-driven coding changes have added nearly $1B in healthcare spending over two years at their companies. The stakes are set for the next round of payer negotiations.
 
In the long term, payers will not allow providers to charge more for providing the same level of care. Payers’ medical loss ratios and consumers' cost exposures will absorb the first hits, but payers will end up arguing for rate cuts across the board: “If you attach secondary diagnoses to every patient, we will pay less for the primary diagnosis.” This makes AI-driven billing gains another pocket of profitability, like hospital outpatient department rates and 340B, that providers cannot afford to get hooked on. Payments are supposed to reflect the resources needed to treat a patient, and AI should, in theory, reduce resource intensity, although not uniformly. Instead of using AI as a revenue maximizer, which payers will try to negotiate away, it will be more effective as a cost minimizer.