En Banc Fifth Circuit Rejects Key QPA Rules in TMA III

Key Takeaways

• The en banc Fifth Circuit rejected two QPA rules central to No Surprises Act payment disputes, holding that health plans may no longer include non-negotiated “ghost rates” in QPA calculations or categorically exclude bonus, incentive, risk-sharing and other retrospective payment adjustments.
• The court upheld the exclusion of one-off single-case agreements from the QPA calculation, and the vacatur of the unlawful provisions stands despite arguments about the practical consequences.
• Key implementation questions remain unresolved, including whether corrected QPAs apply to past claims or completed IDR determinations. Agency guidance is expected, and Supreme Court review remains possible.

On Aug. 11, 2026, the en banc Fifth Circuit issued its long-awaited decision in Texas Medical Association v. HHS, commonly known as TMA III.¹ The court held that federal agencies departed from the No Surprises Act (NSA) in two respects central to how health plans calculate the qualifying payment amount (QPA). The court left the district court's vacatur in place as to the provisions held unlawful.

Background

The case took an unusual route. A three-judge appellate panel largely overruled the district court in favor of the agencies in October 2024. The full court then granted rehearing en banc, vacated that opinion and reheard the appeal. Ultimately, 17 judges produced four separate opinions.

The QPA warrants that attention. It generally sets patient cost sharing, and it is a mandatory factor in every independent dispute resolution (IDR) determination. It also does not float. For established plans and services, Congress anchored it to contracted rates recognized on Jan. 31, 2019, adjusted only for inflation. Each plan calculates its own QPAs from its own data, and while plans must disclose the QPA and certain methodology information, providers do not receive the rate-level data needed to reproduce the calculation independently. A methodology error therefore does not wash out; it compounds.

The Three QPA Issues Before the Court

• Ghost rates are out. Plans commonly present providers with a form contract carrying a default fee schedule for every covered service. Providers typically negotiate only the rates they actually bill and leave the rest untouched. The result is a contract full of rates for services the provider never performs, sometimes as low as $0. The July 2021 interim final rule² directed plans to count every rate on the face of the contract, and August 2022 FAQs later carved out $0 rates alone.³ The court rejected that structure entirely. The NSA limits the QPA to rates for an item or service (1) "provided by a provider" and (2) "furnished" in the relevant region, and a rate for a service nobody performs satisfies neither condition. The court saw no principled line between a $0 ghost rate and a $1 ghost rate; either the provider contracted for it or did not. The court did not decide (nor was it asked to decide) how frequently, or over what period, a provider must furnish a service before its rate may be counted.
• Bonus and incentive payments cannot be categorically excluded. The NSA defines each contracted rate as the "total maximum payment" for an item or service. Reading those words at face value, the court held that plans may not categorically strip out risk-sharing, bonus, penalty, incentive-based and retrospective adjustments. The agencies said tying annual lump-sum incentives back to individual service codes is impracticable. The court replied that this was the sort of question the agencies could have worked out in notice and comment, had they not issued the July rule as an interim final rule. The court acknowledged that some payments may not connect to a particular service, and it did not prescribe how plans must allocate every non-fee-for-service payment.
• Single-case agreements remain excluded. A one-off agreement may be negotiated and paid, the court held, and still not be a generally applicable rate recognized "under" the plan. The QPA is meant to approximate generally applicable in-network market rates, not isolate out-of-network arrangements. The court resolved the tension between its treatment of ghost rates and single-case agreements by excluding both, for different reasons.

The Data Behind the QPA Debate

Public debate over IDR turns heavily on awards running several times the QPA. In 2024, providers and facilities prevailed in approximately 85% of payment determinations, and roughly 85% of selected offers exceeded the QPA.⁴ The en banc court read those figures differently than payors have; it observed that IDR volume exceeded agency projections by a factor of 84, that providers prevailed in more than 80% of determinations and that certified IDR entities selected an amount above the QPA in roughly 85% of cases. The court treated those figures as corroborating its conclusion that including non-negotiated ghost rates produced artificially low QPAs.

That is the practical significance of the ratio; a multiple is only as informative as its denominator, and two of the rules that built the denominator have now been held unlawful. The evidence does not point in only one direction.

• A peer-reviewed study of 7,076 disputes involving Elevance Health found median awards at 3.72 times the QPA but 2.04 times the median allowed amount on matched Elevance in-network claims for the same procedures in the same three-digit ZIP code, suggesting that the QPA explained part of the gap but not all of it.⁵
• A Congressional Research Service analysis of emergency-service disputes in 14 states found the median prevailing offer at more than three times the 2025 median in-network rate among IDR-participating providers, with the median QPA slightly above that benchmark.⁶ That comparison has limits, since the QPA is a 2019 rate indexed for inflation and the benchmark is a 2025 rate.

What TMA III establishes is narrower. Two components of the governing methodology were unlawful in ways that tended to depress QPAs.

Vacatur Remains in Place Despite Implementation Concerns

The agencies and insurer amici warned that recalculation would demand considerable time and resources and that immediate vacatur could disrupt NSA implementation. The court was unmoved, reasoning that the agencies can manage the transition through enforcement discretion, as they have throughout the appeal:

“The APA, however, does not embrace a too-big-to-vacate principle. And administrative agencies cannot survive judicial review simply by making mistakes that are so colossal that the sky will fall if a court reviews them.”

The remedial question is not settled nationally. Judge Ho joined the per curiam because Fifth Circuit precedent treats vacatur as the APA's default remedy, but wrote separately to question whether courts may universally vacate agency rules at all. The Supreme Court expressly reserved that question last year in Trump v. CASA.⁷

A Divided Court and What Comes Next

Nine judges joined the per curiam in full. Judge Southwick joined the ghost-rate and single-case-agreement portions. Judge Oldham concurred separately on procedural grounds, concluding that the agencies could not repair a defect in a legislative rule through an informal FAQ. Judge Haynes wrote on behalf of six judges who agreed on single-case agreements but would have upheld the ghost-rate and incentive-payment provisions and would have remanded rather than vacated. The ghost-rate analysis drew at least 10 votes; the incentive-payment holding and the vacatur remedy each rested on nine. Sixteen judges supported excluding single-case agreements.

A close en banc vote on a federal statute with an unresolved APA remedies question makes Supreme Court review plausible, though not assured. Timing matters more immediately. Current enforcement relief permits continued use of earlier QPA methodologies for items and services furnished before Oct. 1, 2026, and the agencies said further guidance would follow the final en banc disposition.⁸ As of the date of this publication, no postdecision guidance has been issued. Plans, providers and facilities may face operational decisions before anyone knows whether this case might head to the Supreme Court.

Putting the possibility of Supreme Court review aside, three implementation questions remain open for future agency guidance: (1) how often a provider must furnish a service before its rate counts; (2) how plans should attribute non-fee-for-service payments to individual services; and (3) whether corrected QPAs apply retrospectively to prior claims, cost sharing or completed determinations. All of this lands while the IDR Operations Final Rule, effective Aug. 3, 2026, is still phasing in.

Practical Steps for Providers

While the postdecision guidance is pending, providers should take the following proactive steps:

• Preserve every QPA disclosure you receive, including any statement identifying the methodology used, and request all available methodology information going forward.
• Document which services your participating providers actually furnish, since that is now the operative question for whether a rate belongs in the calculation.
• Identify open disputes where the QPA may materially affect the parties' submissions.
• Preserve records on recently completed disputes, recognizing that nothing in TMA III currently authorizes reopening a completed determination over QPA methodology.

Providers should also track whether and when their payors revise QPAs, then model the effect on two fronts. A higher QPA may strengthen a valuation presentation in IDR, but it also raises the recognized amount for patient cost sharing. Neither the QPA nor this decision determines what a certified IDR entity will select.

Polsinelli’s No Surprises Act team continues to track No Surprises Act developments. Should you have questions, please reach out to Josh Arters, Rachel Roberson, Olivia Winnett or your preferred Polsinelli attorney.