Challenging the QPA

What the outcomes show

In the second half of 2025 the prevailing offer stood above the plan's own qualifying payment amount in about 87% of payment determinations — 995,270 determinations out of 1,145,039.

How far above depends on the service. CMS breaks it out by CPT category:

A category running 25.85× is not evidence that arbiters are generous. It is evidence that in that category the plan's benchmark sits a long way from what an independent entity finds defensible — which is a statement about the benchmark, not about the arbiter.

What the entity is directed to consider

The certified IDR entity must consider the qualifying payment amount. It must also consider additional information the parties submit, where that information is credible and relates to the specific circumstances. The statute names categories:

• The level of training, experience, and quality and outcomes measurements of the provider or facility — the regulation points specifically at measures endorsed by the consensus-based entity authorised under section 1890 of the Social Security Act.
• The market share held by the provider or the plan in the geographic region.
• The acuity of the patient and the complexity of the service actually furnished.
• The teaching status, case mix and scope of the facility, where applicable.
• Demonstrations of good-faith efforts — or their absence — to enter into network agreements, and any contracted rates over the previous four plan years.

These are the levers. A submission that argues only "the QPA is too low" has given the entity nothing to act on. A submission that documents an unusually acute presentation, a subspecialty qualification the code does not capture, or a payer that has refused to contract, is arguing in the terms the statute set out.

What the entity may not consider

Three things are expressly excluded, and knowing them saves wasted effort:

• Usual and customary charges — including any amount derived from them.
• The amount the provider would have billed absent the No Surprises Act protections.
• Payment or reimbursement rates under public programmes — Medicare, Medicaid, CHIP, TRICARE, the Indian Health Service.

This matters more than it looks. A submission built around billed charges is building on the one foundation the entity is forbidden to stand on. Independent rate data is a recognised category of additional information; your chargemaster is not.

Where QPA calculations are weakest

The QPA is a median of contracted rates, and every median has a denominator. The arguable questions are about how that denominator was built: which contracts were included, whether the geographic region is defined in a way that reflects where the service was actually delivered, whether the service code genuinely matches, and how it was derived where the plan had few or no contracts for that service.

Anaesthesia is the clearest example of the last point, because the benchmark rests on a conversion factor rather than a per-service rate, and how the plan derived that factor is frequently the substance of the dispute. Newer service categories — remote and non-face-to-face services in particular — rest on thinner contracted histories for the same structural reason.

Baseball arbitration changes what a good offer is

Federal IDR is a final-offer process. The entity selects one party's offer or the other's; it cannot split the difference. That has a specific consequence people routinely get wrong: your offer is not an opening position, and inflating it does not create room to negotiate. It creates a risk that the entity picks the plan's number because yours was the less defensible of the two.

The published medians tell you where defensible offers have been landing by service and by claim size. They are not a target to aim at — they are a check on whether the number you can actually support with evidence is in a plausible range. The full outcome data, with sources. For what the 2026 rule changed about deadlines, fees and eligibility, see the 2026 rule breakdown.

What Each Side Actually Anchors Its Offer To

The Departments report how parties justified their offers, and the pattern is consistent: plans and issuers frequently anchored their offers to the QPA. Providers, facilities, and air ambulances frequently anchored theirs to two other things — past out-of-network payments from the same plan or issuer, and past in-network rates with the same plan, or with a different plan in the same state.

Read together with the outcomes, this tells you something structural about the process. Providers, facilities and air ambulance providers prevailed in roughly 85% of payment determinations in the second half of 2025. That means the median winning offer in the published tables is, most of the time, the provider's offer — not the plan's QPA-anchored number.

The practical implication for building an offer is straightforward:

• The plan's QPA-anchored offer is built from contracts you never signed. Your counter-anchor, if you have one, is your own payment history with that plan or with other plans in your state.
• Past out-of-network payments from the same plan are a documented, frequent basis for provider offers. Pull them before you pick a number.
• Past in-network rates — yours with the same plan, or comparable rates with another plan in the same state — are equally documented. A rate from another state is not what the data describes.

This is also the honest framing of the "IDR vs. QPA comparison": the QPA is what the plan says the service is worth; the winning offers in the published data are anchored to what payers have actually paid, in network and out, in the same geography. Where those two numbers diverge — and the published medians show they diverge widely by specialty — the divergence is between the plan's benchmark and the plan's own payment history.

One caveat the Departments do not resolve: the reports describe what parties cited, not causation. There is no published finding that anchoring an offer to past payments causes a win. Treat the anchoring data as evidence of what defensible offers are built from, not as a formula.

The 85% Win Rate Is Two Different Numbers

Almost every summary of federal arbitration quotes a single provider win rate. The Departments' own report does not, and the distinction it draws is the one that matters if you are deciding whether to file.

A default decision is one where a party failed to submit an offer or failed to pay its fees. When that happens the certified IDR entity is required to rule for the party that did submit and did pay — the merits are never reached. In the second half of 2025 default decisions were 17% of all payment determinations, and 90% of them went to providers, facilities and air ambulance providers. Those are wins by forfeit.

Strip them out and the picture changes:

Two things follow, and neither is visible in the headline figure. The contested win rate — what happens when both sides actually show up, submit offers and pay fees — is 84%, not 85%. And it has moved the wrong way: it was 87% in the first half of 2025. The headline rate fell more gently only because the share of forfeits also fell, from 22% to 17%.

The honest reading is that the provider side still prevails in the clear majority of contested determinations, and that the margin is narrowing. A practice deciding whether a claim is worth filing should be working from the contested number, because a dispute the plan actually contests is the dispute it will have.

The Litigation Pause That Distorts Every 2023 Comparison

If you have seen the figure that determinations grew roughly fivefold in 2024 — 1,047,575 against 209,346 in 2023 — treat it with the context the Departments themselves attach. The 2023 number is understated, and the reason is court-driven, not demand-driven.

Two decisions from the Eastern District of Texas suspended determinations for extended periods:

• February 6 – March 17, 2023: determinations paused following TMA II.
• August 4 – September 21, 2023: determinations paused for single disputes, and until December 15, 2023 for batched disputes, following TMA III and TMA IV.

On August 3, 2023, in Texas Medical Association v. HHS (No. 6:23-cv-59-JDK, "TMA IV"), the court vacated the batching provision at 45 CFR 149.510(c)(3)(i)(C) and vacated the $350 per-side administrative fee set by the December 23, 2022 guidance. On August 24, 2023, the same court (No. 6:22-cv-450-JDK, "TMA III") vacated portions of 45 CFR 149.130 and 149.140.

Why this matters for how you read the outcome data:

• Any quarter-over-quarter or year-over-year comparison that crosses the 2023 pause windows measures litigation calendar effects, not changes in arbiter behavior or payer conduct.
• The win-rate and prevailing-offer percentages for 2023 rest on a smaller, partially interrupted base. The 2025 figures — 1,082,247 determinations in the first half and 1,145,039 in the second — are the more stable reference.
• The batching rules were litigated, and the vacated regulation is not in force as written. The practical batching criteria in use today are not something to assume from the old rule text; verify current requirements before batching claims.

The growth is real. But growth measured off a litigation-suppressed 2023 baseline is a growth rate, not a clean measure of volume expansion. Anyone quoting the fivefold figure without the pause explanation is quoting it out of context.

Where the Numbers Are Heading

The published reports now give a six-period view of how the system has actually performed.

The rate rose in every period through the first half of 2025 and then edged down. That turn is worth stating plainly rather than smoothing over: the long climb is not a trend line to extrapolate from, and the contested win rate above shows the same direction more sharply. Some of the movement in either direction reflects changes in which disputes are being initiated and at what claim sizes, not a change in how arbiters decide.

Volume has expanded far faster than the win rate has moved. Disputes initiated rose from 1,186,812 in the first half of 2025 to 1,372,563 in the second. Closures rose from 1,349,343 to 1,449,900. Determinations rose from 1,082,247 to 1,145,039.

The shape of who is initiating has also shifted. In the first half of 2023, the top ten initiating parties accounted for about 78% of all disputes; in the fourth quarter of 2025 that share was roughly 67%. The dispute base is broadening even as the total grows.

The pattern that matters for a practice considering its own posture: the arbitration pathway is being exercised at scale, the prevailing side in the published record is still the provider, and the share of disputes driven by a small number of repeat initiators is declining — which suggests the system is no longer the exclusive province of any single filing vendor or practice category. The figures are descriptive of what has been filed and decided; they do not predict the outcome of any individual dispute.

What is the qualifying payment amount?

The QPA is generally the plan's median contracted rate for the same or a similar service in the geographic area, calculated by the plan. It is the starting benchmark in federal IDR, not a determination of market value.

Can an IDR entity consider my billed charges?

No. The statute expressly excludes usual and customary charges, the amount that would have been billed absent the No Surprises Act protections, and public programme reimbursement rates such as Medicare and Medicaid.

What can an IDR entity consider besides the QPA?

Credible additional information relating to the circumstances: the provider's training, experience and quality outcomes; market share of either party; patient acuity and service complexity; facility teaching status and case mix; and demonstrations of good-faith efforts to contract, including contracted rates over the previous four plan years.

How often does the prevailing offer exceed the QPA?

In about 87% of payment determinations in the second half of 2025, according to CMS — 995,270 of 1,145,039.