Why We Did This

Congress passed the No Surprises Act (NSA) in 2020 to protect patients from surprise balance bills, which can occur when receiving emergency care or unexpected out-of-network care at an in-network facility. The NSA created a federal Independent Dispute Resolution (IDR) process to mediate payment disputes between insurers and providers.

Policymakers expected the law would lower provider prices, with the Congressional Budget Office projecting that it would reduce premiums by roughly 1 percent as well as federal deficits. Instead, the Center on Health Insurance Reforms estimates that cost increases as a result of the IDR process have jumped from $5 billion in 2022-2024 to $22.4 billion through 2025. Those increased costs—which include administrative expenses and additional plan expenditures—are ultimately borne by Americans and their employers through higher premiums and spending.

This paper examines how the IDR process came to be, how it created incentives that lead to less affordable care, and which policy options can mitigate these consequences while still protecting patients.

What We Found

Total dispute volume is much higher than Congress expected. Total dispute initiations reached 2.56 million in 2025—115 times the government’s initial projection.

Total dispute volume grew rapidly from 2023 through 2025. Disputed line items (the individual services that make up a dispute) decided in federal IDR increased about 17-fold.

Providers win the vast majority of the time, with win rates increasing from 2023 through 2025. In 2025, providers prevailed over insurers in 86.4 percent of disputed line items, up from 78.1 percent in 2023.

Arbitration awards are much higher than expected and growing. In 2025, the median award across all disputed line items was almost four times the Qualifying Payment Amount (QPA)—essentially an adjusted median in-network rate—and about 5.5 times the Medicare rate for the same services.

Across provider types, win rates are consistently high while awards vary widely. Emergency medicine providers had median awards of 3.5 times the QPA, while neurological surgery providers had median awards of 28 times the QPA, and physician assistants about 26 times the QPA.

Provider offers and award amounts have increased over time. As providers have continued to prevail in roughly 85 percent of disputed line items, their median offers have risen substantially, while insurers’ median offers have remained around the QPA.

The top 10 percent of award amounts have significantly escalated over time. At the beginning of 2023, the 90th percentile of awards was around eight times the QPA; by the end of 2025, it had surged to nearly 18 times the QPA.

Insurers lost the vast majority of the time even if they significantly increased their offers. In disputes decided by the highest-volume arbitration firms, insurer win rates remained low even when they made offers of 600 percent or more of the QPA.

Arbitrators with higher provider win rates handle higher volumes of disputes. Over half of dispute volume is handled by arbitration firms where providers win more than 88 percent of the time.

Dispute filings are highly concentrated, driven by a mix of provider organizations backed by private equity and third-party filers. Four organizations filed about half of all provider-initiated disputes determined in 2025.

Emergency medicine and radiology are the top two provider types involved in disputes. These provider types made up about 29 percent and 18 percent, respectively, of all disputed line items.

Why It Matters

Taken together, these findings show that IDR is increasingly an alternative payment system rather than a backstop for unusual out-of-network disputes. The combination of high provider win rates and awards far above in-network rates makes arbitration more attractive to providers than network contracting, leading to higher payments and putting upward pressure on in-network rates.

Providers, who have a high probability of winning awards that substantially exceed in-network rates, have strong incentives to remain out-of-network and file a high volume of disputes. This also gives them greater leverage to demand higher rates as a condition of staying in-network with an insurer.

The cost trajectory is particularly concerning. Only a small fraction of potentially eligible claims currently enter arbitration, even as dispute volume, provider offers, and awards have grown rapidly. If these trends continue, IDR costs could grow rapidly, while increasingly generous arbitration awards will give providers even stronger incentives to remain out of network.

Because arbitration firms earn more revenue by resolving more disputes, they have incentives to deem more claims eligible. The positive relationship between provider win rates and arbitration firms’ dispute volumes also raises concerns that the current system may reward firms that produce unjustifiably favorable outcomes for providers.

For insurers, offers may have to be so high to win in arbitration that winning more disputes would still increase the cost of the IDR process while risking an increase in in-network rates, which would be passed on to enrollees and employers.

What We Recommend

Congress should substantially reform the IDR process to improve incentives for market participants and prevent dramatic increases in costs while preserving patient protections from surprise bills they cannot reasonably anticipate or avoid.

To accomplish this, Congress should pursue a two-track framework that distinguishes between elective services, where patients have choices, and emergency services, where patients are unable to choose providers or negotiate prices before receiving care.

Congress should eliminate the federal IDR process for elective services. Providers should be permitted to balance bill only when patients receive meaningful advance notice that an out-of-network provider will participate in their care and give affirmative consent to any such charges. Congress and the administration should also ensure full implementation of the NSA’s Advanced Explanation of Benefits requirements.

For emergency services, the prohibition on balance billing should remain intact, and Congress has two viable paths forward to reform the payment dispute process.

Congress could eliminate the federal IDR process altogether, leaving insurers and providers to resolve payment disputes through private arbitration, civil litigation, or state law. This approach would substantially reduce the federal government’s role in determining out-of-network payments, though it presents potential implementation challenges. Alternatively, Congress could retain the federal IDR process but limit that process to emergency services. This approach should be combined with a reasonable limit on awards, changes to how arbitrators are paid, a prompt-pay requirement that penalizes insurers that fail to pay awards on time, and our administrative recommendations that arbitrators receive standardized training, provide written explanations of their decisions, and be subject to routine oversight. Together, these changes would reduce dispute volume, curb outlier awards, and allow network participation decisions to be driven by market considerations rather than incentives created by federal arbitration.

Congress passed the No Surprises Act (NSA)¹ in a well-intentioned attempt to shield patients from surprise out-of-network bills and excessive bills in medical emergencies. Congress created a federal arbitration process—the Independent Dispute Resolution (IDR) process—to determine payments for out-of-network providers. The result is a system that protects patients from surprise bills but has produced far more disputes and substantially higher out-of-network payments than policymakers anticipated, creating incentives that have increased overall health care spending and premiums.

This brief examines the IDR process, its regulatory and litigation history, how it has produced perverse incentives, and the policy options to improve outcomes for Americans. Specifically, this paper proposes that Congress distinguish between elective and emergency services: Patients should remain protected from surprise bills in emergencies, while patients receiving elective services should receive advance notice of all costs so they can make informed choices. It also proposes a set of administrative reforms to provide greater transparency of arbitrators’ decision-making and to minimize the cases that end up in IDR. These changes would help mitigate the negative impact of the IDR process on premiums and network incentives while still maintaining important patient protections.

Balance billing is the practice whereby an insurer does not cover the full amount of an out-of-network bill, and the out-of-network provider bills the patient for the balance. The balance bill is an amount beyond the patient’s typical coinsurance or copay. A balance bill is described colloquially as a “surprise bill” in circumstances in which the patient was not aware that an out-of-network bill would result—in either an emergency situation or for a service provided at an in-network facility.

The NSA aimed to limit surprise balance billing in two situations. First, it protected patients receiving emergency care against balance billing. Second, it prohibited balance billing when a patient scheduled care from an in-network provider at an in-network facility but then received a balance bill from an out-of-network provider (e.g., an anesthesiologist).

Balance Billing Before the NSA

Between 2014 and 2016, roughly 15 percent of admissions at in-network inpatient facilities resulted in out-of-network bills, according to the Department of Health and Human Services.² A report by Charm Economics estimates that in 2021, there were around 5.9 million commercially insured emergency department visits and 1.7 million commercially insured inpatient stays that included at least one out-of-network charge.³ Another study found that in 2014, “20 percent of hospital inpatient admissions that originated in the emergency department (ED), 14 percent of outpatient visits to the ED, and 9 percent of elective inpatient admissions likely led to a surprise medical bill.”⁴ Not all of these out-of-network bills may have been surprise bills that would have fallen under NSA protections, but these estimates provide context for the frequency of unexpected out-of-network charges prior to the NSA.

Prior to the implementation of the NSA, studies found that the magnitude of likely balance bills involving in-network facilities and out-of-network physicians averaged around $623.⁵ A 2021 study by Biener et al. found that privately insured patients who likely received surprise out-of-network bills for ED visits between 2001 and 2016 paid an estimated 30.9 percent of the charged amount.⁶ The study found that in likely balance billing situations, 29 percent of patients paid more than $100 out of pocket to physicians, while a little more than 9 percent paid more than $400.

The Impact of the ACA

The Affordable Care Act (ACA) attempted to reduce balance bill amounts for emergency services by requiring insurers to pay a “reasonable amount” to out-of-network emergency providers.⁷ The ACA defined reasonable amount as the greatest of the Medicare rate, the median in-network rate, or the typical out-of-network rate.⁸ The goal of the “reasonable amount” requirement was to prevent insurers—who were limited in how much cost sharing they could impose on enrollees—from paying providers “unreasonably low” amounts and having enrollees pick up the rest. Such a practice would functionally be an end-run around cost-sharing protections for enrollees.

As shown in Figure 1, the Biener et al. (2021) study, which examined pre- and post-ACA payments for ED visits that likely had surprise bills, detected a significant drop in the percentage of charges paid for emergency services and a drop in the share of ED visits where total charges were paid in full.⁹

These drops corresponded with increased total payments to physicians from both insurers and patients, because even though the paid share of charges fell, total charges from physicians more than doubled.

Market Dynamics Before the NSA

Balance billing was considered one form of leverage for providers in negotiating higher in-network rates with plans, but this leverage was capped by patients’ refusal to pay these charges in full.

A 2020 study argues that a provider who is unlikely to lose patient volume by remaining out of network—such as an anesthesiologist or other provider who is not chosen by patients—could use the threat of remaining out of network to increase his or her in-network rate.¹⁰ The study found that in 2015, median out-of-network charges (the amount billed but not necessarily paid) were roughly twice the amount of median in-network payments.

However, Biener et al.’s data show that patients rarely paid balance bills in full. Because providers frequently collected only a fraction of balance bills, remaining out of network did not guarantee that providers would realize their full billed charges. This limited, to some extent, the provider’s network negotiation leverage associated with remaining out of network. Patients who were not paying the full amounts of the balance bills would be less likely to switch to insurers with wider networks.

Previous State Laws

Before the enactment of the NSA, several states had protections against surprise medical bills. According to the Commonwealth Fund’s survey of state balance billing laws as of February 2021, 18 states had enacted what it classified as “comprehensive” surprise billing protections, while an additional 15 states had adopted partial protections.¹¹ The Commonwealth Fund defined comprehensive protections as laws that shield patients from balance bills in both emergency settings and out-of-network care received at in-network facilities, apply broadly across major insurance products, prohibit providers from billing patients beyond normal in-network cost sharing, and establish payment standards or dispute resolution mechanisms to resolve insurer-provider payment disputes. By contrast, states with partial protections generally addressed only some surprise billing scenarios, dealt with a narrower set of insurance products or providers, or lacked a complete framework for determining out-of-network payment amounts.

State laws were limited by the federal Employee Retirement Income Security Act (ERISA), which preempts state regulation of self-funded employer health plans. This matters because ERISA-regulated plans cover most employees who have employer-sponsored health insurance. As such, state surprise billing protections generally could not reach a large share of privately insured Americans covered through employer-sponsored plans. These limitations helped drive congressional support for a federal solution, culminating in the NSA, which extended protections to self-funded plans and established a nationwide baseline against surprise medical billing.

Prior to the NSA, a patient who underwent scheduled surgery performed by an in-network surgeon at an in-network facility but received services from an out-of-network anesthesiologist could be billed for the out-of-network anesthesiologist’s services. Patients reasonably expect that if they go to in-network facilities, all the providers within those facilities are also in-network.

By passing the NSA, Congress prohibited balance billing in such situations as well as for emergency services at out-of-network facilities and created the IDR process to mediate payment disputes between insurers and providers. Under the IDR process, the two parties each make one offer, and an independent arbitrator picks one—theoretically incentivizing each side to make reasonable offers.

The arbitration process may begin after an initial 30-business-day open negotiation period during which the parties attempt to settle the dispute privately. Parties have three days to agree on an arbitrator. If the parties reject each other’s choices, the Departments of Health and Human Services, Labor, and Treasury (the Departments) will randomly assign one from the remaining arbitrators who were not rejected. The arbitrator first determines whether the dispute is eligible for the IDR process. If it is, each party submits an offer and supporting documentation, and the arbitrator chooses between the two offers. After the arbitrator chooses an offer, the insurer must pay within 30 days. A party that initiates a dispute then enters a 90-day cooling-off period, during which they cannot submit a dispute for the same item or service with the same party, though they can hold onto those claims and submit them as disputes after the 90 days are up.

Arbitrators are certified by the Departments for five-year contracts. The arbitrator is paid a fee (separate from the administration fee paid by both parties) by the losing party after choosing an offer. Fees are charged per dispute, and a dispute may contain multiple services. Fees averaged around $596 in 2025 for emergency and non-emergency disputes, up from $482 in 2023.

It is worth emphasizing that before the NSA, most insurers paid some amount to out-of-network providers, because enrollees demanded some out-of-network coverage to reduce balance bills. The NSA removed the necessity of this partial payment. As such, the arbitration process was intended to reasonably compensate out-of-network providers for the services they provided.

The statute requires an arbitrator to consider the Qualifying Payment Amount (QPA)—which the law defines as the median in-network rate for the relevant item or service in the relevant market as of January 31, 2019, adjusted for inflation—alongside a list of other factors, including provider training and experience, patient acuity, teaching status, case complexity, prior contracted rates, and good-faith contracting efforts. The statute, however, does not specify how much weight the arbitrator should apply to each factor, nor does it instruct parties on how they should calculate their own offers.

The Relevance of the QPA

The Congressional Budget Office (CBO) considered the QPA the most important factor in the IDR process for estimating the budgetary effect of the law—even though it was not made an official benchmark in the NSA. CBO assumed that payments would migrate over time toward the QPA and that by reflecting median in-network rates, the QPA would exert downward pressure on out-of-network payments and eventually on in-network negotiations as well.

CBO’s score for the NSA projected lower provider prices that would reduce premiums by roughly 1 percent and reduce federal deficits by $16.8 billion over the 2021–2030 period. Specifically, CBO assumed that the reduction in prices would lead to lower-than-otherwise premiums and higher taxable wages for people with employer-sponsored plans and reduce expected federal premium subsidies for people with ACA exchange plans.¹²

Based on CBO’s logic, if the average final award remained near the QPA, providers would have an incentive to join networks and to negotiate larger-than-median rates—given that the average in-network rate was 15–20 percent higher than the median rate.¹³ If awards moved well above the QPA, the patient protection from surprise billing would remain, but the incentive to stay in network would disappear, as would the savings projected in the bill. This is not just because out-of-network providers would be paid more but also because out-of-network prices influence in-network prices. The more likely it becomes for a provider to receive a high out-of-network rate through the IDR process, the more leverage a provider has when negotiating in-network rates with insurers. As noted by CBO, providers that can “credibly threaten to stay out-of-network … can bargain for higher prices” for in-network services.¹⁴

The QPA may not be truly representative of pre-NSA in-network rates. Analyzing 2023 data, a Brookings study found that QPAs were significantly lower (more than 30 percent) than estimated pre-NSA average in-network prices. The authors speculate that this may be partially due to medians being inherently lower than averages but also possibly because insurers may not adhere to QPA calculation requirements. There may also be some sort of selection bias for cases with low QPAs.¹⁵

The Departments under the Biden administration attempted to issue rules centering arbitration decisions around the QPA, but the courts rejected these rules. In October 2021, the Departments issued an interim final rule creating a “rebuttable presumption that the QPA is the appropriate payment amount.”¹⁶ Providers sued, and the Eastern District of Texas held that the presumption conflicted with the statute’s direction that arbitrators consider not only the QPA but also the additional circumstances Congress listed, and it vacated the provisions that made the QPA the presumptive amount.¹⁷

As a result of this ruling, the Departments promulgated a final rule in August 2022 that dropped the formal rebuttable presumption, but it still required arbitrators to ground their decisions around the QPA.¹⁸ Providers sued again, and eventually the Fifth Circuit Court of Appeals rejected this structure.¹⁹ The court reasoned that Congress told arbitrators to consider all of the statutory factors and gave no instruction that one factor had to come first or that a factor should be discounted because some aspect of it might already be embedded in the QPA.²⁰ Since then, the QPA has been only one of multiple factors that arbitrators consider.

The Departments’ latest final rule, released on May 28, 2026, focuses on streamlining communication among payers, providers, and arbitration firms, and it makes batching services under dispute more flexible, but it limits batches to 50 qualified IDR items. (Our analysis shows that only 0.13 percent of disputes contained more than 50 line items over 2023-2025.)²¹ The new rule also requires plans and issuers to use specified codes when sending remittance advice to noncontracted providers, clarifies when the open-negotiation clock starts, requires parties to communicate during the open negotiation period, and reduces the administrative fee from $115 to $15 per party per dispute regardless of the amount in dispute or its eligibility.²²

To understand how the IDR process has developed and assess its outcomes, we examined the Centers for Medicare and Medicaid Services (CMS) federal IDR public use files, CMS’s federal IDR supplemental background reports and tables, other public federal data, and independent reports from industry. We have 11 key findings.

Finding 1: Total dispute volume is much higher than expected.

The IDR process has led to far more cases in arbitration than the federal government expected. Figure 2 below compares the federal government’s projected number of initiated disputes to the actual number of initiated disputes. From April to December 2022, the number of initiated disputes was nine times the federal government’s projection for the entire year. By 2025, the number of initiated disputes was 115 times the government’s projection.

Finding 2: Total dispute volume grew rapidly from 2023 through 2025.

Table 1 shows that the total disputed line items—meaning individual services that make up a dispute (a single dispute may contain multiple line items)—that were resolved grew about 17-fold from 2023 to 2025 (the latest data available). Even excluding disputes decided by default, disputed line items grew more than 16-fold.

Given that there was a significant backlog of unresolved disputes in the first two years, some of these resolved disputes were initiated in different years than when they were resolved. However, the same pattern is present when looking at dispute initiations only. CMS reports 1,372,563 disputes initiated²³ in the last six months of 2025,²⁴ which was 61 percent more than in the last six months of 2024 (853,374 disputes).²⁵ The CMS data in that 2025 report represent only total disputes initiated, not line items of services being disputed. CMS reported that providers or their representatives initiated 76 percent of those disputes, and the top 10 initiating parties accounted for about 66 percent of all filings.²⁶

Finding 3: Providers win the vast majority of the time, with win rates increasing from 2023 through 2025.

Table 2 shows that in 2025, providers prevailed in 86.4 percent of all disputed line items when default decisions are included and in 85.2 percent of line items when default decisions are excluded. Provider win rates increased considerably from 2023 to 2025: by 8.3 percentage points for all line items and by 3.1 percentage points for merit-only (decisions excluding defaults) line items.

Figure 3 breaks down the highest provider win rates at the line-item level by provider types with more than 10,000 disputed line items. From 2023 to 2025, most of these provider types prevailed in 80 percent or more of disputed line items, and win rates improved or stayed the same for all but two. For context, the provider types included in Figure 3 represented 53.2 percent of all resolved merit line items from 2023 to 2025. Emergency medicine and radiology alone made up 48.0 percent of resolved merit line items.

Finding 4: Arbitration awards are much higher than expected and growing.

CBO initially assumed that offers from disputing parties—and thus awards—would converge toward the QPA. That has not happened. Table 3 shows that in 2025, the median award was 3.9 times the QPA, while the mean award was 7.6 times the QPA. The award was higher than the QPA in 91.6 percent of line items. Notably, the median award was also about 5.5 times the Medicare rate for the same service.

Multiples of Medicare rates are a useful comparison to IDR awards, and commercial rates are typically two to three times Medicare rates.²⁷ In part because Medicare enrollees represent a significant proportion of total patient volume, the vast majority of providers treat Medicare patients and accept Medicare rates as payment in full. Among the individual clinicians named in IDR disputes determined in 2024, more than three-quarters billed Medicare in 2024. Our analysis found that 94.1 percent of emergency physicians and 98.9 percent of radiologists billed Medicare in 2024.

Figure 4 shows how median awards have grown relative to Medicare rates for the same services. In the first half of 2023, median awards were just over four times Medicare rates, while in the second half of 2025, median awards were nearly six times Medicare rates. Average awards were nearly 10 times Medicare rates in the second half of 2025.

Finding 5: Across provider types, win rates are consistently high while awards vary widely.

Table 4 shows both the overall median award amounts relative to the QPA and the win rates for the 21 provider types with more than 10,000 disputed line items from 2023 to 2025.

Merit win rates are consistently high across all of these provider types, with slight variation around an average of about 84 percent. Of the 21 provider types with the highest disputed line-item volume, 18 have merit win rates between 80 and 93 percent.²⁸

The median award relative to the QPA varies greatly across these provider types. Despite having the highest volume of disputed line items, emergency medicine is not even in the top 15 highest median prevailing awards relative to the QPA, with a median award of 3.5 times the QPA. Orthopedic surgeons, by comparison, receive median awards that are 22.6 times the QPA. Neurological surgeons receive over 28 times the QPA, physician assistants nearly 26 times the QPA, and technologists over 23 times the QPA.

Finding 6: Providers’ offers have increased over time, along with award amounts.

As Figure 5 (originally constructed by the Niskanen Center²⁹ and adapted by Paragon) shows, the median award and median provider offer generally increased in tandem from 2023 to 2025. In Q1 2023, the median provider offer on a given line item was 343 percent of the QPA, and the median award was 269 percent of the QPA. Median awards as a percentage of the QPA peaked at 405 percent in Q4 2024 before dipping to 388 percent by Q4 2025. Meanwhile, after settling around 440 to 455 percent of the QPA for about a year and a half, provider offers jumped to 484 percent of the QPA in 2025. In contrast, insurers’ median offers were consistently at the QPA over this period.

Finding 7: The top 10 percent of award amounts have significantly escalated over time.

While the median award has increased as a percentage of the QPA, the largest increases have occurred at the upper end of the award distribution. Figure 6 demonstrates that the top 10 percent of awards as a percentage of the QPA surged after Q3 2024, leading to significantly higher mean awards even as the median award rate flattened out. In Q1 2023, the 90th percentile of awards was 7.8 times the QPA. In Q4 2025, it was 17.7 times the QPA.

Figure 7 shows the distribution of awards by multiple of the QPA. Roughly 62 percent of awards are at or below five times the QPA, while one in six awards exceeds 10 times the QPA. This long tail is what pulls the mean award far above the median.

Finding 8: Insurers lost the vast majority of the time even if they significantly increased their offers.

Figure 8 compares insurers’ mean offers to their win rates from 2023 to 2025. Notably, even as mean insurer offer rates increased, insurer win rates decreased. This analysis separated Cigna out, because while most insurers’ mean offers increased from around the QPA to only a little over 1.5 times the QPA from Q1 2023 to Q4 2025, Cigna attempted a very different strategy and significantly raised its offers before eventually lowering them.

Unlike other insurers, this is true for both Cigna’s mean and median offers.³⁰ As Figure 8 shows, Cigna’s mean offers significantly increased before lowering, while Table 5 shows that its median offers also increased nearly every quarter before declining in 2025. Cigna’s substantially higher offers did not translate into a meaningfully higher win rate: Its win rate generally declined during the period in which its offers rose the most.

Figure 8 also shows that while Cigna’s strategy did yield a higher win rate compared to other insurers, Cigna is still losing the vast majority of its disputed line items.

Figure 9 shows that in disputed line items decided by the seven highest-volume arbitration firms—representing roughly 76 percent of all dispute volume and 79 percent of all line items—insurers still had low win rates even when offering six times the QPA or more. Even the most favorable of these arbitration firms, Federal Hearings and Appeals Services, sides with providers 66.4 percent of the time when the insurer offers 600 percent or more of the QPA.

Island Peer Review Organization (which handled the fourth-largest number of disputes in Q3 and Q4 2025) actually had higher provider win rates when insurers offered 600 percent or more of the QPA. The company had a provider win rate of 93.3 percent when the insurer offered 100 percent of the QPA—and a 98.3 percent provider win rate for insurer offers of 600 percent or more of the QPA.

Finding 9: Arbitrators with higher provider win rates handle higher volumes of disputes.

In an analysis of arbitration firms’ dispute volumes and win rates, we found a correlation coefficient of 0.53³¹—meaning there is a moderately strong positive association between the volume of disputes a firm decides and the provider win rates at that firm (see Figure 10).

Figure 11 shows that the highest-volume arbitration firms had provider win rates above 80 percent. We found that 52.5 percent of total dispute volume was decided by firms with provider win rates above 88 percent.

Finding 10: Dispute filings are highly concentrated, driven by a mix of provider organizations backed by private equity and third-party filers.

CMS reported that providers, health care facilities, or their representatives initiated more than 99 percent of disputes in the second half of 2025.³² Most filings come from two types of high-volume filers: large physician groups with the backing of private equity and third-party filers that file disputes on behalf of providers and are frequently used by smaller providers. Figure 12 shows that four organizations filed about half of all provider-initiated disputes determined in 2025, and the top 10 filed about 70 percent.

The highest-IDR volume filing organization by far is HaloMD, a company that exists solely to manage dispute resolutions, including filing IDR disputes.³³ The other high-volume third-party filing organizations shown in Figure 12 consist of medical billing companies that file IDR disputes alongside other services.

As shown in Table 6, the highest-IDR volume provider organizations are dominated by corporate- and private-equity-backed groups. The top 10 accounted for about 45 percent of all provider-initiated disputes receiving determinations in 2025. Five of these organizations had evidence of private equity backing,³⁴ and two others of corporate backing.

Finding 11: Emergency medicine and radiology are the top two provider types involved in disputes.

Unsurprisingly, certain types of providers have used and benefited from the IDR process the most. Table 7 shows that emergency medicine, radiology, and anesthesiology ranked first, second, and fifth, respectively, among provider types with the highest line-item volume. (The third and fourth were clinics/centers and general acute care hospitals.) Those three specialties had respective line-item merit win rates of 85.9 percent, 92.7 percent, and 82.6 percent.

Figure 13 shows how disputed line-item volume has risen over time among provider types with the highest merit win rates.³⁵ Disputed line items in emergency medicine—a primary target for the NSA—have grown significantly, as have those in radiology, both reaching their highest in Q2 2025. The downturn of resolved disputed line items for these provider types in the last half of 2025 primarily reflects a reduction in backlogged disputes and contrasts with an increase in overall dispute initiations in both halves of 2025.³⁶ According to CMS, dispute initiations increased 16 percent in the last six months of 2025 (1.37 million disputes) compared to the first six months of 2025 (1.19 million disputes).

The findings point to a fundamental problem with the IDR process: Rather than serving as a backstop for resolving unusual out-of-network payment disputes, IDR is increasingly functioning as an alternative payment system. Dispute volume has grown dramatically, providers prevail in the large majority of cases, and awards substantially exceed both the QPA and Medicare payment rates. These outcomes incentivize arbitration rather than negotiating network contracts—with Americans absorbing the costs through higher premiums and lower wages.

Despite the rapid growth of IDR, only a fraction of potentially eligible claims currently enter arbitration. According to a survey of insurers by America’s Health Insurance Plans (AHIP) and the BlueCross BlueShield Association, only 6 percent of the 19.7 million total qualified IDR claims entered arbitration.³⁷ But the findings above show that dispute volume has grown rapidly, as have provider offers and awards. With providers prevailing in the large majority of disputes and receiving awards well above the QPA, there is substantial financial incentive for providers to bring more claims into IDR. As such, the already significant costs associated with the process could increase considerably even if only a modestly larger share of claims enter arbitration.

Why Are Awards So High?

One of the fundamental problems with IDR outcomes is that awards frequently bear little resemblance to conventional benchmarks for either in-network or historical out-of-network payments.

Arbitrators argue that the QPA should be “treated more cautiously with respect to [out-of-network] bills, which are the sole focus of the NSA,” and that awards should not be judged based on their relationship to the QPA.³⁸ Even granting that pre-NSA out-of-network rates are a better market reference point than the QPA, the awards being given do not appear to reflect anything resembling pre-NSA out-of-network amounts.

Out-of-network charges for some services have long exceeded in-network payment rates. However, as shown by Biener et al. (2021), out-of-network charged rates were usually not paid in full. Therefore, actual pre-NSA payments to out-of-network providers were likely considerably lower than the median IDR award.

The arbitrators’ stance also contradicts the legislative text, which states that the arbitrator “shall consider” the QPA (explicitly based on in-network rates) in decision-making.³⁹ This requirement is given its own section in the statute, followed by a second section containing all the other factors for arbitrators to consider. It is difficult to read the text of the law and not conclude that Congress intended for the QPA to be a significant input in IDR decisions. As the Niskanen Center notes, Congress “explicitly barr[ed] arbitrators from considering usual and customary charges or the amount a provider would have billed absent the law’s protections, the closest analogues to the historical out-of-network payments arbitrators now cite to justify these awards.”⁴⁰

The Direct Costs of IDR

While the NSA protects patients from surprise bills, most Americans nonetheless face higher costs, because the IDR process is increasing overall spending and premiums. The Center on Health Insurance Reforms (CHIR) estimated that the IDR process generated $22.4 billion in total costs through 2025—a substantial jump from an estimated $5 billion through 2024—through a mix of administrative fees and increased plan expenditures (which account for the majority of costs and show up in higher employer spending, premiums, and deductibles).⁴¹ And as disputes increased in 2025, systemwide costs did as well. CHIR estimated that total IDR costs were $16.6 billion in 2025, nearly 3.5 times as high as in 2024.⁴² On Elevance’s Q2 2025 earnings call, the chief financial officer stated that IDR costs accounted for “about 30 percent” of the cost increase in its ACA plans.⁴³ Additionally, UnitedHealthcare Insurance Company of New York reported that federal and New York state IDR program costs added 0.8 percentage points to its requested premium increase for small group plans.⁴⁴ In another example, the city manager of San Antonio primarily attributed a $40 million budget overrun in 2026 to costs incurred in the IDR process.⁴⁵

Current trends in IDR award amounts and volumes suggest that overall plan expenditures on IDR awards (totaling $2.24 billion in 2023 and 2024)⁴⁶ will continue to increase. A recent rule from the Departments that reduced the federal administrative fee by nearly 87 percent—and thus the cost of initiating IDR—will likely increase disputes by 30 percent.⁴⁷

IDR Is Likely Raising In-Network Prices

The largest economic effect of the NSA likely occurs outside of the arbitration process itself. Providers can remain out of network and pursue IDR awards or use the prospect of those awards to demand higher rates during contract negotiations. Either response puts upward pressure on health care prices. This mechanism is particularly important because CBO’s original estimate of savings from the NSA depended on precisely the opposite occurring: Out-of-network payments were expected to move toward the QPA, strengthening providers’ incentives to contract and exerting downward pressure on in-network rates.

The Brookings Institution estimates that pre-NSA in-network prices for emergency care, imaging services, and neonatal/pediatric critical care services were all significantly lower than the mean IDR decisions for those services.⁴⁸ The high IDR award amounts incentivize providers to stay out of network and enable them to demand higher in-network rates. CBO has recently called for more research on the impact of the NSA and the IDR process on prices and insurance networks.⁴⁹

A new paper published by Barwick et al. found that the NSA “reduced provider network participation in the specialties and states most affected by the reform.”⁵⁰ Barwick et al. found that relative to states with comparable protections and specialties not covered by the law, in-network participation declined by 4.7 percentage points for emergency medicine, 3.4 percentage points for anesthesiology, and 1.8 percentage points for radiology.

The results of the IDR process are best understood as the product of the incentives facing the three principal participants: providers, arbitration firms, and insurers. Providers benefit from filing an increasing volume of disputes, because they have a high probability of winning awards that substantially exceed contracted rates. Arbitration firms earn more revenue the more disputes they resolve, with little transparency regarding how they evaluate competing offers and statutory factors. Insurers, meanwhile, risk higher in-network rates if they raise their offers and see little improvement in their win rates when they do. Together, these incentives produce a high volume of disputes, high provider win rates, and above-market awards that have made arbitration preferable to contracting.

Provider Incentives: Win More and Win Bigger

Providers’ incentives in arbitration are clear: They want to maximize their payoffs through either arbitration awards or higher in-network rates that they can achieve by threatening to remain out of network. With providers winning 85 percent of disputes decided in the second half of 2025, they have strong incentives to submit more disputes and higher offers.

Rising awards give providers greater incentive to remain out of network and greater leverage to demand higher rates as a condition of joining or remaining in an insurer’s network. Either outcome will increase costs for payers and patients and ultimately cause premiums to rise. As Table 4 demonstrates, the financial incentives for certain provider types are significant, with median awards for multiple categories exceeding 20 times the QPA.

Importantly, before the NSA, remaining out of network carried real financial risk. Providers had to weigh a larger pool of in-network patients against a lower in-network rate. Out-of-network providers could balance bill patients, but patients frequently paid only a fraction of those charges. This collection risk operated as a practical ceiling on out-of-network billing and was an incentive to contract. IDR has replaced uncertain collections from patients with a federally administered payment mechanism that can generate near-guaranteed awards far above typical in-network rates. For certain provider types, this shift makes remaining out of network much more financially attractive than it was before the NSA.

IDR Entity Incentives: Resolve More Disputes

Conflicts of Interest Exist in the Eligibility Decision Process

Disputes involving ineligible claims may be inflating IDR volume and costs. Arbitrators are not paid until they award disputes. If a claim is deemed ineligible for IDR, the arbitrator is not paid. As such, there is a clear incentive for the arbitrator to declare as many claims as possible eligible for IDR. According to the aforementioned AHIP survey, health plans estimate that around 40 percent of initiated disputes are likely ineligible on either procedural grounds (i.e., the initiating party did not follow statutory or regulatory procedure when filing a dispute) or substantive grounds (i.e., the service in question does not qualify under the NSA). CMS reports, “Noninitiating parties challenged the eligibility of 42% of initiated disputes in the last six months of 2025 (574,128 of 1,372,563), a slight increase from 40% in the first six months of 2025.”⁵¹ These are only challenges, however, and are not the official count of ineligible disputes.

Arbitrators point to CMS data showing that only 19 percent of disputes were found ineligible in the second half of 2025.⁵² However, that data actually comes from the arbitrators themselves.⁵³ CMS does note that official ineligible disputes have declined from 69 percent of all disputes in the first half of 2022. Still, so long as arbitrators are paid upon deciding a dispute, the incentive to allow otherwise-ineligible disputes to go forward remains, and it leads to higher costs—particularly given the high win rate and award levels.

Arbitration Firms Profit from More Disputes

Arbitration firms earn more revenue as they resolve more disputes, and providers initiate nearly all disputes. As such, there is a potential incentive problem that warrants scrutiny, particularly given the relationship between provider win rates and dispute volume across arbitration firms. As discussed, there is a moderately strong correlation between arbitration firms’ dispute volumes and win rates. It is also impossible to ignore that over half of dispute volume is handled by arbitration firms with provider win rates above 88 percent.

Insurer Incentives: Control Costs Across the Network

Winning more disputes benefits insurers only if it leads to lower overall costs from the IDR process, which includes any effect on in-network rates. On their own, offers may have to be so high to win that winning more disputes would increase costs—costs that are passed on to enrollees and employer sponsors.

Simultaneously, insurer incentives for cost control are somewhat blunted by both government regulations such as medical loss ratio requirements (though this applies only to fully insured and individual plans) and the business model of insurance. Plans can pass costs along via higher premiums in fully insured or individual plans, and in a self-funded plan the costs are paid directly by the employer. Insurers do risk losing customers to other insurers that are better able to control costs, particularly in the Administrative Services Only market.

Raising the Offer Makes Little Difference

Insurers have been criticized for keeping their median offers at the QPA. But, as the data show, even if they do raise their offers—as in the example of Cigna—it does not lead to a meaningful increase in win rates. Figure 9 shows that at arbitration firms representing roughly 76 percent of all dispute volume, insurers could not meaningfully improve their win rates even when offering six times the QPA or more. Justified or not, the arbitrators repeatedly reject insurer offers of all sizes.

Insurers Have to Consider In-Network Impacts

Arbitration does not exist in a vacuum. If insurers raise their offers above median in-network rates, they risk incentivizing providers who are in network to go out of network—where they can get better rates—if their in-network rates are not raised. Given that the vast majority of claims are either in network or do not go to arbitration, any willingness to cede higher out-of-network rates for what is currently a small portion of claims risks significant cost increases. This is an example of the complicated incentives that an insurer faces in arbitration relative to the very straightforward incentives providers have.

Taken together, these incentives help explain why IDR has evolved so differently from what policymakers expected. Providers have strong incentives to pursue arbitration and seek large awards, IDR entities benefit from resolving more disputes, and insurers may be unable to improve their outcomes without increasing costs elsewhere in their networks. The result is a system in which the incentives increasingly favor arbitration over network contracting.

The NSA successfully protected patients from surprise medical bills. But it attempted to solve a second, fundamentally different problem—how insurers and providers should determine payment—through a government-imposed and -administered arbitration system. That approach has produced negative unintended consequences, including escalating costs and premiums, incentives for providers to remain out of network, and a rapidly expanding dispute resolution system that Congress never envisioned. The growing volume of disputes, rising costs, and distorted contracting incentives demonstrate that Congress should substantially curtail and reform the IDR process while preserving the NSA’s core patient protections.

Patients should continue to be protected from surprise bills that they cannot reasonably anticipate or avoid. Congress should distinguish between situations where patients have meaningful choices and situations where they do not, pursuing a two-track framework that improves incentives for market participants and leads to more appropriate payments.

Elective Services: Restore Consumer Choice and Eliminate Government Price Arbitration

For elective services performed at in-network facilities, Congress should eliminate the federal IDR process. Providers should be required to provide patients with upfront pricing and network information with meaningful advance notice that out-of-network providers will participate in their care and receive affirmative consent to those charges from patients.

A provider should notify a patient at least 48 hours before a scheduled procedure whenever an out-of-network provider is expected to furnish services. The notice should identify the provider, disclose expected charges through an Advanced Explanation of Benefits (AEOB), and provide enough information for the patient to make an informed decision. If adequate notice is not provided, balance billing should remain prohibited.

Congress already anticipated the need for this type of advance notice and price information. The NSA requires providers to submit good faith estimates to health plans so plans can generate AEOBs before elective procedures. The statute generally requires that the estimate be supplied to the plan within one day of scheduling for procedures scheduled three to nine days in advance and within three days for procedures scheduled 10 or more days in advance. The Biden administration never issued rules to implement these provisions, citing the complexity of the data transmission infrastructure.⁵⁴ The Trump administration should complete implementation of these statutory requirements.

A functioning AEOB process would empower patients by providing them the necessary information to make decisions that are right for them while encouraging hospitals to ensure that physicians practicing in their facilities are in network—or charge prices that patients are willing to accept. Facilities that fail to do so would increasingly lose patients to competitors offering predictable, all-inclusive pricing.

This framework would restore market incentives. Provider participation decisions would once again be driven by consumer choice and market competition rather than expectations about an excessively high government arbitration award.

Emergency Services: Preserve Patient Protections While Reforming Payment Disputes

Emergency services require a different framework, because patient choice and market competition cannot function when patients are unable to choose providers or negotiate prices before receiving care. The prohibition on balance billing should therefore remain for emergency services regardless of how Congress reforms payment disputes. Once the patient is held harmless, only the insurer and provider remain in dispute. Congress has two viable paths forward.

Option #1: Eliminate the Federal IDR Process

Congress could preserve the patient protections from the NSA while eliminating the federal arbitration process. Insurers and providers would resolve payment disputes through private arbitration, civil litigation, or state law. This would eliminate federal disputes over eligibility and administrative fees and reduce incentives for excessive arbitration volume while removing the government-created payment framework that encourages providers to remain out of network.

This approach presents several potential implementation challenges. Independent physician groups involved with emergency medicine may have greater incentives to consolidate with hospitals or private-equity-backed organizations to increase bargaining leverage. However, the current IDR process produces those same incentives.

Another implementation issue is that insurers and out-of-network providers generally lack contractual relationships, so Congress would need to determine whether private arbitration should be mandatory or whether disputes should instead proceed through the courts. This approach would represent a substantial improvement over the current system by limiting government’s role to requiring arbitration rather than dictating the rules governing the process.

Option #2: Retain IDR Only for Emergency Services

If Congress maintains a federal arbitration process, it should limit it to emergency services. This approach would protect patients while dramatically reducing dispute volume and eliminating the government’s influence over network participation decisions for elective care. This option should be combined with a reasonable limit on awards, changes to how arbitrators are paid, and our administrative recommendations (below) that arbitrators receive standardized training, provide written explanations of decisions, and be subject to routine oversight by the Departments. The Departments should also strengthen eligibility screening before disputes enter arbitration. Although this option would retain some opportunities for abuse, they would exist on a smaller scale than under the current system.

Changes to the QPA and Awards

Congress should change the definition of the QPA to the most up-to-date median in-network rates rather than merely adjusting the 2019 median in-network rates for inflation. This can be done using data made available under the Transparency in Coverage rule. While there remain gaps in these data, using them to calculate the QPA would provide additional impetus for the Departments to enforce insurer compliance. These data are also publicly auditable, which would mitigate concerns about how insurance companies are calculating their QPAs.

Further, Congress needs to place a reasonable limit on awards. Some groups have proposed anchoring the award to the QPA, although Congress rejected this approach in the NSA. Anchoring the award to the QPA would essentially force a provider to take the median in-network rate, which flips the balance of power heavily in favor of insurers. Congress should instead attempt to create a system that approximates pre-NSA out-of-network payments, when providers faced a practical constraint on extremely high charges because patients were unlikely to pay large balance bills. This could take the form of an upper limit set at a given percentile of a plan’s in-network rates for a specific service. Congress should choose a limit that represents the upper end of a reasonable range of payments for a service without creating large incentives for providers to remain out of network. If the limit is too high, providers would continue to have additional leverage to remain out of network, which would push in-network rates higher. However, such a limit would likely still mitigate the existing trend of increasing awards and contain outlier awards.⁵⁵ Importantly, our proposal would remove a large number of cases from arbitration entirely—leaving the relevance of the IDR process and such a limit applicable only to emergency services.

Changes to Arbitration Fees

Congress should consider making changes to how arbitrators are paid. First, Congress should require that the arbitrator’s fee be split between both parties rather than borne by the losing party. Fees are significantly lower than awards, but splitting the fee would at least reduce the incentive to flood the system with disputes. If the above reforms work as expected and reduce excessively high provider win rates, providers would have to be more selective about which claims they bring to arbitration.

Second, because arbitrators do not get paid unless they decide awards in disputes, the payment structure creates an incentive to declare more claims eligible in order to maintain a high volume of disputes. Congress should change the statute to ensure that arbitrators are paid regardless of whether they decide that disputes are ineligible. Additionally, in the case of an ineligible dispute, the party that filed the dispute should have to pay the entire arbitration fee (rather than splitting it as outlined above). This would further incentivize providers to be more selective about the disputes they bring to arbitration.

Prompt Pay Requirements for Plans

Under the current statute, plans must pay out awards within 30 days.⁵⁶ However, there are no penalties for failure to pay on time. This has resulted in “hundreds of lawsuits” over delayed payments, according to Georgetown Law’s O’Neill Institute.⁵⁷ Congress should create monetary penalties for insurers that fail to pay on time. A prompt pay requirement needs to be combined with reasonable parameters around awards to protect against abusive outliers.

Administrative Fixes

While congressional action is necessary to fix the core IDR structural deficiencies, the Departments can take several steps to mitigate the problems with the arbitration process. Congress could implement these fixes as well, but the Departments have the authority to do so irrespective of congressional action. These include transparency measures and steps to reduce the number of cases.

Require Transparency of Arbitrators’ Decisions

The Departments should require a written explanation of the statutory factors the arbitrator considered and the relative importance of each factor in the decision. This would allow for greater scrutiny of the program by policymakers. The Departments should continue to make decision rates and awards public by individual arbitrator.

Audit Outlier Results and Eligibility Decisions

The Departments should audit arbitration firms producing outlier outcomes, specifically regarding award amounts and provider win rates. Outlier results are not necessarily proof of improper decisions, but they do indicate potential issues. Provider win rates in excess of 90 percent, for example, warrant scrutiny as to whether arbitrators are consistently following statutory requirements around award decisions.

The Departments should also audit arbitrators on the accuracy of their eligibility determinations. Without audits, the Departments cannot know if arbitrators are declaring ineligible claims eligible merely to increase their volume of disputes. By the same token, if—as we propose—arbitrators are paid regardless of whether claims are eligible, an auditing process would ensure that arbitrators are not improperly declaring claims ineligible in order to get paid while avoiding more work.

The Departments have rarely, if ever, conducted audits of arbitrators and have never published findings from any audits they may have done or contracted for. Audits should occur with some regularity rather than once every five years as a product of the recertification process. Additionally, the Departments should publicize the findings of all audits.

Arbitrators Should Face Consequences for Negligence

The Departments have not yet decertified any arbitration firms, in large part because they have not conducted audits. Arbitration firms have functionally faced no consequences if they awarded claims that were not actually eligible or made improper award decisions that ignored statutory requirements. If an audit finds evidence of impropriety or wrongdoing around eligibility or award determinations, the Departments need to sanction or decertify the arbitration firm in question.

Improve Training for Arbitrators

The Departments should develop standardized training for certified arbitration firms that explains the statutory factors Congress requires arbitrators to consider—with a focus on greater consistency in decision-making. Regular continuing education and performance reviews would help reduce unwarranted variation in awards while preserving arbitrator independence.

Prohibit Conflicts of Interest for Arbitrators

The Departments should strengthen existing conflict of interest rules by prohibiting financial or ownership relationships that create actual or perceived conflicts of interest between certified IDR entities and parties participating in the arbitration process—including sharing common investors. Arbitrators should disclose any potential conflicts before accepting disputes, and the Departments should publish those disclosures while prohibiting arbitrators with conflicts of interest from adjudicating cases to strengthen public confidence in the integrity of the process.

Prohibit Forum Shopping

The current system allows parties to direct disputes toward arbitrators they believe are more likely to produce favorable outcomes. The Departments should assign disputes through a randomized allocation process so neither insurers nor providers can strategically select arbitrators based on prior decision patterns.

Restore the Higher Administrative Filing Fee

The Departments should restore a higher filing fee for initiating arbitration. The recent reduction from $115 to $15 substantially lowers the cost of filing disputes while doing little to discourage meritless cases, particularly for sophisticated organizations filing thousands of claims. The Departments estimate that the lower fee will increase disputes filed by 30 percent.⁵⁸ Despite the Departments’ claims that the $115 was “cost-prohibitive” for small and rural providers,⁵⁹ the prevalence of large filing firms submitting batches of disputes on behalf of providers, the number of disputes, and the size of awards suggest that $115 was not a meaningful barrier. A higher fee would help preserve arbitration for disputes where the amount at issue justifies the administrative burden.

The NSA succeeded in protecting patients from surprise bills. However, by creating a process that applied to non-emergency procedures, did not limit award amounts, and had no meaningful guardrails around arbitration firms, Congress unintentionally created incentives for providers to remain out of network, increased administrative costs, and contributed to higher health care spending and insurance premiums. To restore balance, Congress needs to make major reforms to how arbitration works.

For elective services, patients should receive meaningful advance notice, complete price transparency through fully implemented AEOBs, and the opportunity to make informed decisions before receiving care. Once those protections exist, government arbitration is no longer necessary. Market competition—not federal payment determinations—should shape provider participation, negotiated prices, and network formation.

For emergency services, patients should continue to receive complete protection from surprise bills, because meaningful consumer choice is impossible. Congress should either eliminate the federal IDR process and allow insurers and providers to resolve disputes through private legal mechanisms or, alternatively, retain a substantially improved arbitration process limited solely to emergency services.

In addition, Congress needs to place an upper limit on awards to control for outlier payments with no meaningful relationship to market prices. The Departments need to place stronger guardrails around arbitration firms in order to mitigate the large incentives those firms have to increase dispute volume and provider wins.

The objective of these reforms is not to weaken patient protections; it is to restore market incentives wherever market competition can reasonably function while preserving strong protections where it cannot. A two-track framework would better align provider incentives, encourage broader network participation, reduce unnecessary disputes, and moderate long-run health care costs—all while continuing to protect patients from the surprise bills the NSA was intended to eliminate.

Network Impacts

Evidence of network impacts is limited. Knowing if more claims are being made either in or out of network would be helpful in understanding the full impact of the law. A 2026 Government Accountability Office report looked at the percentage change in in-network claims for the four largest specialties most likely to be impacted by surprise billing—emergency medicine, anesthesiology, air ambulance, and radiology—from 2019 to 2023.⁶⁰ It found that the percentage of in-network emergency medicine and anesthesiology claims dropped from 2019 through 2021 before rising in 2022 and 2023, though emergency medicine claims did not return to their 2019 levels, while anesthesiology claims did. The percentage of in-network radiology claims was flat for the entire period, while the percentage of in-network air ambulance claims fluctuated but ultimately rose between 2019 and 2023. Barwick et al. (2026) find reduced network participation by providers most affected by arbitration, though this only applies to three types of providers. More research is needed into how network participation by other provider types has been impacted by the NSA. Barwick et al. also find indirect evidence of premium increases but do not link those premium increases directly to increased network rates.⁶¹ More research is also needed on the pricing impacts of the decreased network participation that has resulted from the NSA.