No Surprises Act independent dispute resolution outcomes for emergency services

Abstract

The No Surprises Act banned surprise billing and established a final-offer arbitration system, independent dispute resolution (IDR), to resolve disagreements between health plans and providers. One factor that arbiters must consider in the IDR process is the qualifying payment amount (QPA), the median contracted rate for the same or similar service in the same market as computed by health plans. We analyzed public IDR data from 2023 for the most common disputed professional service: evaluation and management of a moderate to severe emergency medicine visit. Providers won 86% of cases, with mean decisions 2.7 times the QPA. Private equity-backed providers won more often and higher monetary awards than other providers. The mean QPA was 2.4 times Medicare payments. Disputes were dominated by a small group of health plans and providers, so payments may not reflect the overall market for emergency services.

Introduction

The No Surprises Act (NSA) banned surprise billing and established a final-offer arbitration system, independent dispute resolution (IDR), to resolve payment disagreements between health plans and out-of-network (OON) providers. Arbiters must choose either the health plan's offer or the provider's offer. They must consider several factors when choosing a winning offer, including a metric computed by health plans as the median of contracted rates for the same or similar service in the given market: the qualifying payment amount (QPA). Arbiters must also consider clinical case details, past contracts between the health plan and provider (if applicable), as well as their market shares. We analyzed public IDR data from 2023 for the most common disputed professional service, moderate to severe emergency medicine visit, and Current Procedural Terminology (CPT) code 99 284.

We found that providers won the vast majority of cases, with decisions averaging 2.65 times the relevant QPA (Figure 1). This finding appears driven by private equity (PE)-backed physician staffing companies winning 90% of their disputes vs just 39% for other emergency physician groups, generating an average IDR payment 63% higher relative to the QPA than non-PE groups (Table 1).

Figure 1.

Ratio of arbitration outcomes to QPAs by winning party. Source: Authors’ analysis of CMS IDR Data (non-monetary file). Abbreviations: CMS, Centers for Medicare & Medicaid Services; IDR, independent dispute resolution; QPA, qualifying payment amount.

Table 1.

Dispute descriptive statistics, total, and by PE investment status.

	Non-private equity		Private equity		Total
Dispute lines	3293		43 612		46 905
Ratio provider offer to QPA	2.79		2.89		2.89
Ratio health plan offer to QPA	0.89		1.34		1.31
Ratio winning offer to QPA	1.67		2.73		2.65
Both offers >QPA
No	2654	80.60%	16 160	37.05%	18 814	40.11%
Yes	639	19.40%	27 452	62.95%	28 091	59.89%
Outcome
Health plan wins	2008	60.98%	4354	9.98%	6362	13.56%
Provider wins	1285	39.02%	39 234	89.96%	40 519	86.39%
Split decision	0	0.00%	24	0.06%	24	0.05%

Source: Authors’ analysis of CMS IDR Data for 2023 (non-monetary file) merged with National Uniform Claim Committee's Health Care Provider Taxonomy Code Set and CMS National Plan & Provider Enumeration System NPI files by NPI and name/affiliation text matching.

Disputes are for professional clinician claims that were not decided by a default decision and exclude cases with negative winning offer to QPA ratios.

Abbreviation: QPA, qualifying payment amount.

Data and methods

The primary data for this study are the IDR reports published by the Centers for Medicare & Medicaid Services (CMS).¹ Detailed public use files are available for 2023, presented in quarterly spreadsheets. We use data from 2 spreadsheet files: (1) detailed observations describing each dispute line for non-air-ambulance cases, including the identities of the provider and health plan and offer amounts as a percent of the QPA or certain medians (non-monetary file), and (2) detailed information about the monetary amounts in each dispute (monetary file), but without the identities of the provider or health plan.

We supplemented these data with the National Uniform Claim Committee's Health Care Provider Taxonomy Code Set and CMS's National Plan & Provider Enumeration System National Provider Identifier (NPI) files to refine our identification of providers and distinguish disputes involving facilities and disputes involving professionals.^2,3 In the non-monetary data, in which providers can be identified, we also distinguish professionals associated with PE-backed staffing companies vs those who are not.

In the monetary data, providers cannot be identified, so we cannot directly distinguish disputes involving facilities from those involving professionals. However, emergency facility payments are typically much higher than emergency professional payments. Therefore, we use the facility/professional distribution from the non-monetary data to create a proxy facility identifier in the monetary data at the CPT level based on the QPA distribution and where the dispute's QPA falls within the distribution. For example, for procedure 99 284, 10.2% of disputes involve facilities, so we designate disputes with a QPA exceeding the 89.8th percentile as facility disputes in the monetary data and exclude these from most of our analyses.

We calculate various descriptive statistics for procedure 99 284, the most common emergency evaluation and management code, focusing on disputes involving professionals resolved through IDR arbitration (ie, that did not involve a default judgment). Additional data processing steps and descriptive statistics for other procedures can be found in Appendix S1.

Results

For CPT 99 284, there were 47 897 dispute lines in the monetary file and 46 899 dispute lines in the non-monetary file that were not default decisions and involved professional services. We also excluded dispute lines from the monetary data in which the winning offer was ≤0, the QPA was ≤0, or the winning offer was masked. In the non-monetary data, we excluded dispute lines in which the winning offer to QPA ratio was ≤0. The provider's offer was selected in 86.4% of disputes and the mean winning offer was 2.65 times the QPA (Figure 1 and Table 1). At the mean, providers’ offers were 2.89 times the QPA and health plans’ offers were 1.31 times the QPA. In the majority (59.9%) of disputes, both the provider's offer and the health plan's offer were greater than the QPA.

Variation in IDR decisions and QPAs

We observed significant variance in decisions relative to the QPA (Figure 1). We also find that 6.3% of QPAs in our analytic sample are below the Medicare-allowed amount (Figure 2). The median ratio of the QPA to the Medicare-allowed amount was 1.5 and the mean was 2.4, adjusting for geographic market. This is lower than the mean in-network commercial payments for emergency medicine services reported in prior literature, which range from 2.5 to 2.7 times Medicare.^4-6

Figure 2.

Ratio of QPA to Medicare. Source: Authors’ analysis of CMS IDR data for 2023 (monetary file) merged with Medicare fee schedule by geographic market. Note: There are 4494 dispute lines with ratio values > 4.0 that are not shown. Horizontal axis labels represent the lower limit of the bar range. Abbreviations: CMS, Centers for Medicare & Medicaid Services; IDR, independent dispute resolution; QPA, qualifying payment amount.

Quarterly trends in offers and QPA

We evaluated the median health plan offer, provider offer, QPA, and winning offer for 99 284 by quarter of 2023. Over the course of 2023, median QPAs increased from $179 in quarter 1 to $237 in quarter 4 (Figure 3). Median provider offers decreased from $581 to $538 between the first and fourth quarters. Health plan offers increased from $179 in quarter 1 to $316 in quarter 4, at the median. Median winning offers fell slightly from $560 in the first quarter to $536 in the last quarter and tracked very closely with median provider offers. The share of disputes won by providers increased from 82.82% in the first quarter to 91.19% in the fourth quarter (Table S5).

Figure 3.

Quarterly median health plan offer, provider offer, winning offer, and QPA. Source: Authors’ analysis of CMS IDR data for 2023 (monetary file). Abbreviations: CMS, Centers for Medicare & Medicaid Services; IDR, independent dispute resolution; QPA, qualifying payment amount.

Providers and health plans

Private equity-backed provider groups dominated the IDR process in case volume, win rates, and winning amounts in 2023 (Figures 1 and 4). Only 8% of professional 99 284 dispute lines involved a provider group without a majority ownership by PE investment (USACS, an organization that formerly had PE backing, accounts for 1% of dispute lines). TeamHealth, a PE-backed provider organization, accounted for 54% of dispute lines. Other PE-backed provider organizations that accounted for dispute lines are SCP Health (28% of dispute lines), Envision (5%), American Physician Partners (3%), and Sound Physicians (3%) (Figure 4). At the mean, PE-backed provider organizations have higher offers relative to the QPA (2.89 times QPA) than non-PE provider organizations (2.79 times QPA) in IDR (Table 1). Mean health plan offers relative to the QPA are also higher in IDR disputes involving a PE-backed provider organization (1.34 times QPA) than in disputes involving a provider organization without PE investment (0.89 times QPA).

Figure 4.

Number of dispute lines by health plan and emergency medicine provider organization. Source: Authors’ analysis of CMS IDR Data for 2023 (non-monetary file) merged with National Uniform Claim Committee's Health Care Provider Taxonomy Code Set and CMS National Plan & Provider Enumeration System NPI files by NPI and name/affiliation text matching. Abbreviations: BCBS, blue cross blue shield; CMS, Centers for Medicare & Medicaid Services; IDR, independent dispute resolution; NPI, National Provider Identifier; PE, private equity; QPA, qualifying payment amount; UHC, UnitedHealthcare.

UnitedHealthcare (UHC) is the most frequently involved health plan, accounting for 55% of dispute lines, with cases also involving Cigna (12%), Anthem (9%), Aetna (7%), and non-Anthem Blues plans (TN: 7%, TX: 3%, FL: 3%, other: 2%) (Figure 4). Three-quarters of the dispute lines are attributable to just 6 specific dyads of health plan and provider organizations (Table 2). Disputes between TeamHealth and UHC comprise 46.2% of dispute lines, and disputes between TeamHealth and Blue Cross Blue Shield of Tenessee (BCBS TN) account for 5.3% of dispute lines. SCP Health disputing with Cigna (8.2%), Anthem (7.4%), UHC (5.7%), and BCBS FL (2.4%) comprises 23.7% of dispute lines.

Table 2.

Dispute lines by dyads of health plan and emergency medicine provider organization.

Health plan parent	Provider parent	Number of dispute lines	Percent of dispute lines
UHC	TeamHealth	21 670	46.21
Cigna	SCP Health	3845	8.20
Anthem	SCP Health	3473	7.41
UHC	SCP Health	2668	5.69
BCBS TN	TeamHealth	2475	5.28
BCBS TX	Non-PE	1265	2.70
BCBS FL	SCP Health	1137	2.42
UHC	Envision	1122	2.39
Aetna	SCP Health	1112	2.37
Aetna	TeamHealth	849	1.81
Aetna	Non-PE	672	1.43
BCBS TN	Sound Physicians	586	1.25
Other non-Anthem Blue	SCP Health	577	1.23
Cigna	Non-PE	575	1.23
Cigna	American Physician Partners	547	1.17
Aetna	Envision	465	0.99
Anthem	TeamHealth	447	0.95
Cigna	Envision	353	0.75
UHC	Non-PE	261	0.56
Cigna	US Acute Care Solutions	239	0.51
Anthem	Envision	223	0.48
Other	Non-PE	210	0.45
Aetna	American Physician Partners	207	0.44
BCBS TX	Sound Physicians	203	0.43
Other non-Anthem Blue	American Physician Partners	197	0.42
Other	SCP Health	196	0.42
Other non-Anthem Blue	Sound Physicians	177	0.38
Cigna	Sound Physicians	176	0.38
UHC	American Physician Partners	175	0.37
BCBS TN	Non-PE	152	0.32
Other non-Anthem Blue	Non-PE	141	0.30
BCBS TX	American Physician Partners	115	0.25
Other	US Acute Care Solutions	104	0.22

Source: Authors’ analysis of CMS IDR Data for 2023 (non-monetary file) merged with National Uniform Claim Committee's Health Care Provider Taxonomy Code Set and CMS National Plan & Provider Enumeration System NPI files by NPI and name/affiliation text matching.

There are 285 dispute lines attributable to additional dyads not shown on this table. In these dyads that are omitted, there are fewer than 100 dispute lines per dyad.

Abbreviations: BCBS, blue cross blue shield; PE, private equity; UHC, UnitedHealthcare.

Discussion

Public data on IDR arbitration outcomes in 2023 reveal that arbiters are ruling in favor of providers in the vast majority of disputes and the prevailing offer is much higher than the QPA on average. This contrasts markedly with the expectations of the Congressional Budget Office at the time of enactment that, on average, IDR awards would come out close to the QPA (which undergirded their prediction that the NSA would reduce insurance premiums by between 0.5% and 1%).⁷ However, there is evidence that QPAs may be underestimated in some cases, as a substantial share of QPAs observed in the IDR data are below commercial in-network rates reported in the prior literature and a fraction are even below Medicare-allowed amounts. This may indicate that health plans are underestimating QPAs or it could simply reflect selection into the IDR process if providers are more likely to object to the health plan's initial payment when QPAs are lower than the mean QPA for all out-of-network reimbursements. Notably, other research indicates that IDR outcomes are also substantially higher than independent estimates of pre-NSA average in-network prices, not just the QPAs.⁸

Disputing parties likely weigh risk reward when presenting final offers in arbitration; lower offers by health plans and higher offers by providers may indicate a greater tolerance for risk of losing in arbitration. Independent dispute resolution also occurs within a broader context of contract negotiations, including failures to reach a contract. As such, offers made in arbitration may be part of a broader dynamic between parties.

The 2023 provider IDR wins and health plan losses are not evenly distributed. Health plans were much more likely to lose an IDR hearing when facing a PE-backed physician staffing company, despite the fact that both the provider's offer and the health plan's offer exceeded the QPA in two-thirds of these cases. The average prevailing offer in IDR hearings with PE staffing companies was 63% higher than IDR hearings with non-PE emergency professionals. Our findings suggest that health plans may be responding to their lack of IDR success by raising their offers over time.

The implications of these findings for the NSA's prospects of containing health care costs are unclear. It is possible that the prevailing reimbursement rates in the IDR data do not reflect overall out-of-network reimbursement rates. A majority of the 99 284 disputes are with UHC and the vast majority of disputes involve PE-backed physician staffing companies. Almost half of the disputes reflected in the 99 284 IDR data are between United Healthcare and TeamHealth. Florida, Georgia, Tennessee, and Texas are also disproportionately represented in the IDR data, whereas populous states like California, Illinois, and New York represent relatively few disputes; see Appendix S1.

It is also unclear how NSA arbitration decisions will evolve, whether providers and health plans will adjust their offers in response to these data, and how these arbitration decisions will impact markets for physician services. If PE staffing companies continue to obtain much higher prices than competitors from IDR, it could create additional incentives for other emergency medicine practices to sell to their PE-backed competitors, accelerating the growth of PE-backed companies and reducing the number of independent physicians. Furthermore, provider and health plan offers may change over time if the volume of services arbitrated or regulatory context for disputes evolves in response to litigation and market dynamics.

Limitations

These arbitration decisions may not reflect OON payments generally because cases brought to arbitration likely differ from those settled before arbitration and unchallenged initial payments. In particular, more sophisticated provider groups are disproportionately represented in the disputes, reflecting their heavy use of the IDR process, and their relative success may not be representative of all OON providers. Furthermore, these are the first 2 releases of detailed dispute data. Providers, health plans, and arbiters are likely to use this data to revise their offers and decisions in future disputes. Analyses of these public data are limited by masking and omission of some geographies, exact dates of service and IDR filings, health plan details, and arbiter identities in the public data files released by CMS.

Supplementary material

Supplementary material is available at Health Affairs Scholar online.

Funding

The authors acknowledge support from Arnold Ventures.

Conflicts of interest

Please see ICMJE form(s) for author conflicts of interest. These have been provided as supplementary material.

Data availability

Independent Dispute Resolution data used in this study are publicly available on the CMS website.