No Surprises Act Independent Dispute Resolution: A Five-Gate Evidentiary Framework for Evaluating Reform Claims.

backgroundThe No Surprises Act (NSA) removes patients from many out-of-network payment disputes. Debate continues over whether the Federal Independent Dispute Resolution (IDR) process is being abused and should be restricted or rebenchmarked. We evaluate whether commonly cited Federal IDR statistics support the policy inferences drawn from them.

methodsWe apply a five-gate evidentiary framework (examining the Denominator, Comparator, Selection, Causation, and Attribution) to 12 recurring Federal IDR reform claims using statutory, regulatory, judicial, federal, and published sources. We supplement this framework with a line-item analysis of Centers for Medicare & Medicaid Services (CMS) Federal IDR public-use files for the first half of 2025.

resultsFederal IDR utilization greatly exceeds the original forecast of 22,232 disputes annually. CMS reports 6,336,032 disputes initiated through May 2026, of which 1,066,645 (16.83%) are found ineligible. Challenge rates represent a distinct and materially higher quantity, running approximately 40%-42% across individual CMS reporting periods, each with its own denominator. CMS reports that providers prevail in 87% of non-default, fully contested determinations in the first half of 2025. Among 1,694,788 analyzed line items, providers prevail in 1,476,967 (87.15%), and 1,446,720 (97.95%) of provider-prevailing line items have payments above the Qualifying Payment Amount (QPA). Among these same line items, however, 97.67% of all provider offers exceed QPA regardless of outcome, indicating that the observed coupling between provider victory and above-QPA payment is substantially consistent with the underlying positioning of provider offers. This observation does not explain the underlying win-rate asymmetry. The primary specification showed little evidence of a material residual association after stratification (Mantel-Haenszel odds ratio (OR) 1.0064, 95% cluster-bootstrap confidence interval (CI) 0.973-1.045; 2,000 replicates), as did a wider sensitivity specification (OR 0.9948, 95% CI 0.961-1.032). An exact-QPA post-hoc sensitivity specification showed a small positive association (OR 1.0367, 95% CI 1.002-1.080). Because multiple related sensitivity specifications were examined without multiplicity adjustment, this marginal estimate is exploratory rather than confirmatory. The cause of the overall adjudicatory asymmetry remains unresolved.

conclusionsThe examined statistics establish substantial Federal IDR utilization, ineligibility, filing concentration, adjudicatory asymmetry, and payment divergence from QPA, but do not independently establish provider abuse, benchmark validity, network erosion, price or premium inflation, or incremental arbitration cost. Applied symmetrically, the same evidentiary standard that prevents provider win rates from proving provider virtue also prevents those rates from proving provider abuse. Policy decisions increasing QPA's determinative role require evidence capable of distinguishing case selection, payer behavior, benchmark performance, statutory factors, and adjudicatory effects.