Prior authorization has become one of the most closely scrutinized practices in managed care.

Patients and providers have raised concerns about delays in care, administrative burden and denials. Regulators and legislators are responding with new requirements intended to increase transparency and accountability. At the same time, health plans continue to rely on utilization management as an important mechanism for evaluating medical necessity and appropriate use of health care resources.

The challenge is moving the discussion from anecdotes and individual experiences to objective data.

That is why a new KFF analysis of prior authorization data is particularly important for organizations accredited by NCQA.

For 2026, NCQA expanded the data organizations must collect and analyze under Health Plan Accreditation UM 1. Until now, however, organizations have had relatively little external data to help put their own results into context.

The new KFF analysis begins to change that.

For the First Time, We Have Broader Comparative Data

Under a 2024 CMS final rule, Medicare Advantage, Medicaid and CHIP programs and insurers participating in the federally facilitated ACA Marketplace are required to publicly report specified prior authorization metrics annually.

Insurers reported their first set of data—covering calendar year 2025—by March 31, 2026.

KFF subsequently collected and analyzed publicly reported data from 14 major insurers. The analysis represents approximately 71 million enrollees across Medicare Advantage, Medicaid managed care and the federally facilitated ACA Marketplace.

The results give the industry a new window into prior authorization performance.

For standard prior authorization requests, KFF found average denial rates of:

• 12% — Medicare Advantage
• 14% — Medicaid managed care
• 18% — ACA federally facilitated Marketplace

Denial rates for expedited requests were somewhat lower: 10%, 12% and 16%, respectively.

Those averages are informative.

The variation behind them may be even more important.

Denial Rates Vary Substantially

KFF found wide variation among major insurers.

For standard prior authorization requests:

• Medicare Advantage insurer denial rates ranged from 5% to 17%.
• Medicaid managed care insurer denial rates ranged from 2% to 23%.
• ACA Marketplace insurer denial rates ranged from 3% to 25%.

Even the same insurer could have significantly different results across lines of business.

For example, KFF reported that UnitedHealth Group denied 17% of standard requests in Medicare Advantage, 11% in Medicaid managed care and 21% in the ACA Marketplace.

That is an important reminder for health plans reviewing their own UM data.

A number by itself does not explain performance.

Different populations, benefits, services requiring prior authorization and other characteristics may affect the results. Comparisons therefore need to be made thoughtfully.

But substantial variation should lead organizations to ask questions.

And that is where the KFF findings intersect with NCQA's 2026 UM 1 requirements.

NCQA Is Requiring Greater Visibility Into UM Performance

NCQA expanded its 2026 UM 1 requirements to give organizations greater insight into how their utilization management processes are performing.

The required data address important aspects of the UM process, including approval and denial activity, reasons for denials, appeals and overturns, and timeliness.

The purpose of collecting these data should extend beyond demonstrating compliance with an accreditation requirement.

The data allows organizations to begin asking:

What is happening in our UM program, and what do the results tell us?

Historically, one challenge has been determining what an organization's results mean in a broader context.

If an organization calculates a denial rate, for example, what should leadership make of it?

Until sufficient standardized comparative information becomes available, there has been no simple answer.

The KFF analysis does not establish NCQA benchmarks or normative thresholds, and organizations should not use these numbers that way.

It does, however, provide valuable external comparative context.

Appeal Overturn Rates May Be Even More Thought-Provoking

The KFF findings on appeals deserve particular attention.

KFF found that prior authorization denials are infrequently appealed. But when appeals occur, a substantial percentage of the original denials are overturned.

For standard prior authorization requests that were appealed:

• 67% were overturned in Medicare Advantage.
• 47% were overturned in Medicaid managed care.
• 43% were overturned in the ACA Marketplace.

The variation among individual insurers was again substantial.

In Medicare Advantage, overturn rates among the insurers studied ranged from approximately 40% to more than 90%. Among Medicaid managed care insurers included in the comparison, reported overturn rates ranged from 22% to 81%.

These results should not automatically be interpreted to mean that the initial decisions were incorrect.

As KFF appropriately points out, a denial may be overturned because additional information was submitted during the appeal that was not available when the original determination was made.

That distinction matters.

But it also leads to exactly the type of analysis organizations should be performing.

If an organization has a substantial appeal overturn rate, why?

Are initial requests missing necessary clinical information?

Are particular denial reasons more frequently overturned?

Are there differences by service or delegate?

Are providers unclear about what documentation is required?

Are there patterns that warrant additional analysis?

The metric identifies the signal. The organization's analysis must determine what is behind it.

Timeliness Data Provide Another Important Comparison

KFF's analysis also provides useful information about prior authorization decision timeframes.

Across the three markets studied, the median response time for standard prior authorization requests was approximately one day.

For expedited requests, median response times were approximately:

• 4 days—just under 10 hours—in Medicare Advantage
• 8 days—approximately 19 hours—in Medicaid managed care
• 1 day in the ACA Marketplace

Individual insurer performance again varied.

For Medicaid managed care, for example, median standard response times among the insurers studied ranged from one to four days.

These data provide useful context, but they also demonstrate one of the limitations of aggregate reporting: a median cannot tell us what happened to every member.

By definition, some requests take longer than the median. KFF notes that current reporting does not provide sufficient information to determine how frequently decisions exceed applicable timeframes.

For organizations evaluating their own UM performance, therefore, meeting an aggregate or median performance level should not replace analysis of whether required timeframes are consistently being met.

The 2025 KFF Medicaid Findings Provided an Early Warning

The new KFF analysis builds on concerns identified in its August 2025 report on Medicaid managed care prior authorization.

That earlier analysis described variation in state requirements for prior authorization decision timeframes, denial notices and access to independent external medical review. It also highlighted prior federal findings showing relatively high Medicaid MCO denial rates, low appeal rates and significant overturns when denials were appealed.

At that point, however, comparable national reporting across multiple insurance markets was limited.

The 2026 KFF analysis is different.

CMS's public reporting requirements have now produced a new body of data that allows researchers—and eventually health plans themselves—to begin examining variation across markets and insurers.

That makes the timing of NCQA's expanded UM measurement requirements particularly relevant.

There Are Still Significant Limitations in the Data

KFF cautions against overinterpreting the new information, and health plans should do the same.

Among the limitations KFF identified:

• Insurers are not required to report numeric counts for each required metric.
• Data are generally aggregated across medical items and services.
• Reporting does not identify which service categories are driving approval or denial rates.
• Insurers have used different reporting formats.
• Response-time reporting does not show the full range of response times.
• Differences in populations and service mix make comparisons across lines of business difficult.

These limitations reinforce an important principle for organizations analyzing their own NCQA UM data:

Benchmarking should initiate analysis—not replace it.

A plan should not conclude that its UM program is performing well simply because its denial or overturn rate is lower than another organization's. Nor should a higher rate automatically be interpreted as evidence of poor performance.

The appropriate question is:

What is driving our result?

The Direction Is Clear: More Transparency Is Coming

The regulatory trajectory also deserves attention.

CMS updated its prior authorization metrics reporting guidance in July 2026 to address some of the limitations identified in the first reporting cycle. Among other changes, CMS emphasized the expected use of numerators and denominators and addressed greater consistency in reporting response times.

Additional federal proposals would expand reporting further by requiring numeric counts, standardizing denominators and adding additional prior authorization metrics.

For Medicaid managed care specifically, beginning in July 2026, states are also required to collect prior authorization metrics for inclusion in Managed Care Program Annual Reports.

The message for health plans is clear:

Prior authorization performance is becoming increasingly measurable, transparent and comparable.

What Should Health Plans Do With This Information?

For NCQA-accredited organizations, the answer is not to substitute the KFF findings for the organization's own analysis.

Instead, use the emerging external data as context.

Know your UM data.

Understand how it changes over time.

Investigate significant variation.

Understand what is driving denials.

Understand what is driving appeals.

Understand why denials are overturned.

Evaluate timeliness.

And where the data identify an unexpected pattern, investigate further rather than assuming the number itself provides the answer.

This is where NCQA's expanded UM 1 requirements can provide value beyond accreditation.

From Compliance to Quality Improvement

Prior authorization is unlikely to disappear from the national health care debate anytime soon.

Nearly seven in ten insured adults surveyed by KFF say prior authorization is a burden when obtaining health care. At the same time, insurers use prior authorization as one of their primary tools to reduce unnecessary or low-value utilization.

Those competing realities make objective measurement increasingly important.

NCQA's expanded 2026 UM 1 data requirements give accredited organizations a framework for examining their own utilization management performance.

CMS reporting requirements are making more information publicly available.

And KFF's analysis now gives the industry its first broader look at what those data show across major insurers and markets.

We still do not have definitive normative benchmarks that tell an organization what its UM results "should" be.

But we now have something we did not have before:

meaningful comparative information.

For health plans, the opportunity is to use that information appropriately—not to chase a particular industry average, but to ask better questions about their own performance.

Because ultimately, the goal should not simply be to collect the data NCQA requires.

The goal should be to understand what the data are telling you about the quality, consistency and effectiveness of your utilization management program.

Is your organization prepared to turn its utilization management data into meaningful analysis?

Turn Your UM Data Into Meaningful Insight

Collecting the data is only the beginning.

The real value comes from understanding what your results are telling you about denials, appeals, overturns, timeliness, and variation across your utilization management program.

MHR can help your team evaluate its findings, identify areas that warrant deeper analysis, and prepare for NCQA’s expanded 2026 UM 1 requirements.

Ready to take a closer look at what your UM data are telling you?

Schedule a Discovery Call with MHR

References

KFF. Prior Authorization Metrics Provide New Insights Into Insurer Practices, But Gaps Remain. August 2026. KFF analysis of publicly available 2025 prior authorization data reported by insurers pursuant to CMS prior authorization reporting requirements.

KFF. Prior Authorization Process Policies in Medicaid Managed Care: Findings from a Survey of State Medicaid Programs. Published August 7, 2025; updated August 21, 2025.

Centers for Medicare & Medicaid Services (CMS). CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F). 2024.

National Committee for Quality Assurance (NCQA). 2026 Health Plan Accreditation Standards and Guidelines, Utilization Management (UM), UM 1.

Copyright © 2026 Managed Healthcare Resources, Inc. All Rights Reserved.