An insurer’s denial rate is useful only when the numerator, denominator, time period, product, and type of denial are clear. A statistic about prior authorization requests cannot be compared directly with one about paid medical claims. Before drawing conclusions, ask what counted as a denial, whether repeat submissions were counted, and how many decisions were appealed or reversed.
What exactly is being denied?
Health insurance produces several different decisions: a prior authorization for future treatment, a pharmacy benefit request, a submitted claim, a line item within a claim, or a post-payment audit. A payer can reject an electronic claim before adjudication because required information is missing; that is not necessarily a medical-necessity denial. A single hospital claim can also contain many service lines. Counting claims, lines, requests, or patients yields different rates. Good analysis states the unit of measurement.
Why the denominator changes the story
Suppose two hypothetical insurers each deny 100 requests. If the first receives 1,000 requests, its measured denial share is 10%; if the second receives 10,000, it is 1%. Those arithmetic examples do not prove either insurer treats patients better. Differences in patient populations, service types, authorization requirements, coding quality, network design, and claims-processing rules can distort comparisons. The best denominator is the population of comparable decisions, not an unrelated enrollment figure or dollar amount.
How to read an appeal reversal statistic
An appeal reversal rate often uses only the denials that someone actually appealed. If relatively few denials are appealed, a high reversal percentage does not mean most original denials were overturned. Imagine 1,000 denials, 50 appeals, and 30 reversals: 60% of appealed denials were reversed, but 30 of the original 1,000 denials were overturned through those recorded appeals. That gap raises important questions about patient knowledge, burden, and access to review; it does not alone establish whether each unappealed decision was correct.
Separate payment delays from outright denials
Providers should also measure days in accounts receivable, first-pass acceptance, clean-claim payment time, underpayments against contracted rates, recoupments, and the share of denials eventually resolved. An insurer that pays late but rarely issues formal denials may still create serious financial harm. A useful provider dashboard groups cases by payer, product, service, date received, denial reason, initial payment, final payment, and appeal outcome. Keep adjustments distinct from cash received.
What data should journalists and advocates request?
Ask for the methodology, time window, sample size, plan type, geographic scope, exclusions, and definitions. Request separate counts for submitted requests, approved requests, denied requests, appealed denials, upheld decisions, reversals, and cases with no final disposition. Look for whether decisions were made by the payer or delegated vendor. If comparing insurers, specify how case mix was addressed. The absence of these details is a reason for caution, not permission to fill in a more dramatic conclusion.
How can a patient use data without losing sight of their case?
Population statistics may support a question about patterns, but a particular appeal needs the actual notice, coverage language, medical evidence, and timeline. Our evidence-based appeal guide explains how to address the stated reason for an individual decision. A patient should ask whether a denial was procedural, contractual, or clinical, and which route can change it. A physician’s documentation can be decisive for clinical criteria; it cannot by itself rewrite a coverage exclusion.
How should public figures be described responsibly?
Use precise labels such as “share of prior authorization requests denied in this dataset” rather than “share of care denied to patients.” Say whether a reversal rate is among appealed cases. Avoid inferring illegal conduct from a high denial percentage alone. Compare documented practices against governing rules and actual notices. Investigative accountability becomes stronger when the claim is narrow enough to test.
Frequently asked questions
Does a low denial rate prove a good insurer? No; access, benefit design, payment timeliness, and service mix also matter.
Are overturned denials automatically unlawful? No; reversals can follow new information or a different review.
Can providers publish their own data? They can analyze de-identified operational trends, but must protect patient information and explain selection bias.
Research starting points
Review CMS’s prior authorization reporting framework for the types of metrics certain payers must disclose, and CMS’s explanation of Medicare Advantage organization determinations to distinguish decisions. For private-plan appeals, HealthCare.gov explains available review routes.
Leave a Reply