Prior Authorization

Prior Authorization

Prior authorization is a health plan's advance review of whether a specified service, medicine or item meets its coverage criteria before the person receives it. It can be used to manage use and spending, but its requests, delays and denials also affect patients and care teams. This page follows the decision pathway, distinguishes authorization from treatment and payment, and shows how to measure both administrative burden and access.

What the advance decision covers

The plan sets criteria for a covered benefit, and a clinician or supplier submits information for a particular patient and request. The plan may approve, deny or seek more information, with rules and review routes varying by payer, benefit and jurisdiction. A clinician's recommendation expresses clinical judgment; the payer's authorization concerns coverage under the plan and does not replace that judgment.

An approval is not a guarantee that a patient will receive the service or that every claim will be paid. Eligibility, network, coding, benefit limits and the circumstances of delivery can still matter. A denial likewise does not prove that the care is clinically inappropriate; criteria, documentation and case details need examination.

Step	What happens	What should be recorded
Requirement check	The team determines whether authorization applies.	Plan, benefit, requested item and applicable rule.
Submission	Clinical information and supporting documents are sent.	Complete timestamp, staff time and documents.
Review	The payer evaluates the request against stated criteria.	Decision date, reason and requests for more information.
Reconsideration or appeal	A denied request may be challenged under applicable rules.	Appeal eligibility, timing, evidence and outcome.
Care and claim	An approved item may be delivered and billed.	Actual receipt, patient cost and final payment.

Urgent requests and exceptions may require different handling. The applicable timeframe and rights must be checked against the specific current contract and rules rather than assumed from one country or payer's guidance.

Why plans and providers care about the process

Payers may intend to reduce low-value, duplicative or inappropriate use and direct patients to covered alternatives. Providers may have to gather records, answer questions and repeat a request when an authorization expires or the treatment changes. The process can also affect clinical timing and whether someone abandons recommended care.

Evaluation requires a credible comparison. Lower spending after introducing authorization does not by itself mean waste was removed: care may have been delayed, shifted to another setting or forgone. Conversely, a high approval rate does not establish the process is harmless if nearly every request imposes time and uncertainty before approval.

Measure requests, decisions and actual care separately

Use a clearly defined unit, such as initial requests for a specific benefit in a specified period. Avoid mixing requests, patients, services and appeal decisions in one denominator. Stratify decisions and waiting times by urgency, condition and relevant population groups where possible.

Suppose a fictional plan receives 1,000 initial requests. It approves 800 initially and denies 200; 100 of the denied requests are appealed, and 60 appeals succeed. The initial approval rate is $800/1{,}000=80%$; the appeal-success rate among appeals is $60/100=60%$; approvals after these appeals total $800+60=860$, or 86% of initial requests. The remaining 140 include 100 not appealed and 40 unsuccessful appeals.

Measure	Illustrative formula	Result and denominator
Initial approval rate	=800/1000	80% of initial requests.
Initial denial rate	=200/1000	20% of initial requests.
Appeal rate among denials	=100/200	50% of initial denials.
Appeal success	=60/100	60% of appealed denials.
Approval after observed appeals	=(800+60)/1000	86% of initial requests.
No recorded approval after this stage	=(100+40)/1000	14% of initial requests.

The 60 successful appeals do not show why the initial decision was reversed, and unappealed denials cannot be labelled appropriate by default. Nor does 86% mean 860 patients received care: one person could have multiple requests, and care might not follow approval. Report time from first submission to final decision and time to delivered care as separate measures.

Count administrative and patient consequences

The payer's review costs and the provider's documentation time are resources, even if neither appears in a medicine's acquisition price. Patients may incur additional appointments, travel, time off work or health consequences from delayed care. An economic analysis should specify perspective, account for these components once, and distinguish financial transfers from real staff time and health effects.

In the fictional cohort, suppose an initial request costs £20 of provider administrative time and an appeal costs an additional £50. The provider's assigned administrative cost is $1{,}000\times £20+100\times £50=£25{,}000$. This excludes payer staff, delays, clinical changes and patient costs; it does not establish whether the authorization policy is worthwhile.

Spreadsheet item	Illustrative formula	Result
Initial request administration	=1000*20	£20,000.
Appeals administration	=100*50	£5,000 incremental to initial requests.
Provider administration total	=20000+5000	£25,000 under stated assumptions.

Compare the policy with a plausible alternative such as no prior authorization, an automated review or a narrower set of criteria. Include any change in inappropriate use and spending alongside health outcomes, delay, abandonment and administrative cost. A payer saving is not necessarily a health-system saving when the workload moves to clinicians or patients.

Fairness, transparency and quality checks

Rules should be clear enough for clinicians and patients to know what evidence is required and how to challenge an adverse decision. Automation may reduce clerical effort but can propagate opaque or incorrect criteria if not monitored. Examine whether people with rare conditions, complex histories, limited language access or less support are disproportionately delayed or deterred from appealing.

• Specify the benefit: Drug and non-drug services can have different authorization rules and data sources.
• Trace timing: Measure the full time to care, not only the payer's time after a complete submission.
• Keep denominators clear: Initial requests, appealed denials, patients and delivered treatments are distinct populations.
• Examine reasons: Approval and denial percentages alone cannot establish medical appropriateness.
• Count shifted burden: Provider, payer and patient time and spending should be attributed to the right perspective.
• Check current rules: Response deadlines, exceptions and appeals depend on jurisdiction, payer and date.

Sources and further reading

The US HealthCare.gov prior authorization definition describes the plan's advance coverage decision, while its preauthorization explanation cautions that advance review is not a promise of payment. The US Centers for Medicare & Medicaid Services interoperability and prior authorization rule page describes specified US payer obligations and implementation timing; those provisions do not apply universally. The request counts and pound-denominated costs above are original teaching examples, not current programme data.