Prior authorization is a plan’s advance permission, obtained before a prescription qualifies for payment. For this drug class it has gone from an occasional formality to the ordinary case in about a year, and it did so while coverage was widening rather than narrowing. Those two movements are usually reported separately, which makes the second one look like good news on its own. The broader question of how a covered price compares with a cash one is worked through in the coverage article.
The gate moved, it did not open
An analysis of quarterly CMS Basic Drugs Formulary File data from 2020 quarter two through 2024 quarter three, covering 54,358 Medicare Advantage prescription drug plan and 15,895 standalone plan quarter-plan observations, tracked both lines at once. Coverage of injectable semaglutide remained above 90% from 2021; tirzepatide went from 26.5% of plans in 2022 quarter three to 92.9% by 2024 quarter three, and oral semaglutide from 40.3% to 91.8%.[1]
Prior authorization moved the other way and faster. It stayed below 25% until 2023 quarter three, then rose sharply to 83.6%, 83.2% and 83.0% by 2024 quarter three for injectable semaglutide, tirzepatide and oral semaglutide respectively.[1] Being on the formulary stopped being the question at almost exactly the point nearly every plan carried the drug.
What a request typically has to carry
The structure is visible in the Medicare Part D rules, which set out the parts a request is assembled from and are more explicit than most commercial equivalents. A coverage determination is the plan’s decision on whether a drug is covered for an enrollee. Where the request is an exception — asking the plan to depart from its own rules — the clock does not start on the patient’s request at all. It starts on receipt of the prescriber’s supporting statement, and if that statement has not arrived within 14 calendar days the plan must decide anyway.[2]
That is the practical shape of most criteria sets in this class: a documented diagnosis and a body-mass index threshold, often a comorbidity, frequently a documented trial of lifestyle intervention or of another agent, and a reauthorization requirement tied to weight response. The specific thresholds are written by each plan and each state program rather than by any national rule, so the criteria document governing a particular request is the only place they can be read. A plan design that makes the diagnosis itself the question is discussed in the article on the two indications.
How long it is allowed to take, and how long it takes
The legal maxima are short and public. A Part D sponsor must decide a request for a drug benefit no later than 72 hours after receipt, or no later than 24 hours where an expedited determination is approved.[2] Under Medicaid, if a state requires prior authorization for covered outpatient drugs, a response must be provided within 24 hours of the request.[3]
Employer and commercial coverage runs on a slower clock. The Department of Labor claims-procedure regulation gives a group health plan until 15 days after receipt to decide a pre-service claim, extendable one time by up to 15 further days.[4] The same prescription can therefore have a three-day ceiling under one benefit and a thirty-day ceiling under another.
None of those figures is a typical wait. The 2025 AMA prior authorization physician survey, a 44-question web-based instrument administered in December 2025 to a sample of 1,000 practicing physicians drawn from a Medscape panel (40% primary care, 60% specialists, all screened to provide at least 20 hours of patient care a week and to complete these requests in a typical week), found that 95% report the process delays access to necessary care — 13% always, 42% often, 40% sometimes.[5]
The same survey found practices complete an average of 40 such requests per physician per week, consuming 13 hours of physician and staff time, and that 32% of physicians report requests are often or always denied while 74% say denials have increased over the past five years.[5] Burden was rated high or extremely high most often for Medicare Advantage at 69%, then commercial coverage at 63%, Medicaid at 62% and Medicare fee-for-service at 47%.
What physicians report the delay produces
The same survey asked about outcomes rather than hours. 79% of respondents report that issues related to the process can at least sometimes lead patients to abandon their recommended course of treatment — 1% always, 20% often, 58% sometimes — and 92% describe the overall impact on patient clinical outcomes as somewhat or significantly negative.[5]
More than one in four physicians, 26%, report that the process has led to a serious adverse event for a patient in their care. Broken out, 20% report a hospitalization, 22% a life-threatening event or one requiring intervention to prevent permanent impairment, and 8% a patient’s disability, permanent bodily damage, congenital anomaly or death.[5] More than a third, 35%, report that the criteria applied are rarely or never evidence-based.
These are physician reports rather than adjudicated events, and the instrument is a panel survey rather than a chart review, which is the limit worth stating alongside the numbers. They describe how the process is experienced by the people who run it, not a measured incidence.
The federal reform that does not apply here
A 2024 CMS final rule is widely cited as having fixed prior-authorization turnaround, requiring affected payers from January 1, 2026 to decide standard requests within seven calendar days and expedited requests within 72 hours.[6] Two carve-outs in that rule remove most readers of this page from it.
The first is stated in a single sentence beside the timeframe itself: “Note that these timeframes do not apply to any drugs, as discussed in section I.D. of this final rule”, and section I.D. is headed “Exclusion of Drugs”. CMS explains that it excluded drugs because the standards and processes for issuing authorizations for drugs differ from those for medical items and services.[6] The second is that qualified health plans on the federally facilitated exchanges are not included in the timeframe policy at all.[6]
So the rule that made headlines about faster decisions does not govern a prescription request, and the clocks that do are the older ones above. That distinction is worth holding onto, because it is the difference between expecting a week and being entitled to nothing faster than a month.
An approval is not a fill
The most complete measurement of what clearing this gate is actually worth comes from a different expensive drug class. A study of pharmacy transaction data for 45,029 patients newly prescribed a PCSK9 inhibitor in the United States between August 2015 and July 2016 found that 20.8% received approval on the first day and 47.2% were ever approved.[7]
Then the part that reframes the gate. Of those approved, 65.3% filled the prescription, so 30.9% of everyone prescribed the drug ever received it. Abandonment was predicted almost entirely by out-of-pocket cost, running from 7.5% at a $0 copay to more than 75% where the copay exceeded $350.[7] Approval rates also varied nearly threefold among the ten largest pharmacy benefit managers, and patients with government rather than commercial insurance were more likely to be approved (odds ratio 3.3, 95% CI 2.8 to 3.8).
A systematic review of 44 studies reached the same conclusion from the other direction: higher patient cost sharing reduces specialty drug use by increasing abandonment, with cost sharing above $100 associated with abandonment rates reaching 75% for some drugs, while prior authorization itself produced significant delays in initiation.[8] Winning the authorization returns the patient to the price, and the price is what most often ends it.
What a cash purchase does and does not avoid
A cash purchase has no prior authorization because there is no claim. No criteria document applies, no supporting statement is filed, no clock runs, and no denial is possible, because nothing is being asked of a payer. That is a genuine removal of the delay measured above, and it is the main reason this market exists in the form it does.
What it substitutes is the entire price, paid by the patient, with none of the cost-sharing structure that the abandonment data show to be decisive. It also substitutes the product. Most of the cash market dispenses compounded preparations, which are not FDA-approved and are not reviewed by the FDA for safety, effectiveness or quality before dispensing, so the comparison is not between two prices for one thing — a point argued in the comparison article. The ordered cash field is the price board, and a quoted figure can be run out across a year in the cost calculator.
What this page cannot tell you
Nothing above states what any particular plan requires. Criteria are written plan by plan and state by state, are revised at least annually, and are the only document that answers the question for a specific prescription. The rates quoted are population estimates from named datasets and surveys, not predictions about one request.
A denial is also not the end of the sequence, and the levels, deadlines and published overturn rates are set out in the appeals article. Who decides the criteria in the first place is covered in the employer article, and the public-program rules are in the public-coverage article.