Prior authorization: the paperwork that decides care

Interviewers like this topic because it is where health economics meets the exam room, and because applicants who have shadowed will have seen it without necessarily understanding it. The nurse on the phone arguing with an insurer, the scan that happens next week instead of today, the drug swapped for a cheaper one — that is prior authorization, and being able to name and explain it marks out someone who has paid attention.

Here is the version worth holding. There is a real problem underneath it: American medicine does deliver a great deal of low-value care, and someone has to say no. Prior authorization is one way of saying no. The argument is about whether it says no to the right things, at an acceptable cost in delay and administration — and the appeal-overturn rate is the sharpest evidence in that argument.

A physician decides a patient needs something — an MRI, a biologic, a surgical procedure, a skilled nursing placement. If the payer requires prior authorization for that item, the practice submits clinical documentation showing it meets the payer’s criteria. Historically this went by fax and telephone; increasingly it is electronic, though far from universally. The payer reviews it against internal criteria, sometimes automatically, sometimes by a nurse reviewer, and only where a denial is proposed is it typically reviewed by a physician.

The decision comes back as an approval, a denial, or a request for more information. A denial can be appealed, and appeals often succeed. Meanwhile the patient waits. In an emergency the rules differ — emergency care generally cannot be subject to prior authorization — but for everything scheduled, the clock runs. In Medicare Advantage plans, which now cover more than half of Medicare enrollees, prior authorization is used far more heavily than in traditional Medicare, and that shift is a large part of why the issue has grown politically.

An answer that only argues one way is easy to dismantle. Hold both of these at once.

The case for it. A meaningful fraction of American medical spending goes on care that does not help — imaging that changes no decision, brand drugs where a generic performs identically, procedures with weak evidence. Fee-for-service pays clinicians more for doing more, so someone outside the transaction has to check it. Prior authorization is that check, and payers point to real examples of it preventing harm as well as cost.

The case against how it works. The burden falls on the clinician, not the payer: physician surveys report practices spending upwards of a dozen hours a week on it, and it ranks among the leading contributors to burnout. Delay is a clinical variable — for oncology, cardiology and transplantation, weeks matter. Criteria are often proprietary, so a physician cannot see the standard they are being judged against. And the overturn rate on appeal suggests many denials were wrong in the first place, which means the process is generating delay and administrative cost without a corresponding filter on waste.

Three things are in motion. The federal rule finalised in January 2024 applies to Medicare Advantage, Medicaid and Children’s Health Insurance Program managed care, and qualified health plans on the federal exchange. From 2026 those payers must meet decision deadlines — seven calendar days for standard requests, 72 hours for expedited — give a specific reason for denials and publicly report their authorization statistics, and from 2027 support a standardised electronic prior authorization interface. It is a process reform: faster, more transparent, still there.

State legislation has moved in parallel, with many states passing so-called gold-carding laws that exempt clinicians with consistently high approval rates from prior authorization for particular services. Voluntary commitments from major insurers in 2025 to reduce the number of services requiring authorization are the third strand, and the honest note to add is that voluntary commitments are not enforceable.

None of this addresses the underlying question, which is whether utilization management should sit with the payer at all, or with the clinician under a payment model that does not reward volume. That is where a strong answer can go if pushed.

This arrives in three shapes. The direct one: "What do you think of prior authorization?" The shadowing one, where you are asked what surprised you about clinical practice. And the disguised one — "What contributes to physician burnout?" or "How should a health system control costs?"

For the direct question, give the purpose before the criticism. For the shadowing question, this is one of the best specific answers available if you saw it. For the disguised question, use it as your worked example of a cost control with a clinical price.

Start with the CMS pages on the Interoperability and Prior Authorization final rule for what changes and when, then the AMA’s annual prior authorization physician survey for the burden figures — reading it knowing it is advocacy, which does not make it wrong. KFF’s analyses of Medicare Advantage prior authorization give you the denial and appeal numbers from a neutral source.

Two pieces here give you the surrounding system. How American healthcare is actually structured explains why payers sit between clinician and patient at all, and drug prices is the other place cost control meets the exam room. For the interview formats, see our US interview guides.