Thirteen hours a week, and what is finally changing

The current burden, measured

The AMA's most recent physician survey, fielded in December 2025 across 1,000 practising physicians, reports a practice average of 40 prior authorizations per physician per week, consuming 13 hours of physician and staff time. 40% of physicians have staff working exclusively on prior authorization. (AMA, 2025 survey)

The clinical findings from the same survey are the ones that get quoted, and they should be: 95% report care delays, 26% report a prior authorization leading to a serious adverse event for a patient in their care, and 79% say it can lead to treatment abandonment.

Two operational details matter more for a practice trying to plan. First, only 24% of physicians say their EHR offers electronic prior authorization even for prescription medications, and the telephone remains the most common method for medical services. Second, 21% always appeal an adverse decision, and 59% of those who don't say it is because they don't believe an appeal will succeed.

That second statistic sits awkwardly next to the Medicare Advantage data. KFF found that of the 4.1 million MA prior authorization requests denied in 2024, only 11.5% were appealed — and 80.7% of appeals were partially or fully overturned. The overturn rate has exceeded 80% every year since 2019. (KFF, 2026)

What the federal rule actually requires, and when

The CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F) is the substantive change. It is frequently described imprecisely, so the specifics:

Who it covers: Medicare Advantage organisations, state Medicaid and CHIP fee-for-service programmes, Medicaid and CHIP managed care, and qualified health plan issuers on the federally-facilitated exchanges. It does not cover commercial or employer plans, standalone Part D, or state-based exchange plans.

What lands in 2026: decision timeframes of 72 hours for expedited and 7 calendar days for standard requests (FFE qualified health plans are excluded from the timeframe requirement); a specific reason for every denial sent to the provider; and public posting of aggregated prior-authorization metrics annually by 31 March.

What lands in 2027: four HL7 FHIR APIs — Patient Access, Provider Access, Payer-to-Payer, and a Prior Authorization API that lets a provider system ask whether authorisation is required, ask what documentation is needed, and submit and receive the request and response with denial reasons. CMS delayed these a year from the proposed dates.

What lands in 2027–29: an Electronic Prior Authorization measure in the MIPS Promoting Interoperability category, beginning with the CY2027 performance period. (CMS fact sheet)

The insurer pledge: what was promised

In June 2025, AHIP and the Blue Cross Blue Shield Association announced commitments from roughly fifty plans covering, by their count, 257 million Americans. The six promises: standardised electronic submissions via FHIR APIs operational by 1 January 2027; plan-specific reductions in the scope of services requiring authorisation demonstrated by 1 January 2026; a 90-day continuity period honouring existing authorisations when a patient changes plans; clearer determination explanations; at least 80% of electronic approvals answered in real time in 2027; and medical review of all clinical non-approvals, described as already in effect. (AHIP, 2025)

In April 2026 the industry reported an 11% reduction in prior authorizations and 6.5 million fewer requests. That is a plan self-report from an industry survey, not an audited figure.

Physicians are unconvinced. Only 33% think the commitments will make a meaningful difference. On the one promise already supposedly in force — medical review of non-approvals — only 24% agree that medical-necessity denials are being reviewed by a licensed and qualified clinician, and only 16% of physicians in peer-to-peer reviews say the plan's "peer" often or always has appropriate qualifications. (AMA, 2026)

Historical note worth carrying: a nearly identical set of promises was made in a January 2018 consensus statement signed by the same organisations. Gold-carding penetration today sits at about 5%.

Payer-side AI, and the law arriving around it

Physicians are not imagining the asymmetry. 60% are concerned that AI increases or will increase denial rates — and the administrative-transaction index reports that more than 50% of health plans use AI tools in administrative workflows, against 25% of provider organisations. (CAQH Index, 2026)

The disputes are live rather than settled, and it is worth being precise about their status. A federal court in Minnesota allowed breach-of-contract claims to proceed against UnitedHealth over its nH Predict tool, dismissing several state-law theories as preempted, and in March 2026 ordered broad class-wide discovery reaching back to 2017. Nothing has been decided on the merits. (case summary) A parallel case against Cigna over its PXDX claims-review system — a post-service review process, not prior authorization — survived dismissal in part in March 2025.

CMS has been unusually direct in guidance. An algorithm "can be used to assist" a Medicare Advantage plan, but an algorithm that decides coverage from a larger data set "instead of the individual patient's medical history, the physician's recommendations, or clinical notes would not be compliant"; a predicted length of stay "alone cannot be used as the basis to terminate post-acute care services"; and because published coverage criteria are static, "artificial intelligence cannot be used to shift the coverage criteria over time." (CMS FAQ, 2024)

States have moved faster than Congress. California's SB 1120, effective January 2025, requires that a licensed physician or competent health professional make the final medical-necessity determination and that AI tools work from the individual patient's circumstances. Washington, Georgia, Alabama, Utah, Indiana, Texas, Arizona, Nebraska and Maryland have all enacted variations since. We are not aware of a publicly reported enforcement action under any of them yet.

What a practice can actually do now

Waiting for 2027 is not a plan. Three things are available today.

Appeal more

With an 80% overturn rate on Medicare Advantage appeals and an 11.5% appeal rate, the arithmetic is not subtle. The binding constraint is staff time per appeal, which is precisely where drafting assistance helps — an appeal letter that assembles itself from the note, the order and the denial reason turns a forty-minute task into a five-minute review.

Make the clinical justification a by-product of the visit

Most of the delay in an authorisation request is not the payer. It is the gap between the visit and someone reassembling why the service was ordered. If the medical necessity narrative is already in the note, structured and traceable, the request starts from something.

Track which payers actually cost you

Burden is not evenly distributed — 75% of physicians rate the burden from one national insurer as high or extremely high, against 56% for another. From March 2026, covered payers must publish their prior-authorization metrics annually. That is contract-negotiation material.