The American physician now spends about 13 hours a week asking insurance companies for permission to treat patients. The process is called prior authorization, and it has grown from a tool reserved for unusually expensive interventions into a routine feature of nearly every clinical decision. Physicians submit 39 such requests a week, on average. Many are for medications and imaging studies that are almost always approved in the end. The waiting, however, is not always benign: in the American Medical Association’s 2024 survey of a thousand physicians, 93 percent reported that the process delays patient care, 82 percent saw patients abandon recommended treatment because of it, and nearly one in four—24 percent—reported that a delay or denial caused a patient in their practice serious harm, including hospitalization or death.
This is the system that Health and Human Services Secretary Robert F. Kennedy Jr. and Centers for Medicare and Medicaid Services Administrator Mehmet Oz set out to repair in June of last years when they convened the country’s largest insurers—UnitedHealthcare, Aetna, Cigna, Humana, Elevance, and roughly forty-five others, covering some 260 million Americans—to announce six voluntary commitments. Dr. Oz was candid about what he had secured. “This is not a mandate,” he said. “It’s not a bill or rule. This is not legislated.”
He was right about that. He was also describing the problem.
The insurance industry signed a nearly identical voluntary statement in 2018. The same lobbying associations pledged the same kinds of reforms: fewer authorizations, faster decisions, more transparency. The intervening years are the data. The new pledge appears to be on the same trajectory: eight months in, Healthcare Dive reported that signatories had eliminated only 11 percent of targeted prior authorization requirements, and when KFF Health News surveyed more than a dozen of them in February, half could not name a single service they had actually removed from review.
It would be too easy to read this as bad faith. The more useful diagnosis is structural. Prior authorization is not, primarily, a clerical tool. It is a profit lever. The HHS Office of Inspector General found that 13 percent of Medicare Advantage prior authorization denials were for services that traditional Medicare would have covered. The AMA reports that 81.7 percent of denials, when appealed, are overturned in whole or in part. The original “no” was wrong in roughly four cases out of five. Insurers know this. They also know that most patients and physicians, worn down, never appeal. A voluntary pledge to behave better cannot displace an involuntary incentive to behave worse; not at scale, not durably, not without something behind it.
What can stand behind it is statute, and one is already drafted.
The Improving Seniors’ Timely Access to Care Act of 2025, reintroduced last May by Senators Roger Marshall, a Kansas Republican and physician, and Mark Warner, a Virginia Democrat, would do four things. First, it would require Medicare Advantage plans to adopt electronic prior authorization integrated with physicians’ electronic health records, replacing the fax-and-portal patchwork that today carries the majority of requests and creates much of the delay. Second, it would compel public reporting of denial rates, appeal rates, and turnaround times by service category, so that patterns of overuse become visible to regulators, employers, and the patients choosing among plans. Third, it would authorize CMS to designate categories of “routinely approved” services and require real-time decisions on them; ending the multi-day wait for items that almost always get a yes. Fourth, it would set hard timelines on appeals so that delay itself stops functioning as denial. The previous version of the bill passed the House unanimously in 2022 before stalling in the Senate over a $16 billion CBO score. The 2025 version, after a CMS interoperability rule absorbed most of that cost, is expected to be cost neutral. It has 248 House cosponsors and 64 in the Senate, which comprise a supermajority in both chambers. What is missing is floor time.
A few complementary reforms belong in the same package, and they work through similarly concrete mechanisms. The Medicare Advantage Prompt Pay Act would impose statutory deadlines on the payment of clean claims, removing one of the quieter levers insurers use to manage cash flow at providers’ expense. A federal gold-card framework, modeled on Texas’s 2021 law, would exempt physicians whose approval rates exceed 90 percent on a given service from prior authorization for that service, thinning volume by trusting the clinicians who are demonstrably right. And as insurers move rapidly to adopt algorithmic denial tools, Congress should require disclosure of the clinical evidence underlying any model used in coverage decisions.
The natural objection is that none of this requires legislation and that the current administration could write a stronger CMS rule tomorrow. It could, but the next administration could rescind it the day after. Regulation in this area has proven a poor substitute for law: the rules turn over with each administration, and litigation in the post-Loper Bright environment has made agency action without clear statutory authority increasingly fragile. A statute is what makes a reform last beyond an election cycle, and it is also what gives any future CMS rule firmer legal ground.
There is no real ideological obstacle here. Republicans who favor markets should want the rules visible. Markets do not function in the dark. Democrats who favor patient protection and access should want the moment of denial brought into view. The bill is finished. The hearings have been held. The cost has been priced. The cosponsors are counted. The remaining variable is political will.
The pledge will fail again. The bill, if it is passed, won’t.