A lawn is never finished. It gets mowed on a schedule and fed by the season, and the care shifts with the conditions — reseeded where drought thinned it, patched where tires cut it and grubs chewed it, and sometimes rebuilt outright after a flood or a fire. Some of the work is routine; some is prompted by events. Seeing the crew out front means the lawn is being tended — not that it has failed.
Formularies and medical policies change constantly — along with the prior authorization criteria and step therapy requirements that apply to both — and some critics read the churn as proof that utilization management (UM) is broken. But those policies operate in a changing clinical, economic, competitive, regulatory, and political environment. Patients and providers who feel the friction push back. Policies built for an environment like that should change. Standing still, not revising, would be the warning sign.
Why policies change
Not every policy is right, and no one who runs UM well pretends otherwise. Evidence evolves, guidelines get rewritten, products launch and lose exclusivity, utilization shifts, and prices and contracts change. Some policies are wrong from the start and need correcting. Others were well targeted three years ago but are off target today without anyone having erred because the target moved. Either way, the policies were never designed to hold still.
How policies change
The change process is built in. P&T and medical policy committees meet regularly to revisit formularies and coverage policies, while major developments can trigger off-cycle reviews. Appeals provide another source of feedback; sustained overturn patterns can point to criteria or documentation requirements that need revision.
The changes you can see
Formularies and medical policies leave a visible trail of annual and midyear updates. The flux is no secret. Clarivate’s DRG Fingertip Formulary and MMIT’s alert services track formulary and medical policy changes. Tufts Medical Center researchers likewise have tracked how 17 large commercial plans changed coverage for hundreds of specific drug uses through year-over-year additions, removals, and revised restrictions. Nobody builds tracking infrastructure around documents that hold still.
Midyear changes draw the sharpest complaints, and the concern is real when they disrupt a stable patient’s therapy. Continuity protections and transition fills exist for exactly that case. But midyear changes are the exception, not the norm, in part because they create member and provider abrasion and real communication costs. They can also enhance coverage by adding drugs, loosening requirements, or substituting lower-cost alternatives. The best national evidence is dated: When the Government Accountability Office reviewed Medicare Part D midyear formulary changes in 2008 and 2009, nearly nine in ten were enhancements rather than restrictions.
Some changes are larger and more visible. UnitedHealthcare said in May 2026 that it will eliminate another 30% of prior authorization requirements by the end of the year, three years after it cut prior authorization volume by about 20%. Optum Rx, also part of UnitedHealth Group, has eliminated 33% of drug reauthorizations since the beginning of 2025, covering nearly 270 medications for chronic conditions — a list it says it will keep expanding. And Humana said in July 2025 that it would eliminate roughly a third of prior authorizations for outpatient services by the start of 2026, and has since reported removing requirements for more than 340 codes. KFF’s Larry Levitt read the industry’s cuts as driven by the backlash from patients and providers, and the pressure plainly shaped their size and timing. UnitedHealthcare did not abandon the rationale for review — its chief executive framed prior authorization as “an essential safeguard” to be used when it truly protects patients. Similarly, Humana’s chief executive said, “Today’s healthcare system is too complex, frustrating, and difficult to navigate, and we must do better,” even as Humana’s announcement made clear that it will continue to review high-cost, high-risk treatments.
There is another reason to read the cutbacks carefully: What gets cut is not random. Plans have strong incentives to start with requirements providing the least benefit relative to their burden — those that save little, add little clinical value, or create disproportionate friction. The requirements being dropped are therefore especially likely to look unnecessary in retrospect. A 30% reduction speaks most directly to the 30% selected for removal; it does not establish whether the remaining 70% are warranted.
Regulators are tightening the rules around prior authorization while adding targeted review where they see waste. Under CMS’s interoperability and prior authorization rule, Medicare Advantage, Medicaid, and CHIP plans now face decision deadlines and must explain every denial and publish their prior authorization metrics — a rule CMS framed as cutting waiting and paperwork. And traditional Medicare launched the WISeR model in January 2026, a six-year test applying review to services with documented histories of waste, to keep the program from paying for care that does not help. Both revise UM; neither abolishes it.
Repair is not failure
Analysts read the prior authorization cutbacks in different ways. When Optum Rx pledged in early 2025 to scale back prior authorization requirements, one commentator — no friend of PBMs — offered two readings: The PBM had recognized a problem with its own requirements, or the announcement was just public relations. Both are plausible, but both cast the cuts as embarrassment — error or spin. A third reading, missing from that frame, might be that plans and PBMs adjust their policies all the time, and shedding requirements that no longer serve their purpose is what a working process does. Levitt offered a more specific economic reading — an assessment rather than an accusation. He read the industry’s voluntary limits as a signal that “these denials may not be saving much money in the end” — that the requirements may no longer generate enough savings to justify their costs.
Change itself does not deliver a verdict on its own. In the aggregate, change is predictable, because the clinical, economic, competitive, regulatory, and political environment around these policies is constantly changing. The reason for any particular change is harder to read from the outside. A revision can mean a policy that was wrong from the start, a target moved, incentives shifted, or pressures mounted — or, sometimes, several of these at once. Levitt’s point goes to a question that a functioning UM system should keep revisiting: whether a requirement’s benefits — in safety, appropriateness, or savings — still justify its costs and burden. Optum Rx made that judgment in public, dropping reauthorization requirements for effective, lifelong treatments because the requirements were of “minimal additional value.”
A formulary or medical policy that never changed while everything around it changed would be the real failure. Dropping a requirement today does not establish that it was wrong yesterday — sometimes it was, and sometimes the conditions around it changed. A healthy UM system changes with those conditions; the change is visible, but what it says about the policy that came before may not be.