Payer-Specific Prior Auth Requirements for Buy-and-Bill Biologics

Buy-and-bill biologics work in one direction only: the practice buys the drug, infuses the patient, then bills the payer under the medical benefit and waits to get paid back. Reimbursement lands after the money's already gone out the door, so a denied claim is a hole in the bank account for a drug that's already been used up. The vial is empty and there's no getting it back, and the billing team's approach to this fact is what this whole business runs on.

The dollar amounts make this unforgiving. TNF inhibitors, IL-17s, IL-23s, IL-12/23s, integrin inhibitors like vedolizumab and natalizumab: these run through the medical benefit, not the pharmacy benefit, and nearly all of them need prior authorization from commercial payers, Medicare Advantage, and Medicaid managed care alike. Medical benefit PA has its own forms, its own reviewers, its own appeals process, separate from what a pharmacy-benefit self-injectable would go through. Missing a step there costs thousands, sometimes tens of thousands, of dollars in drug spend the practice never recovers.

How much payer requirements vary, and why treating them as uniform creates predictable losses

A lot of infusion teams run on a simple assumption: prior auth is prior auth. Submit the clinical notes, wait, get an answer. That assumption is flat wrong, and it's the expensive kind of wrong, the kind that looks fine for months until one payer's denial rate quietly outpaces every other line on the P&L.

Step therapy sequences, preferred drug lists, required lab work, documentation thresholds, even which site of care is allowed: all of it shifts from payer to payer. UnitedHealthcare and OptumRx publish detailed clinical policies by biologic class, and citing the specific policy number in a submission signals the reviewer that the practice did its homework. Aetna runs on Clinical Policy Bulletins, public documents that should be cited by bulletin number. Cigna's Coverage Policies go further and name specific lab thresholds, like DAS28 scores for rheumatoid arthritis biologics, that must appear in the chart notes.

Medicare Advantage runs on a different rulebook entirely: Local Coverage Determinations or National Coverage Determinations. The submission needs to cite the LCD number and walk through each criterion the patient meets, one by one. Medicaid piles its own layer of variation on top, since preferred drug lists and step therapy sequences get set state by state and don't line up with either commercial plans or Medicare Advantage.

Formularies aren't static, either. Payers update them, sometimes every quarter, so the preferred TNF inhibitor from three months ago might not be preferred anymore. Submit against last quarter's list and the denial writes itself, entirely avoidable with a current formulary check that takes five minutes.

RxCheckUp reports that practices submitting thorough, documentation-heavy prior auths on the first pass see approval rates run 25 to 35 percentage points higher than practices submitting bare-bones forms. That gap has almost nothing to do with clinical complexity. It comes down to knowing, payer by payer, what's actually required before the drug goes into the patient. Tracking denials as one blended number lets a practice look fine on paper while a single payer is quietly denying a much larger share of its claims. Breaking the number out by payer reveals the real leak.

Step therapy: what each payer requires before a biologic will be covered

RxCheckUp's 2026 findings put step therapy denials at the top of the list. The concept itself is simple: payers want proof the patient tried and failed cheaper conventional therapy before moving to a biologic. What counts as "tried and failed" is where things go sideways, because that definition changes by payer and by condition.

For rheumatoid arthritis, that usually means methotrexate or another conventional DMARD, at an adequate dose, for an adequate stretch of time. For psoriasis, it's typically topical treatments plus phototherapy or a systemic agent like methotrexate, cyclosporine, or acitretin. For inflammatory bowel disease, mesalamine and immunomodulators such as azathioprine or 6-MP are the common required steps.

First-submission denial rates for biologics and other high-cost specialty drugs run somewhere between 25% and 40%, and step therapy documentation drives most of that number. The failure pattern repeats itself almost exactly: the chart shows the patient was on methotrexate at some point, but doesn't say what dose, for how long, or what happened when they stopped it. Payers rarely deny because the therapy is missing. They deny because it's undocumented as adequate, and fixing that on appeal takes a physician sitting down and writing specifics, not an administrative resubmission of the same thin note.

A second common denial trigger is requesting a non-preferred biologic when the payer has a preferred alternative on file, which loops right back to the formulary timing issue discussed above. A handful of other triggers occur constantly, including missing TB tests, missing hepatitis B or C screening, missing CBC or liver panels, an unspecified ICD-10 code where the payer wants a specific one, and a site-of-care mismatch where the payer now wants home infusion or an ambulatory center instead of hospital outpatient.

None of this is a clinical failure. It's a documentation and intelligence failure, which is the good news, because it means the fix is tracking.

Biosimilar mandates and non-medical switching as a new PA variable

RxCheckUp reports that most major commercial payers had rolled out biosimilar-first policies for adalimumab, infliximab, and rituximab by 2026. For a patient starting fresh on one of these molecules, the payer steers toward the biosimilar almost automatically. Submit a PA for the reference product without a documented reason the biosimilar won't work, and the denial is close to guaranteed.

The trickier scenario hits mid-treatment, not at the start. Payers increasingly require patients already stable on a reference biologic to switch to the biosimilar at renewal just to keep coverage, which turns a routine renewal into a live risk sitting inside an otherwise working treatment plan.

Medicare's biosimilar reimbursement structure gives providers financial incentives that commercial payer economics don't mirror, so the incentive landscape looks different depending on who's paying. Assuming Medicare logic applies everywhere is the mistake to avoid here.

A forced switch can be fought, but only with real clinical justification. A medical exception can be supported when the rationale is specific and tied to the payer's own policy language, not a generic preference for staying the course. Where clinical literature supports concern about switching risk for a specific patient, that evidence belongs directly in the exception request, cited by name. None of this holds still. Biosimilar policy shifts by payer and by renewal cycle, so tracking which payers have flipped to non-medical switching, and for which drugs, is a permanent line item in infusion PA management now.

Reauthorization timing and the authorization expiration denial pattern

Infusion PA doesn't work like a one-and-done approval. Authorizations cover a set number of sessions or a fixed window of time, and once that window closes, so does the coverage behind it.

That creates a specific failure pattern. If the reauthorization doesn't go in before the old one expires, any session delivered after that point has no valid authorization behind it. The denial that follows is purely administrative, and it often lands on a claim where the drug has already been infused and paid for out of the practice's own pocket.

Specialty drug PA decisions can take multiple weeks, since they route through medical benefit review and site-of-care policy checks. Wait until the authorization is about to lapse before submitting the renewal, and there's no cushion left if anything slips. Payers don't make this easier: they update PA requirement lists throughout the year, and not always with a clear heads-up to infusion centers. A drug that didn't need PA last month might need it this month, mid-course of treatment, with zero warning attached.

Most authorization expiration denials trace back to a monitoring gap. The fix is straightforward in concept if not always in execution: a standing report showing every active authorization, how many visits are left, how many days until expiration, sorted by which ones are closest to running out and which ones carry the most dollars at risk. There's a quieter version of this same problem in J-code and CPT mismatches, where an authorization approved for one code doesn't technically cover a slightly different code used at billing, even when the clinical intent has stayed the same. That mismatch tends to become visible right at reauthorization, when a dosing change or a different drug lot enters the picture. Providers already spend close to two full business days a week on authorization tasks in general, and renewal tracking is constantly competing with new-start PA work for the same limited hours.

What the 2026 regulatory changes fix and don't fix for infusion PA

CMS-0057-F took effect January 1, 2026, and it does tighten some things. Standard PA decisions now have to come back within 7 calendar days, down from 14, and payers have to give a specific reason for every denial. It applies across Medicare Advantage, state Medicaid and CHIP fee-for-service, Medicaid managed care, CHIP managed care entities, and qualified health plans on the federal exchange.

Drugs are carved out of it, and that carve-out is the whole story for buy-and-bill infusion providers. Part B and Part D drug PA runs on its own separate timeline, untouched by CMS-0057-F, which is exactly the gap that matters most here. KFF data shows 99% of Medicare Advantage enrollees are in plans that require PA for some services, and 94% are in plans requiring PA specifically for Part B physician-administered drugs. The new rule leaves that population's biggest pain point sitting right where it was.

There's a real transparency gain starting March 31, 2026: payers affected by the rule began publishing PA metrics publicly, approval and denial rates by service category, average turnaround times. That's a genuinely new data source infusion revenue cycle teams can use to benchmark how a given payer behaves, separate from anecdote and gut feeling.

CMS-0062-P, proposed April 10, 2026, is the first federal proposal aimed at electronic PA for provider-administered drugs under the medical benefit. If it gets finalized, most of it wouldn't kick in until October 1, 2027, and it's still working through comment and finalization, so none of it is locked in yet. Separately, payers are required to stand up a FHIR-based Prior Authorization API by January 1, 2027. If that system actually communicates documentation requirements at the point of order, it could shrink the back-and-forth of records requests considerably. That's a 2027 story, though, not a 2026 one, and betting the practice's workflow on it now is premature.

Gold carding programs are further along, with a growing number of states running them as of 2026, including Texas. Texas HB 3812, effective September 1, 2026, sets the bar at a minimum of five PA requests for a specific service with at least a 90% approval rate over a full year. For infusion practices whose volume for any single drug and payer combination might not reach five requests a year, that bar is nearly impossible to clear. Gold carding will help hospital systems long before it helps a smaller infusion practice, and pretending otherwise just delays the real fix.

Put together, the 2026 regulatory picture tightens timelines and adds transparency, but it leaves step therapy rules, biosimilar mandates, documentation thresholds, and reauthorization clocks exactly where they were. None of that is going away on its own. It has to be managed inside the practice, not waited out on the assumption that federal rulemaking will eventually cover it.

Building a payer-specific PA intelligence operation for infusion billing

The scale of this work is bigger than most people expect. Industry data shows the average practice handles 39 prior authorizations per physician every week, with staff logging around 13 hours weekly on the paperwork behind them. Without a system that routes each request by payer-specific rules, those hours produce wildly inconsistent results, and inconsistency at this volume turns straight into revenue loss.

Denial tracking has to happen payer by payer. A blended number hides the real problem every time: a practice's overall denial rate can look perfectly fine while one payer is denying a disproportionate chunk of its claims. Payer-by-payer tracking reveals whether the actual issue is documentation, coding, step therapy, or a policy shift nobody caught in time.

Every submission should cite the specific clinical policy number or LCD/NCD for that payer. Building a payer policy library, organized by biologic class, and keeping it current is not optional busywork. It's the single biggest lever available. Generic, one-size-fits-all submissions that ignore payer-specific criteria are the biggest driver of first-pass failures, full stop. A documentation checklist tuned to payer and indication needs to cover the same ground every time: confirmed diagnosis with a specific ICD-10 code, step therapy notes that include dose, duration, and outcome rather than just a drug name, current disease activity scores plus the required labs (TB, hepatitis B/C, CBC, liver panels as needed), a physician letter addressing that payer's exact medical necessity language, and for biosimilar exceptions, clinical rationale backed by literature and tied to the payer's own exception policy.

Authorization expiration needs its own standing report: every active authorization tracked by visits remaining, days until expiration, and dollar value at stake, with renewals started well ahead of the deadline instead of at the last possible moment. Most of the recoverable damage, though, gets baked in at scheduling and intake, long before a claim ever reaches submission. The moment to catch a denial has already passed by the time it appears on the claim.

Infusion-specific revenue cycle operations exist for exactly this reason. General-purpose billing platforms don't carry the payer-specific workflow logic or the judgment calls this work demands: deciding whether a biosimilar exception rationale is strong enough, catching a step therapy documentation gap before it becomes a denial, recognizing that a spike in denials from one payer means a policy changed quietly in the background. A rules engine alone can't make those calls, no matter how well it's built.

The scale of what's lost to this across the industry isn't small, either. 247 Medical Billing Services put the number at $23 to $31 billion a year in revenue lost to prior authorization inefficiency across the healthcare system in 2026. That loss doesn't spread evenly. It piles up hardest in exactly the drug categories covered here, where payer rules are most tangled and least standardized, and where a single missed reauthorization can mean a five-figure loss on a drug that's already been infused into a patient who's doing fine on it. The rules aren't going to simplify themselves. Managing them payer by payer is the job, and there's no shortcut around it.