Spread Economics for High-Volume Infusion Biologics

Spread in buy-and-bill infusion looks simple on a spreadsheet. A practice buys a biologic at acquisition cost, bills the payer at a contracted rate, Medicare pays ASP plus a percentage add-on, commercial payers usually pay more, and keeps the difference. That difference is the spread, and on paper it's a clean number.

In practice, the paper number and the collected number rarely match. Denials chip away at it, underpayments shave dollars off individual claims, authorization lapses stop the infusion schedule cold, and coding errors turn paid claims into rejected ones. None of these are edge cases. In a high-volume practice running dozens of infusions a week, the gap between theoretical and realized margin gets big enough to change whether the biologics line is actually profitable.

Administration fees aren't part of the spread at all, and that's another thing the simple formula leaves out. CPT 96413 (initial hour) and 96415 (each additional hour), for chemotherapy or highly complex biologic agent administration under the CMS 2026 Physician Fee Schedule, are a separate revenue stream from drug margin, often conflated with it, and must be tracked independently. Practices that lump admin fees and drug margin together lose the ability to diagnose where money is actually being made or lost. The two need separate ledgers.

None of this means spread economics is unmanageable. It means realizing spread at scale requires actively managing product mix, authorization status, billing accuracy, underpayment detection, and AR follow-through. Each one is a lever, and each one can quietly erase margin if left alone. The rest of this piece works through them in the order they actually break.

How drug margin varies across a biologics portfolio

Margin varies by product and by whether it's an originator or biosimilar, so the drug mix, not patient volume alone, sets the ceiling on total drug margin before any claim is billed.

The fastrvu.com rheumatology infusion analysis lays out the pattern in dollar terms. Infliximab biosimilar margins tend to run meaningfully higher than the originator, both percentage-wise and in absolute dollars per infusion. Remicade itself, billed under J1745, sits on the compressed end. Rituxan (J9312) shows a similar pattern, with the originator lagging its biosimilar counterpart by a wide margin. Actemra (J3262) is in the middle of the portfolio, and Orencia (J0129) is among the more modest earners, both as a percentage and in absolute terms.

The clearest illustration of how much this matters is the conversion math. Moving a patient from originator Remicade to an infliximab biosimilar, holding reimbursement roughly steady, can lift drug margin from $250 to $500 per infusion, and at scale the cumulative gain is substantial without adding a single patient. Product mix, in other words, is a lever a practice can pull on its own, independent of payer negotiations or patient volume growth.

None of that margin materializes, though, if the infusions don't happen on schedule. And whether they happen on schedule depends on something upstream of billing entirely: prior authorization.

Volume at scale depends on prior authorization running without lapses

Drug margin is a volume business. Every infusion that happens generates margin; every infusion that doesn't happen generates none. And volume, in a recurring infusion practice, hinges almost entirely on whether prior authorizations stay current, because a lapsed reauthorization doesn't just delay a payment. It stops the infusion outright and breaks the treatment schedule that produces volume in the first place.

Timelines make this fragile. Prior authorization approval times vary widely by payer, with some Medicare Advantage cases taking two months or longer. A reauthorization submitted even a few days late can leave a patient without active coverage mid-treatment cycle, and the broken cadence that results reduces the number of infusions a patient receives that year, cutting total drug margin across a full panel.

Step therapy compounds the risk. Biologics including adalimumab (J0139), infliximab (J1745), and tocilizumab (J3262) carry layered step-therapy and authorization rules that differ by commercial payer. If the authorization on file doesn't match the specific biologic actually administered, the claim triggers a step-therapy denial, even for a patient who's been stable on that exact drug for months.

Some of this risk is starting to ease through policy, at least in a few states. Starting January 1, 2026, New York bars insurers from requiring step therapy for longer than 30 days across all conditions. Connecticut bans step therapy outright for conditions including multiple sclerosis and rheumatoid arthritis, letting providers declare a treatment ineffective and move straight to the originally prescribed drug. For practices heavy in those states, on those diagnoses, the authorization math gets a little less punishing.

Biosimilar conversion introduces its own authorization failure mode. When the billed NDC doesn't match the drug name on the authorization, the authorization is void and the claim denies on arrival. That risk grows as practices lean harder into the biosimilar conversions shown earlier to be financially worthwhile. The two goals, converting to higher-margin biosimilars and keeping authorizations valid, have to be managed together, not sequentially.

None of this is solved by reacting faster to denials once they land. It requires ongoing tracking of expiration dates and resubmission windows for every active patient, since reactive management can't keep pace with weekly reauthorization volume.

How J-code billing complexity creates denial and underpayment exposure

Getting the authorization right only protects the right to bill. Whether the claim itself gets paid depends on getting the billing mechanics right, and infusion claims carry more mechanical complexity, and more dollars riding on that complexity, than most other claim types in a practice.

Infusion visits routinely generate thousands of dollars in drug charges on a single claim. A single wrong unit count, missing modifier, or mismatched NDC triggers a full denial, and given the dollar values involved that denial is a cash-flow event. Denial rates across the industry had already climbed to 38% by 2024, stacking the odds against catching an error unless billing is built to prevent it.

J-code billing itself demands precision at several layers at once. Units billed have to match the actual dose administered against each J-code's specific unit definition. NDC reporting is mandatory at the line-item level. Infusion hierarchy rules govern which service counts as "initial" versus "additional," and billing more than one initial service per encounter is a known audit trigger. These baseline requirements sit alongside a set of 2026-specific compliance pressure points that add further exposure.

2026 adds several fresh pressure points on top of that baseline. JW and JZ modifier enforcement, covering drug waste and no-waste reporting, tightened this year, and getting those modifiers wrong now carries both a reimbursement hit and audit exposure. New biosimilar HCPCS codes enter the system quarterly, and using a not-otherwise-classified code when a specific J-code already exists, or the reverse, creates billing validity issues. CMS is also refining how it collects and audits ASP data in 2026, and discrepancies can mean delayed claims or retroactive adjustments, making accurate acquisition-cost recordkeeping a requirement.

Site of billing adds another layer. Under the CY2026 OPPS and ASC Final Rule, CMS extended site-neutral payment policy to drug administration services delivered at excepted off-campus provider-based outpatient departments. The identical infusion, same drug, same dose, reimburses differently depending on where it's billed from and under which entity. A multi-site practice that doesn't manage this centrally will systematically over-bill some locations and under-bill others, unnoticed until an audit or payer review flags it.

Some product categories offer a calmer environment inside all this. Most neurology infusion biologics stay shielded from Medicare price negotiation through 2029-2033, and IVIG stays excluded as a plasma-derived product under the IRA's statutory language. Practices weighted toward those categories get a more stable spread to plan around, at least for the next several years.

Payer scrutiny isn't spread evenly across claim types either. Commercial payers concentrate review of coding, modifiers, drug cost reporting, place of service, authorization, and medical necessity documentation on high-cost therapies specifically. The claims generating the most spread are the same claims drawing the most scrutiny, so billing execution on infusion claims can't be treated as routine back-office work, and it's where a real share of the spread gets captured or lost.

Even a claim submitted correctly on every one of these fronts can still come back paid short. That's a separate problem, and it's a lot harder to catch.

Why underpayments disappear silently

Denials announce themselves. A denial comes back rejected, sits in a queue, and demands a response before a filing deadline closes it out for good. Underpayments don't work that way. An underpaid claim comes back looking like a normal remittance: paid, posted, closed. The shortfall gets filed as a contractual adjustment, indistinguishable from a legitimate write-down, and becomes a permanent loss unless someone checks it against the contracted rate before reconciliation.

Payers routinely post underpayments as contractual adjustments that look identical to legitimate ones unless checked line by line against the contract. Most in-house billing teams don't run that comparison on every remittance. Most in-house billing teams run that comparison only on claims that already look wrong, so underpayments that look right slip through untouched.

Catching them requires infrastructure most practices lack: a database of contracted rates by payer, matched systematically against every EOB, not just disputed ones. That's a data problem and a discipline problem at the same time. The data has to exist and be current, and someone has to actually run the match every time, not only when a claim looks suspicious.

In a multi-specialty or multi-payer infusion group, uncorrected underpayments can account for a meaningful share of annual collections, the Medical Billers and Coders analysis found. Aggregate financial reporting never captures those losses. Total collections can look healthy while underpaid claims quietly erode margin, recoverable only if caught at the remittance stage before the filing window closes.

Infusion claims make this worse than it would be elsewhere. Because a single biologic claim can be worth thousands of dollars, even a small percentage underpayment becomes a real dollar loss. An error barely noticeable on a low-value office visit becomes a significant write-off on an infusion claim, repeated across every patient on that drug, every cycle.

How AR aging in infusion differs from standard AR

Denials and underpayments both feed the same downstream problem: aging accounts receivable. Infusion claims carry higher dollar values, less forgiving filing deadlines relative to those values, and heavier appeal documentation, since disputes demand substantial clinical and authorization backup.

Managing that requires a defined cadence, not ad hoc follow-up. A workable pursuit cycle sets a first follow-up at 30 days, a second at 45, an escalation point at 60, and a clear disposition protocol for anything approaching a payer's filing deadline. Without that structure, claims sit aging quietly until they cross into write-off territory and the money is gone for good.

Claims that age past a certain point need different handling: payer-specific escalation paths, appeal packets with original authorization and clinical documentation, and secondary billing strategies where applicable. A standard billing workflow built around routine claims generally can't execute that case-by-case escalation consistently at volume.

The pressure on this workflow is only building. The 2025 State of Claims Report from Experian Health found denial rates climbed from an already significant 2022 baseline, rising further in 2024 and again in 2025, with prior authorization denials alone accounting for roughly a third of all denials. That's a growing volume of disputed claims funneling into the appeal queue, all needing to be worked before their individual filing windows close.

Practices that shift to a structured revenue cycle engagement see the payoff show up in the numbers. The Medical Billers and Coders RCM analysis found AR days drop and denial rates trend toward a low single-digit range within the first several months of that structured approach. The improvement comes from replacing scattered, reactive follow-up with a defined cadence applied consistently, not from any single fix.

MGMA's January 2026 Stat poll confirms denials and appeals rank as the single biggest revenue cycle leak, ahead of front-end issues, and AR problems arise when denials and underpayments aren't caught and worked in time.

The operating model for managing these variables together

Every variable covered (product mix, authorization status, billing accuracy, underpayment detection, AR pursuit) moves independently, and each can erase margin on its own. A perfect product mix doesn't matter if authorizations lapse mid-cycle. Flawless authorizations don't matter if the claim gets coded wrong. A correctly coded claim doesn't matter if it's underpaid and nobody checks the remittance against the contract. None of it matters if the AR sits unworked past a filing deadline.

That's the case for treating these as one operating system rather than five separate departments. Product mix decisions favoring biosimilar conversion need authorization requirements factored in, since an NDC mismatch on a converted patient voids the authorization instantly. Authorization tracking needs to run on expiration dates proactively, ahead of any denial landing. Billing needs built-in checks for 2026-specific failure points, including JW/JZ modifiers and site-neutral rate differences by location. Underpayment detection needs a live contracted-rate database checked against every remittance. And AR needs a defined 30/45/60-day cadence with escalation paths ready before a claim ages into write-off range.

None of these functions can compensate for failure in another. A practice with excellent billing accuracy but no underpayment reconciliation still bleeds margin quietly, claim by claim. A practice with airtight authorization tracking but no structured AR follow-up still lets aged claims become permanent losses. The spread appears in full only when every one of these variables holds at once, which makes infusion economics an operations problem rather than a pricing one.