In 2023, nine biosimilar versions of Humira launched in the United States. Adalimumab had been one of the best-selling medicines in the world for years, and the arrival of competition was supposed to be the moment the American biosimilar market finally grew up.
For most of that year, not much happened. Humira kept the bulk of prescriptions. Then, on April 1, 2024, CVS Caremark removed Humira from its major national commercial template formularies and covered biosimilars instead. Biosimilar use jumped almost overnight. Nothing about the science had changed in those months. What changed was who decided.
That is the argument worth making about biosimilars: their savings are not determined by the molecules, the regulators or even the prices. They are determined by who controls the switch, and the savings tend to land with whoever that is. Health systems that want the savings need to own the switch.
The science was settled long ago
It is worth clearing away one objection first. Biosimilars are not experimental. Europe approved its first biosimilar, Sandoz’s growth hormone Omnitrope, in 2006, under a regulatory pathway the EU had set up the year before, and has accumulated two decades of experience since. The question of whether patients stable on an originator can safely switch was tested directly in Norway’s NOR-SWITCH trial, published in The Lancet in 2017. It randomized 482 patients across six conditions, from Crohn’s disease to rheumatoid arthritis, either to stay on originator infliximab or to switch to the biosimilar CT-P13, and followed them for 52 weeks. Switching was not inferior to staying put, and serious adverse events were similar in both arms.
Regulators on both sides of the Atlantic require biosimilars to show no clinically meaningful differences from the reference product. For health professionals, the clinical debate is largely over. The remaining debates are about money, trust and logistics.
The switch that didn’t stick
A study published this year in Arthritis Care & Research, reported by Healio, used a US rheumatology registry covering more than 22,000 patients to look at what happened when CVS changed its formulary. Among commercially insured new users, prescriptions for the preferred adalimumab biosimilar rose from close to zero to 24 percent in the month after the change. The next month, that share fell by roughly 14 percentage points.
The authors pointed to the nocebo effect, where negative expectations lead patients to experience or perceive worse symptoms, along with administrative friction and too little education for patients about biosimilar equivalence.
That finding matters more than the headline savings. It says that a switch imposed by a payer, without clinicians and patients prepared for it, partly unwinds. Patients who feel moved against their will report problems, and clinicians under time pressure find it easier to request an exception than to explain.
Follow the money
Here is the uncomfortable part of the American story. Biosimilar savings in the US do not flow in a straight line from lower price to lower cost for patients.
Much of the savings depends on the gap between a drug’s list price and its net price after rebates. Originator manufacturers historically paid large rebates to pharmacy benefit managers in exchange for formulary position. A biosimilar with a lower list price and smaller rebates can look cheaper to a patient paying coinsurance on list price, while looking less attractive to an intermediary whose revenue depends on rebates.
The market’s response has been telling. Large PBMs launched their own private-label biosimilars, including CVS’s Cordavis arrangement for a version of adalimumab. Once an intermediary owns the preferred product, its incentive to switch patients becomes much stronger. Adoption follows.
That can be good for payers. It concentrates market power further, and it raises a fair question: are biosimilars creating a competitive market, or being absorbed into the existing one?
A worked example of where the money goes
A stylised calculation makes the mechanism clearer. The figures below are round numbers chosen for illustration, not real prices.
Suppose an originator biologic lists at 100 units a year per patient, and after a confidential rebate the plan pays a net 60. The 40 in between is shared out among the intermediary, the plan and, through premiums, employers. Now a biosimilar arrives listing at 50 with a small rebate, netting the plan 45.
From the plan’s point of view the saving is 15 units per patient, a quarter of the net cost. From the patient’s point of view, if coinsurance is charged on list price, the saving is much larger, because their bill is tied to 100 falling to 50. From the intermediary’s point of view, the rebate pool that funded part of its business shrinks from 40 to 5. Same molecule, same clinical result, three very different verdicts.
That is why who controls the switch matters so much. An intermediary paid out of rebates will be lukewarm about a low-list-price biosimilar until it has a way to capture a share of the margin, for instance by owning the private label. A public buyer in a tender system has no rebate business to protect and sees only the net saving. A hospital paid a percentage of drug cost on some reimbursement models may even lose revenue when it switches to the cheaper product. Each of these actors is responding rationally to the incentives in front of it.
Notice, too, what the example says about the patient. The person with the most to gain from the list-price drop, the one paying coinsurance, has the least say in the decision. They do not see the rebate, cannot negotiate the formulary and usually learn about the switch from a letter. Any design that expects them to embrace the change without explanation is asking the least informed party to absorb the most uncertainty. That is a recipe for the rebound the registry data show.
What Europe does differently
Several European health systems achieved high biosimilar uptake through public tendering and clinical leadership rather than intermediaries. National or regional buyers ran tenders for a molecule, awarded supply to the winning product, and worked with specialist societies to switch patients in an organised way, often with nurse-led education.
The savings in those systems accrued to public budgets, and in some cases were explicitly reinvested in clinical services. That reinvestment matters. When clinicians see the money they helped save funding more nurses in their own department, cooperation with the next switch comes more easily.
The model is not free of problems. Winner-takes-all tenders can drive out suppliers and create future supply risk. Frequent re-tendering can mean patients are switched repeatedly. But the core idea, that the health system rather than an intermediary owns the decision and the savings, is the one to copy.
The strongest objection
The best argument against the “own the switch” position is that centralised switching overrides the clinician-patient relationship. A patient with a hard-won remission on one product, managed by a rheumatologist they trust, is told by a distant buyer that their medicine is changing for budget reasons. Even if the evidence says the outcome will be the same, the experience is of being treated as a line item. On this view, a non-medical switch is a small breach of trust, and those breaches accumulate.
This objection partly holds. The nocebo rebound in the CVS data is exactly what that breach looks like at population scale. Patients are not wrong to notice when a decision about their body is made elsewhere, and “the evidence is settled” is not an answer they are obliged to accept.
But the objection argues for a better switch, not for no switch. The European experience suggests that when clinicians lead the conversation, when patients get a named nurse to call, and when savings visibly return to their own clinic, most accept the change. The alternative, leaving the switch to whoever has the market power to impose it, delivers the same override with none of the preparation. Ownership by the health system is the version of the switch in which clinicians actually have a voice.
Second-order effects
Getting biosimilar economics right has consequences beyond this year’s budget.
The first is on future competition. Biosimilar developers decide which molecules to pursue years in advance, based on whether they expect to win market share. If uptake depends on a handful of intermediaries choosing a private-label partner, independent developers face a market with one door, and fewer of them will bother. A market that looks competitive today can thin out tomorrow.
The second is on supply. Aggressive tendering that pushes prices toward the floor can leave one or two manufacturers standing, which is fine until a plant has a quality problem. Buyers who treat supply resilience as part of the price, by splitting awards or keeping a second supplier warm, pay a little more now to avoid shortages later.
The third is on the next wave. Many high-cost biologics face loss of exclusivity over the coming years. Health systems that build switching capability now, with protocols, education materials and pharmacy workflows, will be ready to capture those savings quickly. Those that treat each switch as a one-off will relearn the same lessons each time.
What specific actors should do
Ministries and national payers should decide whether they want biosimilar savings to accrue publicly and, if so, design procurement to make that happen: tenders with multiple winners, clinical input into switching timelines, and explicit rules for reinvesting a share of savings in the services that deliver them.
Hospital CEOs should look closely at how their reimbursement treats cheaper drugs. If the hospital loses margin when it switches, the incentive problem sits inside the building, and no amount of clinical education will fix it. Negotiating gain-sharing with payers is often the unlock.
Commercial payers and employers should ask their intermediaries a direct question: when a biosimilar is preferred, who captures the difference between old net price and new net price? The answer should be in the contract, not inferred.
Clinicians and pharmacists should treat the switch conversation as a clinical skill. A short, consistent explanation from a trusted nurse or pharmacist does more against nocebo than any policy memo, and specialist societies can supply shared scripts so the message does not depend on who happens to be on shift.
Funders and founders in biosimilar development should weigh channel risk as heavily as manufacturing risk. A product with excellent comparability data and no route past the gatekeepers of formulary position is a stranded asset.
What to watch over the next two years
Watch whether other large US intermediaries follow the private-label model and whether independent biosimilar makers respond by partnering or exiting. Watch the persistence data: whether follow-up studies show the post-switch rebound fading as patients and clinicians grow used to biosimilars, or settling at a stubborn level. Watch whether any US state or public payer experiments with European-style tendering for a biologic. And in Europe, watch whether supply disruptions in heavily tendered markets prompt buyers to redesign awards around resilience rather than price alone.
The takeaway
Biosimilars were supposed to bring competition to the most expensive corner of the drug market, and in part they have. But the savings are not automatic. They go to whoever controls the switch, and they leak away when patients are switched without being brought along. The science is done. The organising work is not.
Has your organisation run a biosimilar switch that stuck? What did you do to keep patients on board? Tell us in the comments.
