The Bigger They Are, the Faster They Grow. A new white paper from Health Capital Group, sponsored by Johnson & Johnson, analyzed HRSA’s OPAIS database from 2017 through 2023 and found something many observers of the 340B program have suspected for years; size is the dominant variable.
Hospitals in the top quartile by bed size represent about 33% of all 340B hospital participants but captured 81% of the growth in child sites and 60% of the growth in contract pharmacy relationships over that period.
Academic medical centers push the gap even further. Major teaching hospitals now operate nearly twice as many child sites and contract pharmacy relationships as similarly large non-teaching hospitals. And the difference is growing. Teaching hospitals had 66% more child sites than non-teaching hospitals in 2017. By 2023 that gap had widened to 95%.
As you may know, child sites are how hospitals extend their 340B footprint. Prior research has shown that many of these child sites cluster in higher income areas with commercially insured patients rather than underserved communities. Contract pharmacy relationships allow hospitals to dispense 340B drugs through outside pharmacies while capturing the difference between acquisition cost and reimbursement.
For manufacturers, this concentration helps explain why the contract pharmacy fight has escalated. When a relatively small group of very large systems captures most program growth, the financial stakes rise quickly. What began as a program meant to help safety net providers stretch limited resources increasingly looks like a story about scale, spread revenue, and who ultimately captures the benefit.
Do Less, Prove the Same. On March 9, the Food and Drug Administration (FDA) released draft guidance that would eliminate the requirement for three-way pharmacokinetic studies in biosimilar development when sponsors can demonstrate the comparison is scientifically unnecessary. The move builds on October guidance that removed the requirement for comparative efficacy studies, a change estimated to save sponsors about $24 million and one to three years per application.
The agency’s logic is straightforward. If a sponsor has already demonstrated biosimilarity through molecular characterization and two-way pharmacokinetic comparisons, requiring a third arm simply to confirm the same conclusion adds cost without adding meaningful scientific insight.
The FDA estimates eliminating the requirement could cut up to half of pharmacokinetic development costs, which often run around $20 million per application.
There are now 82 FDA-approved biosimilars. And reducing unnecessary development costs changes the economics of entry, but the marketplace is still a work in progress. With Medicare negotiated prices adding instability to a market that is already wobbly with the reliance on private-label biosimilars, it is uncertain what the future of biosimilars will be in the United States. If the goal is more competitive biologics markets, lowering the cost of getting to the market isn’t a bad thing, but it isn’t the only thing.
Show Me the Outcomes. On March 7, the Center for Medicare and Medicaid Innovation staff published a perspective in the New England Journal of Medicine outlining where the agency plans to take the Cell and Gene Therapy (CGT) Access Model.
The model launched in 2025 across 32 states, Washington DC, and Puerto Rico, initially focused on gene therapies for sickle cell disease. It is outcomes-based contracting at scale. If a Medicaid patient does not benefit from a covered therapy, the state receives a rebate from the manufacturer.
The design addresses a real barrier; state Medicaid programs are understandably cautious about paying $1 million to $4 million upfront for a therapy when long-term durability is still being evaluated. The CGT model attempts to shift part of that uncertainty back to manufacturers while giving states a predictable framework for coverage.
The article outlines three priorities for the next phase of the model: prioritize therapies with enough Medicaid patients to allow meaningful evaluation, encourage competition rather than single product arrangements where only one therapy is available, and address provider-level barriers.
Hospitals and infusion centers can be reluctant to administer therapies that cost several million dollars when reimbursement does not reliably cover acquisition costs. Most providers cannot easily float that kind of exposure while waiting for payment.
The stakes are growing quickly. Analysts project that 85 gene therapy indications could reach the market by 2033, with annual U.S. spending potentially reaching $21 billion by 2034. The CGT model represents an attempt to build payment infrastructure before that wave arrives.
Trust (Maybe) But Verify. On March 3, seven Democratic senators led by Senate Finance Committee Ranking Member Ron Wyden sent a letter to GSK CEO Luke Miels asking a specific question: are the prices offered through the company’s GENEROUS model actually lower than what Medicaid already pays?
That question goes to the heart of the model’s public messaging. Medicaid already pays extremely low net prices because of statutory rebates. A Congressional Budget Office analysis found that average net prices for top selling brand drugs were about $118 in Medicaid compared with $343 in Medicare Part D.
The senators also noted that the arrangement appears to include potential advantages for the manufacturer, including tariff relief and the possibility of FDA priority review.
There is also a transparency issue built into the model. The request for applications notes that terms may differ from what the Centers for Medicare & Medicaid Services (CMS) publicly released, which means there is no clear external benchmark for evaluating the final deals.
The senators asked GSK to provide drug-by-drug pricing details by March 23. Whether that response materializes, and what it shows, will determine whether the model represents meaningful savings or primarily a new narrative.
Now Is Always the Right Time to Do the Right Thing. On March 10, Representative Jake Auchincloss of Massachusetts published a Substack outlining a bipartisan framework for Alzheimer’s policy developed with a Republican colleague from Arizona.
By 2050 as many as 15 million Americans could be living with Alzheimer’s disease, requiring an estimated 60 million caregivers and generating annual costs approaching $1 trillion. Those numbers are not ideological; they describe a looming fiscal and human burden.
The proposal has three components: push funding through expanded federal research investment, pull through Medicare reimbursement policies that encourage development, and partner through a national clinical trial network.
The proposal also suggests expanding Medicare coverage beginning at age 50 for diagnosis and preventive interventions, not simply treatment after symptoms appear. (Which would be so cool. Medicare is going to have to absorb those costs eventually anyway, so why not intervene before they compound. We should be doing this for more chronic conditions.)
Auchincloss also highlights a structural issue with the Inflation Reduction Act’s (IRA’s) drug negotiation framework. Small molecules face negotiation timelines earlier than biologics. In Alzheimer’s disease that distinction matters because many large molecules struggle to cross the blood-brain barrier.
If the pipeline for brain diseases relies heavily on small molecule therapies, earlier price pressure could discourage investment. That argument is beginning to surface across several disease areas, but Alzheimer’s makes the stakes visible.
It Worked, It Really Worked. A new study published in JAMA Internal Medicine finds that the IRA provisions eliminating the 5% catastrophic coinsurance requirement and expanding low-income subsidies measurably improved medication adherence among Medicare beneficiaries.
Researchers compared 1,454 Medicare beneficiaries with 3,797 privately insured individuals. Cost-related medication nonadherence among Medicare beneficiaries fell by about 4.9 percentage points. Among patients managing multiple chronic conditions, the reduction reached 7.8 percentage points.
None of that is surprising in direction. Lower out-of-pocket (OOP) costs generally improve adherence. What makes the study meaningful is the magnitude of the effect.
The OOP cap is amazing, but it gave me pause because the researchers did not find broad improvements in overall financial strain among beneficiaries. People were more likely to take their medications, but the underlying affordability pressures did not disappear.
More research needs to be done. Especially as we see the OOP cap go up year over year.
Is He Disappointed? CMS Administrator Dr. Oz has suggested publicly that Affordable Care Act marketplace enrollment is artificially inflated. He has argued that the current figure of roughly 23 million enrollees could fall to about 19 million once improper enrollments are removed.
The estimate he cites comes from the Paragon Health Institute, which argues that 4 to 5 million people may be enrolled improperly. Other analysts strongly dispute that scale. KFF’s Cynthia Cox says fraud and enrollment errors do occur but are likely to affect hundreds of thousands of people, not millions. Brookings scholar Richard Frank described the higher estimate as implausible.
At the same time CMS is proposing to expand eligibility for catastrophic marketplace plans beyond the current under-age-30 thresholds. These plans typically carry deductibles of around $10,600 for individuals and $21,200 for families.
The policy argument is that younger and healthier consumers are increasingly priced out of traditional marketplace plans and catastrophic coverage offers at least some form of insurance.
But a plan with a $10,600 deductible raises an obvious question — for people who cannot realistically absorb that level of cost exposure, how different is that from having no insurance at all. You could maybe tie it with a GoFundMe if a catastrophic event occurred which, to be fair, is a real American healthcare strategy at this point — just not a good one.
If enrollment levels fall and the remaining coverage options shift toward high-deductible catastrophic plans, the debate may become less about how many people are covered and more about what that coverage delivers.
Reimbursement Fundamentals — BALANCE of Power: CMS Tests a Narrow Path to GLP-1 Coverage
CMS announced details of the BALANCE Model this week. The headline version is that Medicare is testing coverage of GLP-1s for weight management. The fine print version is more interesting and more honest about what CMS is and isn’t willing to do here.
Here’s the statutory problem they’re working around: The Social Security Act prohibits Medicare Part D plans from covering drugs “used for anorexia, weight loss, or weight gain.” That exclusion is why Wegovy isn’t covered when prescribed for obesity, even though the same molecule (semaglutide) is covered when prescribed for type 2 diabetes. BALANCE doesn’t repeal that exclusion; what it does is test whether CMS can thread the needle: cover drugs widely known as obesity medications by framing eligibility around cardiometabolic disease.
Most beneficiaries entering the model won’t qualify because they’re obese. They’ll qualify because they have type 2 diabetes, or MASH with moderate to advanced fibrosis, obstructive sleep apnea, or heart failure with preserved ejection fraction, or chronic kidney disease stage 3 or higher with obesity as a co-condition. The closest thing to pure obesity treatment in the model is the BMI ≥35 pathway, which requires active lifestyle modification participation. Even there, obesity is being treated as a risk factor managed through behavioral intervention, not a standalone disease.
That’s a deliberate design choice.
The cardiometabolic framing does a few things at once. Clinically, it targets the populations where the evidence for GLP-1 benefit is strongest and where downstream cost savings because of reduced hospitalizations and fewer cardiovascular events are most plausible. Politically, it potentially keeps the model from blowing up the Medicare budget. Broad coverage for obesity alone could add tens of billions annually to federal spending. Narrow comorbidity criteria make the eligible population manageable. And legally, it keeps CMS on the right side of the statute by treating these drugs as treatments for recognized disease states rather than weight loss per se.
The structure also mirrors what commercial insurers already do; use prior authorization tied to cardiovascular risk or metabolic disease. BALANCE embeds those conditions directly into eligibility rather than layering them on top as administrative gatekeeping.
If the data show improved outcomes and manageable spending among these high-risk populations, that’s the argument for broadening coverage. If costs climb without clear offsets, it reinforces the case for keeping the exclusion in place. CMS has structured the experiment to give themselves the best shot at a positive result by including high-risk patients, strong clinical rationale, tight eligibility. That’s smart design, but it also means the results won’t generalize cleanly to broader obesity coverage.
And the patients who are obese but don’t clear the comorbidity thresholds? They’re not in this model. They’re still on the other side of a statutory exclusion that BALANCE doesn’t touch.
That’s the thing to hold onto. This isn’t Medicare covering obesity drugs. It’s Medicare testing whether cardiometabolic disease can be treated with drugs that also produce weight loss. Those are related but not the same policy and conflating them will cause problems down the road.
The model creates a structured federal test of whether these therapies fit into Medicare. But the design tells you exactly how much political and legal weight CMS thinks it can carry right now. More to come when we see who enrolls and at what cost.
