Dam. What a few weeks. I Whistlered while I worked, but the avalanche of news has been hard to keep up with. (I mean, I found it funny.)

If you’re interested, I’m hosting a free webinar on April 23 at 12 EST on Most Favored Nation. Quick 30-minute dive. Link to register is here: https://us06web.zoom.us/webinar/register/WN_iPlbMuS5RmeGSbcB8RSi-g

Loose Scripts Sink Ships. This week, AbbVie sued the Health Resources and Services Administration (HRSA), asking a federal court to answer a basic question that has been left hanging for three decades: who counts as a “patient” under 340B?

That is not rhetorical. The 72-page complaint, filed in the U.S. District Court for the District of Columbia, is a statutory interpretation fight over HRSA’s 1996 patient definition guidance, which the agency has never meaningfully updated. AbbVie argues that the guidance is so loose that it no longer reflects the best reading of the statute, and the examples in the complaint are designed to make that point impossible to ignore.

According to AbbVie, one nurse practitioner at a small San Antonio federally qualified health center wrote 225 prescriptions for Skyrizi in a single quarter, putting that prescriber above more than 99 percent of prescribers nationwide. The complaint also says that three quarters of that health center’s 340B purchases were dispensed through pharmacies outside Texas. And it isn’t like San Antonio is close to other states.  In another example, three separate Mount Sinai hospitals allegedly claimed 340B discounts on the same prescription, on the same date of service, for the same National Drug Code.

AbbVie is not asking HRSA to rewrite its guidance. It is asking the court to say that HRSA’s 1996 definition is not the best reading of the law, and that manufacturers should be allowed to rely on a narrower standard when they audit covered entities. HRSA’s current position is effectively the opposite: manufacturers may conduct audits, but the agency will not enforce findings that apply a stricter patient definition than the one in the 1996 guidance.

AbbVie says the proper standard should require four things: a direct connection between the prescription and the covered entity’s care, a substantive medical encounter rather than a perfunctory one, a 12-month time limit, and direct provider oversight of the condition being treated.

None of that is radical, but it would narrow a great deal of current 340B volume. That is exactly why this case matters, not just for AbbVie, but for manufacturers across the program.

PIVOT. This week, Axios reported that states are already absorbing the impact of last year’s federal Medicaid cuts, well ahead of the biggest changes taking effect in 2027.

The numbers are stark. State budgets are due to shrink by $664 billion over the next decade. Idaho legislators approved $22 million in cuts to Medicaid disability services. Iowa passed a new tax on certain health insurers to cover an existing shortfall. Colorado is debating benefit cuts, lower provider payments, and a proposed employer tax on companies with part-time Medicaid enrollees. Similar bills are being floated in Washington state and New Jersey. As one Colorado legislator told Axios: “everyone is pissed and everyone is stressed.”

Also this week, a study out of Beth Israel Deaconess Medical Center found that among 16.5 million Medicaid beneficiaries ages 19 to 64, 50.4% — or 8.3 million people — would be at risk of disenrollment once work requirements are fully implemented. That’s how work requirements work when a significant share of enrollees work non-traditional hours, are caregivers, or face barriers to documentation.

And KFF Health News reported that states are paying Deloitte, Accenture, and Optum millions to build improved eligibility technology systems to implement those work requirements. The administrative cost of cutting Medicaid is itself a significant budget line.

As the National Conference of State Legislatures put it: Medicaid programs are like massive ocean liners. They’re already in the process of pivoting. The question isn’t whether cuts are coming. It’s who absorbs the shock first.

Weighing the Math. This week, a research letter published in JAMA Internal Medicine found that the administration’s projected savings from the BALANCE Model would cover costs for an estimated 4.4% of newly eligible Medicare patients in year one.

The administration’s pitch for the BALANCE Model has been that lower negotiated prices offset the cost of expanding GLP-1 coverage to millions of newly eligible Medicare patients. No net cost to taxpayers. Win-win.

The JAMA analysis puts a number on the gap. Savings in year one (~$900 million) would cover costs for that 4.4% of patients. Not the other 95.6%.

4.4%.

The April 20 deadline for insurers to decide whether to join the optional program is days away. The administration has said it won’t proceed unless insurers covering 80% of the Medicare population sign on. That threshold hasn’t been confirmed as met. The JAMA numbers don’t make the pitch easier. And honestly, I’m not entirely sure plans are up for it.

I do believe GLP-1s are magic — and there are reasonable arguments that year-one costs don’t capture the long-run picture. If GLP-1s reduce hospitalizations, cardiovascular events, and downstream spending, the math changes. But that’s a different argument than the one being made.

What I’m curious about is that the Bridge program runs no matter what – so beneficiaries will have coverage from July to the end of the year and then what?

It Must Be True.  This week, Aetna released a provider survey finding that 65% of providers agree prior authorization has a legitimate role in the health system. Interesting framing from a payer, because it validates the tool while suggesting the 35% who disagree are the outliers. Meanwhile, less than half of those same providers (44%) said their current payers prioritize patient well-being and clear information. Only 36% believe insurers follow through on their commitments.

So providers accept that PA exists. They just don’t trust the people running it.

Show Me the Coupon. This week, a research letter published in JAMA found that manufacturer coupon use among commercially-insured patients dropped from 18% in 2017 to 13.9% in 2024, based on analysis of more than 55 million pharmacy claims.

Two things underneath that number are more interesting. First, per-claim coupon amounts went up: median $60 in 2017 to $90 in 2024. Manufacturers are spending more per coupon while reaching fewer patients. That reflects higher underlying cost-sharing requirements that coupons need to offset, and copay accumulator programs squeezing coupon effectiveness.

Second, the GLP-1 story. Coupon use for obesity drugs collapsed from 54.6% in 2017 to 2.5% in 2024. As coverage expanded (unevenly), manufacturers pulled back cash-pay coupon programs. By contrast, coupon use for immunomodulators shot up from 4.2% to 23.8% over the same period.

The immunomodulator increase is worth watching. Formulary exclusions and step therapy are getting more aggressive in that class. When coupons are the affordability bridge and payers are restricting the formulary lane, patients end up in the middle. That’s not a drug pricing solution. It’s a different kind of access problem.

If I had a Trillion Dollars. At the end of March, David Cutler and Lev Klarnet released a Brookings Papers on Economic Activity conference draft asking whether the U.S. has bent the health care cost curve. Their answer is yes, and the numbers are striking. Medical spending in 2024 was $977 billion below what CMS actuaries projected back in 2010. Cumulative gap since 2010: $6.7 trillion. Health spending came in at 18% of GDP versus a projected 21.2%.

They attribute the slowdown to five things: technology maturation, patent expirations, site-of-care shifts (major surgery moving from inpatient to outpatient), demand-side changes (prior auth, high-deductible plans, ACOs), and slower price growth. Prescription drugs account for 26% of the spending gap, driven largely by loss of exclusivity and slower branded drug growth.

The caveat: spending growth picked back up in 2023-24. And they’re clear the curve hasn’t bent enough. Higher cost-sharing and insurer restrictions that drove some of the slowdown have also created access problems.

I mention this paper now because almost every mechanism they identify as having bent the curve is currently under pressure. Tariffs threaten supply chain economics. Medicaid cuts change the demand-side picture. IRA litigation creates uncertainty around price negotiation. GLP-1 coverage expansion could reverse some of the pharmaceutical spending slowdown. The curve bent for reasons. The question is whether those reasons survive what’s coming.

Reimbursement Fundamentals: Tariff-ied and Confused

Last week, the Trump administration announced 100% tariffs on imported patented brand-name drugs and APIs under a Section 232 national security proclamation. The tariff applies to products, specific HTS codes listed in Annex I, not to companies or deals. What a signed agreement changes is the rate you pay on those products. Sign both an MFN pricing deal with the Department of Health and Human Services (HHS) and an onshoring plan with Commerce, and your rate drops to 0% through January 20, 2029. Sign just an onshoring plan, and you’re at 20%, rising to 100% by 2030. Sign nothing, and you’re at 100% starting July 31 if you’re on the Annex III large-company list, or September 29 if you’re not. The tariff is the floor. The negotiation determines how far above it you sit.

But honestly, this week’s announcement is less a starting gun than a deadline. The real story started in July 2025, when the administration sent demand letters to 17 large pharmaceutical manufacturers. By early February 2026, 16 of those 17 had agreed to bilateral MFN pricing agreements. They’re the early movers. Annex II of the proclamation lists 13 companies that formalized agreements between December 2025 and March 2026, so even the “signed” group has layers depending on which moment in time you’re counting.

And now we have a proclamation structured to pull the rest of the industry into the tent. For the largest manufacturers in Annex III, the 100% rate hits July 31. For everyone else, September 29.  There’s another angle that needs to be figured out: companies with approved domestic onshoring plans that haven’t signed MFN pricing agreements can access a 20% rate instead of 100%, but that rate climbs back to 100% by April 2030. It’s not binary. It’s a sliding scale with a clock.

Now. What do you actually get if you sign? And remember – tariffs are on a product, not company basis so deals already signed will likely only provide relief for the drugs specifically included in them.

The obvious answer is tariff relief, a 0% rate through January 20, 2029, if you execute both an MFN pricing deal and a Commerce-approved onshoring plan. But the less-discussed piece is the FDA angle. Participating manufacturers were offered FDA Commissioner National Priority Vouchers (CNPVs) that can reportedly reduce drug or biologic application review times from the standard 10 to 12 months down to 1 to 2 months. I want to be careful here because I haven’t seen full public documentation on this, so flag it as unconfirmed, but if accurate it’s a significant competitive advantage for getting pipeline assets to market ahead of non-participating competitors. That’s a carrot that has nothing to do with pricing and everything to do with market timing.

The investment commitments are also worth naming. Companies collectively pledged somewhere between $150 billion and $400 billion (the range is wide, which tells you something about the precision of these agreements) in U.S. manufacturing infrastructure, R&D, and capital expenditures. They also agreed to donate active pharmaceutical ingredients to a Strategic API Reserve. The manufacturing pledge piece connects back to the onshoring carveout, though the criteria for what counts as a sufficient onshoring plan haven’t been published in the Federal Register yet. Finished dose manufacturing and API manufacturing are different things, and that distinction matters enormously for what companies can actually deliver.

The bilateral deal structure is the clearest window into where this goes internationally. The U.K. deal was formalized alongside the April proclamation and is the template. The U.K. secured tariff-free access to the U.S. market for British-made medicines for at least three years. In exchange, they accepted higher prices for U.S. drugs and agreed to shift NICE’s appraisal framework. For the first time since 1999, the QALY threshold will rise to £25,000–£35,000, which effectively allows the NHS to pay up to 25% more for innovative treatments. The U.K. agreed in principle to a framework shift without specifying what it means in practice. The vagueness is telling.

Switzerland is in conversations. Australia is holding, for now. (CSL, Australia’s largest pharmaceutical exporter, just completed a $1.5 billion manufacturing expansion in Illinois. The timing is not subtle.)

The trade is pretty clean from the U.S. perspective. Trading partners preserve pharmaceutical export access to the U.S. market. In return, they pay more for U.S. drugs, which makes their prices less useful as downward anchors in MFN reference calculations. And HTA methodology shifts open reimbursement lanes for U.S. products that might not have cleared previous cost-effectiveness thresholds. I mean, if you’re trying to build an MFN framework that benchmarks U.S. prices to international ones, and you can simultaneously raise international prices through bilateral deals, the reference basket moves in your favor over time. Whether that’s the explicit design or a convenient side effect is a good question.

Meanwhile, at the end of March, Reuters reported that manufacturers are quietly pausing European launches to avoid creating reference prices that feed back into U.S. MFN calculations. European patients wait longer. The reference basket gets thinner exactly as the policy depends on it being representative. And yet that’s the system we’re building.

Here’s where it gets operationally complicated for companies that have signed or are considering signing.

The IRA collision is live and specific. The negotiated Maximum Fair Price for Novo Nordisk’s semaglutide products under the IRA was set at $274. The voluntary MFN agreement set a Medicare price of $245. Same molecule, two different federal pricing mandates, no clear legal resolution. This isn’t a hypothetical future conflict. It’s happening now.

There’s also an international data-sharing problem that’s getting almost no attention. To comply with MFN benchmarking, manufacturers may have to share detailed international net pricing data with the U.S. government, including confidential rebates in foreign markets. That data-sharing may violate confidentiality obligations with foreign jurisdictions or international distribution partners. If your international contracts have confidentiality clauses covering net pricing, and most do, you have a conflict that needs legal review before you sign anything.

And the commercially reasonable efforts (CRE) problem. Most licensing and commercialization agreements contain CRE clauses requiring manufacturers to maximize a product’s profitability in specific international markets. Once you’ve signed a U.S. MFN agreement, your international pricing is tethered to your U.S. price. Drop a price in Germany, and you may trigger a rebate obligation back in the U.S. That puts you in direct tension with your international partners and potentially in breach of existing contracts. It’s not insurmountable but it requires active management, and the companies I’d worry about most are mid-size manufacturers with complex international licensing arrangements who didn’t have legal teams deeply involved in the deal-making process.

Overwhelmed yet?

The codification piece is where this all converges. The GLOBE and GUARD models, CMS’s mandatory rebate proposals for Part B and Part D drugs benchmarked to international prices, are pending finalization (maybe). Congress is being pushed by the administration to codify the bilateral deals into permanent law, though there’s apparently significant frustration on the Hill about the confidentiality of the deal terms. (Which, fair.) If GLOBE and GUARD finalize, MFN stops being a political deal and becomes a structural feature of Medicare reimbursement. No carveout, no negotiation.

And here’s the thing: a manufacturer that signed a voluntary TrumpRx agreement, accepted the press hit, and agreed to a cash-pay price may find that voluntary price becomes the floor for mandatory rebate calculations under GLOBE. The voluntary becomes the norm. The norm gets codified. The codified price becomes the ceiling for what’s negotiable in the next round. It’s not a conspiracy. It’s just how these things work.

The July 31 and September 29 deadlines are real. The window for favorable terms, with onshoring plan criteria still undefined and bilateral deal structures still forming, is probably better than what’s available after GLOBE finalizes. But signing without legal review of your international contracts, your IRA obligations, and your CRE clauses is how you create problems that are worse than the tariff rate you’re trying to avoid.

Or maybe this a scary TACO waiting to be punted.