The Number Everyone Will Quote About European Launches Rests on Eleven Products
An analysis of European launch data since the Most-Favored-Nation executive order contains one very strong finding and one very fragile one, and the fragile one is travelling further. The pricing signal is clean and hard to explain away. The claim that EU Joint Clinical Assessment is already suppressing launches rests on a subgroup that fell from 48 products to 37, and the same table shows the opposite direction among non-orphan medicines.
- The pricing spillover signal is the durable part. US non-orphan launches were exactly flat at 50 to 50 across matched twelve-month windows, while the same category fell between 32% and 46% across the reference markets that anchor US benchmarks. A general industry slowdown would have moved both.
- The assessment finding is far thinner than its headline. The 22.9% figure attributed to Joint Clinical Assessment is orphan products inside JCA scope falling from 48 launches to 37, a difference of eleven brand-country events over one year.
- The same table points the other way for the larger group. Among non-orphan medicines in the eleven EU states, launches inside JCA scope fell 41.4% against 47.3% outside it. On those numbers, being in scope was associated with a smaller decline, not a larger one.
Both policies plausibly discourage European launches, and the authors are careful about their own limits. The problem is downstream. A finding built on eleven events is being carried into a debate about whether Europe should rethink a framework designed to replace twenty-seven separate national assessments. The pricing evidence can bear that weight. The assessment evidence cannot yet. A reader can stop here with the full picture. The sections below are the detail.
What the analysis is, and what it actually shows
The piece published in the Journal of Comparative Effectiveness Research on 19 June is an editorial rather than a peer-reviewed research article, by Sreeram Ramagopalan of the Centre for Pharmaceutical Medicine Research at King’s College London and Michael Ryan of Koios Enterprises and Consulting. It uses a GlobalData extract to compare brand-country launch events across matched twelve-month windows either side of the executive order of 12 May 2025. The authors declare no competing interests and no financial support.
The mechanism it tests is more specific than the shorthand suggests. Three pilots sit under the order: GENEROUS, a voluntary programme for Medicaid, and GLOBE and GUARD, mandatory proposed rules covering a quarter of Medicare Part B and Part D respectively. GLOBE targets seven therapeutic classes above a $100m Part B spending threshold; GUARD covers seventeen classes above a $69m Part D threshold. Generics, biosimilars and products with an Inflation Reduction Act negotiated price are excluded. There is no orphan carve-out, but the spending thresholds shield most rare disease products in practice.
The headline cohort figures are large. Launches fell 10.5% in the US, 29.9% across the nineteen-country reference basket, and 37.8% in the eleven EU member states exposed to both policies, where the total moved from 527 events to 328.
The pricing finding is clean, and it is the one worth acting on
The most persuasive number in the analysis is the one receiving least attention. US non-orphan launches did not move at all, fifty in the year before the order and fifty in the year after, while the same non-orphan category fell 31.9% across the nineteen-country basket and 46.1% in the eleven EU states.
That asymmetry is difficult to explain by anything other than reference pricing. A general post-2025 slowdown in approvals or commercial appetite would have depressed both sides. Manufacturers holding US launches steady while withdrawing from precisely the markets whose prices feed the US benchmark is the behaviour the policy design predicts.
Brand-country launch events in the eleven EU member states subject to Joint Clinical Assessment, matched twelve-month windows either side of 12 May 2025, as reported in the source table.
J Comp Eff Res 15(7) · Editorial, GlobalData extract · Compiled 21 August 2026
The assessment finding is a difference of eleven products
The paper’s second claim is the one that has driven coverage. Orphan medicines outside JCA scope barely moved, from 73 launches to 72. Orphan medicines inside JCA scope, meaning orphan oncology and orphan advanced therapies, fell from 48 to 37. Because rare disease products are largely shielded from reference pricing by the spending thresholds, the authors read that gap as JCA preparation burden showing through.
The logic is sound and the design is genuinely clever. The difficulty is the denominator. Eleven fewer launch events in a single year, in a category where a handful of programme delays or a single company resequencing a portfolio would produce the same movement, is a thin base for a structural conclusion. No confidence interval is reported for the interaction.
The larger group cuts the other way. Among non-orphan medicines, in-scope launches fell 41.4% and out-of-scope launches fell 47.3%. If JCA burden were the operative mechanism, that ordering is hard to account for. The authors address this by arguing the dominant pricing pressure swamps any marginal assessment effect in US-revenue-heavy products, which is reasonable, but it means the JCA case rests entirely on the smallest cell in the table.
The planning question is real even where the evidence is not yet settled
Nothing here says JCA is costless. The preparation burden described is well documented independently: assessments can generate dozens of population, intervention, comparator and outcome combinations across member states, the dossier must be ready in parallel with EMA review, and a single-asset biotech may lack the capacity to build a health technology assessment package answering twenty-seven sets of questions.
Two dated events make this a planning matter rather than a debating one. JCA extends to all orphan medicines in 2028. Separately, provisions in the EU pharmaceutical package would expose manufacturers to the loss of twelve months of market protection where an approved medicine is not launched across member states within three years. A company can face an evidence burden it is not ready for and a penalty for not launching, in the same year.
The commercial instruction is therefore the same whichever way the JCA question resolves. Build the assessment dossier on the regulatory timeline rather than after it, because both the burden and the launch conditionality land in 2028 regardless of whether the current signal is real. The evidential instruction is different: treat the pricing spillover as established enough to model, and treat the assessment effect as a hypothesis with one year of data and one small subgroup behind it.
What would settle it is a second year. If the orphan in-scope decline persists into the 2026 to 2027 window, on a larger base, the interaction becomes difficult to dismiss. If it reverts, the eleven events were noise. That test arrives before the 2028 expansion does, which is unusually convenient timing for anyone who has to make the decision.