The EU HTA Regulation and Joint Clinical Assessments: What Changed in 2025, and What Arrives in 2028 and 2030

Since 12 January 2025 there has been a new layer of assessment sitting between an EMA marketing authorisation and a national reimbursement decision. Regulation (EU) 2021/2282 created it, and the thing it produces — a joint clinical assessment — is unusual in EU law for being mandatory to consider and deliberately silent on the question everyone actually wants answered.

A joint clinical assessment contains no value judgement and no reimbursement conclusion. It assesses clinical evidence once, at Union level, for all Member States. Whether the technology is worth paying for remains entirely a national matter. Understanding that division is the whole of understanding the Regulation.

The staggered scope, and why 2030 is the date that matters

The Regulation applies from 12 January 2025, but not to everything at once. For medicinal products going through the centralised procedure, the phase-in runs:

• 12 January 2025 — new medicines for the treatment of cancer, and advanced therapy medicinal products.
• 13 January 2028 — orphan medicinal products.
• 13 January 2030 — all new medicinal products.

New therapeutic indications for products that have already been through a joint clinical assessment are also brought into scope, which means a product assessed in the oncology or ATMP wave acquires an ongoing relationship with the process rather than a single encounter.

The 2025 wave is narrow enough that a great many developers have treated the Regulation as a future problem. The practical difficulty with that is the 2030 cohort is every centrally authorised medicine, and the capability it needs — dossiers written to several national evidence questions at once — is not something that can be stood up in the quarter before it is required.

How a joint clinical assessment actually starts

The trigger is the marketing authorisation application, not the authorisation. Before or when an MAA is submitted to the EMA, the developer sends the HTA secretariat the summary of product characteristics and the clinical overview from the application. That material is what the scope of the assessment is built from.

The EMA separately informs the HTA secretariat that the application has been submitted. The assessment formally begins when an assessor and a co-assessor are appointed, and the secretariat then tells the developer. Developers notify the secretariat by email before uploading anything and receive a personalised, product-specific access link to the HTA IT platform; where early information was already submitted, the existing link is reused.

None of that is complicated, but it is sequenced, and the sequence is driven by regulatory milestones a market access team does not control. This is a familiar pattern for anyone who has worked to a device registration timetable — the deadlines belong to someone else's process, as with the staged EUDAMED module mandates and UDI registration deadlines.

Scoping, PICO, and the part that is genuinely hard

The scoping stage is where the work is. The assessor and co-assessor draft an assessment scope proposal setting out the research questions, in PICO form — population, intervention, comparator, outcomes. Members of the joint clinical assessment subgroup then comment on the proposed scope from their own national perspective, with input from individual experts. The subgroup finalises the scope and shares it with the developer in the Commission's initial request.

The timing is tight and tied to the EMA's clock: within 10 days after the CHMP adopts its list of questions, or at the latest 75 days after EMA validation in accelerated procedures and for variations. The developer may request an assessment scope explanation meeting with the subgroup.

The structural consequence is the one worth planning around. Because Member States comment on scope from national perspectives, a single joint assessment can carry multiple PICOs reflecting different standards of care across Europe. One dossier therefore has to answer several comparator questions simultaneously, and it has to do so against a scope that is not final until well after the evidence package was designed. That is a trial design and evidence generation problem long before it is a submission problem.

What Member States must do with the report

National authorities are obliged to give due consideration to published joint clinical assessment reports, and to annex the developer's dossier and the published report to their national HTA documentation. That is the extent of the compulsion.

Member States retain the mandate over pricing and reimbursement, and may add other information at national level — cost-effectiveness analysis being the obvious example. So the Regulation removes the duplication of clinical assessment without touching national decision-making on value. A developer who reads "joint assessment" as "joint decision" will be surprised; a developer who reads it as "the clinical argument is now settled centrally and the economic argument is not" has it right.

Devices work on a different principle entirely

For medical devices and in vitro diagnostics, scope is not defined by category and date. It is defined by selection.

Class IIb and class III devices under the Medical Device Regulation, and class D in vitro diagnostics under the IVDR, may be subject to a joint clinical assessment — but only where the relevant expert panel has issued an opinion or a view. Each quarter the EMA shares the previous three months' expert panel opinions and views with the HTA Secretariat. The Coordination Group may then recommend devices from that list against six criteria, of which a device must meet at least one. The Secretariat tells developers and notified bodies whether a device was recommended, and that information is not public at that stage. Only after certification can the Commission adopt an implementing decision formally selecting the device, which is then published in the joint clinical assessment tracker.

Two things follow. A device manufacturer cannot determine from the Regulation alone whether a product will be assessed, because the answer depends on an expert panel opinion and a discretionary recommendation. And the assessment happens after notified body certification, so it sits alongside rather than inside the conformity assessment route. Device teams already carrying the accountability obligations of the Person Responsible for Regulatory Compliance under MDR Article 15 are now also tracking a selection process they have limited visibility into.

The implementing acts, for anyone building a procedure

Six implementing acts give the Regulation its operational detail, and a compliance procedure that cites only the parent Regulation will be missing most of what it needs:

• joint clinical assessment of medicinal products for human use (OJ L_202401381);
• exchange of information with the European Medicines Agency (OJ L_202402699);
• conflict-of-interest rules — Regulation (EU) 2024/2745;
• joint scientific consultations for medicinal products — Regulation (EU) 2024/3169;
• joint scientific consultations for medical devices and in vitro diagnostics — Regulation (EU) 2025/117;
• joint clinical assessment of medical devices and in vitro diagnostics (OJ L_202502086).

The competence question, stated honestly

The HTA Regulation imposes no training obligation on anyone. It is worth saying so plainly, because the surrounding regimes in life sciences mostly do, and the contrast invites an assumption that is simply wrong.

What the Regulation creates is a competence need without a competence rule. Market access teams now have to work to a scoping timetable anchored on CHMP milestones rather than on national submission windows. Regulatory affairs staff find that the clinical overview in the MAA has acquired a second audience with different questions. Biostatisticians and evidence generation leads face comparator questions set partly by national standards of care they may not have mapped.

That is a different shape of problem from the explicit, auditable training duties elsewhere in the sector — the delegation-based training expectations in ICH E6(R3) good clinical practice, the qualification requirements attaching to the Qualified Person for Pharmacovigilance, or the competence and effectiveness-assessment architecture now embedded in the FDA's Quality Management System Regulation. Those regimes tell you what to evidence. This one does not, which makes it easier to neglect and harder to defend neglecting.

Our suggestions

Three practical points. Work backwards from the CHMP list of questions, because that milestone — not the national submission date — now sets the scoping clock. Treat multiple PICOs as the default assumption in evidence planning rather than a complication to handle if it arises; designing for one comparator and discovering four is the expensive version of this. And if you are in the 2030 cohort, note that the preparation window is the trial you are designing now, not the submission you will make then.

For developers tracking parallel timetables, the Clinical Trials Regulation and CTIS transition is a useful comparison: a staged mandate that was comfortably distant until it was not.