Europe’s fragmented medical device rules fuel calls for a central EU agency

Europe’s system for approving medical devices and in vitro diagnostics is under strain, and a new analysis published in The Lancet Regional Health – Europe argues that the problem is structural. Anett Schönfelder, Tom Melvin, Timo Minssen and Stephen Gilbert trace how the current framework, built around the Medical Device Regulation (MDR) 2017/745 and the In Vitro Diagnostic Medical Device Regulation (IVDR) 2017/746, distributes authority across dozens of actors at EU and national level, and they contend that this fragmentation produces unpredictable approvals, inconsistent evidence standards and a widening gap with the United States. Their proposed answer is a genuinely centralised EU governance structure, potentially a dedicated medical device agency, equipped with binding decision-making power over regulatory strategy and guidance.

The architecture they describe is complex. Manufacturers must demonstrate conformity with MDR or IVDR requirements before applying CE marking, which is currently valid in more than 100 jurisdictions worldwide. For certain devices, that assessment is carried out by designated third-party entities known as Notified Bodies, of which about 53 currently operate across Europe, serving an industry of more than 37,000 MedTech companies. Around them sits a web of oversight: the European Commission chairs the Medical Device Coordination Group, 27 national competent authorities monitor products on the market and inspect Notified Bodies, and the European Medicines Agency supports the expert panels, known as EXPAMED, that deliver scientific opinions during conformity assessment of certain high-risk devices. The authors characterise the arrangement as high on bureaucracy but low on transparency and effective communication.

The consequences are visible in market behaviour. In a representative survey of 211 medical device and IVD manufacturers conducted in 2024, cited in the analysis, companies deprioritised the EU as their first regulatory market. New innovative product launches in the EU fell by 33% for large medical device manufacturers and 19% for small ones, and by 40% for large and 12% for small IVD manufacturers, while the US market gained popularity, particularly among large firms. The same survey recorded a significant decline in innovation activities within small and medium-sized enterprises, with 59% of IVD and 54.4% of medical device manufacturers reporting reduced development of new devices.

Evidence quality is a central concern. Data from the CORE-MD consortium, which examined high-risk devices, showed that only 9% of cardiovascular devices and 0% of orthopaedic implants analysed had any pre-market clinical evidence publicly available at the time of marketing. The authors argue this makes it nearly impossible for clinicians to perform independent risk-benefit assessments before a device is used in a patient. EXPAMED opinions have identified critical methodological flaws in clinical evaluations that Notified Bodies had accepted, ranging from poor study design to insufficient sample sizes and weak post-market evidence plans, yet in all cases the products were nonetheless granted market approval. Empirical studies have also documented inconsistent MDR interpretation across Notified Bodies, including non-uniform standards for clinical evaluation data and divergent handling of conditional certifications for high-risk devices.

Combination technologies sharpen the confusion. An AI-powered closed-loop insulin delivery system, the authors note, falls simultaneously under medicines regulation for the insulin, medical device regulation for the pump and continuous glucose monitor, and artificial intelligence rules for the predictive dosing algorithm, and could additionally touch in vitro diagnostic rules. Manufacturers must comply with each applicable regime, and overlapping frameworks create unclear and often additional requirements, slow timelines and conflicting decisions. Regulatory uncertainty is directly connected to longer approval times and higher costs for innovative devices. Although the MDCG and the Joint Artificial Intelligence Board have issued guidance on avoiding duplication, and the Commission’s MDR/IVDR 2.0 proposal alongside a proposed European Biotech Act aims to clarify categorisation of combined products, the authors find that scientific or clinical development advice remains scarce: Notified Bodies are explicitly restricted from consultancy, and while 57% of national competent authorities offer advice before application, only 11% offer national scientific advice beyond clinical-trial aspects.

The reform debate has gathered momentum across stakeholder groups. Position papers published in 2024 and 2025 called for centralised governance: industry associations including MedTech Europe, AESGP, COCIR, FIDE, EUROM and EEAR in March 2025; the BioMed Alliance representing 35 European medical specialist societies, also in March 2025; 21 national competent authorities through the CAMD and HMA networks and the Notified Body coordination groups in July 2025; and patient and hospital representatives including the European Patients Forum, EAHP and HOPE in September 2025. Members of the European Parliament had already issued a motion for a resolution in October 2024 calling for a central governance structure. Despite differing designs, the common idea is to consolidate responsibilities and eliminate fragmentation, addressing a missing “captain of the ship”, an accountability vacuum documented in a European Commission study in which every MedTech company interviewed described fragmented governance levels, and insufficient coordination leading to inconsistent rule application.

Crucially, the authors distinguish centralisation of governance from centralisation of product authorisation. The position papers advocate central strategic coordination, harmonised guidance and stronger oversight of Notified Bodies, while keeping product authorisation decisions with the Notified Bodies, and the Commission’s simplification proposal published on December 16, 2025 retains that decentralised process. The authors warn that as long as individual Notified Bodies decide whether to approve highly innovative technologies while others reach different conclusions on similar submissions, with limited transparency and no formal precedents, the core problems of disconnected decision-making and unpredictable approvals will persist. The December 2025 proposal, which may come into force in 2027, aims to reduce regulatory burden by strengthening centralisation of conformity assessment, reducing Notified Body involvement for some device types, cutting unannounced audits, promoting pre-submission dialogue and increasing oversight, but it stops short of a single authority.

The case for centralisation draws on evidence from other sectors. Centralised procurement in Denmark and Italy reduced corruption-related inefficiencies, with corruption able to inflate costs by up to 11%, by pooling purchasing power and limiting buyer discretion. The EU trademark system cut the cost of EU-wide protection from about €3,160 under separate national filings to less than €600. EMA’s centralised procedure authorises medicines across all 27 Member States with one submission, and one study found that joining it reduced launch delays by 10.9 months. In banking, an analysis of 745 European banks showed that those moved to centralised supervision under the Single Supervisory Mechanism significantly reduced credit risk exposures while increasing profitability. The European Food Safety Authority, founded in 2002 after the BSE crisis exposed fragmented national assessments, illustrates how centralised risk assessment can improve regulatory decisions and public health.

Yet the authors are careful to note that centralisation is not unconditionally efficient. Decentralisation can be superior where preferences are highly heterogeneous, although they argue this logic fits poorly with CE marking, which grants market access across the whole EU regardless of where a Notified Body sits. They also point to the United States as a cautionary reference: the FDA, a long-standing central federal agency, has recently faced rushed reviews relying on a single pivotal trial instead of two, expert panels described as unbalanced and conflicted, political infiltration of product decisions, and layoffs of experts and leadership turnover. Centralisation itself, they conclude, guarantees nothing; outcomes depend on robust, enforceable frameworks and consistent implementation, and no EU medical device legislation since 1990 has been preceded by a scientific evaluation of policy options’ effects on public health.

Their preferred solution is the most centralised option feasible within EU law: either a new single central EU governance structure, such as a medical device agency, or a well-defined division within an existing agency like the EMA, empowered with high decision-making power over regulatory strategy and guidance. Full legal transplantation of an FDA-style regulator faces obstacles, since Article 168 of the Treaty on European Union frames Union public health action as complementary to national policies, and the Meroni line of case law restricts agencies to clearly defined powers subject to effective oversight, though later jurisprudence such as the ESMA case permits tightly framed delegation. The proposed unit would maintain a transparent database of device classifications, provide binding recorded advice on innovative devices, and ensure approval decisions are evidence-based and propagated as system-level policy. Budget is a known sticking point: EMA’s 2025 annual budget of €600.2 million, 91.5% industry-funded, shows a service-linked fee model is possible. The authors argue a coordination office, the most feasible option, would be the least effective, and that a hybrid half-way house risks combining the worst of both worlds unless its weaknesses are explicitly managed through adaptive governance, academic-led regulatory science and work-sharing models such as Centers of Excellence in Regulatory Science and Innovation.

Subject of Research: Governance reform and centralization options for EU medical device regulation

Article Title: Reforming medical device regulation in Europe: centralization, governance, and the case for an EU medical device agency

Article References: Schönfelder, A., Melvin, T., Minssen, T., & Gilbert, S. (2027). Reforming medical device regulation in Europe: centralization, governance, and the case for an EU medical device agency. The Lancet Regional Health – Europe, 72, Article 101908. https://doi.org/10.1016/j.lanepe.2026.101908

Image Credits: AI Generated

DOI: 10.1016/j.lanepe.2026.101908

Keywords: medical devices, EU regulation, MDR, IVDR, notified bodies, centralization, EMA, FDA, regulatory governance, MedTech, patient safety, innovation