MDR/IVDR under reconstruction: how Europe is recalibrating its medical device framework

The story of the targeted revision of the MDR and IVDR did not start with COM(2025) 1023. It started with a regulatory system whose safety objectives are rarely challenged politically, but whose practical operation has been under increasing pressure for years. Transitional periods had to be extended repeatedly, device availability became a political concern, notified bodies and manufacturers reported high costs and limited predictability, laboratories described difficulties with in-house IVDs, and both research organisations and industry pointed to barriers for clinical development and innovation.

That is why the shorthand description “Brussels wants to simplify the MDR” misses much of what is actually happening. Before the Commission tabled its proposal, the framework went through an unusually broad evaluation and consultation phase. A second round then focused on possible regulatory solutions. Since publication of the proposal, market actors, civil society and political institutions are no longer commenting on an abstract problem. They are reacting to concrete changes in legal text. Keeping those layers separate is essential to understanding the direction of travel.

One thread runs through the entire process. The objectives of the MDR and IVDR are not being reopened. The question is whether the mechanisms chosen to deliver those objectives are proportionate, predictable and sufficiently consistent across Europe to support safety, availability and innovation at the same time. The quality of that recalibration will determine whether the revision genuinely simplifies the system or simply relocates complexity.

The central thread. The political debate is not directed against the safety objectives of the MDR and IVDR. It concerns mechanisms seen as disproportionate, unpredictable or inconsistently applied. Every reform measure therefore has to be tested against simplification, safety and harmonization at the same time.

This story did not begin on 16 December 2025

The MDR and IVDR were adopted in 2017 not as a deregulation exercise, but as a response to specific weaknesses in the former directive-based framework. The reform was intended to strengthen notified-body oversight, clinical and performance evidence, traceability, market surveillance and coordination among Member States. Safety and confidence were the political starting point. Any assessment of today’s revision that reduces the debate to administrative burden loses sight of that origin.

Implementation nevertheless showed that a stronger framework is not automatically a better functioning one. Application of the MDR was postponed in 2020, followed by further adjustments and extensions of the transitional arrangements under Regulations (EU) 2022/112, (EU) 2023/607 and (EU) 2024/1860. Those interventions protected availability and bought time for manufacturers, notified bodies and authorities. They did not answer the structural question of why the system had repeatedly needed transitional relief.

By 2024, the transition problem had become a political system question. On 23 October 2024, the European Parliament adopted resolution P10_TA(2024)0028 calling for urgent revision while expressly maintaining a high level of health protection. A few weeks later, nine Member States presented common reform priorities to the EPSCO Council in Council document 15380/24. In parallel, the Commission brought forward the evaluation that Article 121 MDR required only by May 2027. The policy brief had changed: the task was no longer to manage another transition deadline, but to examine the functioning of the regulatory model itself.

Figure 1. From MDR/IVDR implementation to the current legislative procedure.

What the Commission actually evaluated

The targeted evaluation in SWD(2025) 1051 final covers the period from adoption of the Regulations on 5 April 2017 to 31 December 2024. This matters methodologically because implementation was still incomplete. The Commission therefore does not present the exercise as a final verdict on a mature system. It asks whether the framework is demonstrably moving towards its objectives and whether structural inefficiencies have already emerged that are unlikely to disappear simply with more time.

The evaluation uses the five standard criteria of EU ex-post evaluation: effectiveness, efficiency, coherence, relevance and EU added value. Importantly, the evidence was not drawn from one consultation. Because no source was complete or representative on its own, the Commission triangulated quantitative and qualitative evidence.

The evidence base included a structured public consultation from 12 December 2024 to 21 March 2025 with 332 responses. In parallel, an open Call for Evidence generated a cleaned set of 542 feedback items after removal of duplicates, non-compliant entries, a separately treated campaign and the merger of complementary submissions. The Commission also used targeted surveys of competent authorities, economic operators, notified bodies and the EMA, practical workshops with manufacturers and with healthcare professionals, patients and users, discussions within the MDCG, administrative data, position papers and external studies.

Governance had already been examined before that formal evaluation phase. A 2023/2024 study combined survey responses, interviews and several workshops, including a dedicated SME workshop. The targeted evaluation was therefore not a single lobbying exercise. It was a layered evidence base in which market actors, authorities, notified bodies, healthcare institutions, academia, patients and consumer interests were assessed from different angles.

What policymakers took from that evidence

One of the most important findings is what the Commission did not conclude. The evaluation does not support renationalising medical device regulation or dismantling the safety architecture of the MDR and IVDR. In the structured public consultation, 93% of respondents to the MDR section and 87% of respondents to the IVDR section preferred a single EU regulatory framework to separate national regimes. Despite strong criticism of implementation, the EU added value of the framework remains clear.

The cost signal was equally striking. In early phases of development and conformity assessment, fewer than 10% of MDR respondents considered regulatory or administrative costs acceptable. For the IVDR, the relevant agreement rates across the phases considered were likewise no higher than 10%. Those figures do not show that every regulatory obligation lacks value. They do show how far perceptions of burden and regulatory benefit had diverged.

SWD(2025) 1051 final formulates the diagnosis more carefully. The Regulations have strengthened the infrastructure for safe and performant devices. Requirements for notified bodies, evidence, post-market surveillance, vigilance, transparency and traceability have made the protective framework more robust. At the same time, implementation delays, high and often disproportionate costs, uneven application, complex governance, slow development of guidance and constrained capacity are identified as structural drivers of unpredictability. The Commission is particularly critical of the limited progress towards smooth market functioning, competitiveness, innovation and a level playing field, especially for SMEs.

The policy question therefore shifts. It is no longer whether safety or simplification matters more. The real issue is which obligations produce meaningful safety information, where assessment and documentation are duplicated, and how oversight can be aligned more closely with actual risk, product maturity and compliance history. That reading forms the bridge to the later legislative proposal.

The additional Call for Evidence: from problem diagnosis to regulatory solutions

After the evaluation phase, the Commission opened a second, differently focused consultation in autumn 2025. From 8 September to 6 October, initiative 14808 sought input not primarily on whether the MDR and IVDR were working, but on measures for the targeted revision and on their expected savings or additional costs.

The public portal contained 441 contributions. For the official analysis in SWD(2025) 1050 final, the Commission worked with 427 contributions and 165 attachments after cleaning. The distinction matters: 441 describes the published raw corpus, while 427 is the population used in the Commission’s analysis. The corpus was economically weighted. Companies accounted for 46.6% of contributions and business associations for a further 14.3%. Among participating companies, SMEs represented 64.8%. That helps explain why costs, notified-body performance, time, availability and predictability were prominent.

Substantively, the official Commission synopsis aligns closely with an independent qualitative review of the published raw corpus. Recurring demands included more predictable conformity assessment, earlier structured interaction with notified bodies, transparent fees, risk-based rather than schematic surveillance, more proportionate evidence requirements, workable routes for established and supply-critical devices, more practical rules for in-house IVDs, greater digitalisation and stronger EU-wide harmonisation.

The notable point is therefore not one isolated request but the convergence between market feedback and the official problem definition. The Commission did not simply count and reproduce stakeholder demands. It filtered them through administrative evidence, studies, authority and notified-body input, and the wider competitiveness and simplification agenda. COM(2025) 1023 final is the legislative selection that emerged from that process.

COM(2025) 1023: not a dismantling of the system, but a shift in how it works

The Commission proposal published on 16 December 2025 retains the fundamental architecture of EU medical device regulation. Manufacturers remain responsible for conformity, notified bodies remain central third-party assessors, national authorities retain surveillance and enforcement functions, and CE marking remains the gateway to the internal market. The proposal is not a move away from the New Legislative Framework.

It is nonetheless substantial. The text changes not merely forms or deadlines, but the logic of several lifecycle processes. Time-based repetition is replaced in a number of places by risk-based or trigger-based review. Evidence pathways are intended to become more flexible. Certain device groups receive dedicated regulatory routes. Digital and international work products should be used more effectively. EU-level coordination is strengthened.

The clearest paradigm shift concerns certificates and surveillance. General time limits on certificates would disappear and periodic risk-based review would replace routine recertification. Technical-documentation sampling would become less schematic, audit frequency more differentiated, remote assessment more prominent and unannounced audits more closely linked to triggers. The potential operational relief is significant. The safety question is equally significant: what information will the system continue to generate independently so that increased scrutiny is triggered early enough?

A second block concerns evidence and change. The Commission proposes wider use of non-clinical evidence, in-silico approaches and other modern data sources. Equivalence should become more workable. Predetermined change-control concepts should prevent every foreseeable product change from reopening a full assessment cycle. For software and technology-intensive products with short development cycles, this is one of the most consequential parts of the proposal.

A third focus is availability. Orphan and breakthrough devices would receive prioritised and, in some cases, rolling pathways. The proposed concept of well-established technology devices is intended to make requirements more proportionate for mature technologies. In-house devices, particularly under IVDR Article 5(5), would benefit from material relief. The Commission is clearly responding to the finding that formally identical requirements can have very different consequences in low-volume and supply-critical markets.

The package also covers software classification, clinical investigations and performance studies, digitalisation, eIFU, EUDAMED, international cooperation, fees, structured dialogue and Articles 16 and 17 MDR. Those last areas make an important point: simplification does not automatically create consensus. Some proposed solutions redistribute responsibility, liability, access to data and control. That is where the next phase of the story begins.

Figure 2. Roles and influence channels in the ordinary legislative procedure.

Why the Commission proposal is not “the new MDR”

Publication of COM(2025) 1023 opened ordinary legislative procedure 2025/0404(COD). For regulatory strategy, the institutional roles need to be kept distinct. A Commission proposal is the starting text, not the outcome. It becomes law only when the European Parliament and the Council adopt the same text.

The European Commission, with DG SANTE in the lead, prepared the initiative, coordinated the evaluation and tabled the proposal. During the legislative process it explains the text, provides technical input and participates in negotiations. It does not replace the two co-legislators in the final decision.

In Parliament, the Committee on Public Health (SANT) is responsible. A rapporteur coordinates the dossier on behalf of the committee and the other political groups appoint shadow rapporteurs. The draft report PE787.987v01-00, published on 30 June 2026, is a SANT committee draft in the parliamentary process. IMCO provides an opinion on internal-market and consumer aspects, while BUDG considers budgetary implications.

The Council works in parallel. It is the governments of the Member States, not the national parliaments, that negotiate the Council position. Technical examination takes place in the Working Party on Pharmaceuticals and Medical Devices. The Presidency seeks compromise language among delegations; unresolved issues can move through COREPER to the Council. Only a Council position or negotiating mandate reveals the text Member States can defend in negotiations with Parliament.

Other institutions participate with different legal weight. The European Economic and Social Committee issues an advisory opinion. National parliaments scrutinise subsidiarity and proportionality. Their opinions can matter politically, but they are not the Council position. Associations, manufacturers, notified bodies, hospitals, laboratories, professional societies and patient organisations influence the process through consultations, hearings and transparent contacts. A trilogue can seek a common text only once Parliament and Council have workable negotiating positions.

Feedback on the proposal: the market is not rejecting reform, it is disputing its execution

Following publication of the Commission proposal, another public feedback phase generated detailed reactions to the actual legal text. For this article, 248 contributions published through 2 August 2026, including the available attachments, were qualitatively reviewed. As with the earlier consultations, the corpus is not representative. It shows who chose to engage and which conflicts a concrete draft produces.

The overall pattern is striking. In the qualitative coding, 66.9% of contributions support the direction of reform but call for amendments. A further 19.4% make additional requests without rejecting the reform as a whole. Only a very small share is broadly opposed. The dominant message is therefore not “return to the existing MDR”, but “keep the direction and correct the sharp edges”.

This also changes the nature of the debate. Before the proposal, problems such as costs, time, capacity and administrative duplication dominated. After publication of the legal text, additional questions of legal responsibility, burden of proof, liability, institutional competence and the cumulative effect of multiple simplification measures become visible. A general call for proportionality turns into a debate about how proportionality can be constructed reliably in law.

Figure 3. Qualitative overall stance in feedback on the Commission proposal.

The central system question: how much preventive oversight can be removed?

Each simplification measure can be defended on its own. Open-ended certificates avoid repetitive work. Less schematic technical-documentation sampling can concentrate resources on relevant risks. Longer audit intervals can reward demonstrably stable manufacturers. Remote assessments can improve efficiency. Reduced reporting can remove documentation that does not always create additional information value.

However, the feedback rightly looks at the cumulative effect. Notified bodies, consumer interests and safety-oriented contributors ask what happens if fixed certificate cycles, routine technical-documentation sampling, frequent on-site audits, unannounced audits and selected reporting duties are reduced at the same time.

Two different understandings of a risk-based framework collide. One view is that repeated routine assessment is a poor use of scarce resources when the product, manufacturer and post-market data are stable. The other is that signals only exist if the system continues to generate independent information. A framework that escalates scrutiny only after an incident must not remove the mechanisms that would have detected an early deviation.

The sustainable answer is therefore not simply to preserve legacy frequencies or to move to purely trigger-based oversight. It requires harmonised triggers, a meaningful compliance history, clear minimum information and defined escalation authority. Without that, the same flexibility can produce genuine relief at one notified body and de-facto recertification under another name at a different one.

MDR Article 17: reprocessing is no longer a peripheral issue

Article 17 is arguably the sharpest single controversy in the feedback. The Commission wants to facilitate reprocessing and make single-use designation more explicitly justified. Healthcare institutions, reprocessors and sustainability actors see a route to a more coherent European market, lower resource use and additional supply resilience.

Original manufacturers are much more cautious. Their objections are not simply objections to reprocessing itself. They concern the allocation of knowledge and responsibility. The manufacturer knows the material properties, design limits, ageing behaviour, intended use and original risk assessment. A legal structure that makes reprocessability the starting assumption can, in their view, shift the burden of proof and blur responsibility among the original manufacturer, reprocessor and healthcare institution.

Several regulatory layers therefore intersect: who assumes manufacturer obligations after reprocessing; what information must the original manufacturer provide; how product changes and post-market signals are exchanged; how the reprocessed product is identified; and what validation, sterility, material-integrity and traceability requirements apply. The debate is a good illustration of how an apparently simple simplification can create new interfaces in a lifecycle regime.

MDR Article 61(5): equivalence between proportionality and data rights

The debate on clinical equivalence is similarly fundamental. The Commission is responding to a real practical problem. For certain high-risk devices, equivalence currently depends on contractual access to the technical documentation of the comparator. With a competitor, such access is often unavailable, leaving a formally recognised route effectively closed.

Many contributors therefore support greater flexibility. It may prevent unnecessary duplicate studies and make better use of scientific and regulatory data. Data-intensive manufacturers raise equally material concerns: clinical development programmes are investments; public information does not always reveal technical differences and later design changes; post-market information on the comparator may be incomplete; and the continuing validity of the equivalence claim may become harder to assess.

The conflict is not well described as “innovation versus patient safety”. It concerns the quality and currency of the evidence base, the depth of the equivalence claim and the product-specific information needed to support it. A graduated approach reflecting device risk and the importance of the equivalence claim may be more defensible than a binary choice between mandatory contractual access and unrestricted use of third-party evidence.

Legal certainty of CE marking and the role of EU coordination

Another conflict concerns regulatory status and classification procedures. The Commission seeks stronger coordination of divergent national decisions and greater involvement of expert panels supported by the EMA. The objective is understandable: the same product should not remain classified or qualified differently across Member States.

A problem arises if a CE-marked device can be reopened after completed conformity assessment without a clear threshold. Companies, the EESC and national parliaments argue that this weakens the legal-certainty function of CE marking. Investment and portfolio decisions depend on whether a completed third-party assessment has presumptive stability or can routinely be reopened by a second regulatory layer.

The EMA role is also treated in a differentiated way. Administrative support, coordination and infrastructure are less controversial than an autonomous scientific or technical role in device qualification and classification. The underlying issue is architectural: should a decentralised medical-device framework be substantively steered by an agency whose core mandate is medicinal products, and if so under what competences and procedural safeguards?

Other fault lines that matter

For software, the objective of a more risk-sensitive Rule 11 is widely supported, but the proposed wording is criticised as unclear and in places internally inconsistent. Stakeholders see risks of both continuing over-classification and inappropriate down-classification of clinically relevant software, with the AI Act adding another layer of interaction.

Deleting MDR Article 82 may reduce national divergence in “other” clinical investigations, but academic and clinical stakeholders warn of a regulatory gap for investigator-initiated or non-CE-oriented research. Replacing fragmentation with an unnecessarily heavy or legally uncertain research pathway would not be simplification.

Under Article 16, relief for relabelling and repackaging is partly welcomed, while complete removal of advance manufacturer information is criticised on traceability, vigilance and packaging-integrity grounds. Proposed deletions around liability and financial coverage raise a separate question: why should patient-facing protection mechanisms disappear under an administrative-simplification rationale? Article 10a remains controversial because of definitions, timelines, confidential supply-chain data and parallel IT processes.

By contrast, the direction of travel is much less disputed for in-house IVDs and orphan/breakthrough pathways. Laboratories and healthcare institutions broadly support relief under IVDR Article 5(5), often seeking further flexibility. Dedicated innovation and availability pathways also attract broad support. The debate is more about criteria, thresholds, capacity and consistent implementation than about whether reform is needed.

The political system is already reacting to the same sharp edges

The institutional opinions show that feedback on the Commission proposal is not remaining in a separate consultation silo. Several conflicts identified by market actors reappear almost immediately in political documents.

The European Economic and Social Committee adopted its opinion on 29 April 2026 with 97.9% support and no votes against. It supports the simplification agenda but attaches clear conditions: harmonised criteria for periodic reviews, predictable notified-body practice, effective PMS, vigilance and market surveillance, cost relief that is not offset by higher fees, and caution on reprocessing and in-house devices. Its concrete amendments would limit reopening of the regulatory status of CE-marked notified-body-certified devices to substantiated safety concerns, keep the single-use/reusable decision fundamentally with the manufacturer, and narrow the EMA role in qualification and classification.

National parliaments add different concerns. The Italian Chamber of Deputies accepts the need for reform but questions the breadth of delegated and implementing powers, possible oversight gaps, Article 17 and an expansive EMA role. The French Senate issued a reasoned opinion on subsidiarity and proportionality, focusing on national distributor databases/EUDAMED and on whether essential annexes could be altered too extensively by delegated acts. Two Italian Senate committees also address surveillance, open-ended certificates, reprocessing, CE legal certainty, the orphan threshold and Article 61(5).

The distinction matters: these documents are not “the Council position”. They are circulated through the Council because national parliaments are part of the EU subsidiarity-control framework. The Council position is developed separately by Member State governments. That difference is often lost in public reporting.

The SANT draft: an early attempt at political calibration, not a counter-proposal to reform

The SANT committee draft report PE787.987v01-00, published on 30 June 2026, is particularly informative. Its 131 amendments provide the first broad view of how key conflicts in the Commission proposal might be translated into parliamentary text. Its procedural status must not be overstated: SANT has not adopted the draft and Parliament has not yet adopted a first-reading position. It is nevertheless an important early indicator.

Substantively, the draft does not follow a simple political camp. It maintains important simplification and innovation objectives, expands some of them, and restores safeguards elsewhere. Reopening the status of CE-marked devices would require stronger substantiation. Under Article 17, the decision on single use or reprocessability would be returned more clearly to the manufacturer. Advance manufacturer information under Article 16 would be retained. Liability and financial-coverage mechanisms would be preserved.

At the same time, the draft moves beyond the Commission in selected areas of relief. It proposes a distinct niche-device category for supply-relevant low-volume products, broader use of ISO 13485 and MDSAP audit results, English as a possible common language for information intended exclusively for professional users, and additional relief for in-house devices.

Surveillance illustrates the search for a middle course. A 24-month audit interval would not become an automatic norm, but would depend more strongly on compliance history. Unannounced audits would not be limited to cases where PMS or vigilance has already generated a signal, but would sit within a risk-based audit plan. This is not a return to fixed frequencies. It is an attempt to preserve relief while linking it to proven performance and independent risk information.

Article 61(5) shows a clearer counter-movement. The draft deletes the Commission recital that justified removing contractual access for equivalence and places more emphasis on robust, product-specific clinical evidence. Whether that approach survives committee compromise is open. Politically, however, one point is already clear: one of the proposal’s most contentious changes is not simply being waved through.

Why this matters politically. The draft follows neither a pure deregulation agenda nor a pure control agenda. It preserves simplification as an objective while trying to address several risks identified in market feedback through clearer responsibility and trigger criteria. That combination is what makes it a useful early indicator.

What is happening in the Council in parallel

The Council side is less transparent but no less important. Public meeting and register data show that the Working Party on Pharmaceuticals and Medical Devices examined the proposal by articles and thematic clusters in spring 2026 and continued intensive exchanges throughout July. The file was again on the agenda on 27 and 28 July, with further working-party meetings announced for early September.

The density of meetings does not reveal a substantive Council position because many working documents are not public. It does show intensive technical drafting. No publicly documented general approach exists as of 11 August 2026. The second major reference point for a future trilogue therefore remains missing.

This uncertainty matters for regulatory strategy. The SANT draft indicates what may become politically acceptable in Parliament. It does not tell us what the Council will accept. On competence, market access, language rules, national data obligations and the degree of EU harmonisation, Member State interests can differ materially from those of a parliamentary committee.

Is Europe at risk of “simplifying the system into something worse”?

The concern that political negotiations could load the Commission proposal with so many safeguards, exceptions and new responsibilities that simplification disappears is understandable. But it would be a mistake to treat every amendment to the Commission text as a step backwards. Scrutiny and amendment are precisely what the ordinary legislative procedure is designed to do.

The real risk is additive reform. If existing duties remain while new categories, panels, exemptions and evidence requirements are layered on top, the number of interfaces increases. A new niche-device category can protect supply, but it adds another boundary alongside WET, orphan and breakthrough devices. Structured dialogue can accelerate assessment if its outcome is meaningful; it becomes another loop if the same issues are reopened in full later.

A second risk is relocated uncertainty. Open legal wording can look flexible, but if its application is left to divergent notified-body practice, national authorities, future MDCG guidance and later implementing acts, burden has not disappeared. It has moved from the legal text into interpretation. The experience of the last years shows that interpretive divergence is itself a major cost and time driver.

The third risk sits at the opposite end: removing too much control at once. If preventive mechanisms are reduced simultaneously, the system needs reliable post-market data and escalation mechanisms. Risk-based regulation is a safety model only if data quality, signal detection, responsibilities and escalation thresholds work. Otherwise, “risk-based” becomes shorthand for reacting after something has already gone wrong.

A sound final architecture can therefore be tested against a small number of questions. Which existing process actually disappears? What EU-wide criteria trigger relief? Who decides? What information keeps the system informed despite less routine review? Does digitalisation replace old processes rather than sit beside them? And will guidance, expert panels, EUDAMED functions and notified-body capacity exist before new deadlines begin to run?

Against that test, the political corrections visible so far are not inherently a deterioration. Several address genuine design risks in the Commission proposal. The danger arises where every individual concern is answered with an additional special rule without simplifying the architecture as a whole.

What happens next

As of 11 August 2026, the European Parliament’s Legislative Observatory still lists the file as “awaiting committee decision”. The SANT draft report has been published and the amendment deadline has passed. The next material parliamentary step is compromise-building among the rapporteur and shadow rapporteurs, followed by the committee vote and then Parliament’s first-reading position. No binding timetable for those steps has been published.

The Council resumes technical work after the summer break. Public Consilium data show further meetings of the Working Party on Pharmaceuticals and Medical Devices scheduled for early September. That work may lead to Presidency compromise texts and, eventually, a Council position. Only when both co-legislators have stable mandates can the likely trilogue corridor be assessed with confidence.

Adoption in 2026 is procedurally possible, but it is not a reliable forecast. It is equally plausible that political agreement and formal adoption extend into 2027. Planning should therefore be scenario-based rather than anchored to a presumed date. The more useful milestone is the point at which Parliament and Council begin to converge on a stable text corridor.

What RA, strategy and management should take from this now

The current stage justifies neither inaction nor premature implementation of draft law. The most useful preparation is structured scenario analysis. Organisations should be able to see which parts of their portfolios would be affected by which potential changes and which decisions remain reversible.

For certificates and notified bodies, current audit and assessment cycles, contractual clauses, fee models and escalation mechanisms should be mapped transparently. Potential savings from longer future intervals should not yet be built into budgets. At the same time, organisations can identify products and manufacturers whose strong compliance history would make them natural candidates for risk-based relief.

For clinical strategy, Article 61(5) is an area in which robust alternatives should remain available. Where equivalence is part of the strategy, access to data, technical comparability, currency of information and own-data generation should be assessed separately. The political process already shows that maximum relaxation of contractual access is far from certain.

At portfolio level, potential WET, orphan, breakthrough and niche scenarios should be mapped. The point is not simply formal classification, but economic and supply relevance: which products are disproportionately burdened because of low volumes; where would prioritised assessment be strategically material; and where might a product sit between emerging definitions?

For software and AI, Rule 11 scenarios, change logic and predetermined change-control concepts should be modelled separately. For in-house IVDs, transfer models, ISO 15189 coverage, responsibility boundaries and central laboratory structures deserve attention. For Articles 16 and 17, supply-chain, labelling, information and liability flows should be reviewed because the final text is particularly likely to diverge from the Commission version.

One topic is stable regardless of the final wording: data architecture. EUDAMED, electronic submissions and digital documentation will become more important. Organisations that clarify master-data ownership, interfaces and document structures now are not merely preparing for the revision. They are addressing one of the main causes of regulatory duplication.

The real question is no longer whether simplification will happen

The political direction is now remarkably stable. The Commission, large parts of the market, the EESC, Parliament and many Member States recognise that the MDR and IVDR require structural recalibration. A second line is equally stable: the high level of protection, the CE framework and the European regulatory architecture are not intended to disappear.

The dispute therefore takes place one level deeper. How much assessment is necessary for which risk? Which evidence is sufficiently robust? Which decisions belong with manufacturers, notified bodies, national authorities, expert panels or the Commission? Which obligations genuinely produce safety information and which mainly create repetition? How can Europe harmonise implementation without building another administrative layer?

That is what makes the present phase unusually revealing. The full chain is visible: implementation experience becomes evaluation; evaluation becomes policy conclusions; policy conclusions become a Commission proposal; the proposal creates new concrete conflicts; and those conflicts are already reappearing in institutional opinions and parliamentary amendment directions.

Whether the revision ultimately produces a better system cannot be measured by the number of deleted paragraphs. A proportionate regulatory framework exists when risk, evidence, oversight, responsibility and data flows fit together coherently. That is the standard against which the final text will have to be judged.

Methodological note

The numbers from the different consultation stages refer to different instruments and must not be added together or treated as one sample. The structured public consultation, the Call for Evidence for the evaluation, the additional Call for Evidence for the revision and the later feedback on the concrete Commission proposal serve different functions.

The evaluation phase is described on the basis of the Commission’s official documents SWD(2025) 1050 final, SWD(2025) 1051 final, SWD(2025) 1052 final and Ares(2025)4926888. The 441 published contributions to the additional Call for Evidence were compared substantively with the Commission’s official synopsis. For feedback on the Commission proposal, 248 published contributions and available attachments through 2 August 2026 were qualitatively coded by overall stance and regulatory topic. The percentages describe that published corpus, not the EU population or the market as a whole.

Future procedural dates are scenarios based on the publicly documented status as of 11 August 2026. No binding timetable has been published for the SANT vote, plenary position, Council position, trilogue or formal adoption.

Sources and key procedural documents

Commission documents are identified by exact document reference. Public links are provided to make the procedural status traceable.