Stage 1: Decide what you are bringing to market

Chapter 2

Which market should come first, and on what evidence?

In short#

In the United States, the regulator will answer a company's specific questions before it applies, while in Europe the body that assesses the device may not advise on how to comply. FDA answers in writing through a Pre-Submission, a formal request for feedback. A notified body, the independent organisation that certifies the device, may exchange technical information and regulatory guidance with a company applying to it. It may not provide consultancy, and guidance from the Medical Device Coordination Group (MDCG), the Member States' expert group, treats advice on how to comply as consultancy. For class III devices, the highest-risk class, and certain class IIb devices, the company may consult an expert panel on its clinical development strategy before the study. (MDR Annex VII, s. 1.2.3, 1.2.9; Art. 61(2); MDCG 2019-6, Q&A I.6.1; Q-Submission guidance, s. II.A)

A study run outside Europe can support a European file, because the MDR's definition of clinical data places no condition on the country where an investigation was run. In the other direction, a premarket approval application (PMA), FDA's route for its highest-risk devices, may rest solely on foreign data if, among other conditions, the data apply to the US population and US medical practice. On our reading, a company going to Europe first has to build that condition into its first study if the data are to serve FDA too. (MDR Art. 2(48); 21 CFR 814.15)

Published European and US review times start from different points, so on our reading they cannot be compared directly. The Commission's survey of notified bodies counts from the signing of the written agreement, the contract with the notified body. FDA counts from receipt of a submission that it accepts or files. Authorisation in the United States also does not secure payment. The Centers for Medicare & Medicaid Services (CMS), which run the federal Medicare programme, state that FDA market authorisation alone does not entitle a technology to Medicare coverage, so a market plan has to name the first US payer as well as the first authorisation. (Notified body survey, slide 30; MDUFA V letter, s. VIII; RAPID notice, s. I.B)

Contents

Introduction#

A company with a medical device or an in vitro diagnostic medical device (IVD), a test run on a sample such as blood, has to decide where it sells first: Europe, the United States, or both. The order shapes the evidence plan, the first payer and how long the company's cash has to last.

The first clinical study often settles the order by default, because a protocol designed for one regulator may not meet the other's conditions. A company that wants both markets therefore has to decide which markets the study serves before the protocol is written.

The European rules are the Medical Device Regulation, Regulation (EU) 2017/745 (MDR), and the In Vitro Diagnostic Medical Device Regulation, Regulation (EU) 2017/746 (IVDR). Under them, most devices are assessed by a notified body, an independent organisation designated to certify them. The US rules are those of the Food and Drug Administration (FDA), which authorises devices itself.

Four illustrative products run through the book. The monitor is a wearable cardiac monitor from a US company. The triage tool is artificial intelligence (AI) software from a German company. The implant is a spinal implant system certified under the former Medical Devices Directive. The near-patient test is a cardiac troponin test from a Swiss company. The sources were checked on 29 September 2026.

1. What are the options, and what decides between them?#

Option What it means
Europe first Certify under the MDR or the IVDR, and decide on the United States later
United States first Obtain FDA marketing authorisation, and decide on Europe later
Parallel Build both files at once, on one evidence programme
Staged Enter one market first, with the second market's evidence needs designed into the first study

Four constraints decide between the options. Evidence is what each regulator will accept and when it will say so. Payment is who pays for the device once it is authorised. Capital is how long the company can fund the gap between spending and revenue. Partners are the people a company needs in each market before it may sell there.

The decision applies to devices and IVDs, including software, and to companies inside or outside the Union. Its input is the class and conformity assessment route from chapter 1, Is it a medical device, which class, and who is responsible?. A notified body is a conformity assessment body designated under the MDR or the IVDR. (MDR Art. 2(42), 52; IVDR Art. 2(34), 48)

Chapter 3, How large is the opportunity, and what can the published figures support?, covers market size.

What guidance is worth#

The MDCG is made up of members appointed by each Member State to represent its competent authorities. It is chaired by a representative of the Commission. Each MDCG document cited here states that its views are not legally binding. Each adds that only the Court of Justice of the European Union can give binding interpretations of Union law. (MDR Art. 103(2), (5); MDCG 2025-9, cover page)

FDA's Q-Submission guidance, on asking for feedback before a submission, represents FDA's current thinking. It binds neither FDA nor the public, and uses "should" to mean recommended. (Q-Submission guidance, pp. 1 and 2)

Our reading, the label for an interpretation of a text, is that guidance on both sides shows how the regulator interprets the legal text. A company that departs from the guidance has to justify its position in writing.

2. What will each regulator say before the study?#

The notified body#

Annex VII of the MDR bars a notified body, its top-level management and its assessment staff from any service that may jeopardise confidence in their independence, impartiality or objectivity. They may not provide consultancy "as regards the design, construction, marketing or maintenance of devices or processes under assessment". (MDR Annex VII, s. 1.2.3)

The independence requirements "in no way preclude exchanges of technical information and regulatory guidance" with a manufacturer applying for conformity assessment. The IVDR carries both provisions in substantially the same terms. (MDR Annex VII, s. 1.2.9; IVDR Annex VII, s. 1.2.3, 1.2.9)

Under MDCG 2019-6 Rev.5, exchanges about what the requirements are remain permitted. Advice on solutions for "how to comply" counts as consultancy. Before an application is lodged, a review of clinical data or a partial assessment of the quality management system "are not allowed as they would be regarded as consultancy". (MDCG 2019-6, Q&A I.6.1, I.6.2)

A notified body must have documented procedures requiring it to review pre-application information before it issues any quotation. The review includes a preliminary verification that the product is covered by the regulation, and of its class. The Annex does not entitle the manufacturer to an answer from that review. (MDR Annex VII, s. 4.2(d); IVDR Annex VII, s. 4.2(d))

MDCG 2022-14 encourages structured dialogues on regulatory procedures before and during conformity assessment. It says they should not be considered a consultancy service. MDCG 2019-6 Rev.5 focuses them on "what needs to be fulfilled" rather than "how to fulfill". It says the costs "should not entail extra fees but be integrated" in the fees for the work. (MDCG 2022-14, action item 15; MDCG 2019-6, Q&A I.6.3)

Our reading is that no provision of the MDR or IVDR requires a structured dialogue, so until 25 February 2027 a notified body may decline one.

From that date, Commission Implementing Regulation (EU) 2026/977 applies. A notified body must then have documented procedures making the structured dialogue cover what it needs to issue a quotation. That includes the devices, their intended purpose and their risk class. (Regulation 2026/977, Art. 1(1), (2), 9)

Once the application is in, MDCG 2019-6 Rev.5 lets the two sides exchange views on the sufficiency of the clinical data and the post-market clinical follow-up (PMCF) plan. The exchange can also cover equivalence, the MDR's test for relying on another device's clinical data. Separately, the guidance says the two sides can discuss reusing evidence from assessments in other jurisdictions. It does not say the body will accept it. (MDCG 2019-6, Q&A I.6.3; MDR Annex XIV, Part A, s. 3)

Chapter 8, Which notified body can assess the device, and how does the work with it run?, covers the dialogue boundary, the application and quotations.

The expert panels#

For any class III device, the manufacturer may consult an expert panel on its intended clinical development strategy and its proposals for clinical investigation. So may the manufacturer of a class IIb active device intended to administer or remove a medicinal product. The consultation is optional. The manufacturer has to document in the clinical evaluation report how it considered the panel's views. It "may not invoke any rights" to them in a later conformity assessment. (MDR Art. 54(1), 61(2))

For other classes, the Commission has to facilitate manufacturers' access to expert panel advice on the data set needed to assess conformity, in particular the clinical data. MDCG 2025-9 says a breakthrough device in another class may ask the panels for that advice. (MDR Art. 106(11); MDCG 2025-9, section 10.3)

Our reading is that Article 106(11) places a duty on the Commission and gives a manufacturer no right of its own to the advice. On 29 September 2026, the pilot run by the European Medicines Agency (EMA), described in section 3, accepted advice requests only for the classes covered by Article 61(2). (EMA pilot)

EMA provides the panels' secretariat and levies their fees. Its page states that an EU-based manufacturer or its authorised representative may request the advice. (Regulation (EU) 2022/123, Art. 30; EMA advice page)

Our reading is that a manufacturer outside the Union can request the advice through the authorised representative it has to designate. (MDR Art. 11(1))

The IVDR gives an IVD manufacturer no counterpart to Article 61(2). At class D, the notified body itself consults the experts during assessment, where no common specifications exist and the device type is certified for the first time. Common specifications, a term both regulations use, are technical or clinical requirements, other than a standard, that provide a means of complying with the law. (IVDR Art. 2(74), 48(6); MDR Art. 2(71))

For class III implantable devices, EMA's page also describes parallel advice from the expert panels and the health technology assessment (HTA) bodies, through a joint scientific consultation. Chapter 12, Does Union health technology assessment reach the product?, covers it. (EMA advice page)

FDA#

A Pre-Submission is a voluntary, formal written request for FDA feedback on specific questions before an intended premarket submission. FDA answers in writing, or in writing followed by a meeting. The guidance says a Pre-Submission "is not intended to be a pre-review of an intended submission or a pre-review of data to be provided in a submission". (Q-Submission guidance, s. II.A)

FDA's commitment letter under the Medical Device User Fee Amendments of 2022 (MDUFA V) records the review goals agreed for FDA's fiscal years 2023 to 2027. For a Pre-Submission the goal is written feedback within 70 calendar days, or five days before the meeting if sooner. The goal applies to 90 % of the cohort, for up to 4,300 requests a year in fiscal 2025 to 2027. Above that number, FDA still intends timely feedback for breakthrough-designated products and those in its Safer Technologies Program. Others get feedback as resources permit. (MDUFA V letter, General, s. II.A)

FDA's report of 26 August 2026 shows that goal met for 98.84 %, 98.66 % and 97.49 % of the cohorts for fiscal years 2023 to 2025. Written feedback took an average of about 62 FDA Days, the calendar days a request is under review at FDA. (MDUFA V report, Tables 9.2 and 9.3; MDUFA V letter, s. VIII)

The report does not measure whether FDA's feedback still held when the later submission arrived. FDA intends its feedback not to change if the submission is consistent with the Pre-Submission and no important new issue arises. The guidance gives new scientific findings, or a new public health concern, as reasons FDA may modify its feedback. It also recommends reconfirming advice on significant study design topics that is more than a year old, where the study has not started. (MDUFA V letter, s. II.A; Q-Submission guidance, s. III)

For a study to be run outside the United States, the guidance says the sponsor should consider submitting the whole protocol through a Pre-Submission before the study starts. It also invites sponsors to bring payors, such as CMS staff, into these meetings. (Q-Submission guidance, p. 22)

Figure 2.1. What each system will answer before the study, and what waits for the application Two tracks, the European Union above and the United States below, run left to right through five stages: before application, protocol design, study runs, application or submission, and decision. A vertical line between protocol design and the study marks where the spend commits. On the European track, four items sit before the study. Structured dialogue on the process, covering the contract, documents, timelines and fees, is encouraged by MDCG 2019-6 Rev.5, Q&A I.6.3. A notified body reviews coverage and class under a documented procedure before any quotation, under MDR Annex VII, Section 4.2(d), and MDCG 2019-6 Rev.5, Q&A I.6.2, expects these topics to be discussed. An expert panel may advise on clinical development strategy for class III devices and class IIb active devices administering or removing a medicinal product, under MDR Article 61(2). The IVDR has no counterpart to Article 61(2). A review of clinical data, a gap analysis or a readiness check before application is drawn as a dashed outline, because MDCG 2019-6 Rev.5, Q&A I.6.2, regards it as consultancy. After the application, views may be exchanged on the sufficiency of clinical data, equivalence and the PMCF plan, under Q&A I.6.3. A band across the track records that consultancy on how to comply is barred throughout, under Annex VII, Section 1.2.3(d), as MDCG 2019-6 Rev.5, Q&A I.6.1, reads it. On the US track, a Pre-Submission gives written feedback on specific questions before the study and is not meant as a pre-review of data. For designated breakthrough devices, FDA may agree a protocol in writing as binding, subject to stated exceptions, under 21 U.S.C. 360e-3(e). Review and decision follow the submission. What each system will answer before the study, and what waits for the application. Before application Protocol design Study runs Application or submission Decision The spend commits here EUROPEAN UNION: MDR AND IVDR Structured dialogue on the process Contract, documents, timelines and fees. Encouraged: MDCG 2019-6, Q&A I.6.3 Views on clinical data sufficiency Also equivalence and the PMCF plan, after the application. MDCG 2019-6, Q&A I.6.3 Coverage and class, before quotation Reviewed by the body; discussed in dialogue. Annex VII, 4.2(d); MDCG 2019-6, Q&A I.6.2 Expert panel advice on clinical strategy Class III; class IIb active devices that administer or remove a medicinal product. MDR Article 61(2); no IVDR counterpart to it Review of clinical data, gap analysis Regarded as consultancy before the application: MDCG 2019-6, Q&A I.6.2 Throughout: no consultancy on how to comply. MDR and IVDR Annex VII, Section 1.2.3(d), as MDCG 2019-6, Q&A I.6.1, reads it. UNITED STATES: FDA Pre-Submission: the sponsor's questions Written feedback; not meant to pre-review data. Voluntary. FDA Q-Submission guidance Review and decision 510(k), De Novo request or PMA. 21 CFR Parts 807, 860 and 814 FDA may agree a protocol as binding In writing, for designated breakthrough devices, with exceptions. 21 U.S.C. 360e-3(e) Open at this stage Regarded by guidance as consultancy at this stage MDR: consolidated text of 19 July 2026. IVDR: consolidated text of 10 January 2025. MDCG 2019-6 Rev.5, February 2025, is guidance and not legally binding. FDA Q-Submission guidance, 29 May 2025. Checked 29 September 2026.
Figure 2.1. What each system will answer before the study, and what waits for the application. Open full size

3. How do the breakthrough routes compare?#

MDCG 2025-9 treats a device or IVD as a breakthrough device if it meets two criteria together. One is a high degree of novelty. The other is an expected significant positive clinical impact for a life-threatening or irreversibly debilitating disease or condition. (MDCG 2025-9, sections 2 and 4.1)

Designation runs through an expert panel opinion, which may be requested at any stage of development if there is sufficient supporting data. The panels "will endeavour" to give it within 60 days. Notified bodies should then prioritise the file and engage in early and structured dialogue on clinical evidence, study design and post-market strategies. These expectations come from the guidance, and the MDR itself imposes no such duty on notified bodies. (MDCG 2025-9, sections 10.1 and 11.1)

Our reading is that, on 29 September 2026, designation ran through an EMA pilot launched on 28 April 2026. The pilot page, which shows the phase currently running, expected the pilot to open to all device classes from the first quarter of 2027 and to IVDs from the third quarter, subject to change. (EMA pilot)

In the United States, the breakthrough statute covers a device that provides more effective treatment or diagnosis of a life-threatening or irreversibly debilitating condition. It also has to meet one of four conditions, such as having no approved or cleared alternative. A sponsor may ask for designation at any time before its marketing submission, and FDA decides within 60 calendar days. (21 U.S.C. 360e-3(b))

For a designated device, FDA may agree in writing to clinical protocols it will treat as binding on both sides. A change then needs the written agreement of both, or a written decision by the director of the reviewing office. That decision has to find a substantial scientific issue essential to safety or effectiveness, and can be taken only after the sponsor has been offered a meeting. (21 U.S.C. 360e-3(e))

Our reading is that Europe has no counterpart to a binding protocol, since a manufacturer that consults an expert panel under Article 61(2) may not invoke any rights to the panel's views. (MDR Art. 61(2))

4. How long does review take on each side?#

Europe#

From 25 February 2027, Implementing Regulation (EU) 2026/977 sets maximum timelines for notified bodies, counted from the complete application. Application review and contract signature take up to 30 days. Quality management system auditing takes up to 120 days, and product verification up to 90 days. Decision and certification take up to 20 days. Under Annex IX the audit and the product verification run in parallel, where the audit programme takes account of the technical documentation assessment. (Regulation 2026/977, Art. 2(2))

As a labelled computation, and because product verification runs alongside the audit, the maxima then add to 30 + 120 + 20 = 170 days from complete application to certificate. The Regulation does not say whether these are calendar or working days. Under the Union's general rule on time limits, periods include public holidays, Sundays and Saturdays. The exceptions are periods that expressly exclude them or are expressed in working days. (Regulation 1182/71, Art. 1, 3(3))

Our reading is that the 170 days are therefore calendar days.

The 170 days exclude interruptions. The notified body may interrupt the clock a set number of times so that the manufacturer can answer questions. It has to interrupt it where it needs an opinion from EMA, a regulatory authority, an expert panel or an EU reference laboratory, which the Commission designates under the IVDR. The Regulation sets no maximum length for an interruption. Running past the maxima is not in itself sufficient reason to refuse a certificate. Articles 1 to 3 of the Regulation, which include these timelines and interruptions, do not apply to procedures whose written agreement was signed before 25 February 2027. (Regulation 2026/977, Art. 3, 8(1); IVDR Art. 100(1))

Measured European review times come from the Commission's 20th notified body survey, with data to 28 February 2026. It reports the time from signed written agreement to certificate as shares of the notified bodies that have issued certificates. Those shares describe bodies and not individual applications.

Certificate Bodies answering Time reported, as shares of bodies
MDR, quality management system and product 39 51 % reported 13 to 18 months; 31 % reported 19 to 24 months
IVDR, quality management system and product 12 67 % reported 13 to 18 months; 8 % reported 19 to 24 months

(Notified body survey, slides 30 and 55)

The current values are in the latest edition of the Commission's survey.

The United States#

A 510(k) shows substantial equivalence to a legally marketed predicate device, and the device may not be marketed until FDA says so. A De Novo request classifies a new device into class I or II where no predicate exists. A premarket approval application (PMA) applies, among others, to class III devices not on the market before 28 May 1976. They must also not be substantially equivalent to such a device or to a later device classified into class I or II. (21 CFR 807.100; 21 CFR 860.200; 21 CFR 814.1(c))

Our reading is that US classes do not map one to one to the MDR classes. A European class does not settle the US route.

Total Time to Decision counts calendar days from receipt of an accepted 510(k) or a filed PMA to a decision. The measured figures below are cohort averages of total days, which add FDA days to industry days, the days the submission waits on the sponsor. (MDUFA V letter, s. VIII)

Route MDUFA V goal Average total days, fiscal 2023, 2024 and 2025 cohorts
510(k) A decision within 90 FDA Days for 95 % 141.46, 143.64 and 141.24, with 3 and 98 still pending in the 2024 and 2025 cohorts
De Novo 70 % within 150 FDA Days for 2023 to 2025, 80 % for 2026 and 90 % for 2027 270.07, 247.51 and 278.12, with the 2025 cohort still open
Original PMA without panel review A decision within 180 FDA Days for 90 % 297.94 for the closed 2023 cohort and 320.35 for 2024, with one PMA still pending

(MDUFA V letter, s. II; MDUFA V report, Tables 1.7, 6.4, 6.5, 8.2 and 8.3)

Decisions include negative outcomes, so a met goal means only that FDA issued a decision, and that decision may be a refusal. Of the closed De Novo cohorts, 44.58 % of 2023 decisions and 40.28 % of 2024 decisions were grants. (MDUFA V report, Table 8.4)

Can the European and US figures be compared?#

The European survey starts at the signed written agreement, and the 2027 maxima at the complete application. FDA's total time starts at receipt of a submission FDA accepts or files. Our reading is that the figures cannot be compared as they stand. Any difference read across them would be produced by where each clock starts.

Each figure can still be used as one step in a timeline plan, provided the plan states where that step starts and ends.

Figure 2.2. Three measures of review time, each on its own clock Three panels, each with its own scale, because each measure starts its clock at a different point. Panel A, European Union, from the 20th notified body survey with data to 28 February 2026, measured from signed written agreement to certificate, shows shares of notified bodies by reported time to a new certificate. For MDR certificates covering the quality management system and the product, from 39 bodies, 51 per cent report 13 to 18 months and 31 per cent 19 to 24 months. For MDR quality management system certificates alone, from 47 bodies, 30 per cent report 6 to 12 months and 62 per cent 13 to 18 months. For IVDR certificates covering the system and the product, from 12 bodies, 67 per cent report 13 to 18 months and 8 per cent, one body, 19 to 24 months. The remainder of each bar, labelled other, is shown as other bands, a labelled computation. Panel B, United States, from the FDA quarterly report of 26 August 2026, measured from receipt of an accepted or filed submission to decision, shows average total days, FDA days plus industry days, by receipt cohort. 510(k): 141.46, 143.64 and 141.24 days for fiscal 2023, 2024 and 2025. De Novo requests: 270.07, 247.51 and 278.12 days. Original PMAs without panel review: 297.94, 320.35 and 240.14 days. Cohorts with decisions still pending at 30 June 2026 are open and drawn as dashed outlines: all three fiscal 2025 cohorts, with 98 510(k)s, 12 De Novo requests and 13 PMAs pending, and the fiscal 2024 510(k) and PMA cohorts, with 3 and 1 pending. Panel C, European Union legal maxima under Implementing Regulation (EU) 2026/977 for the Annex IX route, applying to agreements signed from 25 February 2027, measured from complete application: 30 days for application review and contract, 120 days for quality system audit with 90 days of product verification in parallel, and 20 days for decision and certification, which add to 170 days before interruptions, as a labelled computation. The audit and the product verification run in parallel where the audit programme takes account of the technical documentation assessment, as Article 2(2) requires. Running past a maximum is not in itself sufficient reason to refuse a certificate. Three measures of review time. Each clock starts at a different point, so no two share an axis. A. EUROPEAN UNION, MEASURED: SHARE OF NOTIFIED BODIES BY REPORTED TIME TO A NEW CERTIFICATE Clock: signed written agreement to certificate. 20th notified body survey, data to 28 February 2026. MDR, system and product 39 notified bodies 51 % 31 % 18 % other MDR, quality system only 47 notified bodies 30 % 62 % 8 % other IVDR, system and product 12 notified bodies 67 % 8 % 25 % other 6 to 12 months 13 to 18 months 19 to 24 months other bands Shares of notified bodies, not of applications. "Other bands" is 100 % less the shares the slide states, a labelled computation. The IVDR 8 % is one notified body of 12. B. UNITED STATES, MEASURED: AVERAGE TOTAL DAYS TO A MDUFA DECISION, BY RECEIPT COHORT Clock: receipt of an accepted or filed submission to decision, FDA days plus industry days. FDA quarterly report, actions to 30 June 2026. 510(k) FY 2023 141.46 FY 2024 143.64 FY 2025 141.24 De Novo request FY 2023 270.07 FY 2024 247.51 FY 2025 278.12 Original PMA, no panel FY 2023 297.94 FY 2024 320.35 FY 2025 240.14 0 50 100 150 200 250 300 350 days Cohort closed Cohort open, decisions pending Open cohorts can still move and are not a trend. Pending at 30 June 2026: FY 2024, 3 510(k)s and 1 PMA; FY 2025, 98 510(k)s, 12 De Novo requests and 13 PMAs. C. EUROPEAN UNION, LEGAL MAXIMA: IMPLEMENTING REGULATION (EU) 2026/977, ANNEX IX Clock: complete application to certificate. Applies from 25 February 2027, to agreements signed from that date. Application review, contract 30 Quality system audit 120 Product verification, in parallel 90 Decision and certification 20 0 30 60 90 120 150 180 days Labelled computation: 30 + 120 + 20 = 170 days, before interruptions and the manufacturer’s answers. Audit and verification run in parallel where the audit programme takes account of the technical documentation assessment (Article 2(2)). Our reading, on Regulation (EEC, Euratom) No 1182/71, Article 3(3): calendar days. Running past a maximum is not in itself sufficient reason to refuse a certificate. Sources: 20th notified body survey, 2 July 2026, slides 30 and 55; FDA MDUFA V quarterly report, 26 August 2026, Tables 1.5, 1.7, 6.4, 6.5, 8.3 and 8.4; Implementing Regulation (EU) 2026/977, Articles 2, 3 and 8. Checked 29 September 2026.
Figure 2.2. Three measures of review time, each on its own clock. Open full size

5. Can one study serve both markets?#

For implantable devices and class III devices, the MDR requires clinical investigations. The exceptions include a modification of the manufacturer's own marketed device, where the notified body has endorsed the equivalence and the existing clinical evaluation suffices. They also include devices lawfully marketed under the former directives whose clinical evaluation rests on sufficient clinical data and meets any applicable common specification. (MDR Art. 61(4), (6)(a))

Confirming conformity has to rest on clinical data that provide sufficient clinical evidence. The manufacturer specifies and justifies the level of evidence needed for the device and its intended purpose. The MDR's definition of clinical data places no condition on the country where an investigation was run. An application for a multinational investigation names its Member States and third countries. (MDR Art. 2(48), 61(1); Annex XV, Chapter II, s. 1.8)

Our reading is that the country where a study was run is therefore not, in law, a reason to exclude its data. The open question, which the clinical evaluation answers, is whether the data are sufficient for this device, its population and its setting.

A new investigation run for conformity assessment has to confirm or refute the manufacturer's claims, with enough observations for valid conclusions. It needs enough intended users and a clinical environment representative of normal use in the target population. (MDR Art. 62(1); Annex XV, Chapter I, s. 2.1, 2.4)

FDA accepts data from a well-designed and well-conducted investigation outside the United States on conditions. The sponsor states that it followed good clinical practice, which includes ethics committee review and informed consent, and submits the supporting information. FDA must be able to validate the data if it deems that necessary. The sponsor also states whether the studied device is identical to the submitted one, or explains how they differ. (21 CFR 812.28(a), (b))

A PMA may rest solely on foreign data if, among other conditions, those data apply to the US population and US medical practice. FDA encourages a meeting first. (21 CFR 814.15)

Our reading is that a Europe-first evidence plan has to build that condition into the first study from the start, through its patient selection and comparator.

A sponsor established outside the Union that runs an investigation there needs a legal representative established in the Union. A Member State may accept a contact person instead, for an investigation run only on its own territory, alone or with a third country. (MDR Art. 62(2))

The General Data Protection Regulation (GDPR) prohibits processing data concerning health unless an exception in Article 9(2) applies, such as explicit consent. Member States may add further conditions on health, genetic and biometric data. A protocol lawful in one Member State may then need changes in another. A transfer to a third country may rely on a Commission adequacy decision, or on appropriate safeguards such as standard data protection clauses. The Commission publishes the current list of adequacy decisions. (GDPR Art. 9(1), (2), (4), 45, 46)

Our observation, the label for a pattern seen in practice, is that companies commonly treat adding Europe later as a decision about filing. In practice the design of the first study often settles much of that decision. Clinical evidence plans are in chapter 6, What evidence shows the device is safe and works, and how is it kept current?.

6. What changes for a device already on the market under the directives?#

A device certified or self-declared under the former directives can continue to be placed on the market during the move to the MDR or the IVDR only on conditions. They include continued compliance with the directive, no significant change in design or intended purpose and no unacceptable risk to health or safety. The manufacturer must also have had a quality management system, lodged a formal application with a notified body and signed a written agreement with it, each by a deadline. (MDR Art. 120(3a) to (3c); IVDR Art. 110(3a) to (3c))

These rules apply to directive certificates issued from 25 May 2017, still valid on 26 May 2021 and not withdrawn since. One that expired before 20 March 2023 counts as valid only if the manufacturer and a notified body signed a written agreement before it expired. It also counts as valid if, instead, a competent authority granted a derogation or required the manufacturer to carry out the conformity assessment. (MDR Art. 120(2))

As a dated example, on 29 September 2026 class III devices, and class IIb implantable devices outside a listed set, could be placed on the market until 31 December 2027. Other class IIb, class IIa and some class I devices could continue until 31 December 2028. So could a device self-declared under Directive 93/42/EEC that needs a notified body under the MDR, with a declaration of conformity drawn up before 26 May 2021. The deadlines for the application and the written agreement fell in 2024. (MDR Art. 120(3a) to (3c))

Our reading is that a manufacturer that missed those 2024 deadlines cannot now meet the condition.

Units already lawfully placed on the market, under the directives or under these transitional rules, may continue to be made available. (MDR Art. 120(4); IVDR Art. 110(4))

Our reading is that missing a condition stops further units being placed on the market, while stock already placed can still be sold on by distributors.

The IVDR sets its own dates. An IVD certified under Directive 98/79/EC may continue until 31 December 2027. A self-declared IVD that needs a notified body, with a declaration drawn up before 26 May 2022, may continue until 31 December 2027 at class D. At class C the date is 2028, and at class B or sterile class A it is 2029. The quality management system was due by 26 May 2025 for every class. The application and agreement deadlines are staggered by class, and at class C the agreement was due by 26 September 2026. The consolidated IVDR on the Publications Office site carries any later change. (IVDR Art. 110(3) to (3c))

MDCG 2020-6 says data generated under the directives can be taken into account but are not necessarily sufficient clinical evidence under the MDR. New data may be needed where post-market data fall short and equivalence can no longer be shown. Where the evidence does not support the declared intended purpose, the guidance says the manufacturer shall narrow it. (MDCG 2020-6, sections 5, 6.4 and 6.5)

7. Who pays, and what do capital and partners add?#

FDA generally decides on safety and effectiveness, and CMS on whether an item is reasonable and necessary for Medicare patients. CMS states that FDA market authorisation alone does not entitle a technology to Medicare coverage. Coverage comes through a national or local coverage determination, or claim by claim where neither exists. The item also has to fall within a statutory benefit category. (RAPID notice, s. I, I.A, I.B)

As a dated example, CMS's Transitional Coverage for Emerging Technologies (TCET) pathway, set up in August 2024 for certain eligible breakthrough devices, was paused for new candidates on 11 August 2026. The same notice proposed the Regulatory Alignment for Predictable and Immediate Device (RAPID) coverage pathway instead. As proposed, RAPID takes only certain breakthrough-designated devices still at the pre-submission stage of their investigational device exemption (IDE) study. It also excludes IVDs. (TCET notice, s. I; RAPID notice, Summary, s. II.C, II.G)

Our reading is that in the United States the regulator's decision and the payer's decision are separate steps. A market sequence therefore has to name the first US payment route as well as the first authorisation.

Chapter 12 and chapter 13, Which national route does the product take, and who pays at the end?, set out the Union and national routes to payment. Chapter 17, What will European market entry cost, what does it wait on, and who can pay?, prices the work packages.

A manufacturer not established in a Member State may place a device on the Union market only if it designates a sole authorised representative. Contracts for conformity assessment are concluded directly between the manufacturer and the notified body, with no other organisation. (MDR Art. 11(1); Annex VII, s. 4.2(e); IVDR Art. 11(1))

8. How the answer is reached#

The steps depend on one another, in this order:

  1. The class and route come first, from chapter 1, with the company's place of establishment. If the class is still disputed, any sequence built on it carries the same uncertainty. (MDR Art. 52; IVDR Art. 48)
  2. The four options are compared on all four constraints. Review speed alone leaves out payment, capital and partners.
  3. The evidence each route needs is set against what exists. Data from another market have to be sufficient for the device and its intended purpose. The US route follows the US classification. (MDR Art. 61(1); 21 CFR 860.200)
  4. The early advice open to the device's class is identified next, from the pre-quotation check on class to expert panel advice, a Pre-Submission and the breakthrough routes.
  5. The first study's site and the markets it serves are settled before the protocol. The national data rules of each site country apply. A sponsor outside the Union needs a legal representative there. A Member State may accept a contact person instead, for a study run only on its territory, alone or with a third country. (GDPR Art. 9(4); MDR Art. 62(2))
  6. Each timeline is stated with the point its clock starts from and the date of the figure.
  7. The first payer is named in each market, as a payer or as the budget the purchase comes from.
  8. The company's cash runway is set against each option's time to first revenue. Where the runway ends inside an option's range of time to first revenue, that option needs financing before it pays.
  9. The partners each option needs are named, including any investor whose priority market could override the choice.
  10. The date of the analysis is recorded, with the events that would reopen it, such as a new pilot phase or a new survey edition.

9. The four running cases#

The monitor: a wearable cardiac monitor from a US company#

The monitor records heart rhythm continuously for later review by a clinician. A companion application on the patient's phone displays the recording. Chapter 1 classes the hardware as class IIa under Rule 10, on our reading, and the application separately. The application is class IIa if it only records, and class IIb if it analyses the rhythm to guide a physician's diagnosis. The planning assumption is that it only records.

As illustrative assumptions of this book, the US company is already cleared and selling in the United States. US revenue funds the European file.

Question Answer Basis
Sequence United States first, already taken; Europe second, on existing evidence plus any gap study, since the US studies were not designed for Europe Illustrative assumption
Clinical investigation required? No: Article 61(4) covers implantable and class III devices, and the monitor is neither MDR Art. 61(4)
Evidence question Our reading: the US data are clinical data, and the question is whether they, with any gap study, suffice for the intended purpose in a European population and setting; a gap study follows Annex XV MDR Art. 2(48), 61(1); Annex XV, Chapter I, s. 2.4
Early advice A check of the class before the quotation, with the application planned as class IIa; reuse of the FDA evidence open to discussion, with no promise of acceptance; the sufficiency of the clinical data exchanged after the application; no Article 61(2) consultation at class IIa MDR Annex VII, s. 4.2(d); Art. 61(2); MDCG 2019-6, Q&A I.6.3
Partners A sole authorised representative; Union distributors that buy from the company act as its importers, as in chapter 1 MDR Art. 11(1)
First payer In Germany and the Netherlands, entered together, with France later; each country's national route, which chapter 13 covers, names the payer Illustrative assumption
Main uncertainty Whether the notified body accepts class IIa for the application, a class that holds only while the application does not analyse the rhythm Chapter 1

The triage tool: AI software from a German company#

The triage tool suggests how soon each patient should be seen. Triage nurses in adult urgent care centres use it for patients they have already assessed as having no life-threatening condition. The nurse confirms or changes the suggestion. Chapter 1 finds class IIa to III all arguable under Rule 11 and plans it as class IIb, which stays a planning assumption here.

As an illustrative assumption of this book, Germany is the home Member State and the first European market.

The Article 61(2) route is closed at class IIa. It is closed at class IIb too, because the tool is not an active device administering or removing a medicinal product. At class III it opens. Below class III, Article 106(11) and MDCG 2025-9 allow a breakthrough device to seek panel advice on the data set, though the manufacturer has no right to receive it. (MDR Art. 54(1)(b), 61(2), 106(11); MDCG 2025-9, section 10.3)

Our reading is that on 29 September 2026 the EMA pilot's open phase did not cover the tool at class IIb, and would at class III. The page expected all device classes from the first quarter of 2027. (EMA pilot)

For a European first study meant to reach FDA, the guidance suggests a Pre-Submission on the whole protocol. The data then have to meet FDA's rule on data from investigations outside the United States, which requires the good clinical practice statement and that FDA be able to validate the data. That rule applies to 510(k) and De Novo submissions too. (Q-Submission guidance, p. 22; 21 CFR 812.28(a))

As an illustrative assumption, the first US payer is an urgent care operator buying from its own budget, so the first paid order needs no coverage decision. Medicare is a later and conditional route. TCET is paused for new candidates, and RAPID as proposed takes only certain breakthrough-designated devices. The notice does not say whether a benefit category reaches triage software. (RAPID notice, s. I, II.C, II.G)

Chapter 10, Which AI and cybersecurity rules reach the product, and from when?, covers what the AI Act adds.

The capital test below compares, on illustrative figures, each option's time to first revenue with the company's cash runway. Each step is timed on its own clock with a stated start and end. The test leaves out time it does not price, such as the months from the end of the study to a submission.

For the European certificate, the test takes the two time bands reported by the largest shares of notified bodies for MDR system and product certificates. Of the bodies, 51 % reported 13 to 18 months and 31 % reported 19 to 24, so the test uses 13 to 24 months from the written agreement. As a labelled computation, the other 18 % of bodies (100 less 51 less 31) fall outside both bands. The slide's text summary does not place them. The bands describe bodies, so they are a planning assumption for one application. (Notified body survey, slide 30)

The US side assumes a De Novo request, timed from the closed 2023 and 2024 cohorts. Their 20th percentiles of total days were 214 and 178, and their 80th percentiles 329 and 327. As a labelled computation at 30.4 days a month, the test uses about 6 to 11 months. (MDUFA V report, Table 8.3)

About a fifth of decisions took longer, up to 437 days in the 2023 cohort. Our reading, from the matching decision counts in Tables 8.3 and 8.4, is that these figures cover every decision type, including declines and withdrawals. (MDUFA V report, Tables 8.3 and 8.4)

For the Pre-Submission step, the test rounds FDA's average of about 62 days to written feedback up to 2 to 3 months, placed before the study. The last step, from authorisation to first paid order, is an illustrative 3 to 6 months. (MDUFA V report, Table 9.3)

Months, illustrative Europe first United States first Staged, Europe first Parallel
Pre-Submission on the protocol None 2 to 3 2 to 3 2 to 3
First study 12 to 18 12 to 18 12 to 18 12 to 18
First authorisation 16 to 30: 3 to 6 to a written agreement, then 13 to 24 to a certificate 6 to 11, De Novo 16 to 30, as Europe first 6 to 11, De Novo, ahead of the European certificate
Authorisation to first paid order 3 to 6 3 to 6 3 to 6 3 to 6
Time to first revenue 31 to 54 23 to 38 33 to 57 23 to 38
Runway 36 36 36 30, with both files running
Result Runway ends inside the range Runway ends 2 months before the slow end Runway ends inside the range Runway ends inside the range

The team started from a preference for staged entry, Europe first, and the capital test overturns that preference. At the slow end of its range, no option reaches first revenue before the runway ends.

As labelled computations, United States first falls up to 2 months short (38 less 36). Europe first falls up to 18 short (54 less 36), and staged, Europe first, up to 21 (57 less 36). Parallel spends faster, so its 30-month runway falls up to 8 months short (38 less 30).

If the Pre-Submission ran while the study was being set up, United States first would reach first revenue in 21 to 35 months, a labelled computation. That leaves 1 month of runway at the slow end. A step from authorisation to first paid order longer than 7 months would then exhaust the runway (36 less 18 less 11), and so could a De Novo decision among the slowest fifth.

If Medicare were the first US payer, the last step would need a benefit category and a coverage route. No source here gives that step a length.

The outcome is United States first, through a De Novo request, with the Pre-Submission running while the study is set up. The decision on Europe is taken later, and a smaller financing round is planned before the De Novo decision. As a Union manufacturer, the company needs no authorised representative. (MDR Art. 11(1))

The main uncertainty is whether FDA grants the request. Fewer than half of the 2023 and 2024 De Novo cohorts' decisions were grants, and a refusal would leave no runway for a second attempt. (MDUFA V report, Table 8.4)

The implant: a spinal implant system from a Union company with a directive certificate#

The implant is a spinal implant system: an interbody cage with screws, plates, hooks and rods. Chapter 1 classes the cage as class III, and the screws and plates as class IIb. It puts the hooks in class IIb on MDCG 2021-24 rev.1's reading, and leaves rods, wires and pins open.

As illustrative assumptions of this book, the system was CE marked under Directive 93/42/EEC, the Medical Devices Directive, as a non-active implant. It sells in Germany and Italy, and France is not yet entered. The company is seeking its first MDR certificate under the Article 120 transition, and its European revenue must continue until then.

Question Answer Basis
How long each component may stay on the market, on 29 September 2026 Cage, class III: until 31 December 2027. Screws and plates, class IIb implantables on the listed exceptions: until 31 December 2028. Hooks, class IIb and not named in Article 120(3a)(a): 2027. Rods, not named: 2027 at class IIb or III. Wires and pins, named: 2028 at class IIb, 2027 at class III MDR Art. 120(3a)
On what condition The Article 120(3c) conditions, including the 2024 application and agreement dates, read together with the directive certificate's expiry date MDR Art. 120(2), (3c)
Do the 2027 maximum timelines apply? Our reading from the dates: no, if the agreement was signed by 26 September 2024, before 25 February 2027 Regulation 2026/977, Art. 8(1)
Clinical investigation required? No, if the clinical evaluation rests on sufficient clinical data and meets any applicable common specification; screws and plates fall in a separate exception with the same two conditions MDR Art. 61(6)(a), (b)
Evidence risk New data where post-market data fall short and equivalence can no longer be shown, and a narrower intended purpose MDCG 2020-6, sections 6.4 and 6.5
Early advice Consultation of an expert panel on any new investigation of the class III cage MDR Art. 61(2)
US classification Intervertebral body fusion devices are class II with bone grafting material, and class III with PMA where they include a therapeutic biologic; pedicle screw systems, including screws, plates and rods, are class II for stated uses 21 CFR 888.3080, 888.3070

Our reading is that a PMA for the cage follows only if it carries a therapeutic biologic or falls outside that classification regulation. Its European class III does not decide its US route. A PMA, if needed, may rest solely on European data if, among other conditions, they apply to the US population and practice. (21 CFR 814.15)

The outcome is that Europe stays the anchor market until the first MDR certificate, and United States first is closed for that reason. Parallel is deferred until the product code and classification regulation of each component are confirmed. Staged stays open, so that any new investigation of the cage can also be designed for FDA.

The main uncertainty is whether the Article 120(3c) conditions were met by their 2024 dates. The class of wires and pins also matters, since it decides whether their transition ends in 2027 or 2028.

The near-patient test: a cardiac troponin test from a Swiss company#

The near-patient test measures cardiac troponin in blood, near the patient in hospital emergency departments. Chapter 1 gives it two intended purposes. Serial measurement is class C under Rule 3(j). Single measurement is class B under Rule 6 on the guidance's reading, which is contested. Both are planned as class C until a notified body confirms otherwise, and this case works on the serial purpose.

As illustrative assumptions of this book, the Swiss company has no Union entity and sells through distributors in Germany, the Netherlands, Belgium and Austria. It drew up its declaration of conformity under Directive 98/79/EC before 26 May 2022, with no notified body.

Question Answer Basis
Transitional status May be placed on the market at class C until 31 December 2028, if the Article 110(3c) conditions were met IVDR Art. 110(3b)
Conditions A quality management system by 26 May 2025, a formal application by 26 May 2026 and a written agreement by 26 September 2026, all passed on 29 September 2026; no significant change in design or intended purpose IVDR Art. 110(3c)
Early advice The same bar on consultancy, permitted exchanges and pre-quotation check as under the MDR; no expert panel route on development strategy; the expert consultation at class D does not apply at class C; for breakthrough IVDs, MDCG 2025-9 says "a parallel approach may be envisaged" IVDR Annex VII, s. 1.2.3, 1.2.9, 4.2; Art. 48(6); MDCG 2025-9, section 10.3
Partners A sole authorised representative, and Union distributors as importers, as in chapter 1. For a performance study, our reading from the heading of Article 58 is that a sponsor outside the Union needs a legal representative there for the studies Article 58(1) lists; a Member State may accept a contact person instead, for a study run only on its territory, alone or with a third country IVDR Art. 11(1), 58(1), (4)
Time to certificate 13 to 18 months from written agreement, the band 67 % of 12 IVDR bodies reported, which is 8 bodies as a labelled computation Notified body survey, slide 55
US route Product code and route unconfirmed; CMS's proposed RAPID pathway would not accept IVDs RAPID notice, s. II.C

Our reading is that the plan depends on whether the conditions were met. If they were met, distributor revenue continues while the test is certified, and the certificate has to arrive before 31 December 2028. Suppose the agreement was signed on 26 September 2026, the last day allowed. The survey band then puts the certificate between 26 October 2027 and 26 March 2028, a labelled computation inside that limit.

If any condition was missed, the test cannot now be placed on the market under Article 110. Units already lawfully placed on the market may still be made available. The serial-measurement purpose also has to be the purpose declared under the directive, or no significant change from it. (IVDR Art. 110(3c), (4))

Our reading is that distributors can then sell existing stock. No further units can be placed on the market until the test conforms to the IVDR, which at class C means a certificate.

If a notified body accepted class B for the single-measurement purpose, Article 110(3b) would allow it until 31 December 2029. The application would be due by 26 May 2027 and the agreement by 26 September 2027. (IVDR Art. 110(3b), (3c))

The outcome is Europe first, through the authorised representative and distributors, with the United States deferred. The first payer is the hospital budget in each of the four Member States, which chapter 13 covers.

The main uncertainty is whether the company met the Article 110(3c) conditions by 26 September 2026. After that, it is whether the notified body accepts class C for the serial purpose.

10. When specialist help is worth paying for#

  • The device was on the market under the directives, where continued placing on the market depends on the transitional conditions, their dates and the class. (MDR Art. 120(3a) to (3c); IVDR Art. 110(3b), (3c))
  • One study has to serve both markets, where the data have to meet both FDA's conditions for accepting foreign data and the MDR's requirement of sufficient clinical evidence. (21 CFR 812.28; MDR Art. 61(1))
  • The device may qualify as a breakthrough device, since the European and US criteria differ, and the European route rests on guidance and a pilot programme. (MDCG 2025-9; 21 U.S.C. 360e-3)

Conclusion#

The device's class, settled in chapter 1, decides which forms of early advice are open to the company in Europe, and it has no bearing on the route in the United States. There, FDA answers a company's specific questions through a Pre-Submission, a formal request for feedback. In Europe, guidance from the Medical Device Coordination Group (MDCG), the Member States' expert group, treats advice on how to comply as consultancy. A notified body, the independent organisation that certifies the device, may not provide consultancy, so the company cannot look to it for that help.

On our reading, the MDR, the EU Medical Device Regulation, does not exclude clinical data because the study was run outside the Union. What decides whether the data can be used is whether they are sufficient for the European clinical evaluation, the manufacturer's assessment of the device's clinical evidence. In our observation, the design of the first study often settles much of the decision on adding a second market later. Even where the evidence points to one market first, the way the product will be paid for, the capital the company has, and the partners available to it can justify a different order.

The monitor, a wearable heart monitor from a US company selling at home, enters Europe on its US evidence plus any study needed to fill gaps. Its phone application stays at class IIa only while it records the rhythm without analysing it. With no Union establishment, the company must designate a sole authorised representative there before selling. When the illustrative figures set each option's time to first revenue against the company's cash runway, the triage tool, AI software from a German company, goes United States first. It takes a De Novo request, FDA's route for a new lower-risk device with no legally marketed equivalent. Even then the runway is enough only if the Pre-Submission runs while the study is being set up.

The implant, a spinal implant system certified under the former Medical Devices Directive, keeps Europe as its main market until it has its first MDR certificate, and its European revenue has to continue until then. The near-patient test, a cardiac troponin test from a Swiss company, goes Europe first.

Both are sold in Europe under the old directives. New units may be placed on the market only if the company met conditions that include deadlines for applying to a notified body and signing an agreement with it (Article 120(3c) for the implant, Article 110(3c) for the test). On our reading, if a condition was missed, distributors can still sell stock already on the market, while no further units can be placed. For the near-patient test, new units can then follow only once it conforms to the IVDR, the EU regulation for diagnostic tests, which at its class means a certificate.

Sources#

Each statement was checked against the version shown on the date in the last column.

Source Version used Date of that version Link Checked
Regulation (EU) 2017/745 on medical devices (MDR), consolidated text CELEX 02017R0745-20260719, consolidation 007.001, last amendment M8, Delegated Regulation (EU) 2026/1451 19 July 2026 Publications Office 29 September 2026
Regulation (EU) 2017/746 on in vitro diagnostic medical devices (IVDR), consolidated text CELEX 02017R0746-20250110, consolidation 005.001, last amendment M4, Regulation (EU) 2024/1860 10 January 2025 Publications Office 29 September 2026
Commission Implementing Regulation (EU) 2026/977, notified body procedures and maximum timelines As published, OJ L 2026/977; applies from 25 February 2027 5 May 2026 Publications Office 29 September 2026
Regulation (EU) 2022/123, reinforced role for EMA, consolidated text CELEX 02022R0123-20250101, amended by Regulation (EU) 2024/568 1 January 2025 Publications Office 29 September 2026
Regulation (EEC, Euratom) No 1182/71, rules applicable to periods, dates and time limits OJ L 124, 8.6.1971; no amending act recorded 3 June 1971 Publications Office 29 September 2026
Regulation (EU) 2016/679, General Data Protection Regulation OJ L 119, 4.5.2016, with corrigendum OJ L 127, 23.5.2018 4 May 2016 Publications Office 29 September 2026
MDCG 2019-6, questions and answers on notified body requirements Rev.5 February 2025 European Commission 29 September 2026
MDCG 2022-14, notified body capacity and availability of devices Original August 2022 European Commission 29 September 2026
MDCG 2025-9, guidance on breakthrough devices Original December 2025 European Commission 29 September 2026
MDCG 2020-6, clinical evidence for devices certified under the directives Original April 2020 European Commission 29 September 2026
EMA, scientific advice for high-risk medical devices Web page as read 29 September 2026 EMA 29 September 2026
EMA, expert panel support for breakthrough medical devices, pilot Web page as read, last updated 3 August 2026 29 September 2026 EMA 29 September 2026
Commission availability study, 20th notified body survey (MDR and IVDR) Data status 28 February 2026 2 July 2026 European Commission 29 September 2026
FDA, MDUFA V performance goals and procedures, fiscal years 2023 to 2027 Commitment letter FY 2023 to FY 2027 FDA 29 September 2026
FDA, MDUFA V quarterly performance report Actions through 30 June 2026 26 August 2026 FDA 29 September 2026
FDA guidance, the Q-Submission Program Final guidance, supersedes 2 June 2023 29 May 2025 FDA 29 September 2026
21 U.S.C. 360e-3, breakthrough devices United States Code, 2024 edition; no later amendment found in public laws to 29 September 2026 2024 edition govinfo 29 September 2026
21 CFR 812.28, data from clinical investigations outside the United States eCFR, Title 21 25 September 2026 eCFR 29 September 2026
21 CFR Part 807, Subpart E, premarket notification eCFR, Title 21 25 September 2026 eCFR 29 September 2026
21 CFR Part 814, premarket approval eCFR, Title 21 25 September 2026 eCFR 29 September 2026
21 CFR Part 860, Subpart D, De Novo classification eCFR, Title 21 25 September 2026 eCFR 29 September 2026
21 CFR Part 888, Subpart D, prosthetic orthopaedic devices, 888.3070 and 888.3080 eCFR, Title 21 25 September 2026 eCFR 29 September 2026
CMS final notice CMS-3421-FN, Transitional Coverage for Emerging Technologies 89 FR 65724; paused for new candidates on 11 August 2026 12 August 2024 Federal Register 29 September 2026
CMS notice with comment period CMS-3487-NC, Regulatory Alignment for Predictable and Immediate Device (RAPID) Coverage Pathway 91 FR 51710; proposed, comments to 13 October 2026 11 August 2026 Federal Register 29 September 2026