An Investigational Device Exemption, or IDE, is the mechanism the FDA uses to allow a device that is not yet cleared or approved to be used in a clinical study for the purpose of generating safety and effectiveness data to support a future US submission. At a general framing level, an IDE is required when a clinical study on a device in the United States presents significant risk to subjects, and the study must also receive Institutional Review Board (IRB) approval and run under informed consent before any subject is enrolled. Not every FDA submission needs clinical data, and not every clinical study needs an IDE — but when both apply, the IDE is the regulatory container the study runs inside. This post orients EU founders to the concept. For the specifics of IDE content, timelines, and significant risk determinations for your particular device, work with a US regulatory specialist.

By Tibor Zechmeister and Felix Lenhard. Last updated 10 April 2026.

TL;DR

  • An IDE is the FDA mechanism that allows an investigational device to be used in a US clinical study before the device has been cleared or approved.
  • At a general framing level, studies on devices that present significant risk to subjects require an IDE submitted to the FDA; studies on devices that present non-significant risk follow a lighter path centred on IRB oversight.
  • Every IDE study requires IRB approval, informed consent, and sponsor accountability for the conduct of the investigation, regardless of the size of the sponsor.
  • Not every FDA submission needs clinical data. Most 510(k) clearances do not. De Novo and especially PMA submissions usually do.
  • A clinical investigation designed only for MDR Article 62 and Annex XV can miss FDA-critical elements. Plan the clinical evidence strategy with both regulators in mind from the start, or expect to rerun work.

Why this matters the moment the US enters your plan

The question "do we need clinical data for the FDA" usually arrives late. A founder has been working on the CE file, has read MDR Article 61 carefully, has decided whether a clinical investigation under Article 62 and Annex XV is required, and has either budgeted for one or built the clinical evaluation from literature and equivalence under MDCG 2023-7. Then an investor, a board member, or a US advisor asks the follow-up question. "And for the FDA?"

The answer is not symmetrical with the MDR answer. Some devices that need a clinical investigation under MDR Article 61(4) do not need US clinical data for a 510(k). Other devices that avoided a clinical investigation under MDR through equivalence will still need a US study because the FDA does not accept the same equivalence argument. The two regulators are structurally different, and the clinical evidence question gets different answers on each side.

An IDE study is the container the US answer sits in when clinical data is needed. Before committing to an MDR-only clinical investigation, a founder who also wants the US market should understand what an IDE is, when it applies, and how it relates to the MDR clinical investigation framework they may already know. That is the scope of this post.

What an IDE study is at a general framing level

An Investigational Device Exemption, referred to universally as an IDE, is the regulatory mechanism by which the FDA permits a device that has not yet been cleared or approved for commercial distribution to be used in a US clinical study. The exemption is from the normal US premarket requirements, granted specifically so the device can be studied under controlled conditions to generate the evidence that will later support a 510(k), De Novo, or PMA submission.

An IDE study is therefore not a commercial activity. It is a pre-market evidence-generation activity conducted under FDA oversight, with strict controls on who can run it, where it can be run, what data must be collected, how subjects are informed and protected, and how adverse events are reported. The mental model that works for an EU founder is that an IDE is structurally analogous to a clinical investigation under MDR Article 62 and Annex XV — both are regulated studies on an investigational device, both require ethics oversight, both have sponsor obligations, and both generate evidence that feeds the regulatory submission. The two are not interchangeable. An EU clinical investigation authorised under MDR does not automatically satisfy FDA IDE requirements, and vice versa.

At a general framing level, an IDE study needs three things before the first subject is enrolled. First, FDA clearance of the IDE where the study is a significant-risk study (see next section). Second, IRB approval at each site where the study will run. Third, informed consent from every subject. Without all three, the study is not legal, and any data generated is not usable for an FDA submission.

When an IDE is required

The key threshold on the FDA side is the distinction between a significant risk study and a non-significant risk study. At a general framing level, a significant risk study is one where the investigational device presents a potential for serious risk to the health, safety, or welfare of subjects — examples include implants, life-supporting or life-sustaining devices, devices of substantial importance in diagnosing or treating disease, and other devices that otherwise present a serious risk to subjects. A non-significant risk study is one that does not rise to that level.

Significant risk studies require an IDE application to the FDA, FDA authorisation before the study may begin, IRB approval at each site, informed consent, and full sponsor accountability for the conduct of the investigation. Non-significant risk studies do not require an FDA submission but still require IRB oversight, informed consent, and the sponsor's responsibility for subject protection — the IRB itself makes the initial significant risk or non-significant risk determination, and if the IRB decides a study is significant risk, the study cannot proceed until an IDE has been authorised.

Some device studies are exempt from IDE requirements entirely — certain categories of low-risk studies that the FDA has defined as exempt. The specific exemption categories are narrow, and whether a particular study fits an exemption is a question for a US regulatory specialist, not a blog post.

The practical summary for a founder: a genuinely novel implantable or high-risk device studied in the US will almost certainly need an IDE. A lower-risk device studied in the US may not. The determination is not something you should make alone from reading a primer.

Comparison to MDR clinical investigation

The overlap and the gaps between an FDA IDE and an MDR clinical investigation are where EU founders lose the most time. Both frameworks regulate prospective studies on investigational devices. Both require ethics oversight. Both require informed consent. Both hold a sponsor accountable. Both produce evidence that the regulator relies on to decide whether the device can go on the market.

The differences matter as much as the similarities. Under MDR, a clinical investigation is defined in Article 2(45), governed by Article 62 and the framework of Articles 63 to 82 and Annex XV of Regulation (EU) 2017/745, and run under the harmonised Good Clinical Practice standard EN ISO 14155:2020+A11:2024. The study is authorised by the competent authority of the member state where it runs, following the application procedure around Article 70, and reviewed by an ethics committee that operates under national law. The sponsor obligations sit in Article 72 and Annex XV and in EN ISO 14155. The reporting of adverse events follows Article 80 and the end-of-investigation reporting follows Article 77.

On the FDA side, the structure is different. The FDA itself is the authority that authorises a significant risk IDE study. The ethics oversight is provided by an IRB — an Institutional Review Board operating at each study site. The operational standard is a US good clinical practice framework, not EN ISO 14155. Adverse event reporting runs to the FDA and the IRB on US timelines, not to a member state competent authority on MDR timelines. The endpoints the FDA expects are set by US precedent and FDA guidance documents specific to the device category, not by MDR Annex XV.

The failure mode we see is the clinical investigation that was designed only for MDR Article 62 and Annex XV and then gets presented to the FDA in the hope it will support a US submission. Sometimes it does, in part. More often, the FDA identifies gaps: the primary endpoint was not the endpoint the FDA would have required, the comparator arm was absent, the statistical plan does not match FDA expectations for the device category, the reporting did not include data points the FDA needs. The study survives as supporting evidence but cannot stand alone for the US submission. The founder then has to decide whether to run an additional US study or to rebuild the clinical strategy from scratch.

The right move, if the US is in your plan, is to design the clinical evidence strategy with both regulators from day one. That does not mean running two studies. It often means running one study whose protocol is drafted with both MDR Annex XV and FDA expectations in view, so that the same subjects, the same endpoints, and the same data can feed both submissions. This is the kind of planning that pays back the cost of a US specialist many times over.

IRB approval at a general level

Every IDE study — significant risk or non-significant risk — requires approval from an Institutional Review Board before any subject is enrolled at any site. The IRB is the US analogue of the ethics committee in the MDR framework, and it is the gatekeeper for subject protection at the site level.

At a general framing level, the IRB reviews the study protocol, the investigator's brochure, the informed consent documents, the qualifications of the investigator, and the adequacy of the arrangements to protect subjects. The IRB either approves the study, approves it with modifications, or disapproves it. In a significant risk study, the IRB also confirms the significant risk determination that triggers the FDA IDE requirement. In a non-significant risk study, the IRB makes the initial determination and provides the primary oversight mechanism.

IRB approval is not a formality. It is mandatory, it is serial (each site needs its own IRB review unless a central IRB arrangement is used), and it cannot be worked around. Starting a study without IRB approval is a disqualifying compliance failure that will sink any hope of using the data for an FDA submission.

The sponsor must also maintain the IRB approval throughout the study. Continuing review, protocol amendments, adverse event reporting to the IRB, and annual re-approval where required are all part of the sponsor's ongoing obligations. This is familiar territory for any founder who has run an MDR clinical investigation under EN ISO 14155:2020+A11:2024 — the principles of ongoing ethics oversight travel across both systems, even though the specific procedures differ.

The sponsor of an IDE study is the legal entity that takes responsibility for the investigation. At a general framing level, the sponsor is accountable for selecting qualified investigators, providing the investigators with the information needed to conduct the study, ensuring IRB approval is obtained and maintained, ensuring informed consent is obtained from every subject, monitoring the study to ensure it is conducted according to the protocol and applicable regulations, reporting adverse events and unanticipated adverse device effects to the FDA and the IRBs on the required timelines, maintaining records of the investigation, and submitting the final study report.

For an EU founder who has already worked through sponsor obligations under MDR Article 72 and EN ISO 14155, the mental model is familiar. Sponsor obligations do not scale down for small companies. A two-person startup running a single-site US study has the same sponsor accountability as a multinational running a multi-site pivotal trial. What changes is the scope of the study, not the discipline required to run it. Small sponsors can meet these obligations — often with support from a contract research organisation for specific functions — but they cannot ignore them or make them invisible.

The FDA-specific pieces that surprise founders coming from the MDR side are the IDE-specific progress reports and final report requirements, the FDA's expectations on investigator records and financial disclosure, and the fact that the FDA may inspect the sponsor's records and the sites during or after the study. The operational standard for all of this is set by US regulations and FDA guidance, not by EN ISO 14155 — so the QMS clauses that cover clinical operations need to be read with US requirements in mind, not only with the MDR framework.

When EU startups should consider an IDE

Three situations make an IDE genuinely worth planning for.

The US is in the business plan and the device will need clinical data for its US submission. For most De Novo and PMA pathways, clinical data is expected. For some 510(k)s it is required by the product code or by FDA guidance for the device category. In these cases, the clinical strategy has to include a US-valid study, and the IDE is the container.

The EU clinical investigation has been planned but the US has not been factored in. This is the most common case we see, and it is the most expensive one to fix late. Before any Article 62 and Annex XV investigation is locked down, the protocol should be reviewed against FDA expectations for the device category. Adjustments made before the first subject is enrolled cost almost nothing. Adjustments made after the study is running cost the whole study.

The device is a genuine first-in-class innovation. For novel technologies that do not fit existing classifications on either side of the Atlantic, early regulatory dialogue with both agencies — the competent authority on the EU side and the FDA on the US side — is worth the investment. The FDA pre-submission (Q-Sub) process allows sponsors to discuss study design with the FDA before committing to an IDE application, and for first-in-class devices this conversation is often the difference between a study that supports the submission and a study that has to be redone.

None of these situations is a permission slip to default to the most expensive pathway. The Subtract to Ship Evidence Pass applies equally to US work: exhaust literature, existing data, and cheaper pathways before committing to a new study, and design any study that does happen as the minimum viable investigation that answers the specific scientific question the regulator expects answered. See The Subtract to Ship Framework for MDR Compliance for the foundational methodology; the same discipline transfers to FDA work once a US specialist has identified the pathway.

The Subtract to Ship angle on FDA clinical evidence

The pattern that wastes the most money on the US side is the same pattern that wastes the most money on the EU side: running a study before the intended purpose is locked, before the pathway is confirmed, before the evidence already available has been honestly evaluated, and before the scientific question the study must answer has been written down on one line.

The Evidence Pass reverses this. Lock the intended purpose. Lock the US classification and pathway with a specialist. Evaluate existing data — literature, prior studies, data from the EU clinical investigation if one has been run. Identify the specific evidence gaps the FDA will require to be closed. Only then design the IDE study, and design it as the smallest study that closes the specific gaps. Anything larger is bloat; anything smaller will not carry the submission.

This is not advice to skip required work. Where clinical data is required for an FDA submission, the study must happen. The discipline is in not running studies that the submission did not require, and not running larger studies than the submission needed.

Reality Check — Where do you stand on FDA clinical data?

  1. Is the US in your business plan with a target submission date, or is it still a "we will think about it later" item?
  2. Do you know whether your likely FDA pathway (510(k), De Novo, or PMA) is expected to require clinical data for your specific device category?
  3. If you are planning or already running an MDR clinical investigation under Article 62 and Annex XV, has the protocol been reviewed against FDA expectations by a US specialist?
  4. If a US study will be required, have you had the significant risk vs non-significant risk conversation with a US regulatory specialist?
  5. Do you have a candidate IRB arrangement, or are you still assuming the site will handle it?
  6. Have you budgeted for the full sponsor obligations on the US side, including monitoring, adverse event reporting to the FDA and IRB, and final report preparation — not only the site and patient costs?
  7. If the answer to any of the above is "not yet," is the US in your plan at all, or is it a wish?

Frequently Asked Questions

What is an FDA IDE study in one sentence? An Investigational Device Exemption (IDE) is the FDA mechanism that allows an investigational device to be used in a US clinical study, under FDA and IRB oversight, to generate the safety and effectiveness data needed to support a future 510(k), De Novo, or PMA submission.

Does every FDA submission require clinical data? No. Most 510(k) clearances do not require new clinical data and rely on bench testing, biocompatibility, software verification, and predicate comparison. De Novo requests and especially PMA applications usually do require clinical data. Whether a specific 510(k) needs clinical data depends on the device type and FDA precedent in that category.

Can I use my EU clinical investigation data for an FDA submission? Sometimes, partially. Data from a clinical investigation authorised under MDR Article 62 and Annex XV and run under EN ISO 14155:2020+A11:2024 can often be used as supporting evidence in a US submission, but the FDA may identify gaps in endpoints, design, or statistical analysis that require additional US data. Plan the clinical strategy with both regulators in mind from day one to avoid rerunning work.

Is an IDE the same as an MDR clinical investigation? No. They are structurally analogous — both are regulated studies on investigational devices with ethics oversight, informed consent, and sponsor accountability — but the frameworks, authorities, standards, and reporting obligations are different. An MDR clinical investigation is authorised by a member state competent authority and runs under EN ISO 14155. An FDA IDE study (for significant risk studies) is authorised by the FDA and runs under US good clinical practice requirements with IRB oversight at each site.

What is the difference between a significant risk and non-significant risk study? At a general framing level, a significant risk study is one where the investigational device presents a potential for serious risk to the health, safety, or welfare of subjects — examples include implants, life-supporting devices, and other devices of substantial importance in diagnosing or treating disease. A non-significant risk study does not rise to that level. Significant risk studies require an IDE submitted to the FDA; non-significant risk studies are overseen primarily by the IRB without an FDA submission.

Do I need a US agent or specialist to run an IDE? For any serious US regulatory activity, yes. A US regulatory specialist is the right partner for pathway confirmation, significant risk determination, IDE preparation, and FDA interactions. Our authority is the EU MDR and the Notified Body system; the authority on the US side belongs to specialists who practice inside the FDA system daily.

When should an EU founder start thinking about FDA clinical data? As soon as the US is in the business plan, not after CE marking is complete. Clinical evidence strategy is the single area where planning in parallel pays back the fastest, because a protocol drafted once to serve both submissions costs a fraction of two separate studies run in sequence.

Sources

  1. Regulation (EU) 2017/745 of the European Parliament and of the Council of 5 April 2017 on medical devices, Article 2(45) (definition of clinical investigation), Article 61 (clinical evaluation), Article 61(4) (clinical investigations for implantable and Class III devices), Article 62 (general requirements regarding clinical investigations), Article 72 (sponsor obligations), Annex XV (clinical investigations). Official Journal L 117, 5.5.2017.
  2. MDCG 2023-7 — Guidance on exemptions from the requirement to perform clinical investigations pursuant to Article 61(4)-(6) MDR and on 'sufficient levels of access' to data needed to justify claims of equivalence, December 2023.
  3. EN ISO 14155:2020+A11:2024 — Clinical investigation of medical devices for human subjects — Good clinical practice.
  4. U.S. Food and Drug Administration — Center for Devices and Radiological Health (CDRH), public guidance on the Investigational Device Exemption program, referenced at the general framing level only. https://www.fda.gov/medical-devices

This post is part of the FDA & International Market Access series in the Subtract to Ship: MDR blog. Authored by Felix Lenhard and Tibor Zechmeister. A note on the limits of our expertise: Tibor's authority is in the EU MDR and the Notified Body system. This primer orients EU founders to the FDA IDE concept at a general framing level. For the specifics of significant risk determination, IDE application content, IRB selection, pre-submission strategy, and sponsor execution on the US side, work with a US regulatory specialist who practices inside the FDA system daily. The orientation is ours. The US execution is theirs.