Equivalence claims are substantially harder under MDR than they were under the old Medical Device Directive because three things changed at once: MDCG 2020-5 (April 2020) raised the level of detail required across the three pillars in Annex XIV Part A Section 3 of Regulation (EU) 2017/745, Article 61(5) introduced a hard "sufficient levels of access to data" requirement for implantable and Class III devices that MDCG 2023-7 (December 2023) later clarified, and the three pillars of technical, biological, and clinical equivalence are now treated as independent gates rather than a combined judgement. Startups should respond by stress-testing equivalence early against the current guidance, treating competitor-predicate equivalence for implantable and Class III devices as almost never workable, and planning a clean clinical investigation as the default backup rather than the emergency fallback.

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

TL;DR

  • Under the Medical Device Directive, equivalence was governed primarily by MEDDEV 2.7/1 rev 4 (June 2016) and was commonly used. Including against competitor devices. At a level of detail that would not survive an MDR review today.
  • Under MDR, equivalence is governed by Article 61 and Annex XIV Part A Section 3 of Regulation (EU) 2017/745, with MDCG 2020-5 (April 2020) as the primary interpretive guidance. The level of detail expected across technical, biological, and clinical characteristics is substantially higher.
  • Article 61(5) introduced a hard rule for implantable and Class III devices: equivalence across manufacturers requires a contract giving ongoing sufficient levels of access to the predicate's technical documentation. MDCG 2023-7 (December 2023) clarified what that access means in practice.
  • The three pillars are now independent gates. A strong argument on two pillars does not rescue a weak argument on the third, as it sometimes did under MEDDEV practice.
  • The startup response is to run the MDCG 2020-5 three-pillar comparison early, to treat competitor-predicate equivalence as almost never workable for implantable and Class III devices, and to budget a clinical investigation as the default path rather than the disaster recovery plan.

The MDD baseline. What equivalence used to look like

Under the Medical Device Directive (93/42/EEC), equivalence was the workhorse of clinical evaluation. A manufacturer preparing a clinical evaluation report would identify a predicate device on the market, argue that their own device was equivalent in technical, biological, and clinical terms, and use the clinical literature on the predicate as the clinical evidence for their own device. The governing guidance. MEDDEV 2.7/1 rev 4, published in June 2016. Set out the three characteristic groups and the expectation that a CER should cover each of them, but the level of operational detail the guidance described was lower than what MDCG 2020-5 would later demand, and the enforcement pattern across notified bodies was variable.

Several habits grew up inside that baseline. Equivalence was routinely claimed against competitor devices, sometimes on the strength of public information and published literature alone. A well-written narrative comparison was often treated as sufficient in place of a structured feature-by-feature table. A strong argument on the technical pillar was sometimes allowed to carry a lighter treatment of the biological or clinical pillars. The "similar device" category and the "equivalent device" category were sometimes blurred in a single discussion of the clinical context. None of these habits were written into MEDDEV 2.7/1 rev 4 as policy; they grew up as practice around it.

The habits worked, more or less, for legacy devices in a mature regulatory environment where the notified bodies and the manufacturers had been calibrating against each other for years. They stopped working when the MDR took effect and when MDCG 2020-5 turned the operational expectations into explicit guidance.

What the MDR changed

MDR (Regulation (EU) 2017/745) kept the three-pillar structure. Annex XIV Part A Section 3 lists technical, biological, and clinical characteristics, with the same criteria that were familiar from the Directive era. The structural continuity is real. What changed is the level of detail and the hardness of the rules around it.

MDCG 2020-5, published in April 2020, is the single document that captures most of the shift. The guidance is short but dense, and it translates the Annex XIV criteria into concrete expectations that go well beyond the practice many teams had built under MEDDEV 2.7/1 rev 4. The headline moves are four.

First, MDCG 2020-5 expects a feature-by-feature, tabular comparison of the subject device and the predicate against every criterion in the three pillars. A narrative claim without a structured table does not meet the expectation. The guidance does not leave room for the "well-written prose" shortcut that was common in legacy CERs.

Second, the guidance requires that every difference between subject and predicate be identified explicitly and that each difference be justified with evidence. Not asserted. "The difference is not clinically significant" is not a justification on its own. A published study, bench test result, risk analysis, or documented scientific reasoning is.

Third, MDCG 2020-5 separates equivalence from the use of "similar device" data cleanly. Similar-device data has a legitimate place in clinical evaluation for risk analysis, state-of-the-art assessment, and post-market context, but it is not clinical evidence for the subject device and cannot be used as such. Mixing the two is now a documented failure mode rather than a grey area.

Fourth, the guidance reinforces that even a valid equivalence claim does not exempt the subject device from meeting the current state of the art. If the state of the art has moved since the predicate was certified, the equivalence claim does not freeze the clinical evidence at the predicate's era.

The "sufficient levels of access to data" requirement

The second major change from the MDD era sits in the MDR text itself rather than in the guidance. Article 61(4) of Regulation (EU) 2017/745 establishes clinical investigations as the general rule for implantable and Class III devices, listing specific cases where the investigation may be omitted. Article 61(5) governs the use of equivalence across manufacturers for implantable and Class III devices and sets an additional condition on top of the Annex XIV Part A Section 3 criteria: the subject device manufacturer must have a contract in place with the predicate's manufacturer giving ongoing sufficient levels of access to the predicate's technical documentation. Article 61(6) governs the general obligation to justify the non-performance of a clinical investigation in the clinical evaluation where Article 61(4) is relied on.

Under the Directive era, no equivalent contractual access rule existed. A manufacturer could build an equivalence argument against a competitor's device on the basis of public information, published literature, and reasoned inference. Under MDR, that route is closed for implantable and Class III devices. The Regulation requires the access to be contractual and ongoing, which changes the question from "what can we infer about the predicate" to "do we have a relationship with the predicate's manufacturer that would let us prove what we claim."

MDCG 2023-7, published in December 2023, is the authoritative guidance on what "sufficient levels of access to data" means in practice. It addresses the four cases in which implantable and Class III devices may be exempted from the mandatory pre-market clinical investigation under Article 61(4)–(6), and it sets out what a contractual access arrangement has to look like to satisfy the Article 61(5) bar. The practical reading is unforgiving: sufficient access is not public information, not a literature search, not a one-off data purchase, not an NDA signed at a conference. It is an ongoing contract with the predicate's manufacturer over the full technical documentation.

Competitors almost never grant that. The route that worked reliably under the Directive. Equivalence to a competitor's Class III or implantable device on the strength of public data. Now almost never clears Article 61(5).

The implantable and Class III escalation

The combination of Article 61(4), Article 61(5), and MDCG 2023-7 produces a specific escalation for implantable devices and Class III devices that did not exist under the Directive. For these devices, the default under MDR is that a clinical investigation must be performed. Equivalence is one of the listed routes that can displace the default, but it must satisfy both the Annex XIV Part A Section 3 criteria at the MDCG 2020-5 level of detail and the Article 61(5) contractual access requirement. Either test, failed, closes the equivalence route.

For a startup building a novel Class III or implantable device, this has a concrete consequence. The equivalence-to-a-competitor route that used to sit as the default plan in many business cases is effectively closed. Equivalence to the manufacturer's own earlier device still works cleanly. The access is automatic because the manufacturer owns both sides. But that route is only available to companies with a prior device on the market. For a first-product startup in the implantable or Class III space, equivalence against a competitor was the Directive-era fallback that made the clinical evidence plan affordable. Under MDR, that fallback is gone.

The honest response is to stop building business cases on the assumption that competitor-predicate equivalence will be available. Assume it will not. Plan the clinical investigation from day one as the primary route, and treat any later equivalence finding as an upside rather than the base case.

What MDCG 2023-7 added to the picture

MDCG 2023-7 (December 2023) did not invent the Article 61(5) access requirement. The Regulation has carried it since 26 May 2021. What the guidance did was close the interpretive window that had been open in the first two years of MDR application. Several manufacturers and consultants had argued that "sufficient levels of access" could be read broadly, and several notified bodies had been applying stricter interpretations in practice than the text alone seemed to require. MDCG 2023-7 aligned the interpretation with the stricter practice.

Two contributions matter for startups. The guidance confirms that the access must be contractual, ongoing, and directed at the full technical documentation of the predicate, not a subset. And it confirms the four specific cases in which implantable and Class III devices may be exempted from the mandatory pre-market clinical investigation under Article 61(4)–(6), making it easier for a team to check whether any exemption actually applies to their situation rather than assuming one does.

For teams that built their equivalence strategy between 26 May 2021 and December 2023 on the broader reading, MDCG 2023-7 is the document that may force a strategy revision. For teams starting a strategy now, it is the document that sets the expectation from the beginning.

What startups can do

The startup response to the harder MDR bar is not a workaround. It is a different posture toward the clinical evidence question, and it saves time and budget precisely because it stops defending arguments that were never going to survive.

The first move is to run the MDCG 2020-5 three-pillar comparison as a structured, adversarial exercise at the start of the clinical evaluation. Before the CER is written and before sunk costs accumulate. The exercise takes a specific named predicate, compares it to the subject device feature by feature across the technical, biological, and clinical pillars, identifies every difference, and asks whether each difference can be justified with evidence. The question is not "can we write something here". It is "would a sceptical reviewer accept this." If the three pillars hold at the MDCG 2020-5 level of detail, the equivalence route is open. If any one pillar does not hold, the route is closed, and continuing to defend it is waste.

The second move, for implantable and Class III devices, is to treat Article 61(5) as a gate that sits before the three pillars rather than after them. Before investing in a detailed equivalence comparison against a competitor's device, confirm that the contractual access is achievable. If it is not. And for competitors, it almost never is. The equivalence route is closed regardless of what the three pillars would say, and the effort belongs in the clinical investigation plan instead.

The third move is to change the base-case assumption. Under the Directive, many startup business plans treated equivalence as the expected route and a clinical investigation as the contingency. Under MDR, the safer base case is to plan the clinical investigation as the primary route and treat equivalence as the upside if it holds. Budgeting and timeline decisions made under the safer base case are more robust to the discovery that an equivalence argument does not survive review.

The fourth move is to get a second set of eyes on any equivalence claim early. A reviewer who has seen notified bodies reject edge cases will surface failures in hours that would otherwise surface months later. The cost of the early review is a fraction of the cost of a rejected claim discovered late.

When equivalence still works

Equivalence under MDR is harder than under the Directive, but it is not dead. Several patterns still work cleanly and are worth naming explicitly so teams do not over-correct into assuming equivalence is never available.

Equivalence to the manufacturer's own earlier device is the clearest case. Access to the predicate's technical documentation is automatic because the manufacturer owns both sides. Article 61(4) explicitly recognises the case of devices designed by modification of a device already marketed by the same manufacturer. The three pillars still need to be demonstrated at the MDCG 2020-5 level of detail, but the Article 61(5) access gate is satisfied by construction.

Equivalence for non-implantable, non-Class III devices where the access bar does not apply can still work against third-party predicates, provided the three pillars hold at the MDCG 2020-5 level of detail. The rule is the same. Feature-by-feature comparison, justified differences, clean separation from similar-device data. But the Article 61(5) contractual requirement does not add an additional barrier.

Equivalence for modifications of legacy devices under controlled transitional pathways can work where the manufacturer has the data and the pathway allows the legacy evidence to carry forward. This is a narrower case and depends on the specific transitional situation of the device.

The discipline is the same across all three. Run the MDCG 2020-5 comparison honestly, treat the pillars as independent gates, document every difference with evidence, and check whether the access requirement applies before investing in the detail.

The Subtract to Ship angle

The Subtract to Ship framework (see post 065) treats every activity in a regulatory project as something that has to earn its place by tracing to a specific MDR obligation. Equivalence is the highest-leverage subtraction move available in clinical evaluation when it works, and one of the most expensive things to get wrong when it does not. The MDR has raised the bar for when equivalence actually works, and the Subtract to Ship response is to calibrate to the new bar rather than to keep using the MEDDEV-era playbook.

The calibration has two parts. First, stop treating equivalence as a default move and start treating it as a subtraction that must be earned against the current guidance. Run the MDCG 2020-5 three-pillar test honestly and read the result as the answer rather than as a starting point for negotiation. Second, treat a failed equivalence claim discovered early as a win rather than a setback. The sunk cost of a half-built equivalence argument is small compared with the cost of a rejected argument discovered at the notified body, and the runway freed up by cutting the equivalence work is runway that can go into a clean clinical investigation plan instead. The subtraction move is to cut the equivalence work the moment it stops being defensible, not to defend it all the way to the conformity assessment.

Reality Check. Where do you stand?

  1. Is your clinical evidence plan built on an MDCG 2020-5 three-pillar comparison at the current level of detail, or on a MEDDEV 2.7/1 rev 4-era narrative?
  2. If your device is implantable or Class III and your target predicate is a competitor's device, have you confirmed that Article 61(5) contractual access is achievable before investing in the three-pillar detail?
  3. Have you read MDCG 2023-7 and checked whether any of the four exemption cases under Article 61(4)–(6) actually apply to your situation, rather than assuming one does?
  4. Have you treated the three pillars as independent gates, or are you relying on a combined judgement that averages a weak pillar against stronger ones?
  5. Is every difference between subject and predicate identified and justified with evidence, or are any differences acknowledged but unjustified?
  6. Is your base-case business plan built on equivalence as the expected route, or on a clinical investigation as the expected route with equivalence as upside?
  7. Have you had a reviewer who has seen notified bodies reject equivalence edge cases read the argument before you committed to it?

Frequently Asked Questions

Is equivalence easier or harder under MDR than under the MDD? Substantially harder. MDCG 2020-5 (April 2020) raised the level of feature-by-feature detail expected across the three pillars in Annex XIV Part A Section 3, the three pillars are now treated as independent gates rather than a combined judgement, and Article 61(5) added a contractual access requirement for implantable and Class III devices that MDCG 2023-7 (December 2023) later clarified. The habits that worked under MEDDEV 2.7/1 rev 4 practice often do not survive an MDR review today.

Why does Article 61(5) make competitor-predicate equivalence almost impossible for implantable and Class III devices? Because it requires the manufacturer making the equivalence claim to have a contract in place with the predicate's manufacturer giving ongoing sufficient levels of access to the predicate's full technical documentation. Public information, published literature, a one-off data purchase, or an NDA are not sufficient. Competitors almost never grant that level of ongoing contractual access, so the route is closed in practice for most startups building novel Class III or implantable devices against a competitor predicate.

Did MDCG 2023-7 change the rules or just clarify them? It clarified them. The Article 61(5) access requirement has been in force since the MDR became applicable on 26 May 2021. MDCG 2023-7 closed the interpretive window that had been open in the first two years of MDR application by confirming that the access must be contractual, ongoing, and directed at the full technical documentation, and by setting out the four specific exemption cases under Article 61(4)–(6).

Is MEDDEV 2.7/1 rev 4 still relevant under MDR? It is still referenced for continuity in clinical evaluation work, but where it diverges from the MDR text or from MDCG 2020-5, the MDR and the current guidance take precedence. Teams that built their clinical evaluation practice under the Directive should treat MDCG 2020-5 as the operational guidance and MEDDEV 2.7/1 rev 4 as historical context rather than as current policy.

What should a startup building a novel Class III device assume about equivalence? Assume it is not available against a competitor predicate. Plan the clinical investigation as the primary route from day one, budget for it, and design the investigation under Article 62, Annex XV, and the harmonised Good Clinical Practice standard. If an equivalence opportunity emerges later. Typically against the manufacturer's own earlier device or through a narrow transitional pathway. Treat it as upside. Building the base case on equivalence and discovering late that it does not hold is the expensive failure mode.

Sources

  1. Regulation (EU) 2017/745 of the European Parliament and of the Council of 5 April 2017 on medical devices, Article 61(4) (clinical investigations for implantable and Class III devices), Article 61(5) (equivalence across manufacturers and sufficient access to technical documentation), Article 61(6) (justification for non-performance of clinical investigations), Annex XIV Part A Section 3 (demonstration of equivalence). Official Journal L 117, 5.5.2017.
  2. MDCG 2020-5. Clinical Evaluation. Equivalence: A guide for manufacturers and notified bodies, April 2020.
  3. 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.
  4. MEDDEV 2.7/1 revision 4. Clinical Evaluation: A Guide for Manufacturers and Notified Bodies under Directives 93/42/EEC and 90/385/EEC, June 2016 (legacy reference for comparison).

This post is part of the Clinical Evaluation & Clinical Investigations series in the Subtract to Ship: MDR blog. Authored by Felix Lenhard and Tibor Zechmeister. The MDR has quietly closed the equivalence route that used to carry many startup clinical evaluation plans. The sooner a team calibrates to the new bar, the cheaper the recalibration is.