MDR narrowed the clinical evaluation's easiest path

Clinical evaluation under the old Medical Device Directive often ran on a shortcut: find a device close enough in design and materials, cite its published clinical literature, and skip generating new data of your own. EU MDR 2017/745 Article 61 kept that shortcut on the books — equivalence to an already-marketed device is still one of the two recognized routes to sufficient clinical evidence — but Annex XIV and Article 61's own conditions narrowed it enough that a comparison that would have cleared review in 2016 routinely gets rejected now, especially for Class III and implantable devices, where the route is close to a different requirement in practice.

Two routes to sufficient clinical evidence, not one

Article 61(1) requires manufacturers to plan, conduct and document a clinical evaluation confirming conformity with the relevant general safety and performance requirements under normal conditions of use, and to assess the acceptability of the benefit-risk ratio. The clinical data behind that conclusion can come from the device's own clinical investigations, from the scientific literature, or — the route Article 61(3) still permits — from demonstrating equivalence to a device already legally marketed, provided sufficient access exists to the data behind it. A clinical evaluation report has to state, explicitly, which route it's relying on for each conclusion; a CER that blends its own limited data with borrowed literature without stating which one is carrying the weight is hiding the question a reviewer will ask first.

Equivalence is a three-part technical, biological and clinical test

Demonstrating equivalence isn't a family resemblance. It requires showing technical equivalence — similar design, specifications, deployment method and principle of operation; biological equivalence — the same materials in contact with the same tissues for a similar duration; and clinical equivalence — use for the same clinical condition, at a similar severity and stage of disease, in a comparable population. All three have to hold; a device that matches on design and materials but treats a different patient population, or a differently staged version of the same condition, fails the clinical leg regardless of how strong the technical match is. MDCG 2020-5 walks through the three-part test in more detail, and it's the guidance a CER's equivalence section should be checkable against.

For implants and Class III devices, equivalence needs a standing contract, not just a citation

Article 61(5) adds a condition the MDD never carried: for implantable and Class III devices, claiming equivalence to a device marketed by a different manufacturer requires a contract giving the manufacturer relying on the comparison ongoing, full access to that other device's technical documentation — and evidence that the original device's own clinical evaluation was performed to MDR standards in the first place. Absent a common manufacturer or that specific access arrangement, the equivalence route is effectively closed for these device classes, whatever the published literature on the comparator device might say. This is the condition that turns “equivalence is still permitted” into, in practice, a route only a small number of implant and Class III submissions can actually use.

State of the art moved from a citation to a comparison

The literature review a CER performs isn't just a survey of what's been published about the device's own technology — it has to establish the state of the art, the current alternatives and standard of care against which the device's own performance and benefit-risk get judged. A CER that shows the device works, without showing how that performance compares to what's already available for the same clinical purpose, has answered a narrower question than Article 61 actually asks. MEDDEV 2.7/1 Rev. 4 predates the MDR and was written under the MDD, but it remains the methodology most notified bodies still check a CER's literature search and appraisal against, absent MDR-specific replacement guidance covering the same ground.

Where the CER's conclusions have to keep showing up

A clinical evaluation report doesn't close on its own file. Its benefit-risk conclusion is one of the inputs a risk management file's own residual-risk evaluation has to stay consistent with, and the gaps it identifies — a claim the literature doesn't fully support, a population underrepresented in the existing data — are what a PMCF plan under Annex XIV Part B exists to close, feeding directly into the post-market surveillance plan's own class-based reporting cycle. The CER itself sits inside the same technical documentation a design file has to structure around Annex II — not a standalone report filed once and left alone.

A clinical evaluation report outline built around this structure — the two-route data test, the three-part equivalence check, and the Article 61(5) access condition for implants — is previewed in the launch catalog. If your program's equivalence argument runs differently, the shelf takes that correction directly.