How to read a clinical evaluation report

A Clinical Evaluation Report gets treated, sometimes, as a literature review with a cover page — compile the studies, summarize what they found, attach it to the technical file. Under EU MDR, that isn’t what the document is for. A CER has to reach a conclusion: does the clinical evidence assembled actually demonstrate that the device meets the General Safety and Performance Requirements, or doesn’t it.

What the report has to establish, not just contain

Annex XIV Part A sets out the substance: an analysis of the current state of the art in the relevant clinical field, including comparable devices and treatment alternatives; a clinical development plan for how sufficient evidence will be generated or gathered if what exists isn’t enough; and the operative piece — an explicit, reasoned assessment of whether the assembled clinical evidence demonstrates conformity with the relevant General Safety and Performance Requirements. A report that stops at “here is what the literature says” without that final judgment hasn’t actually done what Annex XIV requires, regardless of its length. The same gap between citing evidence and arguing what it means shows up in device submissions generally — it’s the same discipline behind reading a risk management file for its conclusions, not just its hazard tables.

Why the MEDDEV structure still shows up everywhere

MDR itself is silent on the report’s table of contents — Annex XIV states what the evaluation must establish, not how to organize the document that proves it. In practice, almost every CER still follows the section sequence from MEDDEV 2.7/1 Revision 4 (state of the art, device description, equivalence if claimed, clinical data appraisal, analysis, conclusions), because notified bodies are fluent in that structure and it maps cleanly onto Annex XIV’s requirements, even though the guidance predates the regulation it’s now read alongside. Knowing the structure is convention rather than regulatory text matters when a reviewer pushes back on format: the argument to have isn’t about the template, it’s about whether the content Annex XIV actually requires is present. It’s one more place where the gap between the EU and FDA systems shows up in how evidence sufficiency gets argued, not just in submission mechanics.

The sufficiency conclusion is the part that gets challenged

Most of the pushback a CER gets — from a notified body, or internally before it goes out — lands on the conclusion, not the literature review. “Sufficient evidence” has to be argued against the specific GSPRs the device has to meet, not asserted generally, and an equivalence claim to another device has to meet MDR’s own bar — narrower than the old MDD’s — for technical, biological, and clinical equivalence to count as supporting evidence at all. A CER that leans on equivalence without being able to defend each of those three legs is one of the most common reasons a clinical evaluation gets sent back.

A CER that compiles evidence without concluding whether it’s enough hasn’t met Annex XIV’s actual requirement, however complete the literature review looks. The report’s job is the conclusion — argued against the specific requirements the device has to meet — not the collection of sources underneath it.