EU MDR Clinical Evaluation Report: Requirements, Structure and Common Mistakes

Introduction

The Clinical Evaluation Report is the most consequential document in an EU MDR technical file, and the one Notified Bodies most often find inadequate. It causes more delays, more remediation cost and more risk to a CE marking timeline than any other single deliverable.

The reason is not that the requirements are obscure. It is that the MDR raised the standard in three specific places — equivalence, what counts as sufficient clinical evidence, and the obligation to keep the report current — and a great deal of CER practice carried over from the Directive without registering those changes.

This guide covers the framework in Article 61 and Annex XIV, the MDCG guidance that has accumulated around them, the five-stage process, the structure reviewers actually assess against, the equivalence rules, the update frequency by class, and the ten deficiencies that generate the most objections.

Table of Contents

• The framework: Article 61 and Annex XIV
• The guidance that governs the method
• The clinical evaluation process, stage by stage
• CER structure: the ten sections
• Clinical equivalence: the three gates
• The CER and PMCF
• Update frequency, by class
• Evaluator qualifications
• The ten deficiencies that generate objections
• Frequently asked questions
• Conclusions

The framework: Article 61 and Annex XIV

Article 61 establishes the obligation: manufacturers conduct a clinical evaluation to confirm conformity with the relevant general safety and performance requirements in Annex I. It applies to every device, of every class. There is no exemption for low-risk devices — only a difference in the depth the evaluation needs.

Annex XIV divides into two parts: Part A covers the clinical evaluation process, Part B covers post-market clinical follow-up. The two are one system, and treating Part B as a separate exercise is where the most expensive findings originate.

The guidance that governs the method

MEDDEV 2.7/1 Revision 4 predates the MDR and remains the primary methodological reference. The Commission has not replaced it, and Notified Bodies continue to assess CER quality against it. What has changed is that a layer of MDCG guidance now sits on top, tightening specific points the MEDDEV did not anticipate.

MDCG 2020-13 is the most useful document in the list and the least read by manufacturers. It is the template the Notified Body assessor fills in while reviewing your CER. Structuring the CER so that each of its sections can be answered directly from a corresponding section of your report removes an entire category of avoidable question.

The clinical evaluation process, stage by stage

MEDDEV 2.7/1 Rev. 4 structures the evaluation in five stages, numbered 0 to 4. Each has to be documented.

Stage 0 — scope and the CEP. Before any data is collected: the device description covering intended purpose, users, patient population, indications, contraindications and clinical claims; the intended clinical benefits and the outcomes used to demonstrate them; and the Annex I requirements needing clinical evidence. All of it documented in the Clinical Evaluation Plan, which is mandatory under Annex XIV, part of the technical documentation, and updated whenever the device or its intended purpose changes.

Stage 1 — identification of pertinent data. Three sources: clinical investigations conducted under Annex XV, which give the highest quality evidence; literature concerning the device, equivalent devices or the clinical condition, gathered by systematic review methodology; and post-market data — complaints, vigilance, PMS reports, PMCF results, registries and other real-world evidence.

Stage 2 — appraisal. Each item assessed for methodological quality, relevance to the device under evaluation, potential bias and confounding, and then explicitly weighted for its contribution to the evidence base. The appraisal has to be documented in enough detail for an independent reviewer to follow the reasoning. A case report and a randomised controlled trial cannot contribute equally, and a table in which they do is a finding.

Stage 3 — analysis. Does the appraised data demonstrate that the device achieves the intended clinical benefits, that risks are acceptable weighed against them, and that it performs as claimed? The analysis addresses clinical performance, clinical safety, the benefit-risk determination, and the residual risks requiring further PMCF data.

Stage 4 — the report. The CER itself, plus the PMCF plan that follows from the uncertainties it identified.

CER structure: the ten sections

Neither the MDR nor MEDDEV 2.7/1 prescribes a template. The structure below reflects what Notified Body reviewers assess against, and maps onto the MDCG 2020-13 assessment report.

Where this sits physically is Annex II section 6.1 of the technical file, cross-referenced from the GSPR checklist — our guide to the EU MDR Annex II structure covers the placement and the cross-references.

Clinical equivalence: the three gates

Under the Directive, a CER built almost entirely on literature from equivalent or similar devices was normal practice. The MDR closed that route for most devices, and the mechanism is that all three criteria of Annex XIV Part A section 3 must be met simultaneously.

The contract requirement has effectively eliminated equivalence-based CERs for Class III implantables, because competitors do not grant access to their technical documentation. For lower classes equivalence remains available, but the three criteria have to be demonstrated item by item rather than asserted, and MDCG 2020-5 sets out what that demonstration looks like.

The CER and PMCF

The CER and the post-market clinical follow-up plan are one system. The CER identifies the residual clinical uncertainties; the PMCF plan defines the activities that will address them; the PMCF evaluation report reports what those activities found; and that report feeds the next CER update, which either closes the uncertainty or identifies a new one.

The consequence is that a CER concluding there are no residual uncertainties, and therefore no PMCF, will be challenged unless the device has an exceptionally long and well-documented market history. Every device has open questions; a report presenting none has either not looked or not said. Our guide to EU MDR post-market surveillance covers the PMCF plan and report in detail.

Update frequency, by class

The CER is a living document, and the required cadence differs by class. This is frequently overstated as “annually for everyone”, which imposes work the Regulation does not require on lower classes — and understated for the classes where it does.

Regardless of class, five events trigger an update outside the planned cycle: significant post-market safety data; new literature materially affecting the benefit-risk conclusions; a change to the device or its intended purpose; results arriving from a planned PMCF study or registry; and a shift in the state of the art. That last trigger is the one manufacturers miss, and it is the one that can turn an acceptable benefit-risk conclusion unacceptable with nothing having changed about the device.

The review schedule belongs in the ISO 13485 quality system with an assigned owner and defined criteria for an unplanned update. Organisations without one collect findings during surveillance audits with dependable regularity.

Evaluator qualifications

Article 61(6) requires clinical evaluations to be carried out by suitably qualified individuals with expertise in the relevant medical field, and the qualifications have to be documented in the CER — normally through an appended CV. Notified Bodies routinely review those CVs and raise objections where the expertise is not demonstrable.

Three practical consequences follow. A regulatory affairs professional without a clinical background cannot sign a CER as sole evaluator. A clinical expert without knowledge of MDR requirements needs regulatory support alongside them. And many manufacturers engage specialist consultants or CROs to author or co-author, which is entirely acceptable provided the qualifications are documented.

This is the simplest deficiency on the list to prevent, and one of the most common to encounter.

The ten deficiencies that generate objections

The last one is the fastest self-check available and the one most often skipped. Take the IFU, list every clinical claim it makes, and find the specific evidence in the CER supporting each. A claim with nothing behind it is either removed from the labelling or evidenced — and discovering which during a Notified Body review costs a remediation cycle.

Frequently asked questions

Is a CER required for all medical devices under EU MDR?

Yes. Article 61 requires clinical evaluation for every device regardless of class. The depth varies substantially — a Class I non-sterile device requires far less than a Class III implantable — but no device is exempt.

Can a CER be based entirely on literature?

For Class I and Class IIa devices a literature-based CER may be sufficient where the evidence adequately demonstrates conformity with the applicable GSPRs. For Class IIb and Class III, Notified Bodies expect clinical investigation data or a very well-justified rationale for why literature-based evidence suffices. For Class III and implantable devices Article 61(4) makes clinical investigations the default, with reliance on existing data requiring justification.

How often must the CER be updated?

For Class III and implantable devices, at least annually with PMCF data, under Article 61(11). For other classes the frequency is justified in the Clinical Evaluation Plan and is commonly aligned to the PSUR cycle — annual for Class IIb, two-yearly for Class IIa. Five events trigger an update outside the planned cycle regardless of class: significant safety data, material new literature, a device or intended purpose change, PMCF results, and a shift in the state of the art.

What are the three criteria for clinical equivalence?

Technical, biological and clinical, and all three must be met simultaneously under Annex XIV Part A section 3. Technical covers design, materials in body contact, principles of operation, deployment method and specifications. Biological covers materials, tissues and fluids contacted, and duration and frequency of contact. Clinical covers intended purpose, body site, clinical condition, patient population and user. For Class III and implantable devices a contract granting access to the equivalent device’s technical documentation is additionally required.

Who can write a Clinical Evaluation Report?

Suitably qualified individuals with documented expertise in the relevant medical field, per Article 61(6), with their qualifications evidenced by a CV appended to the report. A regulatory affairs professional without clinical background cannot sign as sole evaluator, and a clinical expert without MDR knowledge needs regulatory support. External consultants and CROs are acceptable authors provided the qualifications are documented.

What is the difference between a CER and a clinical evaluation?

The clinical evaluation is the process: the systematic collection, appraisal and analysis of clinical data. The CER is the document recording that process and its conclusions. The report is the output, not the evaluation.

Can one CER cover a device family?

In some circumstances, where the devices share the same intended purpose, clinical mechanism of action and patient population. Notified Bodies vary in how readily they accept family CERs, so the approach should be agreed with the Notified Body before it is adopted rather than presented as a fait accompli.

What happens if the Notified Body rejects the CER?

It issues a list of questions or objections that must be resolved before conformity assessment can proceed. Each remediation cycle adds months and cost to the certification timeline, which is the practical argument for the self-checks above — particularly mapping every clinical claim in the IFU to specific evidence in the report.

Conclusions

The CER is where a technical file states its case, and the MDR raised the standard in three places that carried over from Directive-era practice: equivalence, what counts as sufficient evidence, and the obligation to keep the report current.

Two failures account for most objections, and neither is about the quality of the underlying evidence. The CEP written after the evaluation, which invalidates the sequence the whole method depends on. And the state of the art assembled from favourable comparators, which sets the acceptability benchmark in the manufacturer’s own favour.

The most useful thing to do before submission is also the simplest: read MDCG 2020-13, the assessment template your reviewer will complete, and check that each of its questions can be answered directly from a section of your report. Then take the IFU, list every clinical claim, and find the evidence for each one.

If you are building or updating the documentation, the EU MDR Clinical Documentation Kit covers the CEP, the CER, the PMCF plan and evaluation report and the SSCP, aligned to the MDCG guidance listed above.

Related articles

• EU MDR Technical Documentation Requirements
• EU MDR Annex II: Technical Documentation Structure
• Benefit-Risk Analysis under EU MDR
• EU MDR Post-Market Surveillance: PMS Plan, PSUR and PMCF
• EU MDR GSPR Checklist
• ISO 13485:2016 — The Complete Guide