Clinical Evaluation Under EU MDR: How to Plan and Write a Defensible CER

Clinical evaluation is where many EU MDR submissions are won or lost. Notified bodies routinely raise questions on the clinical evaluation report (CER), and legacy devices that once passed under the old directives often need their clinical evidence rebuilt. This guide explains what the MDR requires, how to plan and write a defensible evaluation, and how it stays alive after CE marking.

What the MDR requires

The MDR defines clinical evaluation as a systematic and planned process to continuously generate, collect, analyse and assess the clinical data about a device, in order to verify its safety and performance, including clinical benefits, when used as intended. Article 61 and Annex XIV set the requirements. Every device, whatever its class, needs a clinical evaluation. What changes with class and novelty is the amount and type of evidence expected.

The clinical evaluation must be planned (a clinical evaluation plan), carried out and documented (the CER), and then updated throughout the device lifecycle using data from post-market surveillance and post-market clinical follow-up (PMCF). For class III and implantable devices, the PMCF evaluation report and, where applicable, the summary of safety and clinical performance are updated at least annually.

The clinical evaluation process

Stage 0: Define the scope and write the plan

The clinical evaluation plan (CEP) should set out at least:

• the intended purpose, target groups, indications and contraindications;
• the general safety and performance requirements (GSPRs) that need clinical data support;
• the intended clinical benefits and how they will be measured, with specific outcome parameters;
• the methods for examining qualitative and quantitative aspects of clinical safety, including residual risks and side effects;
• an indicative list and specification of parameters to determine, based on the state of the art, the benefit-risk acceptability of the device;
• the plan for the literature search and how any equivalence will be handled;
• the clinical development plan, including PMCF.

Stage 1: Identify the clinical data

Clinical data can come from four main places:

1. clinical investigations of the device itself;
2. clinical investigations or other studies in the literature on an equivalent device;
3. published and unpublished reports of other clinical experience with the device or an equivalent one, including post-market data and registries;
4. clinically relevant information from post-market surveillance, especially PMCF.

A systematic literature search needs a written protocol: research questions, databases (such as PubMed, Embase and Cochrane), search strings, inclusion and exclusion criteria, and records of what was screened and why papers were excluded. Reviewers often say that a search is not reproducible because it cannot be re-run from the report.

Stage 2: Appraise the data

Each source is assessed for methodological quality and relevance to your device, population and intended purpose. A weak study should not be given the same weight as a well-designed one. Document your appraisal criteria in advance, and apply them consistently.

Stage 3: Analyse the data and reach conclusions

Combine the evidence and conclude whether it demonstrates conformity with the relevant GSPRs, the safety and performance claims, and an acceptable benefit-risk profile compared with the state of the art. State clearly where evidence is limited and how PMCF will close the gap.

Equivalence: handle it carefully

Using data from an equivalent device is possible, but the MDR is stricter than the earlier directives. You must demonstrate technical, biological and clinical equivalence, in sufficient detail, so that there is no clinically significant difference in safety and performance. For class III and implantable devices, you generally need a contract that gives you ongoing access to the other manufacturer’s technical documentation. Many manufacturers find this impractical and instead rely on their own data and literature on the same device. The MDCG has issued guidance on equivalence, and it is worth reading closely before building any strategy on it.

When a clinical investigation is needed

For implantable and class III devices, the MDR requires clinical investigations unless specific exemptions apply, for example for modifications to a device that the same manufacturer already markets lawfully, or devices based on well-established technology under conditions in the regulation. Where an investigation is needed, it follows ISO 14155 and Articles 62 onwards. Plan early: investigations take time, and the decision whether to run one should come from a documented gap analysis, not from a late reviewer question.

Who should write the evaluation

The evaluation should be performed by people with suitable qualifications and experience: usually relevant clinical or scientific knowledge, training in research methodology and regulatory knowledge of the device and its field. Record the evaluators’ qualifications and declarations of interest. Notified bodies check this.

Connecting the clinical evaluation to the rest of the file

• Risk management: the CER’s benefit-risk conclusion must be consistent with the risk file and with the residual-risk statements. See ISO 14971 risk management and our risk management case studies.
• Labelling and instructions: claims in the instructions for use must be backed by clinical evidence.
• Usability: use-related risks and findings belong in the evaluation; see usability engineering.
• Biological safety: see biocompatibility.
• Technical documentation and SSCP: the CER is a core part of the technical documentation, and for implantable and class III devices it feeds the public summary of safety and clinical performance.

Common reasons reviewers push back

• The device’s intended purpose in the CER differs from the labelling.
• Literature search is not systematic or reproducible.
• Equivalence claims lack detail or access to the other manufacturer’s data.
• Benefit parameters are vague, with no measurable clinical outcomes.
• No state-of-the-art analysis, so the acceptability of the benefit-risk ratio is unsupported.
• Post-market data not included, or the PMCF plan has no concrete objectives.
• The CER is not updated after new complaints, field safety actions or changes in the state of the art.

How to prepare for the career side

Clinical affairs and clinical evaluation specialists are in demand. If you are preparing for interviews, our clinical affairs interview guide has model answers and case studies. For wider EU context, read EU MDR and IVDR explained, and for structured training see our RA Certificate Course.

Frequently asked questions

How often must the CER be updated?

Throughout the lifecycle, whenever relevant new data arrives. For class III and implantable devices, PMCF data must be incorporated at least annually. For other devices, the frequency should be set by your plan based on risk.

Can I use a literature-only evaluation?

Sometimes, if the device or an equivalent device is well documented in the literature and the strategy is justified. For implantable and class III devices, the regulation limits this route.

What is the difference between the CER and the PMCF report?

The CER is the overall clinical evaluation. The PMCF evaluation report analyses the data collected through your PMCF plan, and its findings feed back into the CER.

Is a clinical evaluation needed for class I devices?

Yes. Every device needs one. The depth is proportionate to the risk and to the claims, but the requirement applies to all classes.

Talk to the AptSkill team

Need help with clinical evaluation training, or want to build skills in MDR regulatory work? Send us a message.