How to Structure Your Clinical Evaluation Report Under EU MDR

Most teams know they need a CER. The harder part is knowing what a good one actually looks like — and how it differs from the literature reviews and equivalence declarations that passed Notified Body review under MDD. Under MDR, the CER is a substantive methodological document. It needs to show your work, not just your conclusions.

The structure Annex XIV expects

Annex XIV Part A doesn't mandate a precise section order, but it does define the elements the clinical evaluation must address. In practice, your CER should walk through these in a logical sequence:

1. Scope definition — Define the device, its intended purpose, the target population, and the clinical claims you are evaluating. This is also where you define the clinical evaluation plan (CEP) or reference it. The CEP is often a separate document; the CER summarises its execution.

2. Identification of pertinent data — Document your systematic search strategy: which databases you searched, what search terms you used, which inclusion and exclusion criteria you applied, and what the search produced. This needs to be reproducible. An auditor should be able to re-run your search and get the same results.

3. Appraisal of pertinent data — This is the part many teams underinvest in. Identifying literature is not enough — you have to assess the quality of each source. What kind of study was it? Was the patient population comparable to yours? Were the endpoints clinically relevant? A common weakness here is including papers without critically evaluating them, treating every cited paper as equally valid evidence.

4. Analysis of clinical data — Bring the appraised data together and draw conclusions about safety, performance, and benefit-risk. Where studies show contradictory results, address that directly. Reviewers notice when a CER cherry-picks favourable data and sidesteps unfavourable findings.

5. CER conclusions — Summarise whether the device meets its clinical claims, whether the benefit-risk is positive, and what residual risks require monitoring through PMCF. This section should connect explicitly to the GSPRs it supports.

The Clinical Evaluation Plan

One change MDR brought in that catches teams off guard is the requirement for a Clinical Evaluation Plan (CEP) to be in place before the CER is written. The CEP defines the scope, methodology, and acceptance criteria in advance — not post-hoc. Annex XIV Part A lists the minimum CEP content: the intended purpose, clinical claims, GSPR mapping, the method for assessing equivalence if applicable, and acceptance criteria for safety and performance.

If your team has been writing the CER first and treating the CEP as a formality to fill in afterwards, Notified Bodies are increasingly flagging this. The expectation is that the CEP drives the evaluation, not the other way around.

Equivalence in the CER

If your CER relies on data from an equivalent device, that equivalence claim has to be demonstrated and documented within the CER itself. This means addressing all three criteria — technical, biological, and clinical — and providing evidence for each. Vague statements like "the device is substantially equivalent" don't hold up. If you're relying on a competitor's device, you'll also need to explain how you obtained access to the necessary technical data. This is explored in more detail in the clinical equivalence resource in this category.

Keeping the CER current

A CER is not a static document. MDR requires it to be updated when new clinical data becomes available — typically this means updating it when your PMCF Evaluation Report is finalised. For higher-risk devices (Class III and implantables), the expectation is annual review. For lower-risk devices, a documented rationale for a longer review cycle may be acceptable, but "we haven't reviewed it in three years" is not.

The practical implication: build a review trigger into your post-market surveillance schedule. When PMCF data comes in, it should automatically prompt a CER review. If the new data changes nothing material, document that conclusion. If it changes your benefit-risk picture, the CER needs to be updated before your next Notified Body audit.

What auditors actually check

When your Notified Body reviews your CER, the questions they're working through are roughly: Is the scope clearly defined? Is the literature search methodology sound and documented? Is the appraisal substantive, not just a bibliography? Are the conclusions tied to evidence? Is the benefit-risk analysis balanced? Does the CER connect to the GSPR matrix?

A CER that passes these questions doesn't have to be long — it has to be methodologically sound and traceable. Teams that produce 200-page CERs full of literature summaries but thin on critical appraisal often score worse than teams with tighter, well-argued 60-page reports.