How to Write a CER That Survives Notified Body Review

Written by HATEM RABEH, MD, MSc Ing

Your Clinical Evaluation Expert And Partner

in
S

The Clinical Evaluation Report is not a scientific paper. It is not a literature review. It is not a summary of your technical file.

It is a regulatory document with a defined purpose: to demonstrate that your device meets the general safety and performance requirements through clinical evidence. Everything else is noise.

Yet most CERs I review fail because they treat it as an academic exercise instead of a demonstration of compliance. The authors cite studies, summarize data, and write elegant paragraphs. But they never answer the core question: does the clinical evidence support your claims and your intended use?

This is not a theoretical problem. It is the reason CERs get rejected at Notified Body review. It is why audits drag on. It is why clinical affairs teams rewrite the same sections over and over.

Let me walk you through how to write a CER that actually works.

Understanding What the MDR Requires

Before you write a single word, you need to understand what the regulation asks for.

Article 61 of MDR 2017/745 defines the scope of clinical evaluation. Annex XIV, Part A, defines the content and structure of the Clinical Evaluation Report. MDCG 2020-13 provides detailed guidance on what each section must contain.

The CER must demonstrate three things:

First, that the device achieves its intended performance under normal conditions of use.

Second, that the known and foreseeable risks are acceptable when weighed against the benefits.

Third, that the claims you make in your labeling and IFU are supported by clinical evidence.

This is not a list of suggestions. It is the regulatory standard. If your CER does not explicitly address these three points, it will fail.

Step 1: Define the Scope Before You Start

The biggest mistake I see is starting the CER before the scope is clear.

What is the intended purpose? What are the indications for use? What is the target population? What are the clinical claims?

These are not abstract questions. They define what evidence you need and how you evaluate it.

If your intended purpose is vague, your literature search will be vague. If your claims are not explicit, your appraisal will be arbitrary. If your target population is poorly defined, you cannot assess whether the evidence is relevant.

I have reviewed CERs where the intended purpose changed halfway through the report. I have seen literature searches that included populations not covered by the device. I have seen appraisals that did not match the claims in the labeling.

All of these are major findings. All of them come from not defining the scope first.

Step 2: Conduct a Systematic Literature Search

The literature search is not a formality. It is the foundation of your clinical evidence.

MDCG 2020-13 is clear: the search must be systematic, reproducible, and documented. That means search terms, databases, inclusion and exclusion criteria, and the rationale for each decision.

Most deficiencies I see in literature searches fall into three categories.

First, the search is too narrow. The authors only search for studies on their exact device or their exact indication. But the MDR does not limit you to your device. It requires you to evaluate all relevant clinical data, including data on similar devices and equivalent procedures.

Second, the search is not reproducible. The search terms are not listed. The date ranges are not justified. The databases are not specified. A reviewer reading the report cannot verify that the search was complete.

Third, the inclusion and exclusion criteria are not defined upfront. Studies are included or excluded without a clear rationale. This makes the appraisal look biased.

The literature search must be systematic because the appraisal depends on it. If the search is flawed, the entire CER is flawed.

Step 3: Appraise the Clinical Data

Appraisal is where most CERs fail.

Appraisal is not summarizing studies. It is not listing findings. It is analyzing the relevance, reliability, and weight of each piece of evidence in relation to your device and your claims.

For each study, you must answer three questions:

Is the study relevant to my device, my intended use, and my target population?

Is the study methodologically sound? Are there biases, confounders, or limitations that affect the conclusions?

What weight does this study carry in the overall body of evidence?

If you do not answer these questions explicitly, the appraisal is incomplete.

I have reviewed CERs where studies were cited but never appraised. I have seen CERs where all studies were rated as “high quality” without justification. I have seen CERs where the appraisal was a copy-paste of the study abstract.

None of these pass review.

Step 4: Analyze the Clinical Data

After appraisal comes analysis. This is where you synthesize the evidence and draw conclusions.

The analysis must answer the three questions I mentioned at the start: Does the device achieve its intended performance? Are the risks acceptable? Are the claims supported?

For performance, you must show that the device works as intended under normal conditions of use. This means clinical outcomes, not bench testing. It means data in the target population, not extrapolation from other populations.

For safety, you must identify all known and foreseeable risks, estimate their frequency and severity, and weigh them against the benefits. This is not a generic statement. It is a quantitative or semi-quantitative assessment.

For claims, you must map each claim in your labeling to the evidence that supports it. If you claim your device reduces pain, where is the data? If you claim it is equivalent to a predicate, where is the equivalence demonstration?

The analysis must be explicit. A reviewer should be able to follow your reasoning step by step.

Step 5: Address the General Safety and Performance Requirements

Annex I of the MDR lists the General Safety and Performance Requirements (GSPR). Your CER must demonstrate that these requirements are met through clinical evidence.

Not all GSPR require clinical data. Some are addressed through design, testing, or risk management. But for those that do require clinical evidence, the CER must make that link explicit.

I have reviewed CERs that do not mention the GSPR at all. I have seen CERs where the GSPR are listed in a table with no explanation. I have seen CERs where the connection between the clinical evidence and the GSPR is assumed but never stated.

This is a structural failure. The CER must show how the clinical evidence supports compliance with each relevant GSPR. If that connection is not made, the report is incomplete.

Step 6: Identify Gaps and Plan Post-Market Surveillance

No clinical evaluation is ever complete. There are always gaps in the evidence.

The MDR does not require perfect evidence. It requires sufficient evidence to demonstrate safety and performance, and a plan to address the gaps through post-market clinical follow-up.

Your CER must identify these gaps explicitly. What questions remain unanswered? What risks require ongoing monitoring? What claims need further validation?

Then, your PMCF plan must address these gaps with specific methods, timelines, and endpoints.

I have reviewed CERs where the gap analysis is generic. “More data is needed.” “Long-term outcomes are unknown.” These statements are true for every device. They are not actionable.

A gap analysis must be specific. Which outcomes? Which populations? Which endpoints? And how will you collect that data?

Step 7: Ensure Traceability and Transparency

A CER must be traceable. Every statement must be supported by a reference. Every conclusion must follow from the evidence. Every decision must be justified.

This is not about writing more. It is about writing clearly.

When I review a CER, I should be able to follow the logic from start to finish. I should be able to verify your sources. I should be able to understand why you reached your conclusions.

If I cannot do that, the report fails.

I have seen CERs where references are missing. I have seen CERs where the evidence does not support the conclusion. I have seen CERs where the reasoning is circular.

All of these are avoidable. They come from not reviewing your own work critically before submission.

Step 8: Write for the Reviewer, Not for Yourself

This is the most important step, and the one most often ignored.

Your CER is not written for you. It is written for a Notified Body reviewer who does not know your device, does not know your company, and has 50 other CERs to review this month.

That reviewer needs to understand your device, your evidence, and your reasoning quickly and clearly. If they cannot, they will raise findings.

This means structure matters. Language matters. Clarity matters.

Use headings and subheadings. Use tables and figures where appropriate. Define abbreviations. Write short, direct sentences.

Do not assume the reviewer will read between the lines. Do not assume they will understand your implicit reasoning. State it explicitly.

I have reviewed CERs that are technically correct but structurally confusing. They pass on the second or third revision, after the authors reorganize and clarify. That delay is avoidable.

What Happens After You Submit

The CER is never final. Under the MDR, clinical evaluation is a continuous process.

You will update the CER as new evidence emerges. You will revise it based on PMCF data. You will refine it as your understanding of the device evolves.

But if you follow the structure I have outlined, your initial CER will be solid. It will pass Notified Body review. It will withstand audit. And it will serve as the foundation for all future updates.

The difference between a weak CER and a strong one is not the amount of evidence. It is the quality of the reasoning.

A strong CER demonstrates that you understand your device, that you have evaluated the relevant evidence systematically, and that you have drawn justified conclusions about safety and performance.

That is what the MDR requires. That is what Notified Bodies look for. And that is what separates a CER that survives review from one that triggers major findings.

Write with that standard in mind, and the rest will follow.

Peace,
Hatem
Clinical Evaluation Expert for Medical Devices
Follow me for more insights and practical advice.