Top 5 Notified Body Findings in CER Reviews

Every CER that passes through a Notified Body’s hands generates feedback. Some of that feedback is minor, but a number of CER reviews result in major non-conformities — findings that halt certification and demand substantive remediation. Based on publicly available guidance, MDCG documents, and regulatory intelligence from across the industry, five findings appear with consistency. Here’s what they are, why they matter, and what you can do about them.

Finding #1: Inadequate Literature Search Methodology

A Common finding across CER reviews is a literature search that cannot be reproduced. Notified Bodies expect to see:

1. Multiple databases searched (PubMed, Embase, Cochrane, etc.)
2. Specific search strings with Boolean operators
3. Date ranges
4. Documented inclusion/exclusion criteria
5. A clear overview of how results were narrowed (e.g., PRISMA-style flow diagrams)

Pulling a handful of favorable studies without documenting how you identified them or why unfavorable studies were excluded is a common mistake. A narrative summary of “relevant literature” without a systematic search protocol is not acceptable. If a second reviewer couldn’t replicate your search and reach the same initial dataset, the methodology fails.

Finding #2: Lack of Alignment Between CER and Technical Documentation

Your CER exists within an ecosystem of technical documentation. It must be consistent with, and cross-referenced to, your:

1. Risk management file (ISO 14971)
2. Instructions for Use (IFU) and labelling
3. Device description and specifications
4. Post-Market Surveillance (PMS) plan and reports

Notified Bodies review these documents in relation to each other. Discrepancies - such as clinical claims in the CER that aren’t reflected in the IFU, or risks in the risk file not addressed in the clinical evaluation - are major red flags.

Traceability is not optional. It is the connective tissue of your technical file.

Finding #3: Benefit-Risk Conclusions are not Supported by Evidence

A CER must arrive at a documented benefit-risk conclusion. But many CERs declare the device “safe and effective” without systematically linking that conclusion to the evidence presented.

Notified Bodies look for:

1. Quantified clinical benefits tied to specific evidence
2. Residual risks identified and cross-referenced with the risk management file
3. A benefit-risk determination that acknowledges and accounts for uncertainty
4. Explicit reference to the state of the art for comparable benefit-risk profiles

A conclusion that reads like a marketing claim rather than a scientific determination will be challenged every time.

Finding #4: PMCF Plans are too Generic or Disconnected

Post-Market Clinical Follow-Up is a clinical commitment and - although considered a separate activity - very tightly linked to clinical evaluation. Notified Bodies are increasingly rejecting PMCF plans that are:

1. Generic templates not tailored to the specific device
2. Lacking in defined PMCF objectives tied to residual risks/uncertainties
3. Missing timelines, milestones, or justification of proposed methods
4. Disconnected from the clinical evaluation’s open questions

Under MDR, PMCF must address the specific knowledge gaps identified in your CER. If your PMCF plan looks like it was written for any device rather than your device, it will be flagged. MDCG 2020-7 provides detailed guidance on PMCF plans and evaluation reports. Your PMCF documentation should demonstrate familiarity with — and compliance with — this guidance. 

Finding #5: Unjustified or Poorly Documented Equivalence Claims

Demonstrating equivalence to a predicate device remains one of the most scrutinized areas of any CER. MDR sets a high bar: equivalence must be demonstrated across clinical, technical, and biological characteristics, and the differences must be scientifically justified as clinically non-significant.

The three most common equivalence failures are:

1. Claiming equivalence without a formal equivalence comparison table in accordance with MDCG guidance document MDCG 2020-5.
2. Failing to demonstrate access to the equivalent device’s technical documentation (required under Article 61(5) for implantables and Class III devices)
3. Identifying differences but not justifying why they do not affect clinical safety or performance

Notified Bodies are trained to probe equivalence claims. If yours doesn’t hold up to structured scrutiny, your CER won’t either.

What These Findings Tell Us

These five findings share a common root cause: CERs that were written as compliance documents rather than scientific arguments. The shift MDR demands is not just procedural, it’s cultural. Clinical evaluation must be owned by people who understand both the science, and the clinical side of things, as well as the regulatory expectations linked to it.

Next in the Series

In our final blog, we answer another frequently asked questions in medical device regulatory affairs: when exactly does your CER need updating? Learn more about our clinical evaluation services, or reach out to our team to discuss your challenges or questions.