Common Notified-Body Findings in EU MDR Clinical Evaluation Reports and How to Avoid Them

Clinical Evaluation Reports (CERs) are one of the most scrutinized components of European Union Medical Device Regulation (EU MDR) submissions and one of the most common sources of Notified Body deficiencies. Even experienced manufacturers can find themselves facing costly rework due to avoidable issues like vague safety objectives, incomplete state-of-the-art benchmarking, or unsubstantiated equivalence claims. In this post, we outline the most frequent CER findings cited by Notified Bodies and offer practical strategies to help you avoid them.

‍

Why This Matters

Notified Bodies have repeatedly indicated that clinical evaluations and clinical investigations are the single most frequent source of MDR non-conformities across all device classes, outpacing Quality Management System (QMS), risk management, and labeling issues combined.  A review of recent deficiency letters and white-papers shows the same pattern: Notified-Body reviewers are still flagging predictable, repeatable weaknesses in CERs.  Here, we highlight common NB findings based on our own extensive experience to help shorten your review cycles, lower remediation costs, and protect your launch timelines.

‍

1. Objectives are Too Generic and/or Not Measurable

Notified Bodies note that safety and performance objectives that lack Specific and Measurable Outcomes (SMOs) top the deficiency list.  Vague endpoints (e.g., “demonstrate safety”) are difficult to quantify or measure and Notified Body reviewers may want to see your evidence in more measurable terms. Similarly, safety and performance objectives should be identified with clear acceptance criteria. Reviewers want to see that you understand how your device fits within the therapeutic landscape and clear acceptance criteria helps you show this.  

Example question:

The safety objective is shown below. Would you please provide a more granular and quantified safety objectives, from identified top serious AEs associated with non-vascular procedures, with a quantitative acceptance criteria for each (with justification)?

‍

Ways you can fix this:

• Build SMOs during your planning phase. Clearly define and explain what data you will use to assess your device’s safety and performance. The SMOs should be based on your device’s intended purpose (performance) and risk profile (safety).  

• Establish a quantitative threshold or acceptance criterion for each SMO. These criteria should be justified with an appropriate data source, such as your State of the Art (SOA)

• Document your process for determining acceptance criteria and your methodology for analyzing your subject device data in both the Clinical Evaluation Plan (CEP) and CER.

‍

‍

2. Missing or Inappropriate Benchmark Data

Notified bodies repeatedly ask, “What does good look like?” Essentially, Notified Bodies want to see you think critically about your device and ensure you’re placing it within the currently accepted state of the art. Without a comprehensive State of the Art (SOTA) the reviewer cannot judge if your outcomes are clinically acceptable.  

‍

Practical tips

1. Conduct a dedicated state of the art literature review. This review should address the landscape of the target therapy in which the subject device is intended to be used. A strong SOTA also highlights alternative treatment options and provides a robust discussion of the medical condition for which the subject device is intended to be used.  

2. Leverage guidelines, meta-analyses, reviews, and clinical literature to identify parameters used to assess success (e.g. procedural success rates, re-intervention rates, etc.). Use these sources in developing your SMOs and, as appropriate, acceptance criteria.

3. When appropriate, conduct a specific literature review for similar devices to your device under evaluation. By investigating the SMO data for these benchmark devices, a CER writer can often develop thoroughly justified acceptance criteria.  

‍

‍

3. Incomplete or Overly-Broad Intended Purpose

Devices often have broad clinical indications, but you may only have evidence for a subset of indications. Notified bodies flag that disconnect: “Intended purpose statement is incomplete; clinical indications are unclear,” or “Evidence does not support the full breadth of the intended purpose.”  

‍

Potential Mitigation Strategies:

• Change the presented intended purpose to match the evidence or generate new data before submission.

• Stratify results by indication, population, device variant, and user group in the Data Mapping section so the reviewer sees clear traceability.

‍

‍

4. Inappropriate Evidence Stratification

Reviewers expect the data to map to every combination of patient type, indication, and device variant covered by the label. Failure to stratify your evidence may lead to findings such as “Clinical data have not been appropriately stratified against the intended patient populations, treatment indications, device variants and combinations.”  

‍

Checklist:

• Provide summary tables that display the available SMO data for each indication and patient population, as well as for each device variant.

• Confirm sample sizes are adequate for each subgroup.

• Where gaps exist, justify with risk-based arguments and/or outline confirmatory Post-Market Clinical Follow-up (PMCF) activities.

‍

‍

5. Literature Searches Not Systematic or Reproducible

This common Notified Body finding is often phrased as: “It is not clear that systematic search methods have been used for the literature review.”

What reviewers expect:

1. A prespecified protocol (databases, timeframes, search strings, inclusion/exclusion criteria).

2. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)-style flow diagram with numbers at each screening step.

3. Rationale for any filters or language restrictions.

4. Audit trail of excluded records, including the reason for exclusion.

‍

‍

6. Equivalence Rationale Not Comprehensive or Complete

Medical Device Coordination Group (MDCG) 2020-5 reminds manufacturers that all three pillars—technical, biological, and clinical characteristics—must be similar or the same before equivalence can be claimed, and the CER must show full access to data for the reference device.

Best practice

• Provide a side-by-side equivalence table listing every MDR Annex XIV characteristic and provide justification for every “same” or “similar” assessment. If you cannot fully justify a conclusion of “same” or “similar”, you will need to consider a different path to conformity.  Our team of experts can help!

• Include evidence of access to the equivalent device’s technical file when required.

‍

7. Benefit–Risk Analysis Not Fully Substantiated

Even when your collection of clinical data looks favorable, reviewers frequently conclude: “Benefit–risk conclusion is not justified in relation to the SOTA.”  

Avoid this by:

• Presenting quantitative comparisons (e.g., observed AE rates or procedural success rates in subject device data) against the state of the art .

• Addressing residual risks identified in the Risk Management File as part of the overall risk assessment.

‍

8. PMCF Gaps and Weak Justifications for Not Providing PMCF Data

Deficiencies often cite “The PMCF plan does not address residual risks” or reject blanket statements that no PMCF is necessary.  

What we’ve seen work:

• Tie each open question or data gap directly to a PMCF activity (registry, survey, targeted study).

• Show that the PMCF activity’s endpoints mirror the SMOs needing confirmation.

• If PMCF data is unnecessary, provide a thorough justification for why  proposed ongoing surveillance is sufficient.

‍