The first CER update: what most manufacturers underestimate

Written by HATEM RABEH, MD, MSc Ing

Your Clinical Evaluation Expert And Partner

in
S

I see this pattern repeat across manufacturers who transitioned to MDR. They produce their first CER. They get certification. Then twelve months later, they assume the update is just about adding some new literature and updating a few statistics.

That assumption costs them months in delays.

The annual CER update is not a document refresh. It is a regulatory checkpoint where you must demonstrate, with evidence, that nothing has changed your benefit-risk conclusion. And if something has changed, you must show you handled it correctly.

Most manufacturers underestimate three things about this first update. Let me walk through what I see go wrong in real submissions.

The Scope Expectation Manufacturers Miss

Under the Medical Device Directive, clinical evaluation was often a static document. You updated it when something significant happened. A new study. A field action. Otherwise, it sat unchanged.

MDR Article 61(11) changed that model entirely. It requires manufacturers to update the clinical evaluation throughout the lifecycle. Not when you feel like it. Continuously.

The annual CER update is your formal proof that this continuous process is happening.

Here is what Notified Bodies expect to see in that first annual update:

• Confirmation that your literature search protocol was executed as planned
• All relevant publications identified in the last twelve months, screened, appraised, and analyzed
• PMCF data collected according to your PMCF plan
• PMS data reviewed and integrated into your clinical dataset
• All incidents, trend reports, and field actions assessed for clinical impact
• Reassessment of your benefit-risk profile
• Confirmation that all clinical claims remain substantiated

If any of these elements is missing or treated superficially, you will receive a nonconformity.

This is not about producing more pages. It is about proving a systematic process occurred.

The Literature Search: Why Your First Update Gets Challenged

Your original CER included a literature search. You defined a protocol. You identified databases. You set inclusion and exclusion criteria. You documented your search strategy.

Now, twelve months later, you must execute that same protocol again. Same databases. Same search terms. Same screening process.

What most manufacturers underestimate is the documentation burden.

The Notified Body does not just want to see a list of new papers. They want to see:

• Search strings executed verbatim
• Number of hits per database
• Screening log showing which papers were excluded and why
• Appraisal forms for each included study
• Synthesis showing how the new data integrates with your existing dataset

If you do not have this trail, you cannot prove the search was systematic. And if the search was not systematic, your clinical evaluation is not compliant.

But here is where it gets more complex.

What If New Evidence Contradicts Your Conclusion?

You run your search. You find three new clinical studies. Two show results consistent with your data. One shows an unexpected adverse event rate that is higher than what you reported.

What do you do?

This is the moment manufacturers freeze. They hope the Notified Body will not notice. Or they try to explain it away with a few sentences.

Neither approach works.

If new evidence raises a signal, you must address it systematically. You must:

• Assess the quality and relevance of the conflicting study
• Compare the study population and use conditions to your intended use
• Determine if the finding is clinically significant
• Update your risk management file if needed
• Reassess your benefit-risk ratio
• Adjust your IFU, warnings, or contraindications if necessary

This is not optional. MDCG 2020-13 makes clear that the clinical evaluation must integrate all relevant data, including data that challenges your assumptions.

PMCF Data: The Integration Gap

Your PMCF plan promised specific data collection activities. Surveys. Registries. A clinical investigation. Whatever method you chose, you committed to collecting data that addresses residual risks and uncertainties.

Now, in your first annual update, you must show that data.

This is where the second major underestimation occurs.

Manufacturers treat PMCF data as supplementary. They add a paragraph summarizing some survey results. Or they write: “PMCF activities are ongoing, data will be available next year.”

That does not satisfy MDR requirements.

Article 61(11) explicitly ties the clinical evaluation update to PMCF. The Annex XIV Part B structure includes PMCF data as a core element of the clinical dataset. If you planned to collect data, you must collect it, analyze it, and integrate it into your benefit-risk assessment.

If the data is not yet available, you must explain why and adjust your plan accordingly. But you cannot simply defer it year after year.

What Integration Actually Means

Integration is not summarizing PMCF findings in a separate section.

Integration means:

• Adding PMCF data to your clinical dataset alongside literature and manufacturer data
• Using PMCF data to confirm or refine your safety and performance claims
• Addressing gaps that literature could not cover
• Updating your risk-benefit conclusion based on real-world evidence

If your PMCF data shows a complication rate of 2% but your literature-based estimate was 1%, you must explain the difference. Is it a population difference? A learning curve effect? A detection bias?

You cannot ignore the discrepancy.

PMS and Vigilance: The Forgotten Clinical Input

Your post-market surveillance system generates data continuously. Complaints. Trend reports. Vigilance investigations. Corrective actions.

All of this is clinical data.

Yet most first annual updates treat PMS and vigilance as separate quality system topics. They reference a complaint rate. They mention no recalls occurred. They move on.

This approach fails the clinical evaluation standard.

MDCG 2020-13 makes clear that PMS data must feed into the clinical evaluation. Every complaint that involves a clinical outcome. Every incident that suggests a residual risk was underestimated. Every trend that shows a performance issue.

These data points must be assessed for their impact on your benefit-risk profile.

The Clinical Significance Question

Not every complaint changes your clinical conclusion. A labeling confusion complaint is not the same as an unexpected tissue reaction.

But you must show you assessed each one for clinical significance.

In practice, this means:

• Reviewing your vigilance log for incidents related to safety or performance
• Comparing observed event rates to your predicted rates
• Identifying any new or increased risks
• Updating your clinical evaluation narrative if findings are clinically relevant

If you collected 50 complaints in twelve months and 10 involved clinical events, those 10 must be discussed in your CER update. Not just listed. Discussed.

What was the nature of the event? Was it expected? Was it related to a known risk? Did it change your understanding of device performance?

This level of detail is what separates a compliant update from a document that gets returned with major findings.

The Benefit-Risk Reassessment: Not a Formality

At the end of your first annual update, you must state your conclusion. Is the benefit-risk ratio still acceptable? Do your clinical claims remain valid?

This is not a checkbox statement.

The conclusion must be justified by the data you just reviewed. If your literature search found no new safety concerns, your PMCF data confirmed expected performance, and your PMS showed no adverse trends, then yes, you can conclude the benefit-risk remains favorable.

But if any of those inputs raised a question, your conclusion must address it.

Here is what I see manufacturers miss:

They write: “The benefit-risk ratio remains acceptable.”

The Notified Body asks: “Based on what? You identified three new studies showing higher complication rates. You collected PMCF data showing longer procedure times. Your PMS revealed two complaints about device fracture. How do these findings support your conclusion?”

The manufacturer has no structured answer because they did not connect the evidence to the conclusion.

What This Means for Your Planning

The first annual CER update is a test of your clinical evaluation process. Not just your documentation.

If your process is sound, the update is straightforward. You execute your literature protocol. You collect and analyze PMCF data. You review PMS findings. You synthesize everything into a reassessment.

If your process has gaps, the update exposes them.

The manufacturers who struggle with the first update are usually the ones who treated the initial CER as a one-time document deliverable. They built it to pass certification, not to sustain ongoing evaluation.

The manufacturers who succeed are the ones who designed their CER process with updates in mind. They documented their search protocol precisely. They set up PMCF data collection systems that produce analyzable data. They built feedback loops between PMS, vigilance, and clinical evaluation.

This is the mindset shift MDR requires. Clinical evaluation is not a document. It is a continuous process that your annual update must prove is functioning.

Practical Steps Before Your First Update

If your first annual update is approaching, ask yourself:

• Can you execute your literature search protocol exactly as documented?
• Is your PMCF plan producing the data it promised?
• Have you reviewed all complaints and incidents for clinical relevance?
• Can you trace your benefit-risk conclusion to specific evidence?

If the answer to any of these is no, you have work to do before you submit.

And if you are preparing your initial CER now, think ahead. Design your search protocol to be repeatable. Set up PMCF data collection to produce analyzable results. Build documentation habits that create an audit trail.

The first annual update will come faster than you expect.

Final Thought

The first annual CER update under MDR reveals whether your clinical evaluation process is real or cosmetic.

It is the moment when Notified Bodies see if you can sustain what you built.

The manufacturers who underestimate it treat it like a document refresh. The ones who pass it treat it like proof of an operating system.

Which side you fall on depends entirely on how you set up your clinical evaluation process from the start.

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

The timing of your first update sets the precedent for ongoing compliance. Learn more about how often to update your CER under MDR.