HomeResources › Notified Body Reviewer Webinar

What a Notified Body Clinical Reviewer Actually Looks For

A conversation on clinical evaluation under EU MDR with Dr. Surash Surash (TÜV SÜD Centre for Clinical Excellence) and Dr. Hatem Rabeh (Clinical Evaluation Navigator)

Nobody outside a notified body sees a clinical evaluation being read. Recorded live on 18 August 2026 in front of 300 registrants, this hour is a clinical reviewer describing how he works through a file: what he opens first, where submissions come apart, and what makes one easy to approve. No slides, no pitch, one conversation.

Key Takeaways from This Session

The executive summary sets the tone for the whole review. It is the first thing opened, followed by the device description and the summary of safety and clinical performance. The question being asked from the first page is what the actual intended benefit is, and a summary that does not answer it starts the review badly.
Alignment is what separates a good file from a bad one. The clinical evaluation plan, the report, the PMCF documents and the risk file have to tell one story.“It is all about alignment. Telling the same story, but different chapters of that story.”
Your claims and your thresholds have to compare like with like. The apples-to-apples problem ran for several minutes and it is the clearest teaching moment of the hour: a claim compared against a threshold measured a different way, in a different population, or on a different instrument, is not a comparison at all.
State of the art is not the gold standard. Conservative management counts, and leaving it out of the state of the art removes the very comparator that makes a benefit visible.“It is a very easy decision to operate, but it is a very difficult decision not to operate.”
A literature search is judged on reproducibility. The test applied is whether the reviewer can run your search and land in the same place.“Can I go on to PubMed and put in the search terms and come out with exactly the same results?”
Cast the net wider than ten years. A ten-year window is a convention, not a rule, and it silently discards the landmark papers that established the outcome levels you are being measured against.
AI in the loop needs appropriate human oversight. The useful framing offered was to treat AI in this context as a find function rather than an author, with the firewall problem named as the practical obstacle in a regulated environment.
The reviewer will not consult, and that is the boundary of the hour. What is offered is how a reviewer thinks, not what to write in a specific file. That limit was set on the record at the start and held for the full session.

What We Cover in This Session

What a reviewer opens first, and in what order the file is read
Why the executive summary decides the tone of the whole assessment
Alignment across the CEP, the CER, the PMCF documents and the risk file
Claims and thresholds, and what a like-for-like comparison actually requires
State of the art compared with the gold standard, and where conservative management sits
Literature searching: reproducibility, search terms, and the ten-year window problem
Acceptance criteria: where they come from and where they must not come from
Equivalence, and the difference between equivalent and similar devices
Common failure points in clinical evaluation submissions
AI and appropriate human oversight in regulatory writing

Meet the Speakers

Dr. Surash Surash

FRCS (Neuro.Surg), MD, LLM • Senior Clinical Reviewer, TÜV SÜD Centre for Clinical Excellence

Dr. Surash brings a unique combination of clinical, legal and regulatory expertise.

Neurosurgeon and Legal Expert: A former UK Consultant Neurosurgeon who also holds a Master of Laws (LLM) with expertise in clinical negligence litigation.

Regulatory Affairs: A clinical reviewer under both EU and UK MDR, with subject matter expertise in orphan medical devices and in silico methodologies.

Industry Pioneer: Advisor in a number of major European initiatives (COMBINE, Virtual Human Twin, BRIDGE-IHI) and author of the landmark 2025 white paper, “In Silico methodologies – a Notified Body’s perspective on medical device approval.”

Dr. Surash on LinkedIn →

The Questions We Did Not Reach

Eleven questions were written in the chat during the session and the hour ran out before they were reached. They are answered in full, in writing, on a separate page, with every name, employer and device removed.

Acceptance criteria when the state of the art is thin, equivalence at a different anatomical site, trending thresholds for rare events, vigilance searches in the CER, and clinical data from outside the EU.

Read the Written Answers

Who Should Watch This Session?

This session is for anyone who writes, reviews or owns a clinical evaluation that a notified body will read:

  • Clinical Evaluation Specialists and Medical Writers who want to see the file from the other side of the desk
  • Regulatory Affairs Managers preparing a submission or answering a deficiency round
  • Quality and PMS Managers linking post-market data back into the clinical evaluation
  • MedTech Startups and SMEs building a first clinical evaluation strategy under MDR
  • Consultants who need to explain to a client why a threshold or a claim will not hold
  • Anyone preparing a renewal and wondering what changed since initial certification

Also Watch: More Notified Body Webinars

Three earlier sessions with SZUTEST NB2975: how notified bodies interpret the state of the art during review, what counts as clinical data under the MDR, and how to CE mark AI software under the MDR and the AI Act.

Watch the SOTA Webinar →
Watch the Clinical Data Webinar →
Watch the AI Software Webinar →

Continue the Conversation

Join the Clinical Evaluation Navigator community for weekly expert calls, and work through real clinical evaluation problems with regulatory professionals facing the same ones.

Join the Community

Free to join • Weekly Friday calls • 100+ published articles

This session is educational and general. Dr. Surash speaks in a personal capacity. Nothing said in it is a decision, an approval, a clearance or a position of his employer, and nothing in it binds any notified body. No specific device, company or file under assessment was discussed. Any written answers published alongside this replay are the views of Dr. Hatem Rabeh alone.