Quality of Evidence in Clinical Evaluation: What Regulators Actually Require
Not all clinical data amounts to evidence. Understanding this distinction is what separates an approved dossier from a deficiency request that brings your registration to a halt.

Not all clinical data amounts to evidence. Understanding this distinction is what separates an approved dossier from a deficiency request that brings your registration to a halt.
The problem starts with the definition
One of the most frequent errors I see in technical dossiers submitted to ANVISA, the FDA and European authorities is confusion between clinical data and clinical evidence. A manufacturer may have hundreds of publications and still present insufficient evidence.
Why? Because clinical evidence is measured by quality, relevance and methodological traceability, rather than volume.
What the regulations establish
The three major regulatory references converge on similar requirements:
- EU MDR 2017/745 (Art. 61) - ongoing clinical evaluation, with sufficient data to demonstrate conformity with the general safety and performance requirements. Documentation: CER + PMCF.
- ANVISA RDC 751/2022 - demonstration of safety and performance through clinical data appropriate to the device's risk. Documentation: Clinical Evaluation Report + Technical Dossier.
- FDA 21 CFR 814 / QMSR - valid scientific evidence demonstrating reasonable assurance of safety and effectiveness. Documentation: PMA / 510(k) / De Novo + post-market data.
The five criteria that define quality evidence
Regardless of jurisdiction, regulatory reviewers assess clinical evidence across five dimensions:
1. Suitability for the device
Were the data generated using the exact device or a demonstrably equivalent one? Technological, biological and clinical equivalence must be formally justified.
2. Clinical relevance
Do the endpoints assessed correspond to the intended indications for use? Laboratory performance data alone rarely support clinical claims.
3. Methodological quality
Are the study design, bias control, sample size and follow-up appropriate? Case series without a control group have limited evidentiary weight for devices presenting a high risk.
4. Data traceability and integrity
Are the sources identifiable, verifiable and free from undeclared conflicts of interest? Scrutiny of data generated by manufacturers themselves has increased significantly.
5. Currency
Do the data reflect the state of the art? Literature more than 10 years old, without supplementation by recent post-market data, raises questions about its relevance to emerging technologies.
The mistake I see most often in practice
Manufacturers submit a lengthy CER, 80 or 100 pages, with dozens of bibliographic references, and receive deficiency requests from the regulatory authority for insufficient clinical evidence.
The problem is not the number of pages. It is the absence of a structured clinical argument connecting each data source to the intended indication for use, the risk profile (ISO 14971) and the performance claims on the label.
Clinical evaluation is not a collection of articles. It is a regulatory argument built through a methodical approach.
The evidence lifecycle
Clinical evaluation does not end when registration is approved. The EU MDR, ANVISA and FDA require ongoing updates through:
- Post-market surveillance (PMS/PMCF) - systematic collection of field data.
- Periodic CER review - at a frequency determined by risk and the maturity of the evidence.
- Safety and performance reports (PSUR/SSCP) - mandatory under the EU MDR for Classes IIb and III.
Manufacturers that treat clinical evaluation as a one-off event face growing regulatory vulnerabilities, especially in markets aligned with the IMDRF.
What this means for you
If you are responsible for regulatory affairs or are developing a market access strategy for Brazil, Europe, the US or MDSAP markets, the quality of your clinical evidence is a strategic asset, rather than merely a submission requirement.
It determines the speed of your registration, the strength of your claims and your ability to respond to questions without restarting studies from scratch.
Would you like to discuss how to structure your device's clinical evaluation? Speak to the HSC Global team.
More quality to explore
Each article connects regulatory changes to what they mean for your day-to-day work.
QualityA special process is not audit jargon - it is the factory's most misunderstood risk category
If your process output cannot be verified by inspecting the finished product, you are dealing with a special process - where failure does not show up in the component, but in the patient.
QualityYour sterilisation cycle is validated. But did you validate the right product under the right worst-case conditions?
An approved cycle shows that the equipment delivers the process - not that the most challenging device, in the most demanding load, achieves the required SAL. This confusion is one of the sector's costliest gaps.
QualityPackaging does more than protect the product. It maintains sterility - and that takes different engineering
Physically protecting a product and maintaining a sterile barrier throughout its shelf life are distinct objectives requiring distinct validations. Sterility only reaches the patient if the packaging preserves it.
Comments
No comments yet. Be the first to comment.