What EU MDR clinical evaluation requires

Most clinical evaluation failures do not happen because manufacturers ignored the requirement. They happen because manufacturers misunderstood what it actually asks for. Some treat it as a documentation exercise. Others conflate it with clinical investigation and assume the only way to satisfy the requirement is to run a study. Others underestimate the evidence standard entirely and build a process that looks complete until a notified body starts reviewing it.

None of those approaches hold up at certification.

What clinical evaluation is, and what it is not

Under the European Union Medical Device Regulation (EU MDR), clinical evaluation is a continuous and systematic process manufacturers use to assess, verify, and maintain the safety and performance of their medical devices throughout the entire product lifecycle. It is a confirmation of conformity with the General Safety and Performance Requirements (GSPRs) in Annex I of EU MDR 2017/745. Under normal conditions of intended use, manufacturers must evaluate device safety and performance, assess undesirable side effects, and demonstrate that the benefit-risk ratio is acceptable. That conclusion must be based on objective clinical data, and the data must be sufficient.

The regulation does not suggest this process. It requires that you plan it, conduct it, and document it. The documentation obligation matters more than most manufacturers expect. Well-executed evaluations regularly fail at the certification stage because the process was not recorded in a way that a reviewer can follow and verify.

Clinical evaluation is also not clinical investigation. A clinical investigation under ISO 14155 is one method of generating clinical data. If outstanding safety and performance questions can be answered through systematic literature review, post-market data from an equivalent device, bench testing, or other objective sources, a clinical investigation may not be necessary. The question is whether evidence is sufficient, not whether a formal study was conducted.

The most important reason to take clinical evaluation seriously is not the notified body requirement. Clinical evaluation is the only mechanism for confirming that a device works as intended under real-world conditions. Risk management under ISO 14971 identifies hazards and assigns severity and probability. Clinical evaluation validates those assignments. Without clinical data feeding back into the risk assessment, manufacturers are working from assumptions that have never been tested in practice.

Why it connects to the quality system

Clinical evaluation is a continuous process, not a one-time deliverable. The clinical evaluation report (CER) must be updated on a defined schedule, and every update must answer the same question: is the benefit-risk ratio still acceptable? That update cycle must be built into how the quality management system (QMS) operates. A CER that lives in a folder and gets touched every three years is not functioning as a quality system element.

Understanding how post-market surveillance connects to this feedback loop is essential to running clinical evaluation as an ongoing program rather than a submission task. The post-market clinical follow-up (PMCF) outputs feed back into the CER. CER updates feed back into the risk management file, labeling review, and the periodic safety update report (PSUR).

BONUS RESOURCE: Click here to download your free PMCF Activity Selection Guide.

The five-stage process under Annex XIV Part A

Article 61 of EU MDR defines the clinical evaluation requirements, and Annex XIV describes the expected documentation and methodology. Part A sets out five stages: planning, identification of data, appraisal of data, analysis, and post-market activities that close the feedback loop. The two required outputs are the clinical evaluation plan (CEP) and the clinical evaluation report (CER).

Planning

Planning is where most of the value is created, and where most manufacturers invest too little time. The CEP is the upfront planning document. It defines the scope of the evaluation, the intended purpose of the device, the target patient population, the indications and contraindications, the clinical benefits being claimed, the residual risks requiring address, and the criteria for determining whether evidence is sufficient. It also maps the progression from exploratory to confirmatory data, from early feasibility through post-market clinical follow-up.

A complete CEP also identifies any equivalent devices being referenced, justifies the use of equivalence where applicable, and describes how PMCF will be used to address any clinical data gaps. The connection to design controls is why the plan needs to be in place before verification and validation begin. If the clinical evaluation plan is written after design validation is complete, there will almost certainly be testing that does not align with what the evaluation requires. Starting the plan early is one of the most cost-effective decisions a manufacturer can make.

Identification and appraisal of clinical data

EU MDR defines clinical data as information concerning safety or performance generated from the use of the device. The four main sources are: clinical investigations of the device under evaluation; clinical investigations or other studies reported in scientific literature of a device for which equivalence can be demonstrated; reports published in peer-reviewed scientific literature on other clinical experience of the device or an equivalent; and clinically relevant information from post-market surveillance, in particular PMCF.

Once there is a plan, the data needs to be found. For most devices, that means a systematic literature search. A defined protocol, a clear set of search terms, a specified list of databases, a defined time period, and explicit inclusion and exclusion criteria are all required. After identification comes appraisal: evaluating whether the data is relevant to the specific device, whether bias is present, whether the subject device is sufficiently similar, and how much weight each source should carry. All sources carry different weight, and the CER needs to appraise each one on its own merits. After appraisal comes synthesis: looking for consistency across sources, confirmation of the safety profile, validation of risk assignments, and evidence that the intended purpose is met.

Analysis

The analysis stage produces the CER, the document that draws a conclusion about conformity with GSPRs based on the reviewed data. The CER pulls together all available clinical data, appraises each source for quality and relevance, analyzes the collective evidence, and reaches a clinical evaluation conclusion. It must clearly support the benefit-risk determination for the device. Notified bodies will scrutinize this document. It needs to be written as if the reader has no prior knowledge of the device.

Both the CEP and CER should be reviewed and updated on a defined cadence, typically annually for higher-risk devices (Class III and implantables) and at least every two to five years for lower-risk devices, or sooner if new clinical data emerges that could change the evaluation conclusion.

PMCF and the continuous evaluation loop

Post-market clinical follow-up is not optional for most devices under EU MDR. It is a structured, protocol-driven activity with specific questions derived from the clinical evaluation, not passive data collection. It is the mechanism by which manufacturers continuously collect clinical data once the device is on the market, specifically to confirm safety and performance, identify previously unknown risks, and address any residual uncertainties from the original clinical evaluation.

The PMCF plan is a required component of the technical documentation, and findings from PMCF feed directly back into the CER on every update cycle. PMCF is mandatory unless a manufacturer can justify otherwise, with well-established technology, low device class, and a complete existing evidence base as the most commonly accepted justifications. If a manufacturer claims PMCF is not applicable, that justification needs to be thorough and defensible. Notified bodies scrutinize these justifications heavily, and unsupported claims of non-applicability are a common source of non-conformities during audits. For more detail on how to build and execute a PMCF plan, see our guide to developing a PMCF plan.

Equivalence under EU MDR: the most common point of failure

Equivalence to an existing device is still permitted under EU MDR, but the requirements are considerably stricter than they were under the Medical Device Directive (MDD). To claim equivalence, a manufacturer must demonstrate equivalence across three dimensions:

1. Technical equivalence: the device and the comparator share the same design principles, materials, specification ranges, and similar manufacturing processes, uses similar conditions of use, and uses similar deployment methods and principles of operation. Reviewers need to see CAD data, material specifications, and testing results compared point by point.
2. Biological equivalence: materials in contact with the body are the same, with the same duration and intensity of contact. Similar is not sufficient. The requirement is the same material in contact with the same body tissues or fluids.
3. Clinical equivalence: the same intended purpose, the same clinical conditions, the same indication, the same patient population (accounting for age, gender, anatomy, and physiology), and similar relevant critical performance.

Three equivalence failures come up most often in practice:

1. Lack of data access: equivalence is claimed to a competitor device, but the manufacturer cannot demonstrate actual access to that device's technical and biological data. For Class III and implantable devices, a contract granting explicit data access from the other manufacturer is required. Without that contract, equivalence is not available for these device classes, and a clinical investigation is effectively required.
2. Superficial comparison: presenting visual similarity without specification-level analysis. Reviewers need actual data, not general descriptions.
3. Indication mismatch: the intended use is close but not identical. Applying an adult device to a pediatric population is the example that comes up most often in practice.

For a deeper treatment of the equivalence process, see our guide to the clinical evaluation equivalency process.

Well-established technology and state of the art

If a device uses well-established technology, meaning a design that has been on the market for decades with a stable configuration and a documented safety record, there is a mechanism to reduce the clinical investigation burden. Demonstrating that a specific design fits this characterization requires documented scientific rigor, not just a general assertion that the technology is mature.

State of the art is a related but distinct concept. It encompasses device design, manufacturing processes, and clinical treatment approaches. Two questions need answers: is the device's performance comparable to or better than the state of the art, and are all associated risks acceptable given what alternative treatments exist? Notified bodies will assess whether the evidence base reflects current clinical practice.

Common pitfalls in EU MDR clinical evaluation

Five issues come up repeatedly in notified body feedback on CEPs and CERs:

1. Treating clinical evaluation as a one-time exercise. Manufacturers who build the CER as a submission document, rather than as a living record of clinical evidence, struggle at the first update cycle. The CER needs to be designed for continuous revision from the outset.
2. Weak appraisal methodology. A literature search that is not reproducible, or an inclusion criterion that is not clearly defined, undermines the credibility of the whole evaluation. Notified bodies are looking for a methodology that another reviewer could follow and arrive at the same conclusion.
3. Unsupported equivalence claims. Teams relying on publicly available information about a competitor device, without meeting the stricter EU MDR equivalence criteria or securing the required access contract for Class III and implantables, will have their equivalence rejected.
4. Disconnection from risk management and post-market surveillance (PMS). The CER does not exist in isolation. Risks identified in the risk management file, signals from PMS, complaint trends, and incident reports all need to feed into the clinical evaluation, and any changes in the risk-benefit profile need to be reflected in the CER.
5. Insufficient PMCF planning. A PMCF plan that lacks defined objectives, specified data collection methods, or realistic timelines does not satisfy Article 61 or Annex XIV. For guidance on common PMCF planning mistakes, MDCG 2020-5 provides useful context on equivalence and PMCF expectations.

Where clinical evaluation fits in your broader quality system

EU MDR expects clinical evaluation to be embedded in the manufacturer's quality management system, not handled as a parallel track. The CEP and CER should reference and be referenced by the risk management file, the PMS plan, the PMCF plan, and the technical documentation summary. When this integration is weak, notified bodies see it immediately in the form of inconsistencies between documents, and these inconsistencies are a frequent source of findings.

Building both the quality system processes and the clinical evidence infrastructure early costs significantly less than rebuilding either under audit pressure. Purpose-built medical device quality management system (eQMS) software helps by maintaining traceability between clinical evaluation, risk management, PMS, and design controls in a single environment, so that when new evidence surfaces in PMCF or PMS, it flows into the relevant files automatically rather than depending on manual reconciliation.

For a more detailed look at how clinical evaluation connects to PMCF specifically, see our practical guide to PMCF under EU MDR.

BONUS RESOURCE: Click here to download your free PMCF Activity Selection Guide.

Keep reading

If you are building out your clinical evaluation process, these related guides go deeper on the specific components:

• Clinical evaluation for medical devices: the complete guide
• How to create a clinical evaluation report under MEDDEV and MDR
• Clinical evaluation vs. clinical investigation: understanding the difference
• Post-market clinical follow-up: what EU MDR requires
• How to develop a PMCF plan that holds up in audit
• The clinical evaluation equivalency process explained
• MDCG 2020-5 on equivalence and clinical evaluation
• ISO 14155: planning and running a clinical investigation
• Post-market surveillance for medical devices

Watch the on-demand webinar with Dr. Yuan Li for a detailed walkthrough of common CER rejection patterns and equivalence examples: Clinical evaluation under EU MDR with Dr. Yuan Li.

If your team is preparing a CEP or CER and wants to see how an integrated eQMS can connect clinical evaluation to the rest of your quality system, request a demo of Greenlight Guru.