Back to the Basics - What is a clinical evaluation according to MDR regulations? Fundamentals & Requirements

This blog post provides a concise and practical overview of what a clinical evaluation according to the MDR (Medical Device Regulation) entails, the relevant legal and regulatory frameworks, and how the entire process works step by step. Using the three possible evidence routes as examples, we demonstrate which data is crucial, how to continuously update your clinical evaluation, and which typical mistakes to avoid.

Abbreviations

CEP	Clinical evaluation plan
CERIUM	Clinical Evaluation Report
CIP	Clinical Investigation Plan
MDR	Medical Device Regulation (EU Ordinance 2017/745)
TD	Technical Documentation
Pmcf	Post-Market Clinical Follow-up
Pms	Post-Market Surveillance

Underlying regulations, standards and guidelines

EU Regulation 2017/745 (MDR)

MDCG 2020-1

MDCG 2020-5

MDCG 2020-6

MEDDEV Guideline 2.7/1 Revision 4

Draft of ISO/DIS 18969

1. Introduction

The clinical evaluation is the core of every MDR-compliant Technical Documentation (TD) – it explains why a medical device is safe and effective. Precisely because the MDR places high demands on evidence, timeliness, and traceability, the clinical evaluation must never be considered a one-off final document. Rather, it is a dynamic document that requires continuous updating, linking preclinical and clinical study data with real-world evidence from PMS/PMCF, thus supporting the entire product lifecycle.

For manufacturers, this means that clinical evaluation must be strategically planned from the outset—from the Clinical Evaluation Plan (CEP) and study design to the integration of PMCF and PSUR results. Only when endpoints, data types, and analysis methods are clearly defined and the evidence pathways are transparently documented can claims be formulated effectively, benefit-risk analyses defended, and audits successfully passed.

2. Legal Foundations and Standards

The MDR not only provides the legal framework — it also defines what is meant by "clinical evaluation".

MDR, Article 2:

(44) ‘Clinical evaluation’ means a systematic and planned process of the continuous generation, collection, analysis and evaluation of clinical data relating to a product, which is used to verify the safety and performance, including clinical benefit, of the product when used as intended by the manufacturer.

Article 61 requires manufacturers to plan, conduct, and document clinical evaluations to demonstrate compliance with the essential safety and performance requirements. Article 61, paragraph 11, and Annex XIV, Part B, specify the obligation for life-cycle assessment: PMCF and PMS data must continuously update the CER. PMCF is understood as ongoing clinical follow-up, the results of which must be summarized in an evaluation report and integrated into the CER.

The Medical Device Coordination Group (MDCG) provides the interpretations that are crucial for practical application. MDCG 2020-1 (Clinical evaluation / performance evaluation of Medical Device Software) offers a clear, methodological framework specifically for software products. The document emphasizes that clinical evaluation of software must address three closely related components: valid clinical association (scientific justification that the software output is clinically relevant), technical/analytical performance (verification/validation evidence), and clinical performance (evidence that the software delivers the expected, patient-relevant outcomes in the target population).

MDCG 2020-5 (Guidance on Clinical Evaluation — Equivalence) focuses on the requirements for using equivalence data. The guideline defines which technical, biological, and clinical criteria must be met for a third-party product to be considered "equivalent," and it requires rigorous documentation and justification of the equivalence claim. The key message is that equivalence should not be used as a shortcut without evidence—the manufacturer must demonstrate that differences (e.g., in materials, design, indications, or instructions for use) do not compromise the transferability of the data; in cases of uncertainty, PMCF measures or the manufacturer's own clinical data are required.

MDCG 2020-6 (Clinical evidence needed for legacy devices ) addresses medical devices that were previously CE marked according to the directives (MDD/AIMDD). The guidance requires a systematic gap analysis: manufacturers must assess whether the available historical pre- and post-market data meet the MDR requirements for "sufficient clinical evidence." MDCG 2020-6 describes what a Clinical Evaluation Plan (CEP) adapted to the MDR level should look like for legacy devices, which evidence hierarchy should be established (from high-quality clinical studies to vigilance/PMS data), and when additional PMCF studies or new clinical data are needed. The document explicitly links MEDDEV principles with MDR-specific requirements and provides concrete guidance on the practical aspects of closing the gap.

The MEDDEV guideline 2.7/1 Revision 4 is still operationally helpful. Although it originates from the MDD era and does not replace the MDR, MEDDEV 2.7/1 contains very concrete tools: search strategies, selection and exclusion criteria for literature searches, evaluation grids, and reporting templates. Many continue to use MEDDEV as a practical "how-to" guide to translate the abstract MDR requirements into concrete work steps—always with the caveat that MEDDEV guidance should be adapted to the current MDR/MDCG requirements.

At the international standards level, the ISO/DIS 18969 standard – “Clinical evaluation of medical devices” – currently under development should be mentioned. The draft, currently in inquiry status, aims for more uniform terminology, structure, and evaluation logic and will help to harmonize and standardize clinical evaluation processes internationally. For manufacturers, this means that processes currently set up to comply with the MDR and MDCG should be designed in such a way that they will also comply with a potential ISO standard in the future.

3. Clinical Evaluation Process — Step by Step

Clinical evaluation is not a one-off event, but a clearly structured cycle that begins during product development and is repeated throughout the entire product lifecycle. A practical process, proven in many guidelines (MEDDEV/MDCG) and in practice, can be described in five stages: Scope/Plan (Stage 0), Identification (Stage 1), Appraisal (Stage 2), Analysis (Stage 3), and Report (Stage 4). Importantly, the earlier you begin planning, the clearer the data flow—and the easier it is to decide whether a clinical trial is necessary.

Planning: CEP and Data Route (Stage 0 — Scope)
The clinical evaluation begins with the Clinical Evaluation Plan (CEP). The CEP defines the framework of the study: the intended purpose of the product, the target populations, primary and secondary endpoints, clinically relevant comparators, the planned data sources, the fundamental methodological approaches, and the responsibilities. A crucial question here is which evidence routes are possible and permissible—are literature, preclinical data, and real-world data sufficient, or is a clinical trial necessary? This decision depends on the relevance and quality of existing data, as well as the risk and innovation level of the product. The CEP should therefore be finalized at an early stage of development; it guides search strategies, endpoint definitions, and, if necessary, the study design.

Clinical Data Identification (Stage 1):
Based on the CEP, the systematic identification of all relevant data sources begins. These include preclinical and technical study reports, in-house clinical trials, scientific literature, data from safety databases (vigilance), PMS data from the field, and existing PMCF results.

Appraisal — Critical Evaluation of the Evidence (Stage 2)
Not all data are created equal. In the appraisal phase, each data source is examined for relevance, validity, completeness, and bias. This includes evaluating study designs (RCT vs. observational study), follow-up times, endpoint definitions, population equality (including equivalence questions), statistical robustness, and data quality (e.g., missing data, selection bias). For literature searches, transparent search strategies, inclusion/exclusion criteria, and evaluation rubrics must be documented.

Analysis – Synthesis of Evidence and Benefit-Risk (Stage 3)
In the analysis phase, the reviewed datasets are compiled and systematically evaluated. This involves weighting the evidence, determining relevant key performance indicators, and assessing whether the totality of the data supports the safety and performance claims. The benefit-risk analysis is the core of this stage: Are the benefits and risks for the intended application still within an acceptable balance? If gaps or uncertainties become apparent, a decision is made as to whether and which PMCF measures are necessary, or whether a clinical trial (e.g., a PMCF study) needs to be initiated.

Report — Preparation of the CER (Stage 4)
The final step in this cycle is the Clinical Evaluation Report (CER). The CER documents in a targeted and transparent manner how the data were identified, evaluated, and integrated, and what conclusions were drawn. A good CER includes a concise executive summary, the purpose and scope, a description of the methods (including CEP reference), detailed results with measurable parameters, the benefit-risk analysis, a state-of-the-art (SoTA) assessment, and a clear presentation of any remaining data gaps.

Continuous Loop and Versioning:
Clinical evaluation is not a linear, one-off process. CER, CEP, and PMCF plans are versioned and revised when new data (PMS, PMCF, studies, literature) become available. Conclusions drawn from PMS or PMCF results are fed back into the Identification/Appraisal/Analysis stages and may trigger new studies, IFU changes, or CAPA measures.

In short: an early completion of the CEP defines the data route and avoids costly surprises. Systematic identification, rigorous critical appraisal, sound synthesis of the evidence, and a clearly structured CER are the building blocks that allow you to clearly justify, if necessary, whether or not a clinical trial is required.

4. Routes of clinical evaluation

The MDR allows three fundamentally different ways in which the necessary clinical evidence for the clinical evaluation can be provided. The route you choose significantly determines the structure, scope, and argumentation of your CER—and should therefore be precisely justified in the CEP.

Route 1 — Own Clinical Data
The strongest and easiest-to-defend evidence base is your own clinical data. This can be divided into two subgroups: a) prospective clinical trials initiated and conducted by the manufacturer, and b) clinical data from the literature, i.e., published studies in which the product itself was investigated and evaluated. Prospective studies offer the clearest correlation between effect and product—they allow for precise endpoint definitions, controlled follow-up, and statistical design. Literature data on your own product is useful if high-quality, relevant studies exist; however, their reliability depends on their design, population, and the alignment of the endpoint definitions with your claims. In any case, the study methodology, monitoring, dropout rates, and statistical analyses must be transparently documented and included in the CER.

Route 2 — Equivalence Route:
The equivalence route allows the use of clinical data from another product already on the market—but only under strict conditions. It must be demonstrated that the equivalent product is so closely related to the company's own product in technical, biological, and clinical terms that the external data allows valid conclusions to be drawn for the company's own medical device. Technical equivalence encompasses design principles, materials, and functional parameters; biological equivalence concerns materials and their interaction with the body; clinical equivalence means that indications, user profiles, conditions of use, and expected clinical effects are comparable. The requirements for justification are high—MDCG 2020-5 makes it clear that equivalence must not be used as a standard route but requires a robust, risk-based argument. If a convincing justification for equivalence is lacking, reviewers usually request supplementary data or a separate clinical investigation.

Route 3 — Performance data according to Art. 61(10) — no need for clinical investigation.
For certain products where a clinical investigation would be either technically unreasonable, ethically unacceptable, or disproportionate, the MDR allows the clinical evaluation to be based on analytical/technical performance data and other non-clinical evidence. Typical elements of this route include laboratory/bench tests, in vitro tests, simulation studies, validation of measurement accuracy, robustness and lifetime tests, usability tests, and, where appropriate, comparative tests against a recognized reference method.

When to choose which route? Practical criteria:
The choice depends on the risk profile, the level of innovation, the availability of relevant data, and ethical/practical considerations. For new technologies, ambitious claims, or high risks, a dedicated clinical trial is usually unavoidable. If the product is derived from existing solutions and equivalence can be rigorously demonstrated, the equivalence route can save time and effort—but it requires robust technical and clinical comparative documentation. For simple measuring devices or products where clinical endpoints cannot be meaningfully measured, the performance data route may be adequate—provided that the transferability to clinical practice is justifiable and PMCF measures address any gaps.

Conclusion:
Whichever route is chosen, the justification must be included in the CEP and fully traceable in the CER. The three routes are practical ways to fulfill the CER obligation, guided by risk and evidence requirements. The decisive factor is not the "most convenient" route, but the one that provides the necessary evidence for the product and claim—and which you can defend appropriately and comprehensibly before notified bodies and authorities.

5. Update of the clinical evaluation

The clinical evaluation is not a static document—the MDR explicitly mandates this. Article 61, paragraph 11 stipulates that the clinical evaluation and related documentation must be updated "throughout the product's life cycle based on clinical data" from the PMCF and PMS. For Class III and implantable devices, the MDR goes a step further: the post-market clinical follow-up evaluation report (and, where applicable, the summary report pursuant to Article 32) must be updated at least annually based on this data. In practical terms, this means that for particularly high-risk products, the clinical evaluation itself should also be updated at least annually.

Updating means more than just cosmetic changes to the text. It's a defined process that begins with the systematic collection and analysis of new clinical data—be it PMCF results, findings from PSUR/PMS reports, new literature, registry data, results from user studies, or findings from CAPA and field safety measures. Once relevant data are available, it's essential to assess whether the assumptions regarding safety, performance, or clinical benefit have changed. Crucially, the question here is whether the new findings affect the benefit-risk assessment: Has the incidence of a relevant event increased? Do subgroup analyses reveal new risks? Do field data confirm—or contradict—the claims made?

For Class III / implantable products, this procedure must be carried out at least annually; for other classes, a risk-based, documented rhythm must be defined (in the CEP), supplemented by clear trigger rules for ad-hoc revisions.

Trigger rules are crucial: define which signals automatically trigger a CER revision. Such triggers could include, for example, a statistically significant increase in the complication rate, a cluster in a specific batch, a new serious incident, a relevant study in the literature, or the finding of ineffectiveness in a defined user group.

In short: The clinical evaluation update is a cycle-based, evidence-driven process that must be performed at least annually for Class III/implantable products and must be risk-based for all products. Early definition of trigger rules, rigorous data analysis, and complete traceability are prerequisites for ensuring that CER updates are not merely formal but actually and sustainably safeguard the clinical safety and performance of your product. 

6. Conclusion

Clinical evaluation is the central instrument you use to demonstrate the clinical safety, clinical performance, and clinical benefit of your product under the MDR. It doesn't begin with the completed CER, but with a well-thought-out strategy: an early development of a CEP that defines the data route, endpoints, and responsibilities. Those who strategically plan the CER from the outset avoid costly surprises during later investigations—especially the question of whether a clinical investigation is necessary, as this allows for a well-founded answer.

Crucially, the CER must be viewed as a dynamic process. The MDR explicitly requires that the clinical evaluation be updated throughout the entire lifecycle using PMCF and PMS data; for Class III and implantable products, this review must be conducted at least annually. Updates are not merely editorial changes, but evidence-based reassessments of the benefit-risk balance with clearly documented decisions and traceable actions.

Three routes of evidence are available—your own clinical data, the equivalence route, and the performance data route—and each has its merits. The crucial factor is not the most convenient choice, but the one that provides the necessary evidence for your product and your claims.

7. How we can help you

We provide pragmatic support throughout the entire clinical evaluation lifecycle—from strategic planning to audit-ready documentation. Our goal is to design your CER processes to be MDR-compliant, methodologically sound, and practically applicable. We combine regulatory expertise with practical project and study know-how to transform requirements into genuine, defensible evidence.

Specifically, we can support you with, for example:

• Gap analyses of your existing CEP/CER/PMCF documentation, including prioritized action planning.
• Creation and review of CEPs and CERs.
• Methodology decision & study design: We define endpoints, populations, statistical plans and monitoring concepts for clinical trials or PMCF studies.
• Literature review & evidence appraisal: systematic search, quality assessment and integration into the benefit-risk analysis.
• PMCF conception and implementation.

Want to know more? Contact us for a free initial consultation!

You can get a free initial consultation here: free initial consultation