How to Write a Clinical Evaluation Report (CER) Under EU MDR: A Practical Guide

The Clinical Evaluation Report (CER) is one of the most critical documents in a medical device manufacturer's regulatory submission under the EU Medical Device Regulation (MDR 2017/745). It is the document that demonstrates, through a systematic and planned process, that a medical device achieves its intended clinical benefits, that undesirable side effects are acceptable when weighed against those benefits, and that the device's clinical performance is consistent with the state of the art. The legal basis for clinical evaluation is established primarily in MDR Article 61, which mandates that manufacturers shall confirm the conformity of their devices with the relevant General Safety and Performance Requirements (GSPR) set out in Annex I, based on clinical data providing sufficient clinical evidence. This requirement is further elaborated in Annex XIV Part A, which prescribes the clinical evaluation process itself: a structured methodology for the identification, appraisal, and analysis of clinical data. Unlike the former Medical Device Directive (MDD 93/42/EEC), where clinical evaluation was often treated as a formality for lower-risk devices, the MDR elevates clinical evaluation to a central pillar of the conformity assessment process. Every device, regardless of classification, must have a clinical evaluation, and the depth and rigour of that evaluation must be proportionate to the nature, classification, and intended purpose of the device. The CER is not a static document filed and forgotten; it is a living document that must be updated throughout the device lifecycle as new clinical data emerge from post-market surveillance, post-market clinical follow-up, and the evolving body of published literature.

A common source of confusion among manufacturers, particularly those new to the EU regulatory framework, is the distinction between a clinical evaluation and a clinical investigation. These terms are not interchangeable, and understanding their relationship is fundamental to developing a sound clinical evidence strategy. Clinical evaluation, as defined in the MDR, is the systematic and planned process to continuously generate, collect, analyse, and assess the clinical data pertaining to a device in order to verify its safety, clinical performance, and clinical benefit. It is a desk-based analytical exercise that draws on multiple data sources, including published literature, clinical investigation reports, and post-market surveillance data. A clinical investigation, by contrast, is a systematic study involving human subjects undertaken to assess the safety or performance of a device and is just one potential source of clinical data that feeds into the clinical evaluation. Not every device requires a clinical investigation; for many devices, particularly those with well-established technologies and a substantial body of published literature, the clinical evaluation can be based entirely on literature data and equivalence arguments. However, the MDR significantly raises the bar for when a clinical investigation is required. Article 61(4) explicitly mandates clinical investigations for Class III devices and Class IIb active devices intended to administer or remove a medicinal product, unless the manufacturer can justify reliance on existing clinical data. For implantable devices, the presumption in favour of clinical investigations is particularly strong. Manufacturers must therefore make a documented, risk-based determination early in their regulatory strategy about whether a clinical investigation will be needed, as this decision has profound implications for timelines, budgets, and resource allocation.

MEDDEV 2.7/1 Rev 4, published in September 2016, remains the most detailed and widely referenced guidance document for conducting clinical evaluations and writing Clinical Evaluation Reports for medical devices in the European Union. Although MEDDEV guidance documents were developed under the MDD framework, MEDDEV 2.7/1 Rev 4 has been explicitly endorsed by the Medical Device Coordination Group (MDCG) as remaining applicable and relevant under the MDR, pending the development of MDR-specific clinical evaluation guidance. The document provides a structured methodology for the clinical evaluation process, defines the qualifications required for clinical evaluators, and includes detailed expectations for each stage of the evaluation. MEDDEV 2.7/1 Rev 4 introduced several concepts that were subsequently reinforced and tightened under the MDR, including the requirement for evaluators to possess documented qualifications and relevant clinical experience, the systematic and reproducible nature of literature searches, and the explicit appraisal of individual clinical data sets for quality and relevance. It is worth noting that while MEDDEV 2.7/1 Rev 4 provides the procedural framework, manufacturers must also consider MDCG guidance documents that address specific aspects of clinical evaluation under the MDR. MDCG 2020-6 on sufficient clinical evidence provides additional clarity on what constitutes adequate clinical evidence for different device categories, while MDCG 2020-5 addresses clinical evaluation and equivalence. Manufacturers should treat MEDDEV 2.7/1 Rev 4 as the foundational methodology while overlaying the more recent MDCG guidance to ensure full MDR compliance. Notified Bodies routinely assess CERs against the MEDDEV 2.7/1 Rev 4 structure and expect manufacturers to address each stage explicitly.

The clinical evaluation process, as outlined in MEDDEV 2.7/1 Rev 4 and reinforced by MDR Annex XIV Part A, is structured around five distinct stages. Mastery of each stage is essential for producing a CER that will withstand Notified Body scrutiny. Stage 0 is the definition of scope, where the manufacturer establishes the device description, its intended purpose, the target patient population, the clinical claims being made, the relevant GSPR, and the clinical context including alternative treatments and the current state of the art. This scoping stage is critically important because it determines the boundaries of the subsequent literature search and data analysis. A poorly scoped evaluation will either miss relevant data or waste resources analysing irrelevant evidence. Stage 1 is the identification of pertinent data, where the manufacturer conducts a systematic literature search, identifies clinical investigation data (both favourable and unfavourable), and gathers post-market surveillance and PMCF data. Stage 2 is the appraisal of individual data sets, where each identified data source is assessed for methodological quality, scientific rigour, relevance to the device under evaluation, and contribution to the overall body of evidence. Stage 3 is the analysis of the clinical data, where the appraised data is synthesised to draw conclusions about the device's safety, clinical performance, and clinical benefit, including the benefit-risk determination and comparison to the state of the art. Finally, Stage 4 is the Clinical Evaluation Report itself, the written document that presents the rationale, methodology, data, analysis, and conclusions in a structured and transparent format. Each stage must be documented with sufficient detail that an independent reviewer could understand and verify the process. The CER must also include a Clinical Evaluation Plan (CEP), which is either a standalone document or an integral section of the CER that defines the evaluation methodology before the evaluation is conducted.

One of the most significant changes the MDR introduced relative to the MDD concerns equivalence claims. Under the MDD, manufacturers routinely relied on equivalence to predicate devices to leverage published clinical data from similar devices as the basis for their clinical evaluation. The MDR substantially restricts this practice. Article 61(5) requires that where manufacturers rely on clinical data from an equivalent device, they must demonstrate equivalence across three dimensions: clinical equivalence (same intended purpose, same clinical condition, same patient population, same site of application, and same clinical effect), technical equivalence (same design, same materials, same specifications, same deployment methods, and similar manufacturing processes), and biological equivalence (same materials in contact with the same human tissues or body fluids for a similar duration and exposure). Crucially, the MDR states that these characteristics must be similar to the extent that there would be no clinically significant difference in the safety and clinical performance of the device. This is a higher threshold than the MDD's more general notion of equivalence. Furthermore, the MDR requires that manufacturers claiming equivalence to another manufacturer's device must have a contract in place with that manufacturer giving them sufficient access to the technical documentation of the equivalent device. This provision effectively eliminates the common MDD practice of claiming equivalence to a competitor's device based solely on publicly available information. For many manufacturers, particularly those who historically relied on equivalence to well-known predicate devices, this means that a clinical investigation or a substantially enhanced literature-based clinical evaluation may now be required. Notified Bodies are scrutinising equivalence claims with particular rigour, and insufficiently justified equivalence is one of the most common reasons for CER deficiency letters.

The literature search is the backbone of most Clinical Evaluation Reports, and the MDR and MEDDEV 2.7/1 Rev 4 impose exacting requirements on how it must be conducted, documented, and maintained. The search must be systematic, meaning it follows a pre-defined and documented search protocol. It must be reproducible, meaning another evaluator could execute the same search and obtain substantially the same results. And it must be comprehensive, meaning it is designed to identify all relevant published clinical data, including both favourable and unfavourable evidence. The literature search protocol should define the databases to be searched (PubMed/MEDLINE is the minimum; most Notified Bodies also expect Embase, the Cochrane Library, and potentially additional specialised databases), the search terms and Boolean logic, inclusion and exclusion criteria, the date range, and language restrictions with justification. The search should cover not only the subject device and any claimed equivalent devices, but also the broader therapeutic context, including alternative treatments, benchmark data, and state-of-the-art reviews. Each step of the search must be documented: the number of records retrieved from each database, the screening process (title and abstract screening, full-text review), the reasons for excluding records, and the final set of included publications. A PRISMA-style flow diagram is considered best practice and is expected by many Notified Bodies. Common deficiencies include searches that are too narrow (missing relevant data), searches that are biased towards favourable evidence, failure to include grey literature or conference proceedings where relevant, inadequate documentation of the screening process, and failure to update the search at appropriate intervals. The literature search is not a one-time exercise; it must be repeated or updated as part of the ongoing clinical evaluation, with the frequency determined by the Clinical Evaluation Plan and triggered by factors such as new safety signals, changes to the state of the art, or scheduled CER updates.

A robust Clinical Evaluation Report draws on multiple categories of clinical data, and the MDR explicitly recognises several sources that must be considered. Published scientific literature is typically the primary source and includes peer-reviewed journal articles, systematic reviews, meta-analyses, case reports, and conference proceedings. The evaluator must critically appraise each publication for methodological quality, relevance to the device under evaluation, and the weight of evidence it contributes. Clinical investigation data includes both pre-market clinical investigations conducted by the manufacturer (or sponsor) under the device being evaluated, and investigations conducted on equivalent or similar devices. Pre-market clinical investigation reports must comply with the requirements of MDR Article 62 and Annex XV, and the data must be reported transparently, including adverse events and device deficiencies. Post-market surveillance (PMS) data is a critically important and often underutilised data source. This includes complaint data, trend analyses, field safety corrective actions, vigilance reports, and the outcomes of proactive post-market surveillance activities. The MDR places significantly greater emphasis on PMS data feeding back into the clinical evaluation than the MDD did, and Notified Bodies expect to see this feedback loop clearly documented. Post-market clinical follow-up (PMCF) data is generated through planned studies or registries designed to proactively collect clinical data on the device once it is on the market. PMCF is particularly important for devices with limited pre-market clinical data, novel devices, and high-risk devices. Registry data from national or international device registries (such as joint replacement registries, cardiac device registries, or hernia mesh registries) can provide valuable real-world evidence on long-term performance and safety. Manufacturers should actively identify and incorporate relevant registry data into their CER. Each data source must be appraised individually for quality and relevance, and the overall body of evidence must be assessed holistically to determine whether sufficient clinical evidence exists to confirm the device's safety and performance.

The benefit-risk analysis is the centrepiece of the CER and the point where the clinical evaluator must synthesise all of the appraised data into a definitive conclusion about whether the device's clinical benefits outweigh its residual risks. MDR Article 2(2) defines clinical benefit as the positive impact of a device on the health of an individual, expressed in terms of meaningful, measurable, patient-relevant clinical outcomes, including outcomes related to diagnosis, or a positive impact on patient management or public health. The benefit-risk analysis must be conducted in a structured, transparent, and well-documented manner. It should begin by identifying and quantifying the clinical benefits of the device, supported by the clinical data. Benefits should be expressed in clinically meaningful terms, not merely in terms of technical performance. Next, the analysis should identify and characterise the risks, drawing on the risk management file (ISO 14971), post-market surveillance data, and published literature. Risks include adverse events, complications, device deficiencies, and use errors. For each risk, the analysis should consider the severity, probability, and detectability. The benefit-risk determination then weighs the identified benefits against the residual risks after risk mitigation measures have been applied. This must be done in the context of the current state of the art and available treatment alternatives. A device whose benefit-risk profile is comparable to or better than existing alternatives is generally considered acceptable, while a device that offers no clinical benefit over alternatives but introduces additional risks will face significant regulatory challenges. The MDR also requires that the benefit-risk analysis consider the specific patient population, including any vulnerable subpopulations. For devices where the clinical benefits are indirect (for example, diagnostic devices where the benefit depends on the subsequent treatment decision), the evaluator must carefully articulate the chain of clinical reasoning. Notified Bodies frequently cite inadequate benefit-risk analysis as a CER deficiency, particularly where the analysis is superficial, where benefits are overstated without supporting data, or where the comparison to the state of the art is absent or perfunctory.

Determining and documenting the current state of the art is a requirement that runs throughout the MDR and is particularly important in the context of clinical evaluation. The state of the art represents the current level of technology, clinical knowledge, and accepted medical practice that defines the benchmark against which a device's safety and performance must be assessed. MDR Annex I, General Safety and Performance Requirement 1, states that devices shall achieve the performance intended by their manufacturer and shall be designed and manufactured in such a way that, during normal conditions of use, they are suitable for their intended purpose, taking account of the generally acknowledged state of the art. In the CER, the state-of-the-art assessment serves multiple purposes. It defines the clinical context in which the device operates, including the disease or condition being addressed, its prevalence and severity, the natural history, and the current diagnostic or therapeutic pathways. It identifies the available treatment alternatives, whether other medical devices, pharmaceutical interventions, surgical procedures, or conservative management. It establishes the benchmark for acceptable performance and safety by reference to published literature, clinical guidelines, and consensus standards. And it informs the benefit-risk analysis by providing the comparator against which the device's clinical profile is judged. A thorough state-of-the-art review should draw on clinical practice guidelines from recognised medical societies, systematic reviews and meta-analyses, health technology assessments, and authoritative textbooks or reference works. The assessment must be current and must be updated whenever the CER is updated. A CER that defines the state of the art based on literature from a decade ago, when significant clinical advances have occurred in the interim, will be found deficient. The state of the art is also relevant when assessing the acceptability of residual risks: a risk that was considered acceptable five years ago may no longer be acceptable if newer technologies have reduced that risk. Manufacturers should avoid conflating the state of the art with simply listing competitor devices. The assessment must encompass the broader clinical and technological landscape.

Understanding why Notified Bodies reject or raise deficiency findings against Clinical Evaluation Reports is essential for manufacturers seeking to avoid costly review cycles and certification delays. Drawing on our extensive experience supporting manufacturers through the Notified Body review process, the most frequently encountered CER deficiency categories include the following. Insufficient justification of equivalence is the single most common deficiency under the MDR. Manufacturers who claim equivalence without addressing all three dimensions (clinical, technical, biological) with sufficient specificity and supporting evidence, or who claim equivalence to a competitor's device without an access agreement to the technical documentation, will receive deficiency findings. Inadequate or non-systematic literature searches are another frequent finding. Notified Bodies expect documented, reproducible search protocols, PRISMA-compliant documentation, and searches that cover multiple databases. A PubMed-only search with vague search terms and no documented screening process will not suffice. Superficial benefit-risk analysis, where the manufacturer states that benefits outweigh risks without structured analysis, quantification, or reference to the state of the art, is consistently flagged. Missing or inadequate state-of-the-art assessment, where the clinical context and treatment alternatives are not defined or are outdated, undermines the entire evaluation. Poor data appraisal, where the evaluator fails to critically assess the quality and relevance of individual data sources or treats all publications as equally reliable, is a common weakness. Failure to address unfavourable data is a serious deficiency; the CER must acknowledge and analyse data that is not supportive of the device's safety or performance claims, including adverse event reports and negative clinical outcomes. Insufficient evaluator qualifications, where the CER does not document the qualifications, training, and relevant clinical experience of the evaluator, will be flagged, as MEDDEV 2.7/1 Rev 4 sets explicit requirements for evaluator competence. Finally, a CER that is clearly outdated, with no evidence of periodic updating or incorporation of post-market data, signals to the Notified Body that the manufacturer's clinical evaluation process is not functioning as a living system.

The MDR establishes clear requirements for the ongoing updating of the Clinical Evaluation Report, reinforcing that clinical evaluation is a continuous process, not a one-time pre-market exercise. Article 61(11) requires that the clinical evaluation and its documentation shall be updated throughout the life cycle of the device concerned with clinical data obtained from the implementation of the manufacturer's PMCF plan and the PMS plan. Annex XIV Part A, Section 1(c) further specifies that the clinical evaluation shall be thoroughly documented, including a report that is regularly updated with data obtained from post-market surveillance and PMCF activities. MDCG 2020-13 on clinical evaluation assessment reports provides additional guidance on the frequency and triggers for CER updates. For Class III and implantable devices, the CER should be updated at least annually. For Class IIa and IIb devices, the update frequency should be defined in the Clinical Evaluation Plan and should be proportionate to the risk and the maturity of the clinical evidence base, but updates at least every two to five years are generally expected. Beyond scheduled updates, certain events should trigger an ad hoc CER update: a significant change to the device design, materials, or manufacturing process that could affect clinical safety or performance; new safety signals identified through post-market surveillance or vigilance; publication of significant new clinical data or changes to the state of the art; issuance of new or revised harmonised standards, common specifications, or guidance documents that affect the clinical evaluation; feedback from Notified Body audits or competent authority inspections; and changes to the intended purpose or indications for use. Manufacturers must document their CER update triggers in the Clinical Evaluation Plan and must maintain an auditable record of when and why updates were performed. The CER update should not merely append new data but should re-evaluate the entirety of the clinical evidence in light of the new information, potentially revising conclusions and updating the benefit-risk determination. A CER that has not been updated since the initial submission is a red flag for Notified Bodies and will almost certainly result in a deficiency finding during surveillance audits.

The MDR establishes an integrated framework in which the Clinical Evaluation Report, the Post-Market Clinical Follow-up plan and report, and the Post-Market Surveillance system operate as interconnected components of a continuous evidence lifecycle. Understanding and implementing this interconnection is essential for MDR compliance and is an area where many manufacturers fall short. The CER, as described above, is the document that analyses and synthesises clinical data to reach conclusions about the device's safety, clinical performance, and clinical benefit. The PMS system, governed by MDR Articles 83 to 86 and Annex III, is the manufacturer's framework for proactively and systematically collecting and analysing data on the device's real-world performance, including complaints, incidents, trend analyses, and published literature monitoring. The PMCF, governed by MDR Annex XIV Part B, is a specific component of PMS that involves proactive clinical data collection through PMCF studies, PMCF surveys, or participation in device registries, designed to address specific clinical questions identified in the clinical evaluation. The feedback loop operates as follows: the Clinical Evaluation Plan identifies gaps or uncertainties in the clinical evidence that require further investigation. These gaps inform the PMCF plan, which defines the specific activities (studies, surveys, registry participation) to address those gaps. The PMCF evaluation report documents the findings from PMCF activities. These findings, together with the broader PMS data, feed back into the CER update, which re-evaluates the clinical evidence in light of the new data. The updated CER may, in turn, identify new questions or gaps that require modifications to the PMCF plan. This cyclical process ensures that the clinical evidence base matures over time and that emerging safety or performance signals are systematically captured and evaluated. Common deficiencies include PMCF plans that are generic templates with no clear link to the CER's identified evidence gaps, CERs that do not incorporate available PMS or PMCF data, and PMS systems that operate in isolation from the clinical evaluation process. Manufacturers must ensure that the personnel responsible for clinical evaluation, PMS, and PMCF communicate and coordinate effectively, and that the documentation clearly demonstrates the linkages between these processes.

The MDR imposes heightened requirements for clinical evaluation of Class III and implantable devices, reflecting their inherently higher risk profile and the potential for serious, irreversible patient harm. Article 61(4) establishes a presumption that clinical investigations will be conducted for Class III devices and Class IIb active devices intended to administer or remove a medicinal product, unless the manufacturer can demonstrate that adequate clinical evidence already exists. For implantable devices, this presumption is particularly strong, and Notified Bodies will scrutinise any justification for not conducting a pre-market clinical investigation with considerable rigour. The types of clinical evidence that Notified Bodies expect for high-risk devices include long-term safety and performance data, ideally from clinical investigations or device registries with follow-up periods appropriate to the device's expected service life. For implantable devices with expected implant durations of ten years or more, clinical data covering only one or two years will generally be considered insufficient without a robust PMCF plan to capture longer-term outcomes. Equivalence claims for Class III and implantable devices face the highest level of scrutiny. The three-dimensional equivalence demonstration (clinical, technical, biological) must be exceptionally detailed and well-evidenced, and the MDR's requirement for a technical documentation access contract with the equivalent device manufacturer is strictly enforced. For novel Class III devices with no predicate or equivalent device, a clinical investigation is almost always required. Notified Bodies will also assess whether the device's benefit-risk profile has been adequately characterised for all intended patient subpopulations, including elderly patients, paediatric populations, and patients with comorbidities. The PMCF plan for Class III and implantable devices must be specific, proactive, and proportionate to the risk. Generic PMCF plans that merely commit to monitoring complaints and literature will be found insufficient. Notified Bodies expect to see planned PMCF studies with defined endpoints, follow-up periods, sample sizes, and statistical methodologies. The PMCF evaluation report must demonstrate that the PMCF activities are being conducted as planned and that the results support the device's continued safety and performance. The Periodic Safety Update Report (PSUR) for Class IIa, IIb, and III devices must also incorporate relevant CER conclusions and PMCF findings, creating yet another documentation linkage that must be managed systematically.

Drawing on our experience supporting manufacturers across device classifications and therapeutic areas, we offer the following practical recommendations for writing a CER that will satisfy Notified Body expectations. First, invest adequate time in the Clinical Evaluation Plan before beginning the evaluation itself. The CEP is not a bureaucratic formality; it is the methodological blueprint that determines whether your evaluation will be systematic and defensible. A well-crafted CEP defines the device scope, the clinical claims to be evaluated, the equivalence rationale (if applicable), the literature search strategy, the data appraisal methodology, and the criteria for benefit-risk acceptance. Second, appoint qualified clinical evaluators from the outset. MEDDEV 2.7/1 Rev 4 requires evaluators to have knowledge of research methodology, information management, and regulatory affairs, as well as relevant clinical expertise in the therapeutic area. Document these qualifications in the CER. Third, be transparent and self-critical in your data appraisal. Notified Bodies value intellectual honesty. Acknowledge limitations in the available evidence, discuss the quality of individual studies openly, and address unfavourable data directly rather than ignoring it. Fourth, structure your CER logically and make it navigable. Use clear section headings that align with the MEDDEV 2.7/1 Rev 4 stages, include a table of contents, and provide cross-references to supporting documents (risk management file, IFU, PMS plan, PMCF plan). Fifth, be specific in your benefit-risk analysis. Quantify benefits and risks where possible, cite supporting data for each claim, and explicitly compare your device's profile to the state of the art. Sixth, ensure that your CER addresses all relevant GSPR from Annex I with specific clinical evidence, not just general statements. Seventh, plan for CER maintenance from the beginning. Establish procedures, assign responsibilities, define update triggers, and build the clinical evaluation into your quality management system as a managed process. Finally, consider having an independent clinical evaluation review before Notified Body submission. A fresh set of expert eyes can identify gaps, inconsistencies, and weaknesses that the original evaluator may have overlooked.

Navigating the clinical evaluation requirements of the EU MDR demands a combination of regulatory expertise, clinical knowledge, and practical experience with Notified Body expectations. At Swiss MPC, our clinical affairs team has supported manufacturers of all device classifications, from Class I general medical devices to Class III implantable technologies, in developing clinical evaluation strategies, conducting systematic literature reviews, writing and updating Clinical Evaluation Reports, designing and managing PMCF plans and studies, and responding to Notified Body deficiency findings. Our consultants bring direct experience with leading EU Notified Bodies and understand the specific expectations and review tendencies of organisations such as BSI, TUV SUD, TUV Rheinland, DEKRA, and SGS. We know where CER reviews typically stall, what questions Notified Body reviewers ask, and how to structure documentation that facilitates an efficient review process. Our approach begins with a thorough assessment of your existing clinical evidence and regulatory strategy, followed by a gap analysis that identifies the specific clinical data and documentation needed to support your MDR submission. We then work alongside your team, or independently where preferred, to execute the clinical evaluation plan, conduct the literature search, appraise and analyse the data, and produce a CER that meets the requirements of MEDDEV 2.7/1 Rev 4 and all applicable MDCG guidance. For manufacturers facing CER deficiency findings from their Notified Body, we provide targeted remediation support, addressing the specific deficiencies identified and strengthening the overall clinical evaluation to prevent recurrence. We also offer clinical evaluation training for manufacturer teams who wish to build internal CER competency. If you are preparing for an MDR submission and need expert clinical evaluation support, we invite you to contact us to discuss your specific device, your clinical evidence landscape, and how we can help you achieve a compliant and defensible Clinical Evaluation Report.