Your technical documentation is nearly complete, but the clinical evaluation report is still a 2019 document written to MEDDEV 2.7/1 rev 4, with a literature search that only looked for your own device name. The notified body has booked its clinical reviewer, and you are not sure the report will survive the first round of questions.
This guide explains how to write a clinical evaluation report under MDR (Regulation (EU) 2017/745) that a clinical reviewer can follow from start to finish: what Article 61 and Annex XIV Part A require, how to structure the plan and report, where MEDDEV 2.7/1 rev 4 still helps, what MDCG 2020-6 expects for legacy devices, and how notified bodies use the MDCG 2020-13 template to assess your work.
What the MDR requires of a clinical evaluation report
Article 61(1) MDR requires that conformity with the relevant general safety and performance requirements (GSPRs) is based on clinical data providing sufficient clinical evidence. The manufacturer must specify and justify the level of clinical evidence needed and must plan, conduct and document a clinical evaluation in line with Article 61 and Annex XIV Part A.
Article 2 gives the vocabulary you should use consistently:
- Clinical evaluation (Article 2(44)) — a systematic and planned, continuous process to generate, collect, analyse and assess clinical data to verify safety and performance, including clinical benefits.
- Clinical data (Article 2(48)) — information from clinical investigations of the device, scientific literature on an equivalent device, peer-reviewed reports of clinical experience with the device or an equivalent device, and clinically relevant post-market information, in particular PMCF.
- Clinical evidence (Article 2(51)) — clinical data and evaluation results of sufficient amount and quality to judge safety and the intended clinical benefit.
Article 61(12) requires the clinical evaluation, its results and the clinical evidence to be documented in a clinical evaluation report, which forms part of the technical documentation under Annexes II and III (except for custom-made devices). Annex XIV Part A section 2 requires the evaluation to be thorough and objective and to take account of both favourable and unfavourable data, with depth proportionate to the device’s nature, classification, intended purpose, risks and claims. Correct classification therefore matters before you size the evaluation.
Start with the clinical evaluation plan
Annex XIV Part A section 1(a) makes the clinical evaluation plan (CEP) a legal requirement, not a template nicety. It must include at least:
- the GSPRs that require support from relevant clinical data;
- the intended purpose;
- clearly specified target groups with indications and contra-indications;
- a detailed description of intended clinical benefits with relevant and specified clinical outcome parameters;
- methods for examining qualitative and quantitative aspects of clinical safety, including residual risks and side-effects;
- an indicative list of parameters for judging the acceptability of the benefit-risk ratio, based on the state of the art in medicine;
- how benefit-risk issues of specific components (such as pharmaceutical or non-viable animal or human tissues) are addressed;
- a clinical development plan from exploratory studies to confirmatory investigations and PMCF, with milestones and potential acceptance criteria.
Write the CEP before the search. If your acceptance criteria are defined after you have seen the data, a reviewer will notice.
Structuring the clinical evaluation report under MDR, stage by stage
Annex XIV Part A section 1(b)–(e) sets out the method: identify available clinical data and gaps through a systematic literature review; appraise the data for suitability; generate new data through clinical investigations where gaps remain; and analyse all relevant data to reach conclusions on safety, clinical performance and clinical benefit. Article 61(3) adds that the evaluation must consider currently available alternative treatment options.
MEDDEV 2.7/1 rev 4 (June 2016) was written for the Directives, but its staged structure maps neatly onto Annex XIV: scope (stage 0), identification of pertinent data (stage 1), appraisal (stage 2), analysis (stage 3) and the report itself (stage 4). MDCG 2020-6 Appendix I lists which MEDDEV sections remain useful under the MDR, including section 8 (identification), section 9 (appraisal), section 10 (analysis), and appendices A5 (literature search protocol), A6 (studies lacking scientific validity) and A10 (CER release checklist). Where MEDDEV and the MDR differ, the MDR prevails.
State of the art and the literature search
Run and report separate searches for the device, any equivalent device and the state of the art. MDCG 2020-13 section D states that a search limited to your own product or chosen equivalent could miss important information and is not acceptable. Reviewers expect a pre-defined protocol, more than one database, lists of retrieved and excluded articles with reasons, and full texts available. Section C of the same template checks that benchmarks for safety and performance normally come from aggregate data and that any surrogate endpoints are justified.
Appraisal and analysis
Appraise each dataset for methodological quality, relevance to your device and intended purpose, and contribution to the evidence. Then analyse: does the pooled evidence show the clinical benefits stated in the CEP, is the residual risk acceptable against the state of the art, and which gaps remain? Every claim in the instructions for use (IFU) should trace back to this analysis.
Equivalence, clinical investigations and Article 61(10)
If you rely on data from another device, Annex XIV Part A section 3 and MDCG 2020-5 apply; see our guide to demonstrating equivalence under EU MDR. For implantable and class III devices, Article 61(4) requires clinical investigations unless one of the exemptions in Article 61(4)–(6) applies. Where clinical data are not deemed appropriate, Article 61(10) requires a justification based on risk management and the device–body interaction; MDCG 2020-6 notes this route is not available for class III or implantable devices.
Legacy devices: sufficient clinical evidence under MDCG 2020-6
MDCG 2020-6 (April 2020) addresses devices previously CE marked under Directives 93/42/EEC or 90/385/EEC. Its key messages for CER authors:
- Historical data remain an input, but must be re-assessed against the MDR definition of clinical data.
- Estimates such as incidents or complaints divided by sales generally cannot be considered sufficient proof of safety (section 6.3).
- Appendix III ranks clinical evidence from high-quality clinical investigations (rank 1) down to bench testing and standards compliance (rank 12). Class III and implantable legacy devices that are not well-established technologies should have sufficient clinical data at a minimum at rank 4.
- If the evidence is insufficient, the intended purpose may need to be narrowed (section 6.5e).
How notified bodies read your CER: the MDCG 2020-13 template
MDCG 2020-13 (July 2020) is the clinical evaluation assessment report (CEAR) template notified bodies use to document their review. Writing with it open is the simplest way to anticipate questions. Its sections cover administrative particulars and CER author CVs (A), reviewers (B), device description, CEP, equivalence, data access and state of the art (C), the literature review (D), clinical investigations (E), PMS, PMCF and update plans (F), IFU, summary of safety and clinical performance (SSCP) and labelling (G), a summary and conclusions (H), followed by overall conclusions and sections on the Article 54 consultation procedure (I), Article 61(10) (J) and Article 61(2) expert panel advice (K).
Section G is often underestimated: reviewers check that the intended purpose, population, contraindications and warnings in the IFU match the clinical evidence.
Keeping the CER current
Article 61(11) requires the clinical evaluation and its documentation to be updated throughout the life cycle with data from the manufacturer’s PMCF plan; for class III and implantable devices the PMCF evaluation report, and if indicated the SSCP, must be updated at least annually. Article 32 requires an SSCP, including a summary of the clinical evaluation, for implantable and class III devices. MEDDEV 2.7/1 rev 4 section 6.2.3 suggests annual CER updates for devices with significant risks or not yet well established, and every two to five years, with justification, for well-established lower-risk devices. Build the link to PMCF from day one; our article on the PMCF plan and evaluation report covers this in detail.
Where this goes wrong
- A CEP written after the CER. Benefit-risk acceptance criteria that appear only in the conclusions look reverse-engineered.
- Own-device-only literature searches. MDCG 2020-13 calls this not acceptable. Search the state of the art separately.
- Complaint rates presented as safety proof. MDCG 2020-6 says complaints-to-sales ratios generally cannot prove safety.
- Clinical benefits without outcome parameters. “Improves healing” is not a measurable clinical outcome.
- IFU claims the CER does not support. Every indication, population and user group in the IFU needs evidence; otherwise narrow the claim.
- MEDDEV-era equivalence carried over. Equivalence must now be shown against Annex XIV Part A section 3, which is stricter.
- Evaluators without documented expertise. MEDDEV section 6.4 expects a relevant degree plus five years’ documented experience (or ten years where a degree is not a prerequisite), with a declaration of interests; CEAR section A checks author CVs.
Frequently asked questions
Is MEDDEV 2.7/1 rev 4 still valid under the MDR?
It is not legally binding and was written for the Directives, but MDCG 2020-6 Appendix I lists the sections that remain useful, such as identification, appraisal and analysis of data and the CER checklist. MDCG 2020-5 confirms that where MEDDEV 2.7/1 rev 4 diverges from the MDR, the MDR takes precedence. Use it for method, not for legal requirements.
What is the difference between a clinical evaluation plan and a clinical evaluation report?
The clinical evaluation plan, required by Annex XIV Part A section 1(a), defines in advance the GSPRs needing clinical support, intended purpose, target groups, clinical benefits, outcome parameters and benefit-risk acceptance criteria. The clinical evaluation report, required by Article 61(12), documents the data found, how they were appraised and analysed, and the conclusions against that plan.
How often must a clinical evaluation report be updated under the MDR?
Article 61(11) requires updates throughout the life cycle using PMCF data, and annual updates of the PMCF evaluation report for class III and implantable devices. MEDDEV 2.7/1 rev 4 suggests annual CER updates for higher-risk or less established devices and two to five years, justified, for well-established lower-risk devices. Update earlier when PMS data could change conclusions.
Do class I devices need a clinical evaluation report?
Yes. Article 61 applies to all classes, and Article 61(12) requires the clinical evaluation to be documented in a clinical evaluation report within the technical documentation, except for custom-made devices. The depth is proportionate to the device’s nature, risk and claims, as Annex XIV Part A section 2 states, so a simple class I device will usually need a shorter report.
What is MDCG 2020-13 used for?
MDCG 2020-13 (July 2020) is the clinical evaluation assessment report template that notified bodies use to record their review of your CER. It covers device description, the plan, equivalence, the literature review, clinical investigations, PMS and PMCF, IFU and SSCP consistency, and overall benefit-risk conclusions. Manufacturers can use it as a pre-submission checklist.
Medfins has supported 175+ regulatory projects covering 250+ products for 100+ clients, including clinical evaluation plans and reports written for notified body review. If you need a CER drafted, gap-assessed against MDCG 2020-13 or remediated for a legacy device, see our CE marking under EU MDR service or contact our team.
