Written and maintained by CASRAI Editorial Board
Last updated
A clinical evaluation report (CER) is the document in which a medical device manufacturer assembles and appraises the clinical evidence for a device, then concludes whether that evidence demonstrates conformity with the safety and performance requirements of the EU Medical Device Regulation (MDR, Regulation (EU) 2017/745). It is not a one-time submission artefact. Article 61 frames clinical evaluation as ‘a systematic and planned process to continuously generate, collect, analyse and assess the clinical data’ for a device, and the CER is the point-in-time output of that continuous process — reviewed by the notified body at CE certification, then revisited on an ongoing basis as post-market data accumulates.
This guide covers what a CER actually needs to contain, using the MEDDEV 2.7/1 Rev 4 structure notified bodies still expect in practice, what MDR Annex XIV Part A demands for evidence to count as sufficient, how the equivalence route to a comparator device changed under MDR compared to the old Medical Device Directive (MDD) regime, and how the CER connects to the post-market clinical follow-up (PMCF) plan and evaluation report.
The Legal Basis: Article 61 and Annex XIV Part A
Article 61(1) sets the general principle: clinical evaluation must follow a defined method, proportionate to the device’s classification and intended purpose, and it must be updated throughout the device’s lifecycle. Article 61(3) sets out the sources a clinical evaluation is permitted to draw on:
- Clinical investigations carried out on the device itself;
- Clinical investigations or other studies reported in scientific literature, of a device for which equivalence to the device in question can be demonstrated;
- Reports published in peer-reviewed scientific literature on other clinical experience of either the device itself or an equivalent device; and
- Post-market surveillance data, in particular post-market clinical follow-up.
Annex XIV Part A then specifies what the clinical evaluation itself has to establish: an analysis of the current state of the art in the relevant clinical field (including comparable devices and treatment alternatives), a clinical development plan describing how sufficient clinical evidence will be generated or gathered, and — the operative conclusion — an explicit assessment of whether the clinical evidence assembled is sufficient to demonstrate conformity with the relevant General Safety and Performance Requirements in Annex I. A CER that only compiles literature without reaching a reasoned sufficiency conclusion against Annex I fails on structure, regardless of how much data it contains.
CER Structure: What MEDDEV 2.7/1 Rev 4 Still Gets Manufacturers
MDR itself does not prescribe a CER table of contents; Annex XIV Part A states the required content, not the document’s section headings. In practice, notified bodies and manufacturers have kept using the structure from MEDDEV 2.7/1 Revision 4 (December 2016) — the guidance document written for the predecessor MDD regime — because its logical sequence maps cleanly onto what Annex XIV requires and no MDR-specific replacement has superseded it wholesale. A CER built on the MEDDEV 2.7/1 Rev 4 template typically works through:
- Scope of the evaluation — the device, its variants/accessories, its intended purpose, and the specific claims the evaluation is meant to support;
- Clinical background and state of the art — the clinical condition addressed, current treatment options, and how the device compares to alternatives already on the market;
- Clinical development plan — what evidence exists already and what gaps the manufacturer’s ongoing and planned activity (including PMCF) is meant to close;
- Literature search protocol and literature search report — the search strategy defined in advance (databases, terms, inclusion/exclusion criteria, date range), then the results it actually returned — kept as two distinct outputs, not merged into one narrative, because notified bodies check the protocol against the report to see whether the search was run as planned;
- Appraisal and weighting of the data — each data source scored for methodological quality and relevance to the device and its intended purpose, with the weighting rationale documented, not just a data table;
- Analysis of the clinical data — benefits and risks assessed against the state-of-the-art comparators identified earlier, not in isolation;
- Conclusions — the explicit sufficiency judgment Annex XIV Part A requires, plus residual risks and any labelling/IFU implications; and
- Qualification of the evaluators — documented expertise of whoever performed the evaluation, since Annex XIV Part A also requires evaluators to be suitably qualified in the relevant clinical/scientific field.
Two Medical Device Coordination Group (MDCG) documents supplement this territory under MDR without replacing the core structure: MDCG 2020-6 addresses sufficient clinical evidence for legacy devices already on the market under MDD, and MDCG 2020-13 is the notified body’s own clinical evaluation assessment report template — useful to read because it shows exactly what a reviewer is checking your CER against, section by section.
What “Sufficient” Clinical Evidence Actually Means
Annex XIV Part A does not set a fixed evidentiary bar that applies uniformly across every device; sufficiency is evaluated against the device’s risk classification, its novelty, and the state of the art for its clinical field. A well-established device type with a long safety record and strong equivalent-device literature can sometimes support its Annex I conformity claims through literature and equivalence alone. A novel device, a new intended use, or a device in a clinical area lacking a settled state-of-the-art comparator will generally need clinical investigation data of the device itself — literature alone rarely closes that gap credibly. The state-of-the-art section is where notified bodies focus disproportionate scrutiny, because a CER that understates the available alternatives, or overstates how directly comparable its cited literature actually is, undermines the sufficiency conclusion that depends on it.
The Equivalence Route: Tighter Under MDR Than It Was Under MDD
Article 61(4) sets out what a manufacturer must demonstrate before relying on clinical data generated on a different device: equivalence across three defined characteristics — clinical (same clinical condition/intended purpose, similar severity/stage of disease, comparable patient population and site of use), technical (comparable design, conditions of use, specifications, deployment method), and biological (same materials/substances in contact with the same human tissues or body fluids, comparable release of substances). All three have to be substantiated with real technical/scientific justification, not simply asserted.
MDR also adds a practical constraint that did not exist as sharply under MDD: where the equivalent device is not the manufacturer’s own, Article 61(5) requires the manufacturer to have a level of access to the equivalent device’s technical documentation — normally through a contractual arrangement — sufficient to actually substantiate equivalence. Citing a competitor’s published literature is no longer enough on its own if the manufacturer cannot show it has meaningful visibility into that device’s technical file; this closed a route that was used loosely, and inconsistently policed, in the MDD era.
The other significant tightening sits in Article 61(6): for implantable devices and Class III devices, clinical investigations are required unless specific narrow conditions are met — broadly, the device is a modification of the manufacturer’s own already-CE-marked device with sufficient supporting data, or the equivalence route under Article 61(4)/(5) applies and the notified body agrees clinical investigation is unnecessary, or the device is otherwise covered under legacy/well-established-technology provisions. In practice this means the equivalence-only CER that was routine for many implantables and Class III devices under MDD is no longer a safe default under MDR; manufacturers of this device tier should expect a device-specific clinical investigation to be the baseline assumption, with equivalence as the narrower exception rather than the norm.
The CER-to-PMCF Feedback Loop
A CER is not filed once and left static. Its stated conclusions are only as current as the post-market evidence feeding it, which is where the PMCF plan under Annex XIV Part B comes in. The PMCF plan specifies how the manufacturer will proactively collect post-market clinical data on the device in real-world use; the resulting PMCF evaluation report becomes part of both the CER and the wider technical documentation. Concretely, the loop runs: the CER’s clinical development plan identifies what evidence gaps remain → the PMCF plan defines how those gaps get closed post-market (literature surveillance, registries, or a dedicated PMCF study) → PMCF findings update the CER’s state-of-the-art and sufficiency conclusions at the next revision → and any safety signal serious enough to warrant field action is escalated separately, not folded quietly into the next CER update.
CER updates also feed, and are fed by, two other post-market documents: the broader post-market surveillance (PMS) plan that PMCF sits inside, and the Periodic Safety Update Report (PSUR), which aggregates PMS and PMCF findings on a schedule set by device class. A notified body reviewing a CER at recertification or during a technical documentation audit checks that the CER’s cited PMCF/PMS data is current and consistent with what those parallel documents actually report — a CER that hasn’t been refreshed to reflect a manufacturer’s own PMCF findings is a common audit finding, not a hypothetical one.
Where the CER Sits in the Wider Technical Documentation
The CER doesn’t stand alone; it is assessed alongside, and must stay consistent with, the rest of the device’s technical file under the manufacturer’s ISO 13485 quality management system. Its device-description content should match the design history file and the outputs of design controls; its risk conclusions must line up with the device’s risk management file; and where the device incorporates software, the clinical evaluation needs to stay consistent with the classification logic covered in CASRAI’s guide to Software as a Medical Device (SaMD). Devices relying on unique device identification for post-market traceability, including registry-based PMCF, depend on consistent UDI capture to make that post-market data usable in the first place. For devices also carrying an IVDR performance-evaluation counterpart, see CASRAI’s guide to IVDR classification and notified body requirements — the CER’s structural logic (state of the art, evidence sufficiency, ongoing update obligation) parallels IVDR’s performance evaluation report, though the two are governed by separate regulations with their own article numbering. And where a notified body is involved in reviewing the CER at all, see CASRAI’s guide to notified body designation and audits for how that review relationship actually works day to day.
Frequently Asked Questions
Is a clinical evaluation report required for every MDR-covered device, regardless of class?
Yes. Article 61 and Annex XIV Part A apply across risk classes; what changes by class and novelty is how much clinical investigation data is realistically needed to reach a defensible sufficiency conclusion, and how tightly the equivalence route is constrained (see the implantable/Class III restriction above).
How often does a CER need to be updated?
MDR does not set one fixed interval for every device; the update cadence is proportionate to risk and is generally aligned with the device’s PMS/PMCF and PSUR cycle. Higher-risk devices are reviewed and updated more frequently, and any significant new safety signal or PMCF finding should trigger an update outside the routine schedule rather than waiting for the next planned cycle.
Can I rely on equivalence instead of running my own clinical investigation?
Sometimes, but MDR narrowed this route compared to MDD. You need to substantiate clinical, technical, and biological equivalence under Article 61(4), and if the comparator isn’t your own device, you need contractual access to its technical documentation under Article 61(5). For implantable and Class III devices, Article 61(6) restricts reliance on equivalence to narrow circumstances, making a device-specific clinical investigation the practical default rather than the exception.
Is MEDDEV 2.7/1 Rev 4 still valid guidance under MDR?
MEDDEV 2.7/1 Rev 4 was written for MDD and has not been formally reissued as an MDR-specific document, but its content structure is still widely used by manufacturers and referenced by notified bodies because it operationalises what Annex XIV Part A requires. Treat MDCG 2020-13 (the notified body’s own CER assessment template) as the more current cross-check on what an MDR reviewer is actually looking for.
What’s the practical difference between the CER and the PMCF evaluation report?
The CER is the overall clinical evaluation conclusion for the device, covering pre-market and post-market evidence together. The PMCF evaluation report is a narrower, post-market-only document that reports what the PMCF plan’s activities actually found; its findings feed into and update the CER rather than replacing it.
