State of the art in an MDR clinical evaluation is the consolidated, current clinical knowledge of what good looks like for treating or diagnosing a specific condition. Under MDR Article 61 and Annex XIV, you must define it, benchmark your device against it, and show that your benefit-risk profile is acceptable when compared to currently available alternatives.

By Tibor Zechmeister and Felix Lenhard. Last updated 10 April 2026.

TL;DR

  • MDR Article 61 requires clinical evaluation to establish and verify conformity with relevant General Safety and Performance Requirements under the normal conditions of the intended use of the device, evaluated against the state of the art.
  • Annex XIV, Part A requires the clinical evaluation plan to identify the state of the art in the relevant medical field, including available therapeutic and diagnostic alternatives.
  • "State of the art" in a Clinical Evaluation Report is clinical, not just technical — it is about current medical practice, current outcomes, and currently available alternatives, not only about your device's technology stack.
  • Benefit-risk acceptability is not an internal judgement. It is acceptability against the current standard of care. If another, widely used intervention is safer and more effective, your device has a problem.
  • Startups fail here by defining state of the art as "whatever our device does." Notified Bodies reject that. The state of the art section must be written before you know where your device lands.

Why your CER lives or dies on this section

A clinical evaluator at a Notified Body opens your Clinical Evaluation Report. They do not start at the device description. They do not start at your clinical data. They start at the state of the art section — because that section tells them whether the rest of the document is even written against the right benchmark.

If your state of the art is wrong, everything downstream is wrong. Your benefit-risk analysis is measured against the wrong yardstick. Your clinical data appraisal is filtered against the wrong criteria. Your PMCF plan asks the wrong questions. The reviewer does not need to read the rest of the CER to know it is going back for revision.

This is the single most common structural failure we see in Notified Body submissions. Not missing data. Not weak statistics. A state of the art section that was written as an afterthought, treated as a literature-review formality, and disconnected from the benefit-risk conclusion it was supposed to support.

Separately, post 40 covered how the state of the art principle in Annex I shapes your design decisions. This post is about the other half of the same principle — how it shapes your clinical evaluation and how you demonstrate benefit-risk acceptability.

What the MDR actually says

MDR Article 61(1) requires manufacturers to plan, conduct, and document a clinical evaluation to establish and verify conformity with the relevant general safety and performance requirements under normal conditions of the intended use. The clinical evaluation must follow a defined and methodologically sound procedure based on a critical evaluation of relevant scientific literature, a critical evaluation of the results of all available clinical investigations, and consideration of currently available alternative treatment options. (Regulation (EU) 2017/745, Article 61, paragraph 1.)

Annex XIV, Part A sets out what a clinical evaluation plan must contain, including an identification of available alternative treatment options for that purpose, if any, and specification of methods to be used for examination of qualitative and quantitative aspects of clinical safety with clear reference to the determination of residual risks and side-effects. (Regulation (EU) 2017/745, Annex XIV, Part A.)

Annex I, Section 1 requires that devices achieve the performance intended by their manufacturer and be designed and manufactured in such a way that they are suitable for their intended purpose, with any risks acceptable when weighed against the evaluated benefits to the patient and compatible with a high level of protection of health and safety, taking into account the generally acknowledged state of the art. (Regulation (EU) 2017/745, Annex I, Section 1.)

Annex I, Section 8 requires that all known and foreseeable risks, and any undesirable side-effects, be minimised and acceptable when weighed against the evaluated benefits to the patient and/or user arising from the achieved performance of the device during normal conditions of use. (Regulation (EU) 2017/745, Annex I, Section 8.)

Three phrases carry the entire weight: "generally acknowledged state of the art," "currently available alternative treatment options," and "acceptable when weighed against the evaluated benefits." The rest of this post is about translating those phrases into a CER section a Notified Body will accept.

State of the art is clinical, not technical

The most damaging misunderstanding in startup CERs is treating state of the art as a technology discussion. Startups write paragraphs about sensor resolution, algorithm architectures, or manufacturing tolerances — and call it a state of the art chapter.

A Notified Body clinical reviewer is not reading for technology. They are reading for medicine. The question they are answering is: what is the current clinical approach to this condition, and how does this device compare?

The state of the art in a CER must establish:

  1. The clinical condition or use case — what the device is intended to diagnose, monitor, treat, or prevent, described in clinical terms a physician would recognize.
  2. The current standard of care — what clinicians currently do for patients with this condition, including pharmacological treatments, surgical interventions, other device-based treatments, watchful waiting, and lifestyle modifications where relevant.
  3. Current clinical outcomes with those alternatives — efficacy, safety profiles, complication rates, quality-of-life data, and known limitations, drawn from peer-reviewed literature and clinical practice guidelines.
  4. Known and accepted residual risks in current practice — what complications patients currently accept as part of treatment, because this becomes the reference point for your benefit-risk comparison.
  5. Gaps or unmet needs in current practice — where current alternatives underperform, because this is where a new device can legitimately justify itself.

Only then, against that clinical backdrop, can your device's data be interpreted.

Benchmarking the alternatives

Annex XIV, Part A(a) explicitly requires identification of available alternative treatment options for the intended purpose. This is not a literature-review box to tick. It is a benchmarking exercise with teeth.

For each relevant alternative, the CER should document:

  • What the alternative is — drug, device, surgical procedure, or non-intervention.
  • Indication and patient population — is it used for the same clinical scenario as your device, or a related one?
  • Efficacy data — what outcomes does it achieve, measured against what endpoints, in what populations?
  • Safety profile — what adverse events, complications, contraindications, and residual risks are associated with it?
  • Practical considerations — availability, cost, training requirements, accessibility for the target population.
  • Evidence level — is this data from randomized controlled trials, observational studies, registry data, or clinical consensus?

The output is a defensible picture of what patients currently get. That picture is the baseline against which your device's benefit-risk profile must be judged. MDCG 2020-6 on sufficient clinical evidence for legacy devices, and MEDDEV 2.7/1 revision 4 for methodology, both reinforce that this benchmarking is not optional and not cosmetic.

Benefit-risk acceptability: the framework

Benefit-risk acceptability under MDR is a comparative judgement, not an absolute one. Annex I does not say risks must be zero. It says risks must be acceptable when weighed against the evaluated benefits and compatible with a high level of health protection, taking into account the generally acknowledged state of the art.

Acceptable to whom, compared to what? The answer in practice is: acceptable to an informed clinician and patient, compared to what they would otherwise receive under the current standard of care.

A working framework for the CER:

  1. Quantify the benefits your device delivers, using your own clinical data and any equivalence or similar-device data justified under MDR Article 61 and MDCG 2020-5.
  2. Quantify the risks from your risk management file under EN ISO 14971:2019+A11:2021, including residual risks after all control measures.
  3. Map both against the state of the art baseline — where does your device sit relative to current alternatives on each clinically meaningful endpoint and each clinically meaningful risk?
  4. Identify the trade-off — most devices are not uniformly better. They are better on some dimensions and comparable or worse on others. The honest question is whether the trade-off is acceptable for the target population given what they otherwise get.
  5. State the conclusion explicitly — the CER must contain a reasoned, written conclusion that the benefit-risk profile is acceptable in the context of the state of the art, not leave the reviewer to infer it.

If your device is worse than the current standard of care on the metrics that matter most to patients, the answer is not to hide the comparison. The answer is to narrow the indication, change the target population, or rethink whether the device should ship at all.

Worked example: a wearable heart rhythm monitor

A startup is certifying a wrist-worn continuous cardiac rhythm monitor intended for detection of atrial fibrillation in adults with cryptogenic stroke, as a Class IIa device. Their CER state of the art section needs to do the following work.

Clinical condition: Atrial fibrillation detection post cryptogenic stroke, where identifying paroxysmal AF changes anticoagulation management and reduces recurrent stroke risk.

Alternatives benchmarked: Standard 12-lead ECG at clinic visits, 24-hour Holter monitoring, 7-day Holter monitoring, 30-day event recorders, implantable loop recorders, and other wrist-worn consumer-grade wearables that have been studied clinically.

Efficacy baseline from literature: Published AF detection yields at 6 and 12 months, broken down by monitoring duration and method. Implantable loop recorders set the high-sensitivity ceiling. External patches and event recorders sit in the middle. Intermittent clinical ECGs sit at the floor.

Safety baseline: Implantable loop recorders carry procedural risks — infection, device migration, explant. External methods carry skin reactions and adherence burden. Clinical ECG carries effectively no device risk but very low detection yield.

Trade-off: The wrist wearable will likely show lower sensitivity than an implantable loop recorder but higher sensitivity than intermittent clinic ECG, with essentially zero procedural risk and much higher patient acceptability for long-term wear.

Benefit-risk conclusion: Acceptable for a defined subset of the population for whom implantable monitoring is not clinically indicated or not accepted, and for whom the current alternative is effectively intermittent clinical ECG. Not acceptable to position as a replacement for loop recorders in patients where the clinician would otherwise implant one.

That is a CER conclusion a Notified Body reviewer will engage with. It is specific, it is honest about trade-offs, and it lands on a clearly bounded indication. The intended purpose statement and the target population on the label must then match this conclusion exactly.

Ship: the state of the art section playbook

A repeatable structure for the state of the art section of your CER:

  1. Scope statement — one paragraph, clinical language, defining the condition and the population.
  2. Current standard of care — describe current clinical practice, citing clinical practice guidelines (ESC, AHA, NICE, or equivalent for your field) and current consensus statements.
  3. Alternative treatment options table — a structured comparison across efficacy, safety, evidence level, and practical considerations.
  4. Unmet needs or gaps — supported by literature, not by marketing claims.
  5. Benchmarks for your device — the specific efficacy and safety thresholds your device must meet or exceed to be considered clinically acceptable.
  6. Reference to your risk management file and clinical data sections — so the state of the art section is wired into the rest of the CER, not a standalone essay.
  7. Explicit statement of acceptability criteria — what would make your device's benefit-risk acceptable, defined before you look at your own data.

Write the state of the art section before you write your benefit-risk conclusion. If you write it after, you will subconsciously shape it to fit the conclusion you want, and the Notified Body will notice.

Keep the section living. Under Annex XIV Part B, the clinical evaluation is updated throughout the device lifecycle, fed by PMCF. The state of the art section gets re-examined at every update. New guidelines, new landmark trials, and new alternatives on the market all change the baseline.

Reality Check — Where do you stand?

Answer these honestly before your next CER revision.

  1. Does your state of the art section read like a clinical chapter or like a technology chapter? If technology, rewrite it.
  2. Can you name the three most relevant alternative treatments for your device's indication, with citations to their current efficacy and safety data, without looking anything up?
  3. Is there a structured comparison table, or just narrative paragraphs?
  4. Did you write the state of the art section before or after you drafted your benefit-risk conclusion? Be honest.
  5. If your device is worse than an existing alternative on an important endpoint, does your CER acknowledge that openly and explain why the indication still makes clinical sense?
  6. Is the acceptability threshold you use in your benefit-risk analysis derived from the state of the art section, or did it appear from nowhere?
  7. When was the state of the art section last updated against new guidelines or new published trials? If the answer is "at initial certification," that is a finding waiting to happen.

Frequently Asked Questions

What does "state of the art" mean in an MDR Clinical Evaluation Report? In a CER context, state of the art is the consolidated current clinical knowledge and practice for managing the condition your device addresses. It covers the current standard of care, currently available alternative treatments, their outcomes, their risks, and the accepted benchmarks a new device is measured against. It is a clinical concept, not a technology concept.

Where does MDR require state of the art in clinical evaluation? MDR Article 61(1) requires clinical evaluation to consider currently available alternative treatment options. Annex XIV, Part A requires the clinical evaluation plan to identify available alternative treatment options. Annex I, Section 1 requires benefit-risk judgements to take account of the generally acknowledged state of the art. All three work together.

Is demonstrating state of the art the same as demonstrating equivalence? No. Equivalence under MDCG 2020-5 is a narrow pathway for using clinical data from a specific equivalent device. State of the art is a broader obligation to understand and benchmark against the entire current clinical landscape, regardless of whether you are claiming equivalence to any specific device.

How often does the state of the art section of a CER need to be updated? The clinical evaluation is updated throughout the device lifecycle, with update frequency defined by device risk class and the PMS plan under MDR Article 83. For implantable and Class III devices, Article 61(11) requires annual updates. Each update must reconsider whether the state of the art has moved.

What is the most common Notified Body finding on state of the art in CERs? The state of the art section is generic, not specific to the device's indication and target population, and not linked to the benefit-risk conclusion. Reviewers flag this as a structural failure because it means the benefit-risk analysis is not anchored to a defensible baseline.

Sources

  1. Regulation (EU) 2017/745 of the European Parliament and of the Council of 5 April 2017 on medical devices, consolidated text. Article 61 and Annex XIV. Available via EUR-Lex.
  2. Regulation (EU) 2017/745, Annex I, Section 1 and Section 8 — general safety and performance requirements on benefit-risk and state of the art.
  3. MDCG 2020-5, Clinical Evaluation — Equivalence: A guide for manufacturers and notified bodies, April 2020.
  4. MDCG 2020-6, Regulation (EU) 2017/745: Clinical evidence needed for medical devices previously CE marked under Directives 93/42/EEC or 90/385/EEC, April 2020.
  5. MDCG 2020-13, Clinical evaluation assessment report template, July 2020.
  6. MEDDEV 2.7/1 revision 4, Clinical Evaluation: A Guide for Manufacturers and Notified Bodies, June 2016.
  7. EN ISO 14155:2020 + A11:2024, Clinical investigation of medical devices for human subjects — Good clinical practice.
  8. EN ISO 14971:2019 + A11:2021, Medical devices — Application of risk management to medical devices.

This post is part of the Clinical Evaluation & Investigations series in the Subtract to Ship: MDR blog. Authored by Felix Lenhard and Tibor Zechmeister. For startup-specific regulatory support on clinical evaluation strategy, see Zechmeister Strategic Solutions.