The Class I Fallacy: Why Low Risk Does Not Mean "No Data" Under EU MDR

There is a dangerous, widely held belief circulating among medical device manufacturers:

“Because our device is Class I, we don’t need a real literature search, a State of the Art (SOTA) analysis, or any actual clinical data.”

This assumption is leading companies straight into major audit non-conformities.

While Article 10(9) and Article 83(1) of the EU Medical Device Regulation (MDR 2017/745) state that Quality Management Systems and Post-Market Surveillance must be proportionate to the risk class, proportionate does not mean exempt. Article 61 mandates that every single device—from a Class I reusable surgical scalpel to a Class III drug-eluting stent—requires a clinical evaluation based on sufficient clinical evidence.   

Let’s dismantle the biggest myths surrounding Class I Clinical Evaluation Reports (CERs) and outline what the data analysis differences between Class I and Class III devices actually are.

Myth 1: "Class I Automatically Means an Article 61(10) Non-Clinical Evaluation"

Article 61(10) of the MDR allows a manufacturer to base their clinical evaluation on non-clinical data alone (such as bench testing, performance testing, and usability evaluations). Many manufacturers assume that Class I = Article 61(10).   

This is incorrect.

Class I status does not automatically qualify a device for Article 61(10). To use Article 61(10), you must explicitly justify in your technical documentation why demonstration of conformity based on clinical data is not appropriate based on risk management and device-body interaction.   

Here is the kicker: If there is even a single published article in scientific literature containing clinical data on your subject device, you cannot use Article 61(10).

The presence of clinical data in the public domain invalidates the argument that clinical data is unavailable or inappropriate. Once clinical data for your subject device exists, you are legally required to perform a clinical literature search and execute the full 5 Stages of Clinical Evaluation under MEDDEV 2.7/1 rev 4 or ISO 18969:

1. Stage 0: Plan (Defining the Scope & Clinical Evaluation Plan)

2. Stage 1: Identify (Systematic literature retrieval)

3. Stage 2: Appraise (Evaluating data quality and relevance)

4. Stage 3: Analyze (Synthesizing data to determine compliance)

5. Stage 4: Report (Documenting conclusions in the CER)

You cannot skip Stage 2 or Stage 3 simply because you are a Class I device. If data exists, you must appraise and analyze it.

Myth 2: "The SOTA Search and the Clinical Literature Search Are the Same Thing"

Another common misconception is lumping the State of the Art (SOTA) literature search and the subject-device clinical literature search into a single query. They are two distinct searches with two entirely different regulatory goals:

• The SOTA Search: Examines the broader clinical landscape, benchmark technologies, alternative treatments, and clinical guidelines. Its job is to define the key output variables (performance and safety targets) and identify the known clinical risks for the product type.

• The Subject Device Search: Specifically searches for published clinical experience, case studies, or trials on your specific device (or a fully demonstrated equivalent). Its job is to verify whether your device actually achieves those SOTA targets in clinical practice.

Combining these into one lazy search query obscures the baseline benchmarks, making it impossible to evaluate your device's data objectively.

The Real Distinction: How Class I and Class III Data Analysis Differ

If both Class I and Class III devices require systematic literature searches, key output variables, and formal 5-stage clinical evaluations, where does "proportionality" actually apply?

The difference lies in the depth and granularity of the harm-rate calculations.

For Class III Devices (Granular Harm-Rate Modeling)

For a Class III device, you must conduct deep, quantitative risk-benefit calculations for every single residual risk and benefit. You must establish a "benefit rate" for each specific benefit and specific numerical adverse event rate for each individual harm across three separate data sources:

1. SOTA Literature: What is the baseline success rate for each clinical benefit and incidence rate for each residual Harm X across current medical practice?

2. Subject Device Literature: What is the published success rate for each clinical benefit and incidence rate for each residual Harm X across current medical practice?

3. Internal PMS/Vigilance Data: What is your real-world complaint/incident rate for Harm X post-market?

You must then quantitatively and statistically compare your device's clinical benefit rates and residual harm rates against SOTA benchmarks to demonstrate that your benefit-risk profile remains acceptable.

For Class I Devices (Output Benchmarking & Risk Listing)

For a Class I device, you are not required to execute complex mathematical benefit/harm-rate calculations for every minor adverse event in the literature.

Instead, it is sufficient to:

• Establish Key Output Variables: Identify the crucial performance and safety variables in your SOTA literature (e.g., functional success rates, usability error thresholds, primary safety outputs).

• Identify and List Residual Harms: Document all known residual risks and potential harms in your CER and Risk Management file.

• Analyze Retrieved Data: Analyze your subject device's clinical/bench data against the SOTA key output variables to confirm that your device meets clinical expectations and that no unexpected safety signals have emerged.

The Bottom Line

Low risk reduces the complexity of your proof—it does not eliminate the requirement for proof.

If you manufacture a Class I device:

1. Do not default to Article 61(10) without confirming that zero clinical data exists for your device.   

2. Keep your SOTA literature search separate from your subject device search.

3. If clinical data exists, run a full 5-stage clinical evaluation.

4. Focus your analysis on proving that your device meets the key output variables defined in the SOTA, rather than getting bogged down in granular harm-rate calculations.

By recognizing these distinctions, you can keep your Class I CER compliant, audit-ready, and legitimately proportionate to your device's risk profile.