Medical Device Classification: Implications for Testing Strategy and Preclinical Evidence

A startup is developing software that generates clinical recommendations from image data. The CEO plans to obtain CE marking within twelve months. In month nine, the regulatory consultant determines: the classification of the medical device was never formally carried out. The product falls under Rule 11 and is therefore not in Class I, but at least in Class IIa. Depending on the risk potential of the supported decisions, Class IIb or III may also apply. The consequences are additional evidence requirements, extra effort in planning, and potentially significant delays in the conformity assessment process.

Such scenarios occur in practice, particularly when classification is not integrated into development planning at an early stage. They arise because teams treat the classification of their medical devices not as an initial strategic decision but as an administrative follow-up step. Yet the risk class significantly shapes the scope of testing and influences the evidence strategy, level of documentation, timeline, and resource requirements:

• study design,
• number of test samples,
• involvement of Notified Bodies,
• timeline, and
• budget.

Note on the article series: This article deals exclusively with classification as a planning trigger, not the MDR architecture as a whole. For a complete overview of the regulatory framework, please refer to our article on technical documentation.

Why the classification of medical devices must be clarified early

The most common planning mistake: treating classification as a follow-up step

Manufacturers that address MDR classification only after the design phase risk three specific consequences:

1. Higher evidence requirements: The transition from Class IIa to IIb can significantly increase the requirements for clinical evidence, especially where equivalence arguments are not sufficiently substantiated. The use of equivalence data is interpreted more restrictively, which often makes manufacturer-generated clinical data virtually indispensable.
2. Additional study costs: The jump from Class IIa to IIb or III has a major impact on the budget. This can significantly increase project costs and substantially alter budget assumptions.
3. Time loss: Subsequent design or evidence adjustments can noticeably delay market access.

Device classification should therefore be addressed in the first project phase, ideally before the first protocol is drafted.

Basic logic of classification: Risk-based, but not self-explanatory

The EU MDR 2017/745 regulates the classification of medical devices in Annex VIII with 22 classification rules. These rules assign each product to one of four risk classes: I, IIa, IIb, and III.

The principle: the higher the risk to the patient, the stricter the regulatory requirements. In practice, many manufacturers struggle to apply these correctly, as multiple rules may apply, and in cases of conflict, the stricter classification takes precedence.

Which device characteristics determine the risk class

Five key questions form the core of every classification decision for medical devices. The following guiding questions serve as a pragmatic initial framework for the classification assessment:

This overview provides initial orientation and does not replace a product-specific classification rationale based on intended purpose and Annex VIII MDR.

MDR vs. FDA: When the same class has different consequences

The MDR classification and the FDA device classification follow fundamentally different regulatory logics. The EU uses 22 rule-based criteria in Annex VIII. The FDA works with over 1,700 predefined product codes and a predicate-based system (510(k), De Novo, PMA).

For manufacturers targeting both markets, this creates the challenge of a dual-pathway approval process. Two examples:

Example 1: A metal/polymer hip implant is Class III in the EU—but is often Class II (510(k)) with the FDA. Important exception: The FDA explicitly classifies metal-on-metal systems as Class III due to safety concerns (PMA required).

Example 2: For software under Rule 11, Class IIa is often the starting point for decision-support applications; however, depending on the clinical significance of the decision, Class IIb or III may also be considered. Many software-based medical devices fall under Class II in the U.S.; however, the specific premarket pathway depends on whether a suitable predicate device exists, often 510(k), or De Novo for novel products. This divergence should be built into the evidence and market access strategy early.

Implication Matrix: Classification and Its Typical Effects on Test Planning

The following matrix schematically illustrates how classification typically influences the scope of testing and evidence requirements:

Class I (Is / Im / Ir)	Scope of evidence: typically limited clinical evidence requirements; specific clinical and technical evidence depends on the intended use, risk profile, and product characteristics. Documentation level: technical documentation in accordance with Annex II/III. Notified Body: not required, except Is, Im or Ir.
Class IIa	Scope of evidence: clinical evaluation; depending on the product, possibly using an equivalence approach, supplemented by a PMCF plan and PSUR to the extent required by regulations. Documentation level: complete technical documentation and QMS certification, typically aligned with ISO 13485. Notified Body: required; assessment with a risk-based focus.
Class IIb	Scope of evidence: comprehensive clinical evidence, PMCF plan, and annual PSUR; for Class IIa, the PSUR is required every two years. Documentation level: extended technical documentation; often greater need for independent clinical evidence, while equivalence is assessed more restrictively. Notified Body: required; usually an in-depth review.
Class III	Scope of evidence: high clinical evidence burden; clinical trials are often required. SSCP for Class III as well as implantable products within the intended regulatory scope; CECP only in relevant scenarios. Documentation level: individual review of each technical documentation file and SSCP. Notified Body: required; usually comprehensive case-by-case review.

Class I: Reduced regulatory involvement, but no automatic reduction in evidence effort

Class I medical devices allow the manufacturer to issue the declaration of conformity themselves. There is no obligation to involve the Notified Body, provided that the product is neither sterile (Is), nor does it have a measuring function (Im), nor is it a reusable surgical instrument (Ir). These three subclasses require partial involvement of the Notified Body. The documentation requirements often remain manageable: risk analysis according to ISO 14971, clinical evaluation, and basic verification tests often form the basis; however, the specific scope depends on the product profile.

Class IIa-IIb: The most commonly underestimated category

Class IIa is the entry-level class requiring Notified Body involvement. The effort increases significantly: QMS audit, complete technical documentation, clinical evaluation with a viable equivalence approach where applicable, and a PMCF plan, as required by MDR Annex XIV Part B. However, many manufacturers underestimate the transition to Class IIb.

In practice, there is a particularly high risk of error in the classification decision here. Class IIb products sit between standard conformity expectations and a high-risk evidence burden: The Notified Body conducts more intensive reviews, annual PSURs (Periodic Safety Update Reports) become mandatory (Class IIa: every two years), and reliance on equivalence data is made significantly more difficult by the MDR. This makes it significantly more likely that proprietary clinical data will be required. A ventilator or a surgical laser typically requires a significantly different evidence and testing program than a typical Class IIa product. Misjudging this boundary can result in the loss of months of development time and budget.

Class III: Where the burden of evidence is highest

Class III encompasses the highest risk level: heart valves, active implants, and products with an integrated medicinal component. A high level of clinical evidence is typically required for Class III products; clinical trials are often necessary. The SSCP (Summary of Safety and Clinical Performance) is mandatory for all Class III products as well as for all implantable products, including implantable Class IIa/IIb products such as screws or suture material. The expert panel (Clinical Evaluation Consultation Procedure, CECP under Article 54 MDR), on the other hand, applies only to implantable Class III products and active Class IIb products intended for the administration or removal of medicinal products (Rule 12), and not automatically to every Class III product.

How to properly document the classification rationale

The rationale for the selected classification belongs in the technical documentation pursuant to MDR Annex II and must be clearly documented there. It is not a form, but rather a reasoned document that systematically reviews each applicable rule from Annex VIII and explains why the selected class is correct. Consulting practice shows that this document is often underestimated and frequently leads to follow-up questions during submission and evaluation processes.

What the Notified Body examines in the classification justification

Notified Bodies expect a transparent review of the relevant rules as well as a robust justification for why alternative, stricter rules do not apply:

1. Completeness: Were all 22 rules systematically evaluated, or just the one that seemed to apply? The most common weakness is selective evaluation. Manufacturers often check only one rule and ignore the fact that, for MDR classification, the strictest applicable rule applies.
2. Consistency with the intended use: The classification rationale must align precisely with the documented intended use, marketing materials, and instructions for use. Inconsistencies between the intended use and the selected class are particularly critical during the review and can lead to significant questions, deficiencies, or nonconformities.
3. Traceability: The justification must seamlessly demonstrate the causal path from product characteristics through the applicable rule to the resulting class. In practice, incomplete or inconsistent classification justifications frequently lead to inquiries during the evaluation.

This article is intended for general guidance. For your specific situation, we recommend a personalized consultation.

Frequently Asked Questions About the Classification of Medical Devices

Who makes the final decision on the classification of my medical device?

The manufacturer performs the initial classification. The Notified Body reviews this classification as part of the conformity assessment. In the event of discrepancies, competent authorities may issue a binding classification decision.

Can the classification change retroactively due to a design change?

Yes. Any change in intended use, contact with the human body, duration of use, or active components can result in a change in class. Manufacturers must reassess the medical device classification whenever a significant design change occurs. For example, reclassification from Class IIa to Class IIb or Class III requires an extended conformity assessment procedure.

What is the difference between classification and conformity assessment?

Classification determines the risk class based on the 22 rules in Annex VIII. Conformity assessment is the subsequent process that demonstrates that the product meets all regulatory requirements of the corresponding class. In short: classification is the input; conformity assessment is the process.

Does my MDR classification automatically apply to the FDA as well?

No. The FDA’s Medical Device Classification follows an independent system with its own product codes and classes (I, II, III). There is no automatic transfer. This is particularly evident, for example, with hip implants: conventional metal/polymer systems are often Class II under the FDA, whereas metal-on-metal systems are Class III and require PMA approval. MDR classification and FDA device classification must be determined separately for each target market.

Which class applies to diagnostic software under Rule 11?

For software that provides information for diagnostic or therapeutic decisions, Class IIa is often initially relevant under Rule 11. Class IIb applies only if the supported decisions could result in a serious deterioration of health or a surgical intervention. Class III applies to decisions with the potential for fatal or irreversible harm. According to MDCG 2019-11, the exact classification depends on the combination of the severity of the clinical situation and the significance of the information provided. A blanket assumption that “software = Class IIb” is not correct from a regulatory perspective.

Sources & Further Links

External References

• EU MDR 2017/745
• ISO 14971:2019 – Medical devices — Application of risk management to medical devices
• MDCG 2019-11 rev.1 – Qualification and classification of software – Regulation (EU) 2017/745 and Regulation (EU) 2017/746

Internal Links

• Medical Device Regulation: MDR architecture as the framework for classification
• Class III Medical Devices: A deeper look at high-risk device requirements
• Vascular Device Certification: A concrete example of MDR Class III products
• Preclinical Evidence: Evidence requirements by medical device class
• Animal Studies & FDA: Animal study requirements by device class
• Medical Device Testing: Which test types are required depending on classification

As of: May 2026 | Last reviewed: May 2026